WEBVTT

1
00:02:40.410 --> 00:02:47.480
Uthappa Outskill: Yeah, you can be the greatest, you can be the best, you can be the King Kong
bangin' on your chest, you could beat the world

2
00:02:47.980 --> 00:02:51.820
Uthappa Outskill: You can beat the war, you could talk the guy, go bangin' on his door.

3
00:02:52.070 --> 00:02:53.650
Uthappa Outskill: Lookin' for your hands up.

4
00:02:53.910 --> 00:02:58.149
Uthappa Outskill: You can beat the clock, you can move a mountain, you can break rocks.

5
00:02:58.910 --> 00:03:01.100
Uthappa Outskill: Don't wait for luck.

6
00:03:02.760 --> 00:03:04.750
Uthappa Outskill: Sitting in the Hall of Fame!

7
00:03:24.590 --> 00:03:25.580
Uthappa Outskill: Close to the house!

8
00:03:26.150 --> 00:03:33.770
Uthappa Outskill: You can go the distance, you can run the mile, you can walk straight through
**** with a smile.

9
00:03:33.910 --> 00:03:36.359
Uthappa Outskill: Go I'm gonna make a sneak out there.

10
00:03:36.520 --> 00:03:41.330
Uthappa Outskill: Yeah, do it for your people, do it for your pride.
11
00:03:42.210 --> 00:03:43.160
Uthappa Outskill: even try.

12
00:03:44.990 --> 00:03:46.239
Uthappa Outskill: with your name.

13
00:03:47.120 --> 00:03:48.490
Uthappa Outskill: you and your head

14
00:03:58.810 --> 00:03:59.490
Uthappa Outskill: Cause goodbye!

15
00:04:23.920 --> 00:04:28.809
Uthappa Outskill: We students be teachers, be politicians be preachers.

16
00:04:28.890 --> 00:04:44.959
Uthappa Outskill: Be believers, be leaders, be astronauts, be champions, be truth seekers. Be
students, be teachers, be politicians, be preachers, be preachers. Be believers, be leaders, be
astronauts, be champions.

17
00:04:49.930 --> 00:04:50.630
Uthappa Outskill: The world's…

18
00:05:08.240 --> 00:05:08.960
Uthappa Outskill: He made his way.

19
00:05:44.140 --> 00:06:08.840
Uthappa Outskill: The club isn't the best place to find the lover, so the bar is where I go. Me and
my friends at the table doing shots, tripping fast, and then we talk slow. Come over and start up
a conversation with just me, and trust me, I'll give it a chance now.

20
00:06:08.840 --> 00:06:33.279
Uthappa Outskill: And then we start dancing now, singing like girl, you know I want your love.
Your love was handmade for somebody like me. Comin' now, follow my lead. I may be crazy,
don't mind me, say boy, let's not talk too much. Grab on my waist and put that body on me.
Comin' now, follow my lead, come and now, follow my lead.

21
00:06:33.780 --> 00:06:54.600
Uthappa Outskill: I'm in love with the shape of you. We push and pull like a magnetic. Although
my heart is falling too, I'm in love with your body. And last night you were in my room And now
my bedsheets smell like you. Every day discovering something brand new. I'm in love with your
body.

22
00:06:57.120 --> 00:06:59.599
Uthappa Outskill: I'm in love with your body more.

23
00:07:02.140 --> 00:07:04.610
Uthappa Outskill: I'm in love with your buddy, wow.

24
00:07:07.200 --> 00:07:09.209
Uthappa Outskill: I'm in love with your body.

25
00:07:09.510 --> 00:07:24.129
Uthappa Outskill: You every day discovering something brand new. I'm in love with the shape of
When we care when we let the story begin We're going out on our first date. You're thrifty, so go
all you can eat. Fill up your bag, and I fill up the plate.

26
00:07:24.480 --> 00:07:25.420
Uthappa Outskill: I was supposed…

27
00:08:04.650 --> 00:08:09.010
Uthappa Outskill: Look at the stars Look how they shine.

28
00:08:18.340 --> 00:08:19.359
Uthappa Outskill: Yeah, they were all.

29
00:08:19.600 --> 00:08:20.340
Uthappa Outskill: Yellow!
30
00:08:20.970 --> 00:08:27.169
Uthappa Outskill: I came along, I wrote a song for you

31
00:08:35.390 --> 00:08:37.129
Uthappa Outskill: It was cold yellow.

32
00:08:37.419 --> 00:08:38.840
Uthappa Outskill: Alright.

33
00:08:39.570 --> 00:08:41.490
Uthappa Outskill: Hello everyone, welcome back.

34
00:08:42.559 --> 00:08:48.950
Uthappa Outskill: Let me quickly open up the chat. A quick check if I'm audible visible, let me
know in the chat.

35
00:08:55.680 --> 00:08:56.760
Uthappa Outskill: Alright.

36
00:08:57.260 --> 00:08:59.520
Uthappa Outskill: Great, thank you, thank you, folks, thank you.

37
00:08:59.620 --> 00:09:06.930
Uthappa Outskill: We have a lot of people joining in, so I will let them give… probably I'll give
them a little more time, so that…

38
00:09:07.050 --> 00:09:09.489
Uthappa Outskill: We can all start together.

39
00:09:09.740 --> 00:09:15.680
Uthappa Outskill: So we'll give them 5 more minutes, so we will exactly start in the next, yeah, 5
minutes.

40
00:09:15.910 --> 00:09:31.829
Uthappa Outskill: Until then, what I want to know is, just curious to know, because people have
joined us in different, intervals in the last few weeks. So, let's… let's know, where are you joining
from, and I'm also curious to know what's the time in your place right now?

41
00:09:32.070 --> 00:09:34.730
Uthappa Outskill: So it's 7.30 PM IST here.

42
00:09:34.950 --> 00:09:39.170
Uthappa Outskill: Indian Standard Time. So also, let me know where you're from and what's the
time over there.

43
00:10:15.500 --> 00:10:21.880
Uthappa Outskill: Alright, so let me… It was all yellow.

44
00:10:28.580 --> 00:10:29.570
Uthappa Outskill: Oscar!

45
00:10:31.510 --> 00:10:32.319
Uthappa Outskill: I'm scared!

46
00:10:34.640 --> 00:10:35.360
Uthappa Outskill: God bless you.

47
00:10:36.340 --> 00:10:37.480
Uthappa Outskill: It's a song.

48
00:10:42.340 --> 00:10:44.440
Uthappa Outskill: You know how long this summer.

49
00:10:45.700 --> 00:10:46.390
Uthappa Outskill: Awesome.

50
00:10:48.920 --> 00:10:52.179
Uthappa Outskill: Thank you, thank you folks,

51
00:11:02.890 --> 00:11:04.020
Uthappa Outskill: Oh, there's a book assignment

52
00:11:10.090 --> 00:11:11.059
Uthappa Outskill: Let me know the audience?

53
00:11:12.520 --> 00:11:14.109
Uthappa Outskill: I swam across…

54
00:11:26.600 --> 00:11:29.010
Uthappa Outskill: Cause she will, yellow!

55
00:11:29.400 --> 00:11:31.080
Uthappa Outskill: A truer lie!

56
00:11:32.190 --> 00:11:34.159
Uthappa Outskill: I drew a line

57
00:11:39.030 --> 00:11:39.720
Uthappa Outskill: *****

58
00:11:43.780 --> 00:11:44.589
Uthappa Outskill: It was all…

59
00:11:51.830 --> 00:11:53.100
Uthappa Outskill: Just scared.

60
00:11:55.090 --> 00:11:56.640
Uthappa Outskill: Yes, Steven Bird!

61
00:11:57.320 --> 00:11:58.080
Uthappa Outskill: Chad!

62
00:11:59.590 --> 00:12:01.930
Uthappa Outskill: It's a sun, it's a flow.

63
00:12:05.400 --> 00:12:14.710
Uthappa Outskill: Hey folks, sure, just give us 2 more minutes. Kathy, we can stop the music
right now. I see a few people are asking for it.

64
00:12:21.840 --> 00:12:24.209
Uthappa Outskill: Alright, alright, don't worry, done, done.

65
00:12:28.830 --> 00:12:32.839
Uthappa Outskill: So we'll… another minute, then we'll start, beginning this.

66
00:12:39.060 --> 00:12:53.550
Uthappa Outskill: So, see, this is how we are going to proceed, okay? I'll let you know how we
are going to go with all the future sessions. We are definitely going to give all of you 5 minutes
before we start off with the session, so that everybody is in. We'll let you know…

67
00:12:54.050 --> 00:13:11.719
Uthappa Outskill: on how this whole thing is going to unwrap in the next one minute. Thanks for
joining, everyone, again. I hope all of you are excited to start this journey with us, to become the
AI journalist, the one that you had actually joined this program for. So let's see some excitement
in the,

68
00:13:11.890 --> 00:13:13.040
Uthappa Outskill: crowd.

69
00:13:14.750 --> 00:13:20.009
Uthappa Outskill: Yes, our favorite word, boom. How excited are you in terms of starting this
journey with us?

70
00:13:21.320 --> 00:13:22.540
Uthappa Outskill: Can't wait.
71
00:13:25.110 --> 00:13:25.930
Uthappa Outskill: Awesome.

72
00:13:29.210 --> 00:13:41.990
Uthappa Outskill: Great, great. I mean, love the energy, folks. It's a long journey that we have, 4
months of… sorry, 6 months of journey that we're going to have together. So we need this
energy, and we need this participation towards… for the entire journey that we have.

73
00:13:41.990 --> 00:13:51.989
Uthappa Outskill: So, hoping to see all of you in all the sessions that is coming forward. So,
today is going to be our first day in our long journey of becoming an AI generalist.

74
00:13:52.000 --> 00:13:54.020
Uthappa Outskill: Just a quick check.

75
00:13:55.540 --> 00:14:00.209
Uthappa Outskill: Of the folks that have attended here, I just want to know.

76
00:14:00.340 --> 00:14:14.849
Uthappa Outskill: Only those people who have not attended our… attended our onboarding call,
or watched our onboarding call recordings. Can you just let me know in the chat? I just want to
know how many of you have not attended or watched our onboarding call?

77
00:14:33.260 --> 00:14:42.800
Uthappa Outskill: Okay, I see, most of you have attended. Very few of you who would have
missed it. Let me just pause the chat here. Thank you, folks.

78
00:14:44.170 --> 00:15:02.979
Uthappa Outskill: Don't worry about it, the recordings are available, and then, yes, we have
notified every learner about the onboarding call. If you would have missed the onboarding, it
would be because you have joined us in the last two to… one or two days. So we have these
recordings uploaded on our dashboard, so go over it after this session. Try to get an
understanding of how we are going to proceed with our program.

79
00:15:02.990 --> 00:15:08.439
Uthappa Outskill: Well, Karthik, if you can stop presenting the screen, let me present my screen
and start off with the session without any further delay.

80
00:15:11.940 --> 00:15:28.120
Uthappa Outskill: So, while I do this, folks, a request to all of you here, please try to keep your
videos turned on, especially for this session. It's the beginning of our cohort. We need to know…
you need to know whom are you, interacting with. The mentor needs to feel that energy from the
crowd as well.

81
00:15:28.120 --> 00:15:39.000
Uthappa Outskill: So, as much as possible, please always keep your videos turned on, not just
for this session, for all the upcoming sessions. I'll give you that a minute while I pull up my deck,
and then get going with this.

82
00:15:46.810 --> 00:16:00.360
Uthappa Outskill: Yep. So, there we begin, our journey towards becoming an AI generalist, a
6-month fellowship program. A very hearty congratulations to all of the folks who are here in this
call today. And…

83
00:16:00.450 --> 00:16:10.439
Uthappa Outskill: I hope to have all of you present in all the sessions going forward, so that way
you can get the complete value of whatever you have signed up for.

84
00:16:10.870 --> 00:16:11.750
Uthappa Outskill: No.

85
00:16:12.260 --> 00:16:28.909
Uthappa Outskill: this is the roadmap, just to kind of refresh your memory on a couple of things
before we get going with our first session. This was the roadmap that we discussed during our
onboarding call. You're going to go through all these 6 levels for the next 6 months, and,
achieve,

86
00:16:28.910 --> 00:16:40.770
Uthappa Outskill: a definite objective in each of these levels, and at the same time, by the end
of 6 weeks, sorry, 6 months, you'll also be in a position where you can build products with AI,
and become an AI journalist as well.
87
00:16:40.870 --> 00:16:45.299
Uthappa Outskill: So, what are we starting today? We are starting with Level 1.

88
00:16:45.470 --> 00:16:49.179
Uthappa Outskill: Right? So this level is going to go on for…

89
00:16:49.370 --> 00:17:04.460
Uthappa Outskill: the two weeks, so this week as well as next week. I'll also show which are the
slides that are com- sorry, the sessions that are coming up for you as part of this level. So this is
going to be about fundamentals of, generative AI, which also includes a perspective

90
00:17:04.670 --> 00:17:05.530
Uthappa Outskill: of…

91
00:17:06.000 --> 00:17:19.890
Uthappa Outskill: how an AI generalist works, and also, we will help you in this particular
session today. Also, we're going to go about giving you certain frameworks, that you can start
using, to think of AI. For example.

92
00:17:20.130 --> 00:17:37.639
Uthappa Outskill: How do you… where do you use AI? In what situations do you use AI? Or in
what capacity, or how do you basically use AI? We have designed certain templates and
frameworks that you can kind of use and apply in your day-to-day use cases, alright? So, so
we're going to have…

93
00:17:37.970 --> 00:17:46.330
Uthappa Outskill: four sessions as part of this level. Today is the first session, which is going to
be Fundamentals. Tomorrow, you will have your second session, which is going to be Prompt
Engineering Part 1.

94
00:17:46.670 --> 00:18:05.580
Uthappa Outskill: So you'll understand the fundamental different techniques of good prompt,
good prompts, bad prompts, and different techniques that are available. And then, next
Saturday, you're going to have the second session, which is the advanced prompt engineering
session, advanced techniques involved in prompt engineering, and then we'll close this level
with an applied AI session.

95
00:18:05.580 --> 00:18:10.819
Uthappa Outskill: So this is how Level 1 is going to be for you, and today we start with the first
session.

96
00:18:11.040 --> 00:18:16.660
Uthappa Outskill: Alright? Before we begin, we did discuss about it during our onboarding call.

97
00:18:17.190 --> 00:18:20.300
Uthappa Outskill: But we are also going to kind of refresh your memory here.

98
00:18:20.770 --> 00:18:38.140
Uthappa Outskill: this session particularly, we will take the questions, all your questions that you
have. We want to get done with the teaching part, in the first few hours, so that all your
questions can be parked for a dedicated Q&A session that we'll have towards the end of this
call.

99
00:18:38.140 --> 00:18:43.070
Uthappa Outskill: Right? So, second, make sure not to spam the chart.

100
00:18:43.070 --> 00:19:06.290
Uthappa Outskill: Because whatever you put in, you might also be disrupting the flow of not just
the mentor, but also your fellow learners as well. So ensure not to spam the chat as much as
possible. We recommend it. If we find, too many spams in the chat, then we will also have to…
we'll be forced to kind of remove you from the session as well, just to ensure the session
experience is great for the rest.

101
00:19:06.840 --> 00:19:22.250
Uthappa Outskill: Any links or resources that we discuss here, or we're going to talk about it, we
will be putting it in our post-sheets after the session is done, which is going to be tomorrow. So
make sure that you don't have to ask for any links or resources. We will ensure that it is added
in your postsheets after the session.

102
00:19:22.380 --> 00:19:23.450
Uthappa Outskill: Right?

103
00:19:23.690 --> 00:19:29.120
Uthappa Outskill: Fourth important thing, we did discuss about it in length during our onboarding
call.

104
00:19:29.940 --> 00:19:38.920
Uthappa Outskill: We all are wearing an experienced hat, vast amount of experience that we
come here with. With that also comes certain set of, notions or thoughts or perspectives.

105
00:19:39.490 --> 00:19:59.109
Uthappa Outskill: for the reminder of the program, for the next 6 months, a request to all of you,
let us all keep those experienced hats aside, and absorb what is being discussed in this
particular program, because most of it might be new to you, you'll also kind of figure out how to
kind of adapt to it. So keep that experienced hat aside, and absorb whatever is being given to
you.

106
00:19:59.310 --> 00:20:05.470
Uthappa Outskill: And today's session, and every Saturday, Sunday session that we're going to
have, is going to be a watch and take notes session.

107
00:20:05.640 --> 00:20:22.470
Uthappa Outskill: It is not going to be a do-it-along session. For that, we will have, let's see if
you have any… you practice it after the session is done, and then if you have any doubts, you
come to our, if you find any blockers, you come to our office hours and get it resolved. Alright?
So, with this expectation.

108
00:20:22.790 --> 00:20:26.029
Uthappa Outskill: Let me introduce our mentor for the day.

109
00:20:27.700 --> 00:20:29.519
Uthappa Outskill: Quickly check… yep.

110
00:20:29.790 --> 00:20:47.279
Uthappa Outskill: So, our mentor for the day is Mr. Dilip. I'm sure some of you would have
already met him in different capacities. He's been a leader in multiple education startups for the
last 6 years, and also a very experienced AI researcher as well. He heads AI initiatives for us at
Outskill.

111
00:20:47.280 --> 00:20:57.330
Uthappa Outskill: With that short intro, Dilip, if you're here, I'll hand over the call to you, and the
next few hours is up to yours. So let's all welcome Dilip in the chat, folks.

112
00:20:58.330 --> 00:21:00.220
Dileep Karri: Hi everyone, how is everyone doing?

113
00:21:06.560 --> 00:21:19.970
Dileep Karri: Good, good, great, great. Nice to see you all, and thanks for all the love and the
wishes. And I think I remember taking a few sessions, and I think some of you might have seen
me, some of you

114
00:21:19.970 --> 00:21:35.700
Dileep Karri: might not have, seen me. So, for those of you who have not seen me, right, my
name is, Dilip, and, I have been an, I think I've been an AI researcher, AI generalist, etc, a lot of.

115
00:21:35.760 --> 00:21:53.719
Dileep Karri: stuff I have been doing all through, and I have been, in that. So what I keep doing
is, I will keep turning the chat on and off, so that I keep that interaction on, and everything else.
Of course, the reactions will be on, you can keep showing your reactions, etc, for what I'm
doing, and we will take it forward from there.

116
00:21:53.720 --> 00:21:56.150
Dileep Karri: So, first, let me…

117
00:22:20.410 --> 00:22:29.390
Dileep Karri: There you go. So, today's session, in fact, I'll tell you this. Today's session is
something we haven't done for any cohort so far.

118
00:22:29.390 --> 00:22:44.060
Dileep Karri: So you guys will be the first cohort to experience this session, and this session is a
reflection, and I would call it a re-invention based on previous cohort's feedback.
119
00:22:44.340 --> 00:23:02.269
Dileep Karri: So, in the previous cohorts, we tried to understand what were some of the gaps
they had in the basic understanding, and then we tried to… we actually did this session in bits
and pieces in between the cohort, but then we realized that doing this session at the beginning
gives

120
00:23:02.350 --> 00:23:22.330
Dileep Karri: a lot more clarity for folks to understand. And that's why it's a new session,
absolutely new session, so you'll be the first to experience this. Would love to have your
thoughts and your feedback towards the end of the session. So the main part of this session,
what I'm going to address is answer the question of who is an AI generalist.

121
00:23:22.330 --> 00:23:37.630
Dileep Karri: And how do you do problem solving with AI, okay? How, when you have problems,
how do you solve it with AI, okay? And in a way, these are self-explanatory in the sense that an
AI generalist is nothing but a person who solves problems with AI.

122
00:23:37.630 --> 00:23:40.310
Dileep Karri: So, how do you approach that? What do you do?

123
00:23:40.320 --> 00:23:43.660
Dileep Karri: All that we will try to address, so far.

124
00:23:44.290 --> 00:24:05.109
Dileep Karri: Okay? So, who is an AI generalist? If you ever had this doubt, we have spoken
about it in length during the mastermind, when you'd have seen Wybot talk in detail about it,
funny talk in detail, and a lot of folks would have talked in detail about it. But the summary, the
sum and substance of it is this.

125
00:24:05.110 --> 00:24:08.530
Dileep Karri: An AI generalist is a problem solver.

126
00:24:08.530 --> 00:24:13.780
Dileep Karri: who uses AI as leverage, and this is the key part, a problem solver.

127
00:24:13.850 --> 00:24:21.470
Dileep Karri: If you say, I know how to use AI tool, I know how to use that AI tool, I know how to
use this AI tool.

128
00:24:21.800 --> 00:24:32.390
Dileep Karri: Okay? When you say that knowledge, when you show off your knowledge saying
that, I know 10 tools, but what problems have you solved? If it is 0, you are not an AI generalist.

129
00:24:32.480 --> 00:24:43.470
Dileep Karri: Okay? That is a first and foremost thing that you need to understand, and this
mindset will have multiple pathways and multiple downstream effects that I'm going to talk
about.

130
00:24:43.820 --> 00:25:01.780
Dileep Karri: Okay? So, an AI generalist is not an AI ML specialist who knows only technical AI.
Okay? He is not a tools expert. He or she is definitely not a tools expert, where you say that I
have collected 20 tools, and I have used it for AI.

131
00:25:02.340 --> 00:25:10.909
Dileep Karri: They are not AI-dependent, that means they outsource their thinking, okay? And
this is the point I'm going to talk about a lot.

132
00:25:11.080 --> 00:25:17.589
Dileep Karri: They are not just a prompt engineer, okay, who only knows how to talk to AI.

133
00:25:18.020 --> 00:25:42.500
Dileep Karri: Okay? So, an AI generalist is not a specialist in technical AI, not a tools expert who
knows who collects multiple apps, who is AI-dependent, that is, they outsource their thinking to
AI, and a prompt engineer only who talks to AI. Okay? There might be a bit of all of this, like,
there might be a bit of a tools expert, there might be a bit of AI-dependent, there might be a bit
of prompt

134
00:25:42.500 --> 00:25:45.739
Dileep Karri: engineering, but definitely not just that.

135
00:25:45.740 --> 00:25:51.209
Dileep Karri: Okay? Think of this picture as the blind man and the elephant kind of analogy.

136
00:25:51.240 --> 00:26:01.369
Dileep Karri: Okay? In the blind man and the elephant analogy, for those of you who don't know,
six blind men were taken towards an elephant and asked, please describe the elephant to me.

137
00:26:01.370 --> 00:26:15.959
Dileep Karri: One person hold the trunk and said, elephant is cylindrical in nature. Somebody
hold the ears and said, it is flat in nature. Somebody held the stomach and said, it is big in
nature. Somebody held the tail and said, it is thin in nature.

138
00:26:15.960 --> 00:26:25.929
Dileep Karri: All of them are right, and all of them are wrong, okay? The unified picture of what
an AI generalist does is what I am trying to mention.

139
00:26:25.970 --> 00:26:40.699
Dileep Karri: Okay, so before I move ahead, just a quick check so that everybody is on the same
page. Folks, are you able to follow me clearly? My accent is good, you are not having any
difficulties understanding? Yes, everybody clear?

140
00:26:42.840 --> 00:26:44.210
Dileep Karri: Okay, awesome.

141
00:26:46.720 --> 00:26:57.450
Dileep Karri: Right, now what I would request the team to do is drop in the Slido of the question,
so that whenever I turn off the chat, I will have the team drop the Slido.

142
00:26:57.450 --> 00:27:19.350
Dileep Karri: The reason being, if you have got any questions, you can open the Slido and put in
those questions, and if you saw a question that somebody else has asked, and it's the same
question you have, you can upload those questions, so that I know for a fact that these
questions are very, very, important, and these are the questions that most of the cohort has.

143
00:27:19.350 --> 00:27:22.019
Dileep Karri: My guess and my hope is
144
00:27:22.020 --> 00:27:33.649
Dileep Karri: By the time I finish my delivery, a lot questions would have been answered already,
but in case some are not answered, I will ensure that those are all taken up and answered.

145
00:27:34.180 --> 00:27:35.170
Dileep Karri: Okay?

146
00:27:36.260 --> 00:27:39.160
Dileep Karri: Now, who is an AI generalist?

147
00:27:39.600 --> 00:27:55.369
Dileep Karri: AI generalist is a problem solver, okay, who uses AI as leverage. Now, one of the
earlier definitions I had was, AI generalist is a problem solver using AI. That was a definition
even I started with.

148
00:27:55.380 --> 00:28:11.299
Dileep Karri: But now I have changed the definition to saying who uses AI as a leverage, and
why am I using this particular thing, is because there is this critical judgment coming into picture.
What it means is, a person

149
00:28:12.040 --> 00:28:18.839
Dileep Karri: is an AI generalist, even when they know that, okay, you know what, I don't need
AI for this problem.

150
00:28:19.670 --> 00:28:28.319
Dileep Karri: If I know that I do not require AI to solve this problem, or AI is an overkill not
required.

151
00:28:28.640 --> 00:28:32.620
Dileep Karri: Or, this problem cannot be solved even with AI.

152
00:28:32.720 --> 00:28:48.270
Dileep Karri: So that wisdom is important, and that's why I'm calling an AI generatorist as a
problem solver who uses AI as a leverage. So problem solving is the key part. Happen to use AI
overall. A domain expert who amplifies with AI.
153
00:28:48.330 --> 00:29:09.239
Dileep Karri: Okay? A critical thinker who knows when not to use AI. That is very, very
important. It is not just about saying, okay, you know what, I know how to use AI. It is also very,
very important to understand when not to use AI. And how do you decide when not to use AI,
etc? We are going to discuss the frameworks.

154
00:29:09.240 --> 00:29:14.080
Dileep Karri: And a value creator at the intersection of human and AI.

155
00:29:14.080 --> 00:29:22.099
Dileep Karri: And the formula, I think, is domain expertise plus AI capabilities plus critical
judgment equal to an AI generalist.

156
00:29:22.130 --> 00:29:31.800
Dileep Karri: So, if you're trying to go towards AI generalist, the domain expertise, AI
capabilities, and critical judgment becomes an AI generalist. Now, what does this mean?

157
00:29:31.970 --> 00:29:51.029
Dileep Karri: Domain expertise, okay? So, let me get an understanding of your domain
expertise. Think for a few seconds what is your main domain expertise, and write in 2-3 words
on the chat what your domain expertise is. Okay, I'm just opening up the chat. 2-3 words, your
domain expertise.

158
00:29:54.290 --> 00:30:12.930
Dileep Karri: Insurance, coaching, wealth management, agriculture, banking and finance,
design, business ops, finance, project management, sales, digital media, customer success,
wow. I see a plethora of domain expertise, okay? Now, domain expertise is something you
already have.

159
00:30:12.950 --> 00:30:26.429
Dileep Karri: This is not something that we will teach you in this program, okay? So what that
means is we are not going to go and try to solve problems in your domain, that is your expertise.

160
00:30:26.430 --> 00:30:36.229
Dileep Karri: Where we will put in a lot of effort is over here, which is the AI capabilities, and a
bit of critical judgment also.

161
00:30:36.230 --> 00:30:39.519
Dileep Karri: So, we are here to make you an AI generalist.

162
00:30:39.520 --> 00:30:59.960
Dileep Karri: We are not here to make you an AI finance expert, an AI design expert, an AI
marketing expert, AI insurance expert, or any AI dash expert. That dash is something that you
already have as domain expertise, and it's your responsibility to pick that domain expertise and
apply to your problems.

163
00:30:59.960 --> 00:31:04.189
Dileep Karri: What we will teach is the AI capabilities completely through and through.

164
00:31:04.240 --> 00:31:20.980
Dileep Karri: So, what it means also is, when we try to teach you via some examples, we will
take very, very generic examples, which everybody can understand. Everybody might not relate
to it, right? For example, a marketing person might relate to marketing examples.

165
00:31:20.980 --> 00:31:27.909
Dileep Karri: But if I do… if we show only marketing examples, the rest of the people will be like,
okay, others could relate, I could not relate.

166
00:31:27.910 --> 00:31:39.230
Dileep Karri: Or this, if you can show me this example, if you can show me a finance example, if
you can show me an insurance example, right? It is not possible. The name of this program is
AI Generalist

167
00:31:39.230 --> 00:31:49.750
Dileep Karri: fellowship, so what it means is, all the examples, everything will be built around AI.
You will be making you a very good AI capability expert.

168
00:31:49.750 --> 00:32:00.050
Dileep Karri: And that means domain application, domain knowledge is your play, AI expertise is
our play, critical judgment, we meet somewhere in between.
169
00:32:00.050 --> 00:32:14.659
Dileep Karri: As in, AI critical judgment of what is the critical judgment you need in AI is
something we will give you frameworks for. In your domain expertise, that critical judgment you
will have to bring through, and somewhere you'll have to meet in the middle.

170
00:32:14.790 --> 00:32:30.029
Dileep Karri: Okay? So this is very, very important. This is not an AI for marketing, AI for
insurance, AI for events. I'm telling this because in the previous cohorts, one of the feedback
points was, hey, you gave us more that domain example, you should give us this domain
example.

171
00:32:30.030 --> 00:32:40.440
Dileep Karri: That will be a very tough balancing act, because how many countless tons of
examples we give, not everybody in the cohort will be satisfied.

172
00:32:40.560 --> 00:32:53.549
Dileep Karri: But what we can do instead is, we will tell you how you can understand your own
examples, those frameworks we can show you, okay? And I'll talk about those also as we move
on.

173
00:32:54.700 --> 00:33:00.949
Dileep Karri: So… Now, this, in some aspect, you have already seen.

174
00:33:01.340 --> 00:33:08.720
Dileep Karri: There is a solo virtuoso who is very good at one instrument and the best possible
at that instrument.

175
00:33:08.850 --> 00:33:18.239
Dileep Karri: Whereas an orchestra conductor, who is decent in a lot of instruments, And
running them all together.

176
00:33:18.590 --> 00:33:30.360
Dileep Karri: What is your personal preference? Do you want to be a solo instrument specialist,
or do you want to be an orchestra conductor? What is your thoughts? What do you want to be,
folks?
177
00:33:30.590 --> 00:33:33.680
Dileep Karri: Let me in the chat, solo or orchestra?

178
00:33:34.660 --> 00:33:39.500
Dileep Karri: Okay? Orchestra, solo, orchestra, sono… Nice.

179
00:33:39.910 --> 00:33:53.390
Dileep Karri: So, I always love it when I have a mixed response, okay? That means I am forcing
you to think, and there is some confusion there, which I can now clarify.

180
00:33:53.890 --> 00:33:57.450
Dileep Karri: Alright? So… the specialist

181
00:33:57.810 --> 00:34:06.459
Dileep Karri: Who is this specialist? Let us understand in the detail. A specialist is this one
person who can master one instrument to perfect

182
00:34:07.110 --> 00:34:09.500
Dileep Karri: You know, complete perfection.

183
00:34:09.840 --> 00:34:12.590
Dileep Karri: Unmatched depth in their domain.

184
00:34:12.760 --> 00:34:20.010
Dileep Karri: Creates beauty through singular expertise, Vulnerable when the music changes.

185
00:34:20.460 --> 00:34:28.580
Dileep Karri: Okay? For example, there is a developer who is extremely good at Python, and
nothing else.

186
00:34:28.699 --> 00:34:38.829
Dileep Karri: So what happens when AI codes better than Python? That's where you'll have
some problems. But there is a little more nuance into this than what you see on the screen.
187
00:34:39.650 --> 00:34:44.610
Dileep Karri: So, what it means is, it brings to everybody a conundrum.

188
00:34:45.010 --> 00:34:48.999
Dileep Karri: Should I be, like, very good at one thing?

189
00:34:49.530 --> 00:34:53.269
Dileep Karri: Or, you know, the other part. Should I be…

190
00:34:53.440 --> 00:34:57.009
Dileep Karri: be said in many things, okay? So let me…

191
00:34:57.520 --> 00:35:05.080
Dileep Karri: Talk about the orchestra conductor also, and then try to provide some answer or
direction for giving this answer.

192
00:35:05.120 --> 00:35:09.050
Dileep Karri: An orchestra conductor understands every instrument's role.

193
00:35:09.080 --> 00:35:27.370
Dileep Karri: creates harmony from diversity, right? For example, somebody sings at bass,
somebody at alto, somebody at a little higher note, somebody goes to opera, okay? So there is
different notes in a song.

194
00:35:27.930 --> 00:35:39.270
Dileep Karri: Music aficionados, you will understand this very clearly, that a beauty in a
composition comes only when you are having those differences.

195
00:35:39.810 --> 00:35:59.629
Dileep Karri: Okay? So this is the professional who orchestrates AI for different analysis, domain
knowledge for context, and human insight for decisions. So this is what an orchestra conductor
does. He's a generalist, he or she is a generalist who brings it together. Now, coming to the
conundrum.
196
00:35:59.650 --> 00:36:06.069
Dileep Karri: Do I need to become a specialist, or do I need to become a generalist or an
orchestra conductor?

197
00:36:06.220 --> 00:36:10.560
Dileep Karri: This is what I would recommend. All of you are solo

198
00:36:11.250 --> 00:36:21.910
Dileep Karri: in instrument virtuoso in some level, because you have spent so many years in
your domain expertise. You have spent all those years, so you have honed one instrument.

199
00:36:22.310 --> 00:36:36.450
Dileep Karri: Now is the time that you have, since you have become a master of your
instrument, now is the time to take a step into the shoes of the orchestra conductor, where you
are very sharp in that one instrument.

200
00:36:36.450 --> 00:36:43.009
Dileep Karri: What that means is, even if somebody blows air into the instrument, you will know
what note they are catching.

201
00:36:43.530 --> 00:36:49.399
Dileep Karri: Plus, you are decent in other instruments, so that you are able to create a music
and orchestra.

202
00:36:49.650 --> 00:36:52.070
Dileep Karri: Okay? Now…

203
00:36:52.340 --> 00:36:59.089
Dileep Karri: This, again, will be a little question for music folks, okay, or anybody who listens to
music can, do this.

204
00:36:59.820 --> 00:37:15.659
Dileep Karri: every music director, or every music composer, they have a particular signature in
their compositions. That means one or two instruments play a very major part in their music
compositions.
205
00:37:15.760 --> 00:37:23.330
Dileep Karri: The rest of the instruments are there. It's not that they're not there, but they are a
little less. So there's a signature of

206
00:37:23.380 --> 00:37:38.360
Dileep Karri: that music director. That instrument, that arrangement, is that music director's
signature. So, that is the level I want you to reach, wherein you will transform that one or two
instruments you know as your signature.

207
00:37:38.410 --> 00:37:44.200
Dileep Karri: And then do the compositions which carry your signature all through to bring
forward an orchestra.

208
00:37:44.470 --> 00:37:47.890
Dileep Karri: So, Hans Zimmer, one of the very foremost,

209
00:37:48.140 --> 00:37:58.299
Dileep Karri: you know, composers, he has his own peculiar style of, you know, strings, the way
strings sound in his instrument. He is very…

210
00:37:58.520 --> 00:38:12.259
Dileep Karri: I would say cordal in nature, okay? Then, there is A.R. Rahman, who tries to,
orchestrate with highs. His framework is in the heists of his music.

211
00:38:12.330 --> 00:38:28.839
Dileep Karri: Okay? So find that signature, and that signature is nothing than your domain
knowledge. Your domain knowledge is your signature. Learn AI tools so that you become this
signature orchestra conductor. That is the AI generalist I want all of you to think to become.

212
00:38:29.710 --> 00:38:38.300
Dileep Karri: Okay, so the reality check I wanted to give you is AI is compressing 10-year
expertise paths to 10 weeks.

213
00:38:38.470 --> 00:38:54.930
Dileep Karri: Okay? What does it mean? In some cases, not all, in some cases, AI is beating
doctors at diagnosis, it is beating coders at debugging, it is beating lawyers at contract review,
and it is beating analysts at pattern redemption.

214
00:38:55.470 --> 00:39:03.079
Dileep Karri: Okay? Now the question at the back of your mind is, if AI has already beaten these
folks, now where am I into the picture?

215
00:39:03.210 --> 00:39:06.520
Dileep Karri: Okay? The key part, again, nuance. Nuance, folks.

216
00:39:06.710 --> 00:39:11.389
Dileep Karri: AI beats doctor at diagnosis, but

217
00:39:11.910 --> 00:39:24.110
Dileep Karri: It is only in a few cases. AI beats lawyers at contract review, maybe, but in case of
arguments before a judge, a human is involved, AI cannot beat them.

218
00:39:24.240 --> 00:39:44.090
Dileep Karri: AI can beat coders at debugging in writing test cases, but in system design and
architecture, maybe AI is still not that great. AI beats analysts at pattern recognition, but pattern
interpretation, interpretation of the patterns that were discovered, still AI cannot beat humans.

219
00:39:44.130 --> 00:39:53.449
Dileep Karri: Okay? So that is a reality check I wanted to give. What it means is, what AI is
doing good, and what AI is not doing so good.

220
00:39:53.450 --> 00:40:07.380
Dileep Karri: awareness of that, using AI to amplify your work by using it in what it does really
good, and then adding your human expertise on top of that layer, that is what an AI generalist
should do.

221
00:40:08.800 --> 00:40:26.300
Dileep Karri: Okay? So, the four core principles of an AI generalist. Okay? So, this is the hero
slide. Take a screenshot of it. Anyways, the slides will be shared with you, but if there is
something I want all of you to understand, these are the four core principles, and throughout the
program.

222
00:40:26.300 --> 00:40:34.809
Dileep Karri: we will keep harking these four principles. These are the commandments of an AI
generalist, if I can so call them.

223
00:40:34.890 --> 00:40:39.689
Dileep Karri: Number one, I solve problems, not operate tools.

224
00:40:40.240 --> 00:40:49.880
Dileep Karri: What is the repercussion of this? What it means is, I understand the framework of
solving that particular problem. I am not

225
00:40:50.090 --> 00:40:53.189
Dileep Karri: Attached to the tool that is being taught.

226
00:40:53.590 --> 00:41:00.430
Dileep Karri: So, for that one use case, there are 10 tools that can be taught for the same
problem statement.

227
00:41:01.030 --> 00:41:12.919
Dileep Karri: Okay? I can use 10 tools for the same problem statement, but I don't want you to
think that, okay, you know what? Outskill taught us this tool, they didn't teach us that tool.

228
00:41:13.040 --> 00:41:19.980
Dileep Karri: Okay? The idea is the tool is just a means underneath you are solving problems.
That is point number one.

229
00:41:20.300 --> 00:41:23.850
Dileep Karri: Number two, AI amplifies my thinking.

230
00:41:24.000 --> 00:41:35.049
Dileep Karri: What it means is, it is a tool to amplify my thinking overall and give me more
dimensions to think about. It is not going to replace my thinking.
231
00:41:35.180 --> 00:41:44.040
Dileep Karri: Don't think that AI can replace my thinking, I can outsource my thinking to AI. And
we will go into a lot of examples as we do this.

232
00:41:44.550 --> 00:41:51.310
Dileep Karri: Third, I can learn anything just in time. What it means is, I will try to learn

233
00:41:51.310 --> 00:42:07.929
Dileep Karri: when I'm trying to learn something, I will have a bunch of questions which will lead
me to a depth exploration. That means I will have this mind thing that unless I understand this
topic thoroughly till the very end, I am not going to move ahead.

234
00:42:08.420 --> 00:42:17.469
Dileep Karri: Okay? Now, that approach, which is a depth-first approach, is how our earlier
education system used to work, where you are taught to go very deep.

235
00:42:17.640 --> 00:42:33.770
Dileep Karri: But in a field like AI, that approach could be a disaster, because if you're going
deep into one thing, you can spend 6 months just learning one concept, still you will not have
confidence, and then you'll be confused, like, I spent 6 months on this, now where am I in my AI
journey?

236
00:42:33.970 --> 00:42:45.749
Dileep Karri: So, we learn anything just in time. What it means is, we learn as much as needed
to get going, and to start to attempt to solve the problem. I learn only that much.

237
00:42:45.820 --> 00:43:03.230
Dileep Karri: Rest, I will park it aside, and I'll keep revisiting it time and again. When I'm learning
from different mentors, they will shine light on different aspects of AI, and that light will illuminate
a dark corner which I had previously ignored.

238
00:43:03.390 --> 00:43:06.990
Dileep Karri: Okay? Think of it like an exploration of the hill.

239
00:43:07.460 --> 00:43:20.109
Dileep Karri: So, what it means is, when I'm climbing the hill, I should not climb the hill like a
spiral, saying that I will cover each and every footprint of the hill, and become like an expert in
the hill.

240
00:43:20.520 --> 00:43:26.600
Dileep Karri: You climb the hill not to cover its entire area. You climb the hill to reach the summit.

241
00:43:27.270 --> 00:43:45.940
Dileep Karri: What it means is, I take the shortest path, reach there. Once I have reached the
summit, then I can say, okay, let me try a more adventurous path. Now, you know, earlier I wore
a rope and I climbed the hill. I'm confident now I can climb without any protective equipment.
You can do all sorts of gymnastics.

242
00:43:45.940 --> 00:43:48.900
Dileep Karri: Adventures, but not as a learner.

243
00:43:49.010 --> 00:43:54.719
Dileep Karri: Okay? When you are learning any martial art, you don't try to do the flying kick.

244
00:43:54.970 --> 00:44:00.840
Dileep Karri: First, you strengthen your basics, you strengthen your lower body, and then you
build it.

245
00:44:01.170 --> 00:44:22.759
Dileep Karri: Then, the fourth core principle, very, very important, and this is a principle by which
you… we will judge you on whether you have become an AI generalist or not, which is create
unique value at intersections. You have your domain expertise. We teach AI. At the unique
value, create… at the unique intersection that we build, create that unique value.

246
00:44:23.500 --> 00:44:26.279
Dileep Karri: That is very, very, very, very, very important.

247
00:44:27.090 --> 00:44:35.859
Dileep Karri: Okay? These are the four commandments of AI. Utapa and team will keep
reinforcing these commandments as we go
248
00:44:35.980 --> 00:44:43.619
Dileep Karri: into different levels, we try to explore different things. These are the principles you
need to have in mind.

249
00:44:45.700 --> 00:44:55.990
Dileep Karri: Okay? So, folks, are we clear with the definition of an AI generalist and the four
core principles that are needed? Yes, everyone?

250
00:44:58.530 --> 00:44:59.490
Dileep Karri: Okay?

251
00:45:02.090 --> 00:45:02.930
Dileep Karri: Awesome.

252
00:45:04.720 --> 00:45:05.820
Dileep Karri: Now…

253
00:45:05.970 --> 00:45:14.350
Dileep Karri: other part, okay? You will enjoy the sessions, okay? You will definitely enjoy the
sessions, but you are not here to be editing. Very, very different.

254
00:45:14.700 --> 00:45:20.090
Dileep Karri: The truth… unfortunate truth about running is, it is not supposed to be fun.

255
00:45:20.480 --> 00:45:26.489
Dileep Karri: The outcome of learning is supposed to be fun, but the learning process
sometimes is not as fun.

256
00:45:26.670 --> 00:45:38.040
Dileep Karri: It requires effort, a lot of mental sweating. So that means there are no 10-minute
hacks or shortcuts, and you should have the discomfort of not knowing.

257
00:45:38.850 --> 00:45:52.949
Dileep Karri: Okay, so I… I just… give me a thumbs up in the reactions, if you think this is true.
Your greatest learning, your greatest movements of celebration, did they come from points of
discomfort?

258
00:45:53.570 --> 00:46:00.379
Dileep Karri: Thumbs up if you think that they came out of challenge and discomfort. Your
greatest victories.

259
00:46:00.800 --> 00:46:02.700
Dileep Karri: I see one thumbs up.

260
00:46:06.150 --> 00:46:06.960
Dileep Karri: Wait.

261
00:46:08.610 --> 00:46:10.010
Dileep Karri: Why am I not seen?

262
00:46:12.170 --> 00:46:14.189
Dileep Karri: Yes, I see a few more.

263
00:46:14.390 --> 00:46:15.480
Dileep Karri: Others…

264
00:46:18.890 --> 00:46:20.940
Dileep Karri: All right, all right, all right.

265
00:46:23.630 --> 00:46:29.510
Dileep Karri: Cool. So, I hope you got that, thing, that, you know, it is…

266
00:46:29.930 --> 00:46:35.100
Dileep Karri: Requiring a… you need to embrace the discomfort of not knowing things.

267
00:46:35.230 --> 00:46:40.710
Dileep Karri: for you to proceed in learning. Very, very important. So what it means is.
268
00:46:40.960 --> 00:46:43.459
Dileep Karri: At times, you will get stuck.

269
00:46:43.760 --> 00:46:53.699
Dileep Karri: At times, it'll be frustrating, okay? At those moments, understand that you are very
close to a moment of a light bulb moment.

270
00:46:53.700 --> 00:47:05.769
Dileep Karri: Right? A light bulb moment is very close, because after you cross that anxiety, that
frustration, you reach that right light bulb moment, you will never not know that problem again.

271
00:47:06.400 --> 00:47:13.010
Dileep Karri: Okay? So, the thing I wanted to kind of summarize in the last point is you cannot
outsource push-ups.

272
00:47:13.050 --> 00:47:13.970
Dileep Karri: Okay?

273
00:47:13.970 --> 00:47:37.770
Dileep Karri: For physical fitness or mental fitness, right? You… like, physical fitness, you can't
watch workout videos and get fit. You cannot have someone else do your reps. There is no pain,
no gain. So therefore, in mental fitness, you just can't watch these sessions and become an AI
expert. You cannot have AI do your thinking. You will have to put in the effort to apply AI to solve
your problem.

274
00:47:37.770 --> 00:47:41.689
Dileep Karri: And that's where it all moves towards.

275
00:47:43.000 --> 00:47:50.819
Dileep Karri: Okay? So, these are, again, your irreplaceable assets that protect your journey.

276
00:47:50.860 --> 00:48:04.619
Dileep Karri: Those are, number one, your critical judgment, okay? Hone your critical judgment
skill as we move across the AI Generalist Fellowship. Again, this will be something you will learn
as we go through multiple sessions.

277
00:48:05.160 --> 00:48:08.569
Dileep Karri: Get that critical judgment muscle very well.

278
00:48:08.880 --> 00:48:18.319
Dileep Karri: Kind of combining the disparate ideas uniquely, that is creative synthesis and
empathetic connection, that is, reading between the human lines.

279
00:48:18.350 --> 00:48:38.139
Dileep Karri: So, these are the three things that you need to be very, very protective of. Don't
outsource it completely to AI. One is judgment. You cannot tell AI, hey, choose this or that for
me. Don't ever outsource your judgment. You can take directional inputs of judgment from AI.

280
00:48:38.140 --> 00:48:42.789
Dileep Karri: But end of the day, you should be responsible for your own judgment.

281
00:48:42.850 --> 00:48:49.169
Dileep Karri: Second, there will be multiple ideas, bringing them unique ideas together.

282
00:48:49.870 --> 00:49:02.499
Dileep Karri: That is your job. You… it is not AI's job to synthesize multiple ideas together. That
has to be your work. What it means is, if AI has shown you insights.

283
00:49:03.020 --> 00:49:09.829
Dileep Karri: In your mind, how can you thread those insights to come to a decision? That is
your job.

284
00:49:10.160 --> 00:49:27.989
Dileep Karri: Then, empathetic human connection, where you're reading between the human
lines. So, suppose you're doing something with AI which has impact on human. That empathy
angle is something you have to bring it up, and this, again, cannot be outsourced to AI.

285
00:49:28.680 --> 00:49:36.990
Dileep Karri: Okay? So, one thing I want to kind of… little slight detour, but we'll come back. It is
that writing is equal to thinking.

286
00:49:37.070 --> 00:50:01.030
Dileep Karri: And writing isn't just output, it is how you learn to think. So while we will provide
you all the notes and everything, I want all of you in every session to at least write down not
what the mentor is saying, but what is that moment in your head that resonated with you, and
make a quick note of that, because it forces linear logic, clarifies your thoughts, and reveals gap
in your reasoning.

287
00:50:01.470 --> 00:50:10.279
Dileep Karri: So, one of the biggest AI traps people fall into is they put in a prompt, they get an
AI draft, they do a light edit, and then they submit.

288
00:50:10.330 --> 00:50:13.440
Dileep Karri: Okay? But the ideal way to do it is think.

289
00:50:13.460 --> 00:50:32.240
Dileep Karri: write your core rough thoughts, then you use the AI enhancement, and then do the
review. That is the way you should do it. So basically, don't outsource your thinking that much,
and any kind of AI-led writing you're doing, ensure that you're doing this, and this comes to all
tasks.

290
00:50:32.240 --> 00:50:47.189
Dileep Karri: If the task is not clear in your head, if you're not able to put it on paper clearly, how
much ever you are trying to understand AI, etc, you will not be able to bring that idea to life.
Very, very important.

291
00:50:47.500 --> 00:50:56.690
Dileep Karri: Okay? So, once again, this is a very important point, I want to repeat it. Don't fall
into the AI trap where you're outsourcing 90% of thinking to AI.

292
00:50:57.410 --> 00:51:00.640
Dileep Karri: Do thinking for at least 5 to 10 minutes.

293
00:51:00.950 --> 00:51:14.320
Dileep Karri: put your effort to write, like, a bullet point draft, not, like, a clear draft. Use those
bullet points as foundations to enhance, and then review. That is the right way of using AI to
write your own stuff.

294
00:51:14.820 --> 00:51:18.360
Dileep Karri: Okay? For example, when I write my post, I use AI.

295
00:51:18.540 --> 00:51:32.880
Dileep Karri: Okay, but I use it in a second way, where I think about it a lot, and then I… then I
put the prompt and enhancement later, I add a lot of thoughts of other folks, I try to synthesize
those together and try to do it.

296
00:51:33.900 --> 00:51:45.319
Dileep Karri: Okay, so now I'm going to teach you all a very important framework called Spark
Framework, where you will understand when should you use AI.

297
00:51:45.550 --> 00:52:01.199
Dileep Karri: Okay? When should you use AI when you're trying to do anything? And also, I will
show you a practical use case of how I will use this same framework to approach a problem and
try to solve it with AI.

298
00:52:01.200 --> 00:52:11.790
Dileep Karri: Okay, so the framework that I'm going to teach is called Spark Framework,
S-P-A-R-K, H stands for Assemble, Test, this thing.

299
00:52:12.890 --> 00:52:16.809
Dileep Karri: The first thing is structured versus unstructured.

300
00:52:17.280 --> 00:52:22.910
Dileep Karri: Okay, so the idea here is, can you clearly define the input and the output?

301
00:52:23.080 --> 00:52:34.210
Dileep Karri: So, if you're trying to solve a task with AI, are you able to clearly define what are
the inputs you are going to give AI, and what is the output you are expecting from AI?
302
00:52:34.470 --> 00:52:45.329
Dileep Karri: If you can clearly define the inputs and the outputs, then you are in a green zone,
where you say you're formatting this data into a particular table, etc.

303
00:52:46.330 --> 00:53:01.859
Dileep Karri: Yellow zone is somewhere in between, where you are asking things like, help me
structure my strategy. That is okay, by the way, in some cases. So for a very hard problem, you
might want to lean a bit on AI. That is fine.

304
00:53:02.710 --> 00:53:13.139
Dileep Karri: For tough problems, do that, but try to weigh more out towards the green zone,
where you are clearly defining the inputs and the outputs. Red zone is, don't say.

305
00:53:14.380 --> 00:53:20.819
Dileep Karri: AI, I have this problem, tell me what to think about this. That is the completely
wrong usage.

306
00:53:21.460 --> 00:53:34.059
Dileep Karri: Alright, so S of the SPARK framework stands for structured and unstructured, and
the test is, can you clearly define the inputs and outputs for your given task?

307
00:53:35.850 --> 00:53:41.739
Dileep Karri: Next is P, or past patterns versus future innovation.

308
00:53:42.070 --> 00:53:57.629
Dileep Karri: Does the answer already exist somewhere? For a problem that you're trying to
solve, does the answer already exist somewhere? If it is a very common problem that you know
would be, like, a common staple problem.

309
00:53:57.630 --> 00:54:04.980
Dileep Karri: It will always be there, and therefore it is green zone, where you're asking the
question, what are the common solutions to X?

310
00:54:05.610 --> 00:54:11.910
Dileep Karri: Then, I have the yellow zone, where you give the context and say what might work
in my context.

311
00:54:12.110 --> 00:54:16.229
Dileep Karri: Red zone is… What breakthrough should I invent?

312
00:54:16.820 --> 00:54:22.640
Dileep Karri: Okay? I will talk about this in the example of scientific innovation.

313
00:54:22.800 --> 00:54:23.810
Dileep Karri: Right?

314
00:54:24.170 --> 00:54:28.190
Dileep Karri: Thomas Edison, suppose he was in 2025.

315
00:54:28.670 --> 00:54:44.090
Dileep Karri: He can tell those 99 times he has failed, feed it to AI, and ask it what is the pattern
of failure you observe, what are the recurring mistakes that I am seeing, etc, and all that. That
part can be done.

316
00:54:44.340 --> 00:54:47.740
Dileep Karri: But Thomas Edison cannot say, hey.

317
00:54:48.080 --> 00:54:55.719
Dileep Karri: build a bulb for me. The bulb should do exactly this. Invent the bulb for me. AI can
never do that.

318
00:54:56.280 --> 00:55:02.749
Dileep Karri: Okay? So that is the difference of P or past patterns versus future innovation.

319
00:55:02.940 --> 00:55:09.649
Dileep Karri: So, S is for structured and unstructured data, P is for past patterns and future
innovation.

320
00:55:09.780 --> 00:55:13.890
Dileep Karri: Then A is amplification versus abdication.

321
00:55:14.320 --> 00:55:18.419
Dileep Karri: Are you enhancing your thinking, or are you replace it?

322
00:55:18.810 --> 00:55:23.670
Dileep Karri: So, the green zone is generate 10 options for me to evaluate.

323
00:55:23.930 --> 00:55:28.309
Dileep Karri: Yellow zone is help me think through the pros and the cons.

324
00:55:28.480 --> 00:55:32.400
Dileep Karri: And the red zone is, decide what is best for me.

325
00:55:32.900 --> 00:55:38.380
Dileep Karri: In the green zone, you're trying to amplify. You can… even better way to…

326
00:55:39.200 --> 00:55:48.880
Dileep Karri: do this is say that, hey, give me 10 options. I think there are 2 options that I have in
my head. Help me expand this to 10 options.

327
00:55:48.910 --> 00:55:59.190
Dileep Karri: That's a green zone level of thinking. Yellow zone would be more in tile lines of,
okay, I have these three, four options, help me think through the pros and cons. The two is okay.

328
00:55:59.190 --> 00:56:09.440
Dileep Karri: But you can't say, okay, these four are available, now tell me what should I
choose? You cannot outsource your choice to AI. The choice and your judgment has to be
yours.

329
00:56:09.440 --> 00:56:14.440
Dileep Karri: That is the amplification versus abdication A part of your framework.

330
00:56:14.750 --> 00:56:18.200
Dileep Karri: Then is repeatable versus relationship.

331
00:56:18.560 --> 00:56:25.940
Dileep Karri: Does the success depend on a human connection, or is this task repeatable like
anything?

332
00:56:26.490 --> 00:56:37.760
Dileep Karri: For example, in green zone, I will say draft test status update. That's a green zone,
because it is a reputable task. It is something you keep doing over and over and over again.

333
00:56:38.610 --> 00:56:45.489
Dileep Karri: Yellow zone is where I try to suggest ways to face difficult feedback. You say that

334
00:56:45.850 --> 00:56:52.070
Dileep Karri: hey, there is this customer that is having this problem, it is turning out to be very
difficult.

335
00:56:52.210 --> 00:57:11.260
Dileep Karri: how should I approach this difficult conversation? How should I approach is fine,
but you cannot say, okay, there is this problem, handle this emotional conversation. I can't take
it, I'm giving you AI, this voice agent. Hey, voice agent, there is this angry customer, deal with it
and get it done.

336
00:57:11.260 --> 00:57:13.070
Dileep Karri: That is not how you should do.

337
00:57:14.240 --> 00:57:31.040
Dileep Karri: And finally, K, knowledge exists versus creation. If you are in the green zone, you
say that you summarize the research that is already available. Yellow zone is where you ask,
okay, these are these concepts, can you help me connect them?

338
00:57:31.350 --> 00:57:36.340
Dileep Karri: Or red zone is where you say, create a new scientific theory.

339
00:57:38.720 --> 00:57:48.569
Dileep Karri: you say, make this invention. It is very similar to the previous one, only thing here
is, here you are… it is on the research end, you are either asking, summarize the research.

340
00:57:49.230 --> 00:57:54.170
Dileep Karri: create a new scientific theory. Is it synthesis, or is it genesis?

341
00:57:54.300 --> 00:58:02.750
Dileep Karri: Synthesis is something you can do with AI. Genesis, bringing out something new,
is not something that you can do with AI.

342
00:58:03.180 --> 00:58:04.909
Dileep Karri: Once again, let me…

343
00:58:06.540 --> 00:58:24.090
Dileep Karri: quickly walk you through this framework of Spark, which is going to tell you a
framework on whether… when should I use AI. S-P-A-R-K, yes, is structured versus
unstructured, where I can clearly define the inputs and outputs.

344
00:58:24.220 --> 00:58:33.169
Dileep Karri: then… P is whether the answer already exists somewhere, is it common wisdom,
or is it a breakthrough?

345
00:58:33.570 --> 00:58:45.809
Dileep Karri: Then A is amplification versus abdication, where you are asking, either you are
giving AI the chance to give options, or you're asking AI to do the choice.

346
00:58:46.440 --> 00:58:57.529
Dileep Karri: R is repeatable versus relationship, where repeatable task is, like, give the status
update, which is a repeatable task, or something, like, handle this difficult conversation for me.

347
00:58:58.350 --> 00:59:14.330
Dileep Karri: Then, K is whether the knowledge already exists, or is it a new scientific theory? Is
it synthesis or connections, or it is genesis? One key point, patterns and future innovation is
more with common wisdom, common
348
00:59:14.430 --> 00:59:25.090
Dileep Karri: problems kind of stuff, K is more on the research angle. When you're trying to do
research, or you're trying to do something very different, that is, K

349
00:59:27.400 --> 00:59:38.649
Dileep Karri: P is more in terms of pattern recognition, pattern matching for problems and
solutions. So, P is more problem solving, problems and solutions. K is more research.

350
00:59:38.650 --> 00:59:49.290
Dileep Karri: or innovation, which is what you're trying to speak about. That's a subtle difference
over here. In some cases, you can say everything makes sense together, but yeah, this is the
SPARC framework.

351
00:59:49.360 --> 01:00:02.979
Dileep Karri: Okay? Now, before I move ahead, folks, did you understand the Spark framework
clearly? Now, I'm going to take a practical example and try to apply this Spark framework and
see how we are able to do this.

352
01:00:04.130 --> 01:00:20.709
Dileep Karri: Okay, so we will take a practical example and go with this Spark framework, and I'll
use AI as my co-pilot in doing this. Now, a caveat to everybody, okay? I want to tell this point
very clearly.

353
01:00:23.630 --> 01:00:28.490
Dileep Karri: If you are trying to apply this framework.

354
01:00:29.380 --> 01:00:48.710
Dileep Karri: The first time for you to apply, it might feel difficult, and you might feel a little
discomfort when you're trying to do it. Please do not get agitated or feel anxious, because when
I am applying this framework, it is because I have worked a lot in AI. It is years and years of

355
01:00:48.850 --> 01:00:53.110
Dileep Karri: You know, work that has kind of culminated in this moment.

356
01:00:53.240 --> 01:01:09.850
Dileep Karri: For you, if you're taking your first steps into AI, if you're understanding this, it is
fine. If you're applying this and you still feel confused, okay, this vertical R is not clear, that K is
not clear, that A is not clear, etc, right?

357
01:01:09.850 --> 01:01:13.759
Dileep Karri: Don't try to take it in that way. Just think in terms of

358
01:01:13.870 --> 01:01:22.520
Dileep Karri: before this session, how I was with this framework, after this session, how am I
with this framework, and try to take it forward from there.

359
01:01:22.800 --> 01:01:41.150
Dileep Karri: Okay, that's how I would like you all to think about it. Now, we will apply the Spark
framework on a practical example. For that, I want you to propose to me a problem statement,
okay? Propose a problem statement to me, okay? And it can be very random, so take a minute.

360
01:01:41.340 --> 01:01:58.130
Dileep Karri: frame that problem statement in your head, so that when I open the chat, you have
a decent framed problem statement that is available. I will take a well-formed problem
statement, apply the structure, the Spark framework, to decide if I have to apply AI to it or not.

361
01:01:58.200 --> 01:02:07.820
Dileep Karri: Okay? Take a minute to think through this, then I'll open the chat, and I would like
to get some questions from problem statements from you, where I could apply the SPARK
framework.

362
01:02:22.210 --> 01:02:24.140
Dileep Karri: Alright, opening the chat.

363
01:02:24.450 --> 01:02:29.510
Dileep Karri: share your problem statements. I will pick one random problem statement and take
it forward from there.

364
01:02:38.950 --> 01:02:48.869
Uthappa Outskill: Dilip, sorry to interrupt, just while they're putting up their inputs. There has
been one common question that has been popping up in the Slido, which is around.
365
01:02:48.990 --> 01:02:56.799
Uthappa Outskill: Whether the slides will be shared after the session, it will be, but please make
your respective notes as well. That is also necessary.

366
01:02:57.090 --> 01:03:00.550
Uthappa Outskill: So that should answer most of the questions that was answered… asked.

367
01:03:02.700 --> 01:03:08.440
Dileep Karri: All right. I think I have seen some very good, things.

368
01:03:10.570 --> 01:03:12.670
Dileep Karri: Okay, awesome.

369
01:03:19.630 --> 01:03:22.070
Dileep Karri: Got it. So, I have got some…

370
01:03:22.070 --> 01:03:40.229
Dileep Karri: pretty good understanding. So, let me take one thing. I think I'll take this one
question. Regulatory application for new drug application, okay? Can AI write the regulatory
application for a new drug application?

371
01:03:40.840 --> 01:03:49.990
Dileep Karri: Okay, so let me try to tackle it with the Spark framework. I'll take two, three
questions with the Spark framework, so that you understand this part very clearly.

372
01:04:03.340 --> 01:04:08.770
Dileep Karri: Okay, so for my reference and for US, I will write the SPARC framework, essay
structured.

373
01:04:11.400 --> 01:04:15.300
Dileep Karri: Okay? P is pattern, pattern recognition.

374
01:04:15.720 --> 01:04:18.049
Dileep Karri: Previous patterns are there or not.

375
01:04:18.580 --> 01:04:22.499
Dileep Karri: Then A is, amplification.

376
01:04:24.430 --> 01:04:27.840
Dileep Karri: Am I amplifying your thinking, etc?

377
01:04:28.310 --> 01:04:38.970
Dileep Karri: R and K. K stands for knowledge. Does the knowledge already kind of exist? And,
yeah, just…

378
01:04:40.450 --> 01:04:43.200
Dileep Karri: Let me check… What is…

379
01:04:52.430 --> 01:04:55.210
Dileep Karri: R is something, like, repeatable.

380
01:04:56.330 --> 01:05:04.619
Dileep Karri: Is it something that is repeatable? Basically, is it something that I do day-to-day
without any kind of

381
01:05:04.890 --> 01:05:08.499
Dileep Karri: Monotony, etc. How is this particular thing?

382
01:05:08.900 --> 01:05:12.940
Dileep Karri: So, the problem statement that I have chosen

383
01:05:13.090 --> 01:05:18.259
Dileep Karri: Okay, now let me write the… Problem statement clearly.

384
01:05:21.320 --> 01:05:24.390
Dileep Karri: Regulatory application for a new truck.
385
01:05:40.200 --> 01:05:47.700
Dileep Karri: So, regulatory application for a new drug, that is what I'm looking at. Now, each
question. Structured.

386
01:05:47.750 --> 01:06:06.450
Dileep Karri: Okay? So, do I know exactly what is the input for this problem statement, and what
is the output? Okay, so here the output is clear, where I have to formulate an application, where
I'm applying for a compliance for a new drug, so there's an application that will become my
output.

387
01:06:06.450 --> 01:06:11.319
Dileep Karri: My input here would be the various Rules of Drug Act.

388
01:06:11.330 --> 01:06:14.040
Dileep Karri: Okay? The rules of the Drug Act.

389
01:06:14.120 --> 01:06:37.410
Dileep Karri: the framing policies, etc, all that. So, how well am I following that application, etc,
and all that, okay? Again, if some… whoever has put in this question, I am taking some gross
oversimplification over here. I am not telling how to apply this. I am just asking in terms of the
Spark framework.

390
01:06:37.410 --> 01:06:39.400
Dileep Karri: How do I take this problem statement?

391
01:06:39.400 --> 01:06:41.420
Dileep Karri: So, in terms of Spark.

392
01:06:41.550 --> 01:06:47.830
Dileep Karri: There is a clear input and clear output, so therefore, it looks conducive to be
tackled with AI.

393
01:06:48.050 --> 01:06:55.979
Dileep Karri: Then, I look at pattern. Pattern means I'm asking this question, Previously, also.
394
01:06:56.310 --> 01:07:08.199
Dileep Karri: Is this a problem that you think many people would have had, or is it a very unique
problem? Like, is it a problem not anybody in the world would have this problem?

395
01:07:08.200 --> 01:07:25.140
Dileep Karri: Okay? So, there are a lot of drug, regulation, regulatory experts who will be doing
this. So, on terms of pattern also, there will be a lot of people who would have had this. So, in
terms of pattern, that is, repeat, easy tip.

396
01:07:25.490 --> 01:07:30.059
Dileep Karri: problem that many people would have? Yes, the problem is many people would
have.

397
01:07:30.430 --> 01:07:37.739
Dileep Karri: Now, the question is amplification. Am I am amplifying my thinking, or am I
abdicating my thinking?

398
01:07:37.860 --> 01:07:47.820
Dileep Karri: Okay? So here, when I say, write the complete regulatory application for a new
drug, okay, am I,

399
01:07:47.980 --> 01:07:54.229
Dileep Karri: amplifying my thinking or abdicating my thinking. Now, here is where the answer, I
would say, is maybe.

400
01:07:55.070 --> 01:08:01.400
Dileep Karri: somewhere in between, because I just said, write a regulatory application for a
new drug.

401
01:08:01.450 --> 01:08:12.260
Dileep Karri: Here is where I need to think, okay, am I am amplifying my thinking? Do I give
sufficient context? What other information do I need to provide? Okay?

402
01:08:12.280 --> 01:08:31.450
Dileep Karri: Am I saying that I'm not going to put my brain at all and allow only AI to do it, or am
I going to amplify my thinking by giving it more points? So, I would say, if I added some more
points to the question, which gives my own knowledge along with this part, then yes.

403
01:08:31.450 --> 01:08:37.859
Dileep Karri: That is where I would say I would have tried to amplify my thinking, and that is
through additional context.

404
01:08:40.950 --> 01:08:51.569
Dileep Karri: Okay? Then R, is it a repeatable kind of a thing, or is it something, that is, you
know, it is a very hard thing to do.

405
01:08:51.569 --> 01:09:01.359
Dileep Karri: So, it is repeatable, because for every new drug, I will keep writing regulatory
application, so the answer is yes, answer is yes, answer is yes.

406
01:09:01.569 --> 01:09:07.300
Dileep Karri: Here, I have taken a maybe. Now, in terms of knowledge, is it a synthesis or a
genesis?

407
01:09:07.500 --> 01:09:15.259
Dileep Karri: Okay? So, here, again, since I'm writing regulatory applications for a lot of things.

408
01:09:15.260 --> 01:09:27.830
Dileep Karri: It would not be something… it is more of a synthesis, because there have been
previous applications I have written, I just need to adapt it a little. So the answer for this also is
yes.

409
01:09:27.830 --> 01:09:34.550
Dileep Karri: So for this problem statement, for structure, for pattern, for amplification, for
repeatable, for knowledge.

410
01:09:34.550 --> 01:09:42.280
Dileep Karri: I would say a good majority tells that this problem has to be tack… can be tackled
with AI.
411
01:09:43.609 --> 01:09:52.920
Dileep Karri: Okay? This problem can be tackled with AI, that particular problem statement.
Now, let me take another example and see this.

412
01:09:58.730 --> 01:10:01.600
Dileep Karri: Okay, so somebody said, sharing…

413
01:10:06.350 --> 01:10:08.930
Dileep Karri: Photo or event coverage.

414
01:10:13.330 --> 01:10:16.909
Dileep Karri: Okay? Sharing photo and event coverage.

415
01:10:17.340 --> 01:10:20.250
Dileep Karri: With the client in an impressive way.

416
01:10:26.370 --> 01:10:33.630
Dileep Karri: Okay? So, in the Spark framework, is it structured? Now, let me think about my
input and output.

417
01:10:35.470 --> 01:10:48.790
Dileep Karri: Now, when I think about this, if I'm trying to share a photo event coverage, where
are those files? How will I share those files with my system? There'll be terabytes of information.
How do I do that?

418
01:10:48.920 --> 01:11:01.599
Dileep Karri: Then what is the impressive way? What does it mean by impress the client? How
should I do that? If I look at the structure thing, whether I have a well-defined input and output,
the answer is no.

419
01:11:02.270 --> 01:11:08.480
Dileep Karri: Okay? Pattern. Do I know cases where somebody just fed

420
01:11:08.660 --> 01:11:17.429
Dileep Karri: terabytes of information, and do I know cases where an automation system kind of
did this overall?

421
01:11:17.440 --> 01:11:33.860
Dileep Karri: Or is it something that pattern has been there? If I think about it, this is a very hard
problem where technology hasn't committed, and there is definitely huge involvement of human
in the loop. So, in terms of pattern also, answer is no.

422
01:11:33.980 --> 01:11:45.920
Dileep Karri: Is it amplification or abdication? Instead of you putting and telling, this is the editing
thing I need to do, you are abdicating your intelligence, so it is a cross.

423
01:11:46.280 --> 01:12:04.219
Dileep Karri: Okay? Is it repeatable, or is it a one-kind kind of a thing? Every wedding, every
thing, every event is unique in itself, it requires a different strategy, so it is not a repeatable
thing, it is very unique. So, therefore, again, cross.

424
01:12:04.220 --> 01:12:10.320
Dileep Karri: Okay? Again, in terms of knowledge, maybe yes, but you see, right, majority of the
cases.

425
01:12:10.320 --> 01:12:32.610
Dileep Karri: you are able to see that your Spark framework is telling that if you have a photo or
event coverage that you're trying to do it for a client, unless until you rework on this problem
statement, add in more layers of what is the input, what is the output, what does impressive
mean, how do I do it, etc, this would still be like a non-AI kind of task.

426
01:12:33.280 --> 01:12:42.050
Dileep Karri: So, folks, Did you understand how Spark framework is applied just from problem
statements that I've sourced from you?

427
01:12:43.480 --> 01:12:45.150
Dileep Karri: What… how is your this thing?

428
01:12:47.280 --> 01:12:49.009
Dileep Karri: Did you understand this?
429
01:12:51.970 --> 01:12:52.990
Dileep Karri: Okay?

430
01:12:54.390 --> 01:12:55.210
Dileep Karri: Right.

431
01:12:55.340 --> 01:13:05.720
Dileep Karri: Now, I saw a very interesting question. I saw one very interesting question. What if
I don't use this framework? See, again, I would say.

432
01:13:06.750 --> 01:13:10.199
Dileep Karri: You don't have to use this framework always.

433
01:13:10.360 --> 01:13:27.850
Dileep Karri: Okay, I'm not saying we should use only the Spark framework. I am saying the
underlying thing I'm trying to say is whenever you're approaching any problem statement, first,
you should cultivate a wisdom to understand whether you should use AI to solve the problem, or
should you not use AI to solve the problem.

434
01:13:28.090 --> 01:13:40.189
Dileep Karri: Yes, the second problem could be subjective, in a way. Yes, maybe I could have
framed the problem very much better way overall. What I'm trying to show you is a direction, a
framework.

435
01:13:40.380 --> 01:13:44.760
Dileep Karri: How you apply it, how you think about it, how you do it, is up to you.

436
01:13:45.540 --> 01:13:54.310
Dileep Karri: Okay? If you feel that, in your wisdom, you think not the Spark framework, I have
another framework where I decide whether I'll use AI or not.

437
01:13:54.800 --> 01:14:04.519
Dileep Karri: please feel free to do that. In fact, I would be even more happy when you think
critically and say that we will not follow this framework. I have my own framework to do that.
438
01:14:04.520 --> 01:14:17.959
Dileep Karri: But the idea behind doing this exercise for you, in front of you, is to tell you that at
least you should start thinking in these lines before you think, okay, I want to use AI to solve
problems.

439
01:14:17.960 --> 01:14:27.899
Dileep Karri: The first question you had to ask is, do I need to use AI to solve this problem? Is
this problem worth tackling to solve AI with, or use AI to solve the problem?

440
01:14:28.150 --> 01:14:43.980
Dileep Karri: Okay? The key part is you're solving the problem with AI, the problem is not solved
by AI, okay? When you say with, it is a partnership. When you say by, it is delegation. Never
delegate your problem solving to AI.

441
01:14:47.650 --> 01:14:48.550
Dileep Karri: Okay?

442
01:14:52.600 --> 01:15:09.630
Dileep Karri: Others, you might need to practice, you might feel it is a little ambiguous, but I
wanted you to just answer me this question that is the import of this. Do you guys at least have
a small directional level understanding that I need to think about

443
01:15:09.630 --> 01:15:15.299
Dileep Karri: when should I use AI to solve problems? Do you have that directional level
understanding?

444
01:15:16.850 --> 01:15:35.909
Dileep Karri: Okay? Awesome. I see the bunch of thumbs ups, I get it. Thanks a lot for that
feedback. I now know that you have that directional level understanding of how you can think
about it. Okay? You are in the very first session, so some things might appear to you like, okay,
it's a little slightly hard.

445
01:15:35.910 --> 01:15:46.739
Dileep Karri: I have to practice, I have to apply this, etc. Do that, but the main thing I want to
mention is try to, not, you know, don't try to
446
01:15:46.890 --> 01:15:54.510
Dileep Karri: shoehorn yourself into a corner by saying, okay, I tried to apply Spark framework, it
is not working, okay, what should I do?

447
01:15:54.750 --> 01:16:05.109
Dileep Karri: If you're understanding Spark Framework, apply it, make it work well and good. If
you're getting confused by it, at least out of all the five things, pick up one or two things.

448
01:16:05.270 --> 01:16:11.399
Dileep Karri: Okay? S for structured, unstructured is very easy. Inputs and outputs are
structured or not, number one.

449
01:16:11.690 --> 01:16:20.259
Dileep Karri: Then, you think about, is it a knowledge that is already out there, or is it a fresh
problem statement I'm trying to solve? That is also easy to do.

450
01:16:20.260 --> 01:16:34.520
Dileep Karri: So, even if you're not able to pick up that entire 5-spark formula, pick up pieces of
it, at least try to directionally think in terms of understanding, do I need to use AI to solve the
problem or not? That is the main thing I want to push all of you to do towards.

451
01:16:34.920 --> 01:16:38.550
Dileep Karri: Okay, so coming back to our presentation…

452
01:16:45.400 --> 01:16:46.260
Dileep Karri: Alright.

453
01:16:49.200 --> 01:16:50.890
Dileep Karri: Okay? So…

454
01:16:51.010 --> 01:17:07.609
Dileep Karri: the summary of everything is, where does AI create the maximum value? Okay? If
a problem statement solution has high impact, but low criticality, those problem statements, you
should automate.
455
01:17:08.060 --> 01:17:15.019
Dileep Karri: Okay? Where you have high impact and high criticality, use AI to assist you.

456
01:17:15.820 --> 01:17:34.249
Dileep Karri: Okay? Where there is low impact and high criticality, okay, the impact is low, but
the criticality is high, you still have to learn and understand the problem statement deeply before
you apply AI. Wherever that's low impact and low criticality, eliminate.

457
01:17:34.250 --> 01:17:41.170
Dileep Karri: Okay? So these places where you are having high impact, high criticality, where
you assist.

458
01:17:41.170 --> 01:17:48.670
Dileep Karri: high impact, low criticality, where you automate, these are the places where AI
creates a lot of value.

459
01:17:48.860 --> 01:18:05.219
Dileep Karri: Okay? In Learn, what happens is you will have to provide little more depth into the
problem statement. When your impact is low and the criticality is high, that means your problem
statement has to mature a little bit more before you can bring in AI into the picture.

460
01:18:05.960 --> 01:18:22.829
Dileep Karri: Okay? So this is something you need to understand. Now, I will going to teach you
a framework which will help you to work with AI. And this is the framework of clarify, Context,
and Crocate. This is, again, the 3C framework.

461
01:18:22.930 --> 01:18:32.929
Dileep Karri: Okay? It is clarify, Context, and procreate. So, the previous Spark framework
answers the question, should I use AI or should I not use AI?

462
01:18:32.930 --> 01:18:43.479
Dileep Karri: If your answer is yes, you should use AI, how should you use AI? That is the 3C
framework, which is the clarify, Context, and co-create framework.

463
01:18:43.950 --> 01:18:44.850
Dileep Karri: Okay?

464
01:18:46.210 --> 01:18:53.769
Dileep Karri: So, in Clarify, what it does is, you are able to go from a very fuzzy

465
01:18:53.950 --> 01:18:57.960
Dileep Karri: goal to a focused goal, okay? Example is.

466
01:18:58.030 --> 01:19:14.199
Dileep Karri: I need to increase sales. That is the starting goal. It is very fuzzy, it is not very
clear, and everything. You can use AI to clarify it further, and AI can only help when you say
that, you know, what specific metric

467
01:19:14.250 --> 01:19:32.910
Dileep Karri: What time frame, what constraint, what are the constraints that are available, etc.
Okay? Because if you have a template like this, where you have a vague goal, and you are
stuck because of something, help me break this down into specific questions. AI can take the
fuzzy, focused part.

468
01:19:32.980 --> 01:19:40.409
Dileep Karri: And bring that to a clarity, okay? If your question is mostly, like, I need to increase
sales.

469
01:19:40.420 --> 01:19:56.320
Dileep Karri: AI will help you to understand the specific metric, the specific time frame, and the
marker to move, which will give you the focused problem that you need to solve. This is one part
AI is very good at. So the first C is to clarify. Clarify your problem statement properly.

470
01:19:57.200 --> 01:19:58.900
Dileep Karri: Once you clarify.

471
01:19:59.380 --> 01:20:08.909
Dileep Karri: You should try to provide context to AI. And what it means is, provide the
information that you know and the AI does not know.
472
01:20:09.410 --> 01:20:13.759
Dileep Karri: Think about what is the information, or what is the context.

473
01:20:14.260 --> 01:20:18.870
Dileep Karri: that you have, that you know very well, but AI does not know.

474
01:20:19.070 --> 01:20:27.309
Dileep Karri: For example, let us take a previous problem statement, I need to increase sales.
You need… you know what are the constraints, the budget.

475
01:20:27.800 --> 01:20:33.940
Dileep Karri: the team size. You know the environment, you know the culture, you know the
policy, you know the history.

476
01:20:34.100 --> 01:20:42.369
Dileep Karri: You also know the overall goals or the KPI, the key performance indicator that is
given to you in terms of sales.

477
01:20:42.740 --> 01:21:01.420
Dileep Karri: So, all that context that AI does not know, you should try to provide. And a good
template here is, given this, what questions do I need to be answering? Or, what are the
clarifying questions that you might have that I need to further provide?

478
01:21:01.440 --> 01:21:02.750
Dileep Karri: So, context.

479
01:21:03.010 --> 01:21:07.960
Dileep Karri: provide the information that AI does not know. So, first is clarity.

480
01:21:08.140 --> 01:21:12.400
Dileep Karri: Get your… take your vague problem statement, try to get clarity.

481
01:21:12.520 --> 01:21:17.160
Dileep Karri: Context, provide the information that AI does not know.
482
01:21:17.440 --> 01:21:26.259
Dileep Karri: and co-create, wherein you have this process of AI generates options, you
evaluate the feasibility.

483
01:21:26.410 --> 01:21:30.399
Dileep Karri: Refine together, and make the final call.

484
01:21:30.950 --> 01:21:39.340
Dileep Karri: Okay, what it means is AI proposes you dispose the options. You have to build
together, decide alone. This is the key.

485
01:21:39.450 --> 01:21:41.380
Dileep Karri: Okay? You are…

486
01:21:41.670 --> 01:21:48.639
Dileep Karri: doing a lot of brainstorming with the AI, but finally, you have to make that final call.

487
01:21:49.130 --> 01:21:51.179
Dileep Karri: Okay? So the,

488
01:21:51.520 --> 01:22:04.500
Dileep Karri: I have this very great example. Let me talk about that example, and then I will
actually practically show you with this 3C framework how I would go about this problem
statement.

489
01:22:04.770 --> 01:22:15.309
Dileep Karri: So, there was once a machine breakdown problem. Okay, there was a machine
breakdown problem, and a mechanic was brought to

490
01:22:15.400 --> 01:22:29.440
Dileep Karri: You know, to solve the problem. The mechanic turned on the machine, heard the
machine for 5 minutes, switched it off, took a hammer, did one strike at one corner of the
machine.
491
01:22:29.520 --> 01:22:36.700
Dileep Karri: Then turned it back on, the machine was working perfectly well, and he asked the
client, okay, give me $10,000.

492
01:22:37.360 --> 01:22:38.410
Dileep Karri: Okay?

493
01:22:38.840 --> 01:22:44.659
Dileep Karri: Then the guy was, why should I pay $10,000 for a single hammer hit?

494
01:22:45.470 --> 01:22:48.640
Dileep Karri: Why should I pay $10,000 for a single hammer hit?

495
01:22:48.940 --> 01:22:50.560
Dileep Karri: That was the question he had.

496
01:22:51.130 --> 01:22:55.870
Dileep Karri: Then… This person said, okay, let me give you the breakdown.

497
01:22:56.270 --> 01:22:59.799
Dileep Karri: $1 for the hammer hit I have given.

498
01:22:59.910 --> 01:23:08.129
Dileep Karri: And $9,999 for the wisdom to listen to the machine and decide where to hit.

499
01:23:09.090 --> 01:23:22.989
Dileep Karri: You can get a $1.00 technician, they will smash your machine, turn it upside down,
and they would not know where to hit the machine, but I listened to the machine once, and hit it
there.

500
01:23:24.280 --> 01:23:28.669
Dileep Karri: Okay? Now, let me tell you with AI how this works.

501
01:23:29.270 --> 01:23:41.220
Dileep Karri: AI is somebody who will break apart your machine or your problem statement, go
in, go out, and bring it through and through, and confuse you like it.

502
01:23:41.820 --> 01:23:49.410
Dileep Karri: Okay? You should be that AI generalist who knows, okay, I will use AI there and
not here.

503
01:23:50.130 --> 01:23:57.550
Dileep Karri: So that is why this 3C framework. This person had the clarity to understand where
the problem was.

504
01:23:57.820 --> 01:24:14.890
Dileep Karri: Then, he know the context, because that machine's context, he knew which gear
had the exact problem, and then he did the co-creation, or whatever, where he just hit it over
there. Not the exact parallel, I know, but I hope you get the drift, okay, what I'm trying to do.

505
01:24:15.080 --> 01:24:31.900
Dileep Karri: I want all of you to understand this story very clearly, so that when you use AI, think
of this person with the hammer. A person who is inexperienced with AI will try to say, okay, I
have AI, let me dismantle. I will use AI to, again, build it from the ground up.

506
01:24:31.980 --> 01:24:35.280
Dileep Karri: That's a wrong approach. First, listen.

507
01:24:35.390 --> 01:24:40.800
Dileep Karri: First, let the things move across, gather context.

508
01:24:40.920 --> 01:24:54.170
Dileep Karri: Listen. Like, when somebody comes to you with a problem statement, okay, come
on, let's do it. Don't roll your sleeves light away. Take a step back, listen to what is happening,
and then try to solve it. That's the right way to solve the problem.

509
01:24:54.280 --> 01:25:02.639
Dileep Karri: Okay, so now what I'm going to do is, I'm going to use the 3C framework right now
to actually apply and solve the problem.
510
01:25:02.840 --> 01:25:07.050
Dileep Karri: Again, one thing, it will be like watching a mechanic.

511
01:25:07.400 --> 01:25:25.569
Dileep Karri: You will feel like, oh my god, the mechanic had done this easily. When I face it,
how much, you know, effort I have to take, that guy did it with a single stroke. I am getting
confused. How did this person do, and what did that person do? Okay? Understand that there's
a difference between a mentor and a mentee, okay? You are at the starting steps.

512
01:25:25.730 --> 01:25:38.619
Dileep Karri: So, don't get agile, you know, think that, okay, the Hidali breeds so easily. When I
do it, I am kind of stuck there, etc, and all that, okay? So, let's take the thing, and we'll try to
understand how do we do this.

513
01:25:38.750 --> 01:25:52.870
Dileep Karri: So, number one, what I'll do is, I will take a very vague problem statement of an
industry, and this would be the food delivery industry. In India, the top players are Zomato and
Swigi.

514
01:25:53.250 --> 01:25:57.699
Dileep Karri: And in, abroad, the top players are.

515
01:25:58.560 --> 01:26:03.289
Dileep Karri: what is… what do you call it? DoorDash? DoorDash, Uber Eats, Deliveroo.

516
01:26:03.650 --> 01:26:18.800
Dileep Karri: Deliveroo, DoorDash, Uber Eats, okay? Everybody, you are able to understand the
industry we are in, right? Food delivery, right? And you have an app, and you're trying to deliver
the food, especially people abroad. I hope you are able to relate to this example.

517
01:26:19.160 --> 01:26:31.050
Dileep Karri: Okay? Cool. So now, I will take an extremely vague problem statement and try to
use the 3-3 framework to at least attempt to solve it.

518
01:26:31.360 --> 01:26:33.550
Dileep Karri: Alright, so let's go.

519
01:26:42.580 --> 01:26:49.130
Dileep Karri: So… So the problem statement is the food delivery industry.

520
01:26:52.420 --> 01:26:59.119
Dileep Karri: Okay? And examples you can take are Swiggy in India, Zomato in India.

521
01:26:59.250 --> 01:27:07.159
Dileep Karri: Then you have got Uber Eats internationally. Internationally, you have Uber Eats,
or Deliveroo.

522
01:27:08.310 --> 01:27:09.980
Dileep Karri: slash DoorDash.

523
01:27:10.220 --> 01:27:23.350
Dileep Karri: So basically, these are apps where I can order on the app a particular food item
from a restaurant, then a person would pick the order from the restaurant and deliver it to me at
home.

524
01:27:24.150 --> 01:27:28.699
Dileep Karri: Okay? Now, let us suppose I am a head of business.

525
01:27:31.930 --> 01:27:43.060
Dileep Karri: or a head of business at food delivery, comes to you, an AI generalist, an AIG, as I
like to call, comes to you and AIG and say, I want to increase

526
01:27:45.010 --> 01:27:46.360
Dileep Karri: Revenue.

527
01:27:48.460 --> 01:27:50.900
Dileep Karri: And reduce costs.

528
01:27:53.920 --> 01:27:55.989
Dileep Karri: For the food delivery app.

529
01:28:02.880 --> 01:28:06.799
Dileep Karri: come up with a plan of how I can do this.

530
01:28:06.980 --> 01:28:23.669
Dileep Karri: Okay? I want to increase revenue, I want to reduce cost for the food delivery app
company, so I want a revenue-increasing initiative while keeping my costs under control. So,
just for easier sake, I will say, I want a revenue

531
01:28:25.170 --> 01:28:27.320
Dileep Karri: Increasing initiative.

532
01:28:27.950 --> 01:28:30.439
Dileep Karri: While keeping costs under control.

533
01:28:31.580 --> 01:28:36.650
Dileep Karri: So basically, I don't want to burn money, I want to increase revenue.

534
01:28:42.410 --> 01:28:44.250
Dileep Karri: And kind of do this.

535
01:28:44.680 --> 01:29:00.600
Dileep Karri: Okay, so now what I'll do is, I'll first think whether I should use AI or not over here.
First is Spark Framework, wherein I… what is my input and output? Here, again, I know that my
inputs are various

536
01:29:00.650 --> 01:29:10.060
Dileep Karri: context that I need to give, all the context I need to give, etc. The output is a plan
to increase revenue while keeping stocks under control, so it is decent.

537
01:29:10.780 --> 01:29:18.639
Dileep Karri: Are there previous companies which would have gone through this cycle? Every
company would have at least debated or thought of implementing things
538
01:29:19.150 --> 01:29:43.369
Dileep Karri: that are revenue increasing, etc. Is this knowledge kind of available across the
board? Yes, in a lot of business case studies, in a lot of business books, this knowledge should
be kind of available, and am I kind of using, amplifying or abdicating? I am amplifying because
of the way I'm going to use AI to clarify myself, etc.

539
01:29:43.370 --> 01:29:45.769
Dileep Karri: But more or less, what I want to say is.

540
01:29:45.870 --> 01:29:58.229
Dileep Karri: At least in terms of this framework, this is a problem where I can definitely use AI
as my co-pilot and take it forward. Now, I'm going to use the 3C framework, which is clarify.

541
01:29:58.520 --> 01:30:07.579
Dileep Karri: context and co-create, okay? So, let me stop this, and I'll come back and use the
clarify context and co-create to solve this problem.

542
01:30:07.580 --> 01:30:19.460
Dileep Karri: Now, for this, I will use a large language model, and I can use anything. I can use
ChatGPT, I can use Claude, I can use Gemini, I can use DeepSeek, I can use Kimi, K2, etc.

543
01:30:19.660 --> 01:30:28.579
Dileep Karri: Okay? Now, having said that, I will go with the model I personally use day-to-day,
because I will have the comfort level with that.

544
01:30:28.580 --> 01:30:45.020
Dileep Karri: Okay? A hidden C is comfort, but again, you can discard it. I have a comfort level
with that model, and therefore, I am going to use Claude. I am going to use Claude only
because I'm comfortable with it, it is not because you have to use Claude for this. It's a
tool-agnostic framework.

545
01:30:45.870 --> 01:30:48.699
Dileep Karri: Alright? Everybody with me so far?

546
01:30:48.830 --> 01:31:03.150
Dileep Karri: You understood the framework, you understood the things, etc. So I'm going to go
into practical demonstration, where I'll take this very vague problem statement and try to tackle it
with the 3C framework that we have just discussed.

547
01:31:03.390 --> 01:31:08.449
Dileep Karri: Okay? Keep your questions in the Slido, and I will answer them towards the end.

548
01:31:11.120 --> 01:31:12.989
Dileep Karri: Alright, so let's go.

549
01:31:25.130 --> 01:31:38.769
Dileep Karri: Alright, here is Claude. And one more thing, folks, I am using Whisper Flow. So, it
is a dictation AI tool, which even Weibau also uses, so I'm going to use Whisper Flow to dictate
things.

550
01:31:38.770 --> 01:31:48.110
Dileep Karri: So that it makes things easier. Otherwise, there is nothing else I'm doing. I could
have even typed this exact same thing, but to save time, I'm going to do that.

551
01:31:48.460 --> 01:31:50.320
Dileep Karri: Okay, here it goes.

552
01:31:52.380 --> 01:32:02.549
Dileep Karri: I am working for the head of business at a food delivery app like Swiggy slash
Zomato or Uber Eats.

553
01:32:02.690 --> 01:32:17.350
Dileep Karri: And I am the AI generalist in task of solving business problems with AI there. The
pre-business problem that I have got is to increase the revenue of this food delivery app.

554
01:32:17.490 --> 01:32:20.610
Dileep Karri: While keeping the costs under control.

555
01:32:21.040 --> 01:32:38.049
Dileep Karri: So, I want to clarify this problem statement further, and come to a problem
statement that I can tackle further. So, can you help me with those clarifying questions, etc, so
that I can go ahead and try to do this?

556
01:32:42.600 --> 01:32:43.440
Dileep Karri: Okay.

557
01:32:44.640 --> 01:32:45.750
Dileep Karri: One second…

558
01:32:58.710 --> 01:33:09.249
Dileep Karri: Right? So, this is the problem statement I want to clarify. So, this step is where I'm
trying to clarify the problem. Whenever you are trying to do this, okay.

559
01:33:09.750 --> 01:33:20.770
Dileep Karri: keep your model in thinking mode. So, either use GPT-5 thinking model, or Claude
4.1, or Cloud 4 SOAR model. Basically, you have to do thinking model.

560
01:33:21.130 --> 01:33:40.540
Dileep Karri: Okay? So now, great, let us break down this broad business challenge into specific
AI-tractable problems. Here are key clarifying questions organized by category. There are so
many questions that are there. But what I'll do is, I will only pick few questions, and I'm going to
give a commentary, and I'll try to do that.

561
01:33:45.640 --> 01:34:03.079
Dileep Karri: So, my primary revenue streams are through delivery fees and commission from
restaurants, where I charge around 15% commission from my restaurants currently. I don't have
any subscription services, but I do run a few ads.

562
01:34:03.420 --> 01:34:14.490
Dileep Karri: I think the revenue stream that has the most potential for growth, according to me,
is on the advertisement. The current order value is somewhere between $10 to $15.

563
01:34:14.620 --> 01:34:21.450
Dileep Karri: And I see that a user has 2 orders per week, which is the current order frequency.
564
01:34:21.580 --> 01:34:37.109
Dileep Karri: And my specific customer segments are people in Tire 1 metro cities, like Mumbai,
Delhi, Bangalore, and my Tire 3 cities are the ones where they're underperforming a lot.

565
01:34:37.200 --> 01:34:50.120
Dileep Karri: My biggest cost center is my logistics, my delivery logistics, etc, and my… I do give
discounts, but they are not as much as a cost center than the delivery logistics costs.

566
01:34:50.620 --> 01:34:52.240
Dileep Karri: Again,

567
01:34:53.130 --> 01:35:05.439
Dileep Karri: Rest of the things, I would ask you to kind of make suitable assumptions yourself,
and tell me those exact problem statements I should start to tackle and try to solve further.

568
01:35:10.500 --> 01:35:27.880
Dileep Karri: Okay? So, what I'm doing is, I have given the complete information, and I'm trying
to tackle that information with this thing overall. Again, don't think that I'm… I'm not following any
prompting procedures, etc, because you are yet to learn about prompting, which will be from

569
01:35:28.140 --> 01:35:41.980
Dileep Karri: tomorrow's session, I'm just trying to tell you how I take a vague problem statement
in terms of problem solving, and then try to weave in a particular clear problem statement that I
want to use.

570
01:35:43.330 --> 01:35:44.380
Dileep Karri: Okay?

571
01:35:48.750 --> 01:35:58.080
Dileep Karri: So now it is trying to think about that, revenue streams, etc. So, what are the
priority one or different problem statements to tackle?

572
01:35:58.200 --> 01:36:14.569
Dileep Karri: Delivery cost optimization in Tier 1 cities, which is a quick win, where reduce
delivery cost per order through intelligent batching and routing. How can I batch orders to kind
of do that? Tire 3, demand simulation, increase order frequency,

573
01:36:14.570 --> 01:36:31.490
Dileep Karri: through personalized interventions, that is one way I can do. Build an intelligent ad
platform to grow ad revenue, to 15-20% of the revenue. Then, AOV expansion, increase
average order value.

574
01:36:32.050 --> 01:36:38.040
Dileep Karri: through personalized recommendations, etc. So, it has given me multiple

575
01:36:38.720 --> 01:36:48.099
Dileep Karri: questions to kind of do. Delivery cost optimization, Tire 3 demand stimulation,
aerodizing revenue, acceleration, AOV expansion.

576
01:36:48.960 --> 01:37:02.040
Dileep Karri: or anything else overall, okay? I can pick anything I want, so it has given the
priority, but I will have my judgment to decide which one is the one I should pick up. So, I will
pick up the

577
01:37:02.110 --> 01:37:26.110
Dileep Karri: this problem statement, which is intelligent ad platform to grow ad revenue, okay? I
want to run ads on my food delivery platform, so that through those ads, I can get some
additional revenue, where my infra costs of running ads is low. It is the same platform I'm
running, and I'll get some auxiliary revenue, which can kind of improve my revenue by

578
01:37:26.110 --> 01:37:28.539
Dileep Karri: Keeping my ad margins intact.

579
01:37:28.590 --> 01:37:40.689
Dileep Karri: Now, here is where I used my judgment. I… while I didn't know which problem to
exactly tackle, but somewhere in the back of my mind, I had this hint, which got clarified with
this.

580
01:37:41.200 --> 01:37:44.259
Dileep Karri: So now, I will pick this problem statement.

581
01:37:46.770 --> 01:37:49.179
Dileep Karri: Which is to build an ad platform.

582
01:37:50.100 --> 01:37:51.110
Dileep Karri: Okay?

583
01:37:52.160 --> 01:37:55.000
Dileep Karri: Let's… clarify?

584
01:37:55.450 --> 01:37:58.670
Dileep Karri: this… further.

585
01:38:05.240 --> 01:38:20.720
Dileep Karri: Okay, so now I am trying to do this. Let us clarify this further, etc. There are some
bunch of things that are being shown, but what I'll do is, before I even tackle this, okay, I am still
at the clarity level. I want absolute clarity in terms of what is it that I am chasing.

586
01:38:23.300 --> 01:38:24.989
Dileep Karri: Okay, so let it finish.

587
01:38:37.490 --> 01:38:43.869
Dileep Karri: Again, it is giving some very technical things, you don't have to worry about it,
okay? So what I'll do is this.

588
01:38:46.730 --> 01:39:01.389
Dileep Karri: Hey, you kind of jumped the gun here. I… when I said that problem statement, I
wanted you to refine the problem statement with clear goals and metrics, so that I can start
solving the problem. Even now, the problem is quite vague.

589
01:39:05.950 --> 01:39:15.950
Dileep Karri: Okay? So, I kind of treat your AI, like Bible says, as a smart intern, and then try to
get there.
590
01:39:16.510 --> 01:39:18.400
Dileep Karri: Alright, so…

591
01:39:19.670 --> 01:39:30.329
Dileep Karri: Again, psycho fancy of LLM models, that's okay, you can live with it. So, increase
advertising revenue from current 3% of total revenue

592
01:39:30.330 --> 01:39:44.449
Dileep Karri: To 15% within 12 months by building an automated ad platform that helps
restaurants acquire new customers profitably while maintaining user order frequency at 2 orders
per week.

593
01:39:45.240 --> 01:39:49.999
Dileep Karri: Okay? So folks, are we okay with this problem statement?

594
01:39:50.350 --> 01:40:04.180
Dileep Karri: Do you think we achieved some level of clarity right now? Do you think we have
achieved a level of clarity as to what is that initiative we need to chase to kind of get there?
Some level of clarity, yes?

595
01:40:04.590 --> 01:40:05.720
Dileep Karri: Everyone?

596
01:40:06.910 --> 01:40:25.999
Dileep Karri: And are you clear of this process? I tried to drive clarity with very simple
conversation with AI. I didn't use any fancy prompt techniques, etc, though with a good prompt
technique, I can, you know, get it done very well, okay? But yeah, this is a good place to start.

597
01:40:26.360 --> 01:40:30.760
Dileep Karri: Now that I have done this, I will copy this.

598
01:40:31.320 --> 01:40:40.500
Dileep Karri: This is the statement I have clarity towards, now I will build context for it. What are
the things that I know that AI would not know?
599
01:40:41.310 --> 01:40:43.899
Dileep Karri: Alright? So copy the problem statement.

600
01:40:45.440 --> 01:40:52.369
Dileep Karri: I want to… Solve the above problem.

601
01:40:55.210 --> 01:41:02.450
Dileep Karri: Okay? I can ask something like this, what is the unique… information.

602
01:41:03.120 --> 01:41:04.829
Dileep Karri: that I have.

603
01:41:05.370 --> 01:41:07.300
Dileep Karri: But you do not have.

604
01:41:11.020 --> 01:41:18.489
Dileep Karri: You don't have to try to… Tackle, and… Solve the problem.

605
01:41:20.820 --> 01:41:25.190
Dileep Karri: Okay, so now I'm moving from clarity towards context.

606
01:41:25.640 --> 01:41:42.720
Dileep Karri: Okay? So I'm trying to now gauge context, etc. Okay, so this is the unique
information that you have. It is telling me that these are the unique information that you have, I
don't, for solving the problem. And it has given me a lot of information.

607
01:41:46.800 --> 01:41:47.790
Dileep Karri: Okay?

608
01:41:49.020 --> 01:41:55.849
Dileep Karri: there are too many questions to answer, okay? If I keep answering the questions,
that is going to be very, very hard.

609
01:41:56.130 --> 01:42:03.400
Dileep Karri: Okay? I will say, too many… was chimps?

610
01:42:06.650 --> 01:42:10.920
Dileep Karri: Give me the top 5… Questions?

611
01:42:11.590 --> 01:42:17.430
Dileep Karri: Originally, when you try to do this, Maybe, you know…

612
01:42:17.880 --> 01:42:25.930
Dileep Karri: We will try to do it a little better. Again, it's a demo. I will talk about humor, context,
and nuance, etc. So, what are the top questions

613
01:42:26.650 --> 01:42:33.640
Dileep Karri: that… are available. Now, I'll attempt to solve these questions.

614
01:42:36.040 --> 01:42:41.909
Dileep Karri: Hi, so, in terms of the A-B test, where I can…

615
01:42:44.660 --> 01:42:56.649
Dileep Karri: Okay, so the first question on incrementality of ground truth, can I run an A-B test,
etc? I think I can run an A-B test all through, and I think…

616
01:42:56.650 --> 01:43:08.850
Dileep Karri: some of the data is not available, I'm doing it for the very first time, but I definitely
want to design the experiment that can, you know, kind of show or hide restaurant promotions.

617
01:43:09.110 --> 01:43:10.539
Dileep Karri: And

618
01:43:11.090 --> 01:43:24.010
Dileep Karri: In terms of user intent signals, yes, I have the search volumes for each dish, which
I can provide you that particular data. I also have the, restaurant, sorry, I think I went…

619
01:43:25.560 --> 01:43:28.079
Dileep Karri: Sorry, sorry, I think I skipped it.

620
01:43:33.470 --> 01:43:34.380
Dileep Karri: Okay.

621
01:43:37.920 --> 01:43:44.659
Dileep Karri: So, what percentage of orders are truly incremental? I think,

622
01:43:44.660 --> 01:43:59.759
Dileep Karri: I'm not sure how do you look at the A-B test. I'm too early to kind of do that, so I'll
skip that question. Which users actually try new restaurants? I see that 55% of my existing
users

623
01:43:59.760 --> 01:44:15.629
Dileep Karri: tried a new restaurant in the last 30 days, and my insights tell me that most of them
switched, due to either they wanted to try new cuisine, or new restaurants within the same
cuisine.

624
01:44:16.190 --> 01:44:18.800
Dileep Karri: And, I think,

625
01:44:19.000 --> 01:44:28.280
Dileep Karri: The… if the restaurant acquires a new customer, how many times do they reorder?
I think, as I said, a typical reorder is, easily,

626
01:44:28.530 --> 01:44:36.979
Dileep Karri: 3 to 4 times in around 3 to 4 months, and in 6 months, it will be 8 to 10 times there
would be order.

627
01:44:37.100 --> 01:44:53.590
Dileep Karri: So, where can I show ads? I think I can show ad, in the search results by clearly
marking it as an ad. I want to show the top two results as organic, and the third result only, I will
show, like, an ad, listing.

628
01:44:53.590 --> 01:44:57.689
Dileep Karri: And, I have a decent tracking mechanism overall.

629
01:44:58.090 --> 01:45:02.209
Dileep Karri: So that I know if the ad, drove the order.

630
01:45:08.920 --> 01:45:24.050
Dileep Karri: Okay? So, I have given these questions right now. So, basically, right now, what
I've done is, first is clarity the problem statement I had to solve, then the insight, which is with
the problem statement, and the solution approach is kind of coming in.

631
01:45:24.130 --> 01:45:41.300
Dileep Karri: Okay? So the solution approach is, like, predict which of the 55% Explorer users
will try a sponsored restaurant if shown in position number 3, optimizing for restaurants getting
new customers at so-and-so, this thing.

632
01:45:41.300 --> 01:45:57.890
Dileep Karri: Again, I don't need to understand thoroughly. The food experts will actually
understand, the marketing or the business experts will actually understand. My aim is to not
solve this particular problem, but I wanted to show you that… how I will use this process to kind
of

633
01:45:58.380 --> 01:46:22.960
Dileep Karri: you know, come to and build the context. So I've given the context, and what it is
now doing is, it is giving me various inputs, the problem to solve, and now my MVP approach is
coming, okay? At this point in time, what I'm doing is, I have done my clarity stage, I have done
my context stage, where I've given the context, now it is co-creation, where it is saying for MVP
approach for the next

634
01:46:22.960 --> 01:46:23.940
Dileep Karri: 4 weeks.

635
01:46:23.940 --> 01:46:25.509
Dileep Karri: This is what I need to do.

636
01:46:25.700 --> 01:46:44.939
Dileep Karri: Where I want to create a simple model, where I want to show users who tried 2
plus restaurants Hi Explorer, show ads to high explorers initially, charge restaurants fixed $5 per
new customer acquired, start at 50 restaurants, only show ads when their cuisine matches user
search.

637
01:46:44.940 --> 01:46:53.220
Dileep Karri: track some metrics and optimize pricing. So this is a 4-week thing that I… it is
giving. I will take this 4-week.

638
01:46:53.270 --> 01:46:56.340
Dileep Karri: Exactly. Sorry, it's… I picked this up.

639
01:46:59.790 --> 01:47:00.950
Dileep Karri: Copy this.

640
01:47:01.130 --> 01:47:09.359
Dileep Karri: paste this, so I have a 4-week plan AI has recommended. Now, here is where I
have to put my wisdom and try to co-create with AI.

641
01:47:09.610 --> 01:47:11.890
Dileep Karri: Okay? So, I would say…

642
01:47:15.830 --> 01:47:18.429
Dileep Karri: Here, so this is what I'll do.

643
01:47:18.700 --> 01:47:35.310
Dileep Karri: So, the Building the Explorers core, I think I will make it even more simpler,
wherein I will say, within the last 7 days, any user who has taken orders from 2 unique
restaurants.

644
01:47:35.680 --> 01:47:52.069
Dileep Karri: I would, which they have not ordered before, I will think of them as a high explorer.
I want more recency bias, 30 days is too long, and for these people, I want to show ads for
exactly 1 day, and

645
01:47:52.150 --> 01:48:00.629
Dileep Karri: For the restaurants, I want to, I will only work with the top 10 restaurants on my
platform. That means…

646
01:48:00.630 --> 01:48:25.459
Dileep Karri: I will not go for extremely new restaurants. The top 10 restaurants which are
popular on my platform, people might not be visiting them, so I want to run my pilot with only
those 10 restaurants, so the ads I'll run, I'll run only for them, and for them, I will, kind of charge
them only very complimentary rates. It is up to me to go and convince them, and then only I will
make

647
01:48:25.460 --> 01:48:42.139
Dileep Karri: it live. So first, I will take those 10 restaurants, the top restaurants of my platform,
go and talk to them, convince them, and then I would do that. And the testing of the
optimization, pricing, etc, I don't want to do it right away. First, I want to test.

648
01:48:42.140 --> 01:48:56.590
Dileep Karri: If I am running an ad, and I'm charging, say, $2, how much, dollars of orders they
are putting, and just measure that before I do anything else. I want to keep it as simple as
possible.

649
01:48:59.970 --> 01:49:14.640
Dileep Karri: Okay? So, I have kind of, given my inputs to this, so I'm here, I'm co-creator, I'm
actively looking at each of the things that you are doing this, and exactly telling what needs to
be done.

650
01:49:15.090 --> 01:49:31.389
Dileep Karri: Okay, so I have a simplified MVP plan. Last 7 days, ordered from 2 plus
restaurants they have never ordered from before. Add window, show ads for exactly 1 day.
Reset, after seeing ads for 1 day, they need to explore two new restaurants again to re-qualify.

651
01:49:31.390 --> 01:49:42.189
Dileep Karri: Top 10 restaurants I'll pick, brand recognition, etc. If your restaurant pays $2 for the
ad campaign, do they get greater than $20 in new customer orders?

652
01:49:42.190 --> 01:49:52.319
Dileep Karri: then what I need to track, these are the things I need to track, my checklist,
everything, etc, and I can build it further overall, okay? You see this?

653
01:49:52.320 --> 01:49:59.989
Dileep Karri: At the beginning of my chat, I had taken a very vague problem statement, and
towards the end, I think…

654
01:50:00.120 --> 01:50:15.340
Dileep Karri: Almost 20 to 25 minutes I spent on time on this, and I reached a pretty decent
formulation of this problem, which I know that with a small team, I can actually apply. Again,
nuances.

655
01:50:15.340 --> 01:50:30.170
Dileep Karri: I have taken a lot of assumptions and a lot of simplification. My aim is not to solve
this problem, my aim is to tell you how I would apply the 3C framework to tackle any problem
that is out there.

656
01:50:30.310 --> 01:50:49.689
Dileep Karri: Okay? So it is about clarity, where first I get the problem statement's clarity, then
context. I will give a lot of context to my AI, which I have, which AI might not have. Then next is
co-create, where I try to build the plan like a complete sparring partner. I like the term.

657
01:50:50.090 --> 01:51:03.520
Dileep Karri: Always be a sparring partner to your AI. Challenge its assumptions. Make it think
through very hard. Challenge the assumptions, be a sparring partner with AI, and try to come to
a semblance of a plan.

658
01:51:04.090 --> 01:51:16.970
Dileep Karri: Alright, so I hope everybody understood the 3C framework in action. If you have
understood, say yes in the chat, so that you know you have followed the process, taking all the
caveats into account.

659
01:51:21.340 --> 01:51:25.020
Dileep Karri: All right, awesome. Great to see this response, folks.

660
01:51:28.800 --> 01:51:42.300
Dileep Karri: So now I have understood the 3C framework, I know the Spark framework, so we
have done two things, two very important things. We have learned when should you use AI,
when should you not use AI.

661
01:51:42.360 --> 01:51:49.949
Dileep Karri: Then you learned, if you have to use AI, what is a simple framework to get to,
okay? This is an absolute new framework.

662
01:51:50.090 --> 01:51:53.580
Dileep Karri: Now, what I'll do is, I will kind of map

663
01:51:53.610 --> 01:52:04.399
Dileep Karri: a problem statement, and actually tell you if I were to tackle this complete problem
statement, okay? If I were to tackle this problem statement of

664
01:52:04.400 --> 01:52:15.140
Dileep Karri: making this ultra-simple MVP, etc, and I was actually solving the problem. What are
the different AI techniques I might be needing?

665
01:52:15.970 --> 01:52:19.130
Dileep Karri: Okay, so I will copy this.

666
01:52:21.770 --> 01:52:35.240
Dileep Karri: So, basically, this is the last 7 days, 2 plus restaurants have ordered, there's an
add window, and there's a reset, okay? So, very simple understanding of what I will need to do.

667
01:52:54.260 --> 01:53:03.469
Dileep Karri: Okay? So… So, what I need to do is, I need to do a MVP ad platform.

668
01:53:06.260 --> 01:53:09.009
Dileep Karri: Okay? And I am an AI generalist.

669
01:53:09.230 --> 01:53:17.090
Dileep Karri: So, how should I tackle this? So, first is, you know, devising explorer score. So, I
have a dataset.
670
01:53:17.320 --> 01:53:23.120
Dileep Karri: I have a dataset of… Users… Who?

671
01:53:24.310 --> 01:53:27.810
Dileep Karri: ordered… 2 plus times.

672
01:53:31.200 --> 01:53:39.849
Dileep Karri: Previously. Now, I want to build an MVP. Somehow, my data team gave me this
data. Now, I have to define an explorer score.

673
01:53:42.820 --> 01:53:57.320
Dileep Karri: Okay? And to build this Explorer score for my… each of my users, for which I have,
like, a spreadsheet data that is available, I can take two routes. One is, I can calculate the score
on my spreadsheet.

674
01:53:58.460 --> 01:54:08.029
Dileep Karri: directly with Excel, so I will calculate the score on my spreadsheet. Now, for people
who have crossed a threshold score.

675
01:54:11.550 --> 01:54:21.430
Dileep Karri: Okay? Before I even build an ad platform, what I can do is, I can send an email
campaign to them with some

676
01:54:21.430 --> 01:54:37.810
Dileep Karri: voucher or something saying, hey, why can't you try out these new restaurants?
And I'll charge a new restaurant. If the user orders, I will charge the new restaurant $0.5. That is
an absolute, absolute, you know, kind of…

677
01:54:38.180 --> 01:54:41.130
Dileep Karri: the MVP I'm trying to do.

678
01:54:41.270 --> 01:54:48.900
Dileep Karri: Okay? Now, I want to write this email, okay? To write this email, to compose the
copy.
679
01:54:49.640 --> 01:54:52.419
Dileep Karri: I will require a great prompting.

680
01:54:54.570 --> 01:55:05.430
Dileep Karri: I need prompting to compose a good copy of this email. Then, I need a great
banner for an image. For a banner.

681
01:55:07.060 --> 01:55:23.890
Dileep Karri: I will need to use good image and video knowledge. So I have my image and
video. If I want, I want to create an Heijen Avatar video clone of my founder, who is sending a
personalized message to these folks.

682
01:55:24.060 --> 01:55:26.019
Dileep Karri: email. I can do that.

683
01:55:26.450 --> 01:55:49.710
Dileep Karri: So, I can use my audio, image generation for a great personalized banner for
image. I can create something like a personalized video for this person, and if I want to trigger
this threshold and try to do this, I can use automation, and I know most of you have seen the
Mastermind. I can use automation with

684
01:55:49.750 --> 01:55:59.370
Dileep Karri: make or an ATEN to kind of put the modules or the nodes. And then, if I want to
build a very

685
01:55:59.550 --> 01:56:05.280
Dileep Karri: basic ad running analytics platform. I can do the vibe coding.

686
01:56:07.060 --> 01:56:15.459
Dileep Karri: Okay? I can build, like, a simple dashboard analytics with Bold, Lovable, or any
other platform.

687
01:56:15.960 --> 01:56:23.360
Dileep Karri: Okay? So, the idea why I'm breaking this down and I'm telling all this, It should tell
you.

688
01:56:23.690 --> 01:56:31.820
Dileep Karri: that if, as an AIG, I am tackling a problem statement, there are different knowledge
at different levels.

689
01:56:32.310 --> 01:56:37.100
Dileep Karri: Which are going to come into play, and holistically solve the complete problem.

690
01:56:37.370 --> 01:56:50.419
Dileep Karri: You cannot say, hey, I am a, you know, you know what? I am a… I am not a
marketing person, so I don't require to know ads, so I don't require how to create image and
video. Wrong way to think about it.

691
01:56:50.420 --> 01:57:02.379
Dileep Karri: as an AIG, you need to know as good as how to generate images and videos with
AI, because what you're learning this skill is not for your current job, it is for the future.

692
01:57:02.380 --> 01:57:04.660
Dileep Karri: Your future job, when it evolves.

693
01:57:04.770 --> 01:57:12.030
Dileep Karri: Having these other skills with your core skill makes you that music composer with
your own signature.

694
01:57:12.340 --> 01:57:15.189
Dileep Karri: Okay? So the idea why I'm trying to tell you is.

695
01:57:15.800 --> 01:57:23.320
Dileep Karri: Utoppa kind of gave you a vision of different levels, what levels, what you'll tackle. I
wanted to give you, again, I took a

696
01:57:23.320 --> 01:57:47.509
Dileep Karri: sub-problem of a subproblem and told you if an AIG were to touch it, this is how
they will do. If an AIG were to write ad copy, they can do that. If the AIG was to build an ad
mock-up, they could do that. If the AIG wanted to run some automations, they can do that. If the
AIG wants to take the spreadsheet and convert it into a playable dashboard, they can wipe code
and build that. So, every skill that you are going to

697
01:57:47.510 --> 01:58:11.700
Dileep Karri: learn at different levels will kind of come together later, okay? But what it means is,
as you are learning ahead, there will be some pieces which will fall into place, some pieces
which will not fall into place. Whenever you're having that confusion, always come back to this
example and say that for a problem statement to be tackled by an AIG, multiple skills need to
come together.

698
01:58:11.700 --> 01:58:15.619
Dileep Karri: So, I will not question why this skill has been taught.

699
01:58:15.790 --> 01:58:29.589
Dileep Karri: I will take this skill, make my ripper, and as I go along the journey, I will slowly keep
connecting the dots and move towards becoming an AIG who is a problem solver who
leverages AI.

700
01:58:30.400 --> 01:58:42.810
Dileep Karri: Okay? So, having said that, come back to the presentation. So, what I've done is I
kind of skipped a little bit ahead, but I tried to kind of paint that picture for you as to how an AI
generalist would think.

701
01:58:43.320 --> 01:59:07.320
Dileep Karri: Okay, so I had done this complete case study all through. So where, you know, I
tried to apply the Spark framework in the case study. So the… I told the company spreadsheets,
but I didn't explicitly mention that, but that is exactly the use case I tried to solve, where I said
about more revenue. So there are portions of it where you saw it had done without AI,

702
01:59:07.320 --> 01:59:18.660
Dileep Karri: where I had done it with a partner. If I were to do the exact same thing absolutely
without AI, the same thing I had done for 25 minutes, I will have to do at least a one and a half
to two hour

703
01:59:18.660 --> 01:59:21.880
Dileep Karri: Basic brainstorm, I would just scratch the surface.

704
01:59:22.360 --> 01:59:41.239
Dileep Karri: Okay? So, again, some four powers you require as an AI generalist. One is to
power to build without code, the power to automate without programming, the power to create
without traditional skills, and the power to analyze, okay, without complex frameworks. These
are the four powers.

705
01:59:41.270 --> 01:59:53.770
Dileep Karri: And as you learn different levels, you'll be touching upon each power. For example,
the power to create will come in both all the level 1, 2, and 3.

706
01:59:53.770 --> 02:00:17.990
Dileep Karri: The power to automate is completely specific for level 4. The power to build without
code is level 5. The power to analyze, it comes a bit in Level 1. Today, you have started to think
of how you can use AI to analyze, etc. The prompting techniques will help you, how you can
convey what you need to do to the LLM better, so that you can analyze stuff and

707
02:00:17.990 --> 02:00:22.740
Dileep Karri: Analyze any problem statement without complex frameworks, And,

708
02:00:22.740 --> 02:00:42.190
Dileep Karri: you will be able to tackle all of these four powers. So, there are four core principles
of an AIG, and these are the four powers that you will accumulate. You will not accumulate all
four powers equally at each level. Level 1 is where you might maybe focus more on power to
analyze.

709
02:00:42.190 --> 02:00:54.039
Dileep Karri: Then maybe in the image level, you will go more on power to create, then
automation level on power to automate, and the build phase is where you have the power to
build.

710
02:00:54.650 --> 02:01:18.019
Dileep Karri: Okay? So again, this is a 6-month journey overall, basically where you will learn
foundations, where you will try to spend time on power to analyze, then image creation is power
to create, automation, power to automate, this is power to build, and level 6 is where you bring
all of these problems together and be focused and build, solve a problem, being an AIG.

711
02:01:18.140 --> 02:01:32.860
Dileep Karri: So at the end of 6 months, the first session that you have seen through, you will…
trust me, you will get a lot more clarity as to what I have done today, because at the end of 6
months, you will kind of connect all the dots. So this is something, like, looking forward.

712
02:01:32.860 --> 02:01:43.260
Dileep Karri: The conundrum or the problem is you can always connect the dots better looking
back than looking forward, but at least I wanted to give you those hints.

713
02:01:43.580 --> 02:01:54.309
Dileep Karri: Okay? Again, this part we have kind of covered all through, so I won't go into it
much. Basically, month to two, it is about thinking with AI that you will try to learn.

714
02:01:54.540 --> 02:02:12.690
Dileep Karri: visual creation, where your power to create is what you will learn in Level 3.
Automation is where you build a digital orchestra. This is the part where we'll spend quite a bit of
time, because it is a little difficult, frustrating, why, how? We will discuss when we come to it.

715
02:02:12.820 --> 02:02:22.500
Dileep Karri: Then Level 5 is we will make you from a user or a consumer to a builder, where
you'll be able to create applications with the help of AI.

716
02:02:22.760 --> 02:02:30.869
Dileep Karri: And then, we will try to orchestrate the complete symphony by building an
end-to-end AI solution through and through.

717
02:02:31.400 --> 02:02:41.389
Dileep Karri: Okay, that's your journey. So, if I have to map your journey going ahead, let us look
at day 1 versus day 180. Again, this is a slide, keep it

718
02:02:41.980 --> 02:02:55.529
Dileep Karri: keep it in the back of your minds. Day one, you might be thinking, how do I use
ChatGPT? Day 180 is, hey, I have the knowledge to build AI solutions that could save some
dollars. Okay, even if it saves $10, it is fine.

719
02:02:55.530 --> 02:03:03.620
Dileep Karri: AI might take my job if that was something that was there in your head. Day 180
should be, I created a role that does not exist.

720
02:03:03.720 --> 02:03:06.830
Dileep Karri: I created a role that does not exist.

721
02:03:07.500 --> 02:03:23.599
Dileep Karri: Alright, then what is the… I'll need to learn every tool. If you thought day one is
where you say, I need to learn every tool that is out there, then, day 180 is the time where you
would say.

722
02:03:23.600 --> 02:03:29.659
Dileep Karri: I can master any tool, because I am not learning tools here, I am learning
frameworks.

723
02:03:29.660 --> 02:03:36.329
Dileep Karri: Okay? Today, I did not teach any tool. And going ahead also, mentors will focus
more on frameworks.

724
02:03:36.330 --> 02:03:41.480
Dileep Karri: And tool will become just a means to learn that particular framework.

725
02:03:41.560 --> 02:03:55.319
Dileep Karri: Okay? So you'll learn how to master every… any tool overall, all that. So the
protocol is always, in this session, the way to learn is make notes.

726
02:03:55.340 --> 02:04:10.080
Dileep Karri: Not because we don't give you resources so that… but you can absorb the
material better, okay? And try to apply a little bit, even though difficult, and try to not, you know,
after the session is over, I would say.
727
02:04:10.220 --> 02:04:18.919
Dileep Karri: the… don't think, like, okay, what is the assignment, what should I do, etc. The
mentor would have shown something, just replicate it as is.

728
02:04:19.370 --> 02:04:23.420
Dileep Karri: Take the mentor, what they have shown, replicate it as is.

729
02:04:23.420 --> 02:04:39.920
Dileep Karri: And then you will slowly, as you come to different months, and as you move ahead,
apply it on your domain, pick small, small problems, and proceed ahead. Okay? Unless you
apply AI to solve your problems, the learning will not be complete.

730
02:04:40.460 --> 02:04:43.909
Dileep Karri: Alright? So there are 3 learning principles. First.

731
02:04:43.910 --> 02:05:03.550
Dileep Karri: Learn on demand, okay? Not, like, you know, don't go in a very thing that I will
understand this principle properly. Learn as much as you need. Don't try to say, okay, if it is… if I
want to use AI, this is prompting, I need to learn every prompting method that is out there. No.

732
02:05:03.720 --> 02:05:07.470
Dileep Karri: Take those few basic prompting techniques, get started there.

733
02:05:07.590 --> 02:05:20.860
Dileep Karri: Okay? Then teach everything yourself, or your spouse, your partner, etc. Teach
everything what you have learned in your own words, and make it like a good 5-10 minute
summary, not big required.

734
02:05:20.860 --> 02:05:28.440
Dileep Karri: If you are able to explain anybody in your own words, and they understand the gist
of it, that means you have also understood it.

735
02:05:28.440 --> 02:05:36.359
Dileep Karri: Third point, very important, compare yourself only to the yesterday's you. Your only
competition is your past self.
736
02:05:36.400 --> 02:05:52.550
Dileep Karri: There is a point where you'll get imposter syndrome, especially when you see a
member of the cohort asking some questions that you don't even understand. You see a
member of cohort trying to build things and moving ahead of you. Don't ever get into that kind of
a syndrome.

737
02:05:52.590 --> 02:06:02.520
Dileep Karri: Always compare you to yourself, and how you were before this, and how you were
after this, that is the only comparison you need to make.

738
02:06:02.570 --> 02:06:14.369
Dileep Karri: How were you yesterday before you had come to this session? How are you right
now as we come to the little end of the session? That's all you need to think through.

739
02:06:14.590 --> 02:06:19.479
Dileep Karri: Okay? So the generalist manifesto, I call it, again.

740
02:06:19.880 --> 02:06:22.259
Dileep Karri: I will think before I prompt.

741
02:06:22.380 --> 02:06:38.539
Dileep Karri: I will maintain mental fitness, I will not be anxious, and I will not say I'm not
understanding, so I'll give up. I will embrace discomfort, try to show up and get my thing solved.
I will use AI as a leverage and not crutch.

742
02:06:38.540 --> 02:06:43.139
Dileep Karri: I will create unique value. Unique value you might not create right away, but…

743
02:06:43.200 --> 02:06:49.980
Dileep Karri: you can create unique value by, you have learned a Spark framework, go back.

744
02:06:49.980 --> 02:07:06.280
Dileep Karri: and into your office, and pick one problem statement, apply the Spark framework,
and think, okay, can I apply AI or not? That is a unique value you have already created. Go talk
to your manager, go talk to your reportee and say, hey, we can think about AI like this. That itself
is

745
02:07:06.280 --> 02:07:17.429
Dileep Karri: unique value, and help others adapt. When somebody asks a question or anything,
if you know the answer, help them out, because that way you are going to get much better.

746
02:07:17.950 --> 02:07:20.819
Dileep Karri: Alright, so let me see…

747
02:07:37.580 --> 02:07:42.119
Dileep Karri: I think screen sharing suddenly stopped for some reason, let me again resume it,
yeah.

748
02:07:42.170 --> 02:07:58.740
Dileep Karri: So, again, this is the AI Generalist Manifesto. So, think before I prompt, maintain
mental fitness, use AI as leverage and not crutch. Create unique value, help others in your
journey. And one thing is, you know.

749
02:07:59.770 --> 02:08:05.509
Dileep Karri: I think Ghutapa would have already mentioned, you will get support, but don't
expect handholding.

750
02:08:05.510 --> 02:08:23.950
Dileep Karri: Okay, support and hand-holding are different, and if you've got any problem, do
feel free to come to the officers, do ask your questions, we'll have all the support needed, but
don't say, okay, I don't know how to do this, no, you tell me how to do this, and I will… I… make
me do this, okay? Again, you cannot outsource your push-ups.

751
02:08:24.590 --> 02:08:42.740
Dileep Karri: Alright? So, success metrics, okay? And I want you all to explain or tell me the
answer to these questions as we kind of go ahead. So, today's session, folks, are you able to
define an AI generalist, what they do in your own words, yes or no?

752
02:08:42.740 --> 02:08:44.920
Dileep Karri: Can you define an AI generalist?
753
02:08:46.500 --> 02:08:48.200
Dileep Karri: Okay? Awesome.

754
02:08:48.300 --> 02:08:55.959
Dileep Karri: Do you know how do you identify when not to use AI? Do you have that wisdom
right now, a little bit?

755
02:08:56.210 --> 02:08:57.170
Dileep Karri: Yes?

756
02:08:57.290 --> 02:09:06.429
Dileep Karri: Got it. Do you know how you can potentially apply the 3C framework? You might
not know it 100%, but at least you know the direction.

757
02:09:07.680 --> 02:09:20.970
Dileep Karri: Okay? Then, you see your path. Do you guys feel confident of what you've signed
up for, and how the skills will kind of come together and make you a complete AIG?

758
02:09:21.280 --> 02:09:24.130
Dileep Karri: Do you have that clarity right now?

759
02:09:24.890 --> 02:09:31.999
Dileep Karri: Okay? You might get the clarity slowly, you might take some time about it all
through. It's okay completely.

760
02:09:32.110 --> 02:09:42.249
Dileep Karri: This is the most important question I'll ask, okay? The question is, are you better
than your previous self 3 hours back?

761
02:09:43.710 --> 02:09:47.070
Dileep Karri: Are you better than your previous self 3 hours back?

762
02:09:47.810 --> 02:09:54.149
Dileep Karri: Okay? That's it. That is the question you keep asking, that is the thing you keep
doing and keep going ahead.

763
02:09:54.380 --> 02:10:07.280
Dileep Karri: Alright, so I wanted to keep the first session a little lighter, not make it heavy right
away. Tomorrow onwards, you will have slightly more heavier sessions with the mentors. Take it
with a pinch of salt.

764
02:10:07.300 --> 02:10:20.259
Dileep Karri: Okay? And, tomorrow, what we'll do is, we will learn a little skill of prompting. The
mentor will take you through the prompting techniques, etc, and kind of take it forward.

765
02:10:20.530 --> 02:10:28.749
Dileep Karri: Okay? For those of you who feel this is also heavy, this is nothing. You are going to
get even more heavier ahead. Okay? Embrace the discomfort.

766
02:10:29.090 --> 02:10:31.710
Dileep Karri: All right? So, Utapa, over to you.

767
02:10:32.700 --> 02:10:40.269
Uthappa Outskill: Thank you, Dilip. Thank you for setting the right context in the first session.
Folks, I hope you have

768
02:10:40.920 --> 02:10:48.259
Uthappa Outskill: I would say a ground rules, a roadmap on how you should start thinking as a
problem solver, being an AI journalist right now.

769
02:10:48.470 --> 02:10:50.770
Uthappa Outskill: Whenever you kind of…

770
02:10:51.160 --> 02:10:56.529
Uthappa Outskill: Digress or divert from this route, come back, watch these notes, watch this
session, make this…

771
02:10:56.660 --> 02:11:10.270
Uthappa Outskill: this should be, like, your guiding rail for the rest of the 6 months, right? So this
is something that is going to bring you back to the path of thinking like an AI generalist by
solving problems, leveraging AI. Today, basically, you've learned

772
02:11:10.510 --> 02:11:15.039
Uthappa Outskill: When to use AI, or when not to use AI, and most important.

773
02:11:15.360 --> 02:11:30.900
Uthappa Outskill: have to use AI, right? So this is something that you've learned today. As Dilip
mentioned, tomorrow is where, your journey becomes a little… it goes a step higher. You will
start understanding how to talk to AI as well, right? How to kind of,

774
02:11:30.970 --> 02:11:39.399
Uthappa Outskill: articulate whatever you want to communicate to AI, because I think you would
have already heard this in our masterminds of previous session. Garbage in is garbage out.

775
02:11:39.460 --> 02:11:46.740
Uthappa Outskill: So, you know how to talk to AI, it responds in the same manner. So that's
something that you're going to start learning tomorrow. So, we're going to have

776
02:11:46.760 --> 02:11:54.289
Uthappa Outskill: part one of prompt, engineering, the techniques that you're going to use, to
prompt it tomorrow. Some of you would have already

777
02:11:54.290 --> 02:12:07.899
Uthappa Outskill: heard about magic prong formula that we discussed in our mastermind. This is
going to be beyond that, that we're going to cover. Alright? So thanks again, Dilip. We will,
again, have the Q&A section after a short, diversion here.

778
02:12:07.900 --> 02:12:17.749
Uthappa Outskill: Folks, I would be doing this on a every-session-level basis. We would
definitely value all your feedbacks. We want to know how we should, proceed forward, or…

779
02:12:17.750 --> 02:12:30.179
Uthappa Outskill: what improvements we need to keep having. So, I would like to take a poll on
the mentor as well as session content here. Karthik, if you would launch the poll.
780
02:12:31.700 --> 02:12:42.660
Uthappa Outskill: No, we are not finished. Folks, stay back. I have a couple of things to discuss.
We also have Q&A after this section, so stay back for 5 more minutes. Sorry, stay back till the
end of the session, to be precise.

781
02:12:44.160 --> 02:12:49.330
Uthappa Outskill: Right? I'll just take a minute here as you, participate in the poll that has been
launched.

782
02:12:49.450 --> 02:12:53.860
Uthappa Outskill: And… Let me pause the chat here.

783
02:12:55.140 --> 02:12:57.460
Uthappa Outskill: I'm gonna drop my email ID here.

784
02:13:01.560 --> 02:13:07.839
Uthappa Outskill: Okay? So, anybody who has rated it 3 or less than 3,

785
02:13:07.940 --> 02:13:10.719
Uthappa Outskill: I would want to hear your feedback personally.

786
02:13:10.740 --> 02:13:15.120
Uthappa Outskill: So, please go ahead, drop me your feedback, whoever has rated.

787
02:13:15.130 --> 02:13:31.899
Uthappa Outskill: 3 or less than 3, drop me a feedback on our, on my email ID that I've shared,
so that I can understand your perspective better, and have a discussion and, figure out how we
should, address it. So that's something that we'll keep doing, right? So you have my ID there.

788
02:13:32.660 --> 02:13:39.049
Uthappa Outskill: We still have a few more folks to vote in. Folks, I'll keep it open for the next 5
seconds if you can complete it.

789
02:13:46.770 --> 02:13:50.700
Uthappa Outskill: Got it. I already see 80% of you have given your feedback.

790
02:13:51.220 --> 02:13:57.550
Uthappa Outskill: Karthik, you can just… Hold on, I still see people giving it.

791
02:14:04.750 --> 02:14:08.549
Uthappa Outskill: Yeah, you can pause the chat, sorry, the poll, Karthik.

792
02:14:09.350 --> 02:14:16.829
Uthappa Outskill: Thank you, folks, and as I mentioned, all those people who've rated 321,
please drop me an email with your reasons.

793
02:14:17.600 --> 02:14:20.229
Uthappa Outskill: Now, what are we going to do next?

794
02:14:20.460 --> 02:14:25.520
Uthappa Outskill: I'm going to give Dilip a 5-minute break, and also all of you. Do you guys need
a break?

795
02:14:28.850 --> 02:14:30.370
Uthappa Outskill: Yes, Olit.

796
02:14:32.600 --> 02:14:36.410
Uthappa Outskill: So, we'll take a 5-minute break, at this point.

797
02:14:36.770 --> 02:14:47.829
Uthappa Outskill: I'll set a timer as well. Freshen up, come back. Once we are back, firstly, we
are going to begin with a surprise that I have for all of you. After that.

798
02:14:47.970 --> 02:14:59.100
Uthappa Outskill: we are going to have a Q&A. All the questions that you've put into the Slido,
we are going to have a Q&A done, alright? So let's take a quick 5-minute break, come back,
same link.
799
02:14:59.270 --> 02:15:02.669
Uthappa Outskill: And then we start. I'm going to set up a timer right now.

800
02:20:27.710 --> 02:20:30.040
Uthappa Outskill: Alright, I'll be back.

801
02:20:31.550 --> 02:20:33.840
Uthappa Outskill: Folks, let me know if you're back.

802
02:20:34.860 --> 02:20:36.390
Uthappa Outskill: Are we good to start?

803
02:20:47.430 --> 02:20:49.010
Uthappa Outskill: Yes, let's go.

804
02:20:49.260 --> 02:20:59.040
Uthappa Outskill: Okay, I see most of you are back, so let me, again, quickly pause. And this is
something we'll keep doing, okay, so that the flow of the session is maintained. I'll keep
switching on and off the chat.

805
02:21:01.460 --> 02:21:06.469
Uthappa Outskill: So, I was going through the Slido questions. There were some operational
questions which…

806
02:21:06.720 --> 02:21:13.749
Uthappa Outskill: Dilip need not, take it up, so probably I'll answer those questions I've
categorized into 5, questions that you guys have asked.

807
02:21:14.030 --> 02:21:17.639
Uthappa Outskill: And, although it is limited to a few people, I'll still answer it.

808
02:21:17.910 --> 02:21:19.310
Uthappa Outskill: And,
809
02:21:20.080 --> 02:21:29.649
Uthappa Outskill: anything content-related, after this, we'll take it up, alright? So, first thing, I
understand sessions can be longer, and…

810
02:21:30.240 --> 02:21:39.950
Uthappa Outskill: We cannot fix if we can give you a break after every 1 hour. So the question
was basically, can we add breaks after every 1 or 1.5 hours, because the sessions are going to
be long?

811
02:21:40.020 --> 02:21:53.589
Uthappa Outskill: See, it completely depends on the topic we are discussing. At any point of
time where I feel that, the complexity is a little higher, and we might have to take a break so that
you don't, kind of, fatigue.

812
02:21:53.590 --> 02:22:00.369
Uthappa Outskill: That's where I will probably kind of add the break, or else expect it to be just
after the session is done.

813
02:22:00.580 --> 02:22:08.920
Uthappa Outskill: If you have to take a break, feel free to go, take your bio break and come
back, that's completely okay. But as a…

814
02:22:08.970 --> 02:22:27.969
Uthappa Outskill: break that I would be giving, it'll depend upon the topic and the way the
session has been going, but definitely we'll have, one or two breaks, at least in the session,
minimum one, that we can have all of you concentrated on the session, alright? That was one
operational question that I was able to see. A lot of questions on the tools. Firstly.

815
02:22:28.540 --> 02:22:37.210
Uthappa Outskill: During our onboarding call, we discussed about 10 to 12, or in the
mastermind, also, you would have, heard a few tools that you would be provided.

816
02:22:37.820 --> 02:23:00.580
Uthappa Outskill: you will get all the access for that, right? But we're not going to share all the
access at the beginning of the session, or the beginning of the cohort itself. As I said, tools like
Bold, where you build, or you are MVPs, that topic itself is coming around 5th or 6th month. So
there is no way, it'll be helpful for you right now. If I give it to you right now, you'll already be
losing 5 months of the subscription, so we cannot

817
02:23:00.580 --> 02:23:09.139
Uthappa Outskill: go in that manner for a 6-month longer, program. So, for each level, at each
level, we will be unlocking few tools to you.

818
02:23:09.140 --> 02:23:24.160
Uthappa Outskill: So that you can use it. Some of those tools will be directly relevant to the
content that we are going to discuss, and some are, like, a value-added tools that you'll be
getting. The contents of it will not be discussed in the session, but you can still go ahead and
make use of it, right?

819
02:23:24.160 --> 02:23:28.400
Uthappa Outskill: For example, okay, I'll come to the surprise a little after this question.

820
02:23:28.450 --> 02:23:44.050
Uthappa Outskill: There was another question around tool, around, will we get Cloud or
WhisperFlow? No, we don't have it as part of a partnership right now, because if you go and
see, the kind of, API rates or token rates that, Cloud has already, it's

821
02:23:44.050 --> 02:23:53.459
Uthappa Outskill: quite high, I don't think they'll get into a partnership with anybody, soon. So
Cloud or Whisperflow is not part of the tool list that we provide. It is optional, it is your personal
choice.

822
02:23:53.670 --> 02:24:06.670
Uthappa Outskill: We are already giving you, publicity as part of the tools list, right? So if you
want to, get cloud subscription, it's going to be your personal choice. That's completely up to
you, but we don't have it as part of the program that we have.

823
02:24:07.770 --> 02:24:13.090
Uthappa Outskill: Now, what I wanted to announce, since I'm already discussing about pulses,

824
02:24:13.710 --> 02:24:18.409
Uthappa Outskill: So, after this 5 questions that I've answered,
825
02:24:18.950 --> 02:24:27.590
Uthappa Outskill: To all the folks that have… that are part of the cohort right now, we are going
to start unlocking or giving you access for the first set of tools.

826
02:24:28.190 --> 02:24:45.179
Uthappa Outskill: First is going to be perplexity. So, all of you, you would already have access to
perplexity tool, we have, given access to the, we have given you those codes as well. So, after
this session is done, sorry, after the session is done, Karthik, who is your code manager, he will
demonstrate how you're supposed to redeem

827
02:24:45.180 --> 02:24:46.260
Uthappa Outskill: The tool.

828
02:24:46.260 --> 02:25:09.170
Uthappa Outskill: Alright? So this is the first tool that you're going to have access to. Second,
we're also going to give you access to Emily AI, but it is not going to be… access is not going to
be given today, it'll be given in another 2-3 days, we are waiting for some coupon codes to come
in. So you will have access to MLE as well in the next 2-3 days. So, before the next weekend,
you will be able to redeem these two tools, Perplexity and Emily.

829
02:25:09.380 --> 02:25:20.820
Uthappa Outskill: So this way, at each levels, depending upon the relevance of those tools in
each of these levels, we will be unlocking tools at regular intervals. Alright?

830
02:25:21.640 --> 02:25:37.029
Uthappa Outskill: Cool. We will show you how you're supposed to redeem it, and if you find an
issue with the process of redeeming it, we'll also share a Google form which you can fill, and our
team will reach out to you and help you out. As I said in our onboarding call, these are external
partnerships.

831
02:25:37.350 --> 02:25:57.100
Uthappa Outskill: if you find any issue, obviously there is going to be some time that we will
take, at least 1 to 7 days for us to be able to fix this, because we have to reach out to those
teams and get it sorted. So I expect, as we said, we are in a partnership here, you, me,
everyone here, so please allow us that time to be able to solve it, and we'll definitely get to it and
get it addressed.
832
02:25:57.150 --> 02:26:01.509
Uthappa Outskill: Okay? I hope that piece of information helps.

833
02:26:01.520 --> 02:26:11.869
Uthappa Outskill: question around recordings, post-reads, assignments. For this session, there
is no assignment. We will let you know whichever session there is an assignment. For this
session, you just have to go back.

834
02:26:11.870 --> 02:26:24.809
Uthappa Outskill: make a note or create your running doc of all the notes that you have. In that,
highlight the templates or the frameworks that we have discussed today on when to use AI or
how to use AI. Just have those things jotted down.

835
02:26:24.830 --> 02:26:32.629
Uthappa Outskill: Right? Recording, you will have access to the recording, we'll be putting it in
your LMS, by 11am.

836
02:26:32.630 --> 02:26:50.859
Uthappa Outskill: tomorrow. That's going to be the standard that we try to follow for all those
sessions. The next day, by 11am IST, is where you will have your recordings available. Expect
plus or minus 30 minutes, depending upon the availability and other things, but it should be
available the next day itself. It'll also have the post-read material, so not worry on that.

837
02:26:52.030 --> 02:27:01.220
Uthappa Outskill: The pre-read materials for tomorrow's session is already available, if you go to
LMS under, your content, section.

838
02:27:01.850 --> 02:27:04.159
Uthappa Outskill: Let me also show it to you so that there is no…

839
02:27:04.290 --> 02:27:05.969
Uthappa Outskill: Doubts on how to access it.

840
02:27:19.570 --> 02:27:22.430
Uthappa Outskill: Yeah, so… Oh, yore?

841
02:27:23.600 --> 02:27:24.960
Uthappa Outskill: your first space.

842
02:27:25.240 --> 02:27:26.480
Uthappa Outskill: Go to that.

843
02:27:26.960 --> 02:27:38.549
Uthappa Outskill: And then when you go to, so this time, we have added it under Week 1,
Session 1, so you have a pre-read material for your Session 2 available here. You can go
through it.

844
02:27:38.740 --> 02:27:42.079
Uthappa Outskill: Okay? So that is it.

845
02:27:42.190 --> 02:27:50.269
Uthappa Outskill: And then coming to POCs. Who are your POCs? First point of contact,
Karthik. Again, to show where you can find these details.

846
02:27:59.840 --> 02:28:03.420
Uthappa Outskill: So, if you go, to the starter kit.

847
02:28:03.950 --> 02:28:07.409
Uthappa Outskill: you will have an option called as Help and Support Contacts.

848
02:28:07.540 --> 02:28:25.549
Uthappa Outskill: We have listed down the people whom you're supposed to reach out to,
official email ID channel, your first point of contact, Karthik. Karthik, will… is also available on
WhatsApp. Kartik, you can drop your number at the end of the session, or in the cohort itself, so
that they can mention it here, so that they can be able to reach out to you on WhatsApp as well.

849
02:28:26.590 --> 02:28:45.279
Uthappa Outskill: Alright, so Karthik will be your point of contact. You can also write your emails
to [Link], which is our official support channel. Any escalations, or you want
to reach out to me directly, I've dropped my email ID as well in the chat earlier,
[Link]. Please reach out to me as well. This is your POCs.

850
02:28:46.460 --> 02:28:54.719
Uthappa Outskill: Okay, I have finished, those five operational questions, so expect me to skip
those when they, pop up in the

851
02:28:55.100 --> 02:28:56.539
Uthappa Outskill: the Q&A section.

852
02:28:56.820 --> 02:29:03.460
Uthappa Outskill: We will take a short 5-minute section here on how to redeem your tools.

853
02:29:04.590 --> 02:29:22.100
Uthappa Outskill: and questions of it, or we will probably take it up at the end once we finish up
with the content-related Q&A that we can cover here. So, Karthik, if you can quickly take them
through the steps to redeem the tools, and how can they access these steps, it'll be helpful.

854
02:29:25.670 --> 02:29:26.889
Kartik - Outskill: Sure, thanks, Utapa.

855
02:29:27.770 --> 02:29:31.489
Kartik - Outskill: Hi folks, so… just let me know once my screen is visible.

856
02:29:49.360 --> 02:29:50.440
Uthappa Outskill: Jakarti.

857
02:29:50.680 --> 02:29:52.820
Kartik - Outskill: Okay, I hope my screen is a bit.

858
02:29:52.910 --> 02:30:05.309
Kartik - Outskill: Okay, so, folks, we have created a document for all of you, just so that you all
know how to redeem the tools. The first two tools, as Utapa mentioned, we have

859
02:30:05.360 --> 02:30:21.459
Kartik - Outskill: We are, we are distributing, we are, like, unlocking our Emily and Perplexity.
So, firstly, to unlock them, you will have to visit [Link]. I have already added the
link, you can just click over there.

860
02:30:24.120 --> 02:30:28.300
Kartik - Outskill: This is, the page, which will open.

861
02:30:29.430 --> 02:30:41.649
Kartik - Outskill: And once, once this gets opened, you can just go to your tool, let's say it's
Perplexity or Emily, you can just click on Redeem now, and you will have your

862
02:30:41.950 --> 02:30:43.359
Kartik - Outskill: Coupon code ready.

863
02:30:43.850 --> 02:30:45.960
Kartik - Outskill: I hope this is clear.

864
02:30:46.620 --> 02:30:49.720
Kartik - Outskill: Like this, the coupon code will be visible.

865
02:30:52.840 --> 02:30:54.529
Kartik - Outskill: Just give me a second.

866
02:31:00.480 --> 02:31:01.420
Kartik - Outskill: Okay.

867
02:31:01.550 --> 02:31:09.229
Kartik - Outskill: So, that's how you can just, go to your coupon code, go to,
[Link].

868
02:31:09.690 --> 02:31:28.620
Kartik - Outskill: go for your, go for your tool, just select the tool, generate the coupon, copy the
coupon code, and then you can just go ahead with the, with the, with the tool, and you can use
the coupon code to redeem the tool. Now, for perplexity, how to do it?
869
02:31:28.630 --> 02:31:29.960
Kartik - Outskill: I'll just show you.

870
02:31:42.370 --> 02:31:46.449
Kartik - Outskill: So, for perplexity, you need to go to [Link].

871
02:31:46.630 --> 02:31:56.549
Kartik - Outskill: Once you go there, click on this upgrade icon. Upgrade icon. Once you click on
upgrade icon, continue with Pro.

872
02:31:57.350 --> 02:32:07.479
Kartik - Outskill: and then add your billing details. So, see, folks, there are a few tools which will
ask you to add your billing details.

873
02:32:07.600 --> 02:32:26.659
Kartik - Outskill: Okay? So, you just need to put in your, credit card or debit card details over
there, and once you'll put in your, coupon code, that whole amount, the money that it is asking…
asking for, the app is asking for, that will be nullified and it will be zero. Okay, so you need not to
worry about that.

874
02:32:26.670 --> 02:32:38.759
Kartik - Outskill: But, for a few tools, like Perplexity, you will have to fill in your payment details.
So, just make sure of that. Okay, once you fill in the payment detail, over here.

875
02:32:40.220 --> 02:32:57.950
Kartik - Outskill: there is an option called Enter Discount Code. You can enter your discount
code, and once that is done, you will ju- you just need to click on Apply and Subscribe. This total
amount will be… will become zero as soon as you put in the

876
02:32:58.110 --> 02:32:59.940
Kartik - Outskill: You put in your coupon code.

877
02:33:00.810 --> 02:33:05.060
Kartik - Outskill: I hope this makes sense, and this is clear in terms of perplexity.
878
02:33:06.760 --> 02:33:14.010
Uthappa Outskill: So folks, I know there's a few questions that have been asked in the chat,
relevant questions. I'll quickly sum it up once this section is done.

879
02:33:16.210 --> 02:33:35.039
Kartik - Outskill: Okay. Then, the next one is, also we have added, like, what these tools are,
what these tools are useful for. So you can obviously, just go through this document, and you
will get a good understanding, how to use it, and what these tools can be used for.

880
02:33:35.100 --> 02:33:37.130
Kartik - Outskill: Second is Emily.

881
02:33:37.980 --> 02:33:39.080
Kartik - Outskill: So…

882
02:33:39.310 --> 02:33:54.360
Kartik - Outskill: For Emily, again, you need to go to [Link], you need to generate
your coupon code, and once you generate it, copy it, then go to [Link].

883
02:33:55.830 --> 02:33:58.830
Kartik - Outskill: Okay, just go to [Link].

884
02:33:59.170 --> 02:34:11.670
Kartik - Outskill: Emily works as a Chrome extension, so you can just add it, add it to the
Chrome. Once you add it, there will be few questions which will be asked. You can answer
those questions.

885
02:34:11.770 --> 02:34:15.419
Kartik - Outskill: And once done, this page will open.

886
02:34:17.390 --> 02:34:23.840
Kartik - Outskill: And once you are here, you can either click on Upgrade to Pro, or you can click
on subscribe.
887
02:34:24.020 --> 02:34:28.709
Kartik - Outskill: And there you will get an option to redeem a code.

888
02:34:29.260 --> 02:34:36.450
Kartik - Outskill: You'll put in your discount code, you will, you'll click on Get Emily for Free, and
it's done.

889
02:34:36.570 --> 02:34:39.269
Kartik - Outskill: you will have your MLE Pro XS with you.

890
02:34:40.510 --> 02:34:42.409
Kartik - Outskill: I hope this is clear.

891
02:34:43.180 --> 02:34:44.279
Kartik - Outskill: Make sense?

892
02:34:49.000 --> 02:35:02.280
Kartik - Outskill: Okay. So, the two tools that we will be, releasing this week are Perplexity and
Emily, and as I show… as I have shown, this is how you can just redeem them. I will be sharing
this document with all of you.

893
02:35:02.300 --> 02:35:20.879
Kartik - Outskill: On the LMS, so just go through it, and take, take it into help, so that you all are
clear on how to redeem your tools. As Utapa mentioned, Perplexity, we will be releasing first,
and then Emily will take 2-3 more days, so just keep a note of it.

894
02:35:21.390 --> 02:35:24.049
Kartik - Outskill: Utapa, that's it from my side. I think you can take…

895
02:35:24.050 --> 02:35:40.060
Uthappa Outskill: So, folks, as Karthik mentioned, Emily is not yet given access to. Wait for 2-3
days, we'll let you know in the community itself once it is given access, and then you can start
redeeming it. Right now, it might throw an error to you, so it is an expected thing.

896
02:35:40.110 --> 02:35:57.709
Uthappa Outskill: I'll quickly… I'll not spend much time on the tools questions here, because this
is something that we can still take it upon our groups, so we'll quickly move on to the Q&A part.
But to summarize a couple of questions that has been asked, subscription for both the tools is
going to be for one year. You should use for Perplexity Pro, not Max.

897
02:35:57.710 --> 02:36:02.440
Uthappa Outskill: If you already have a Pro subscription, it will not get extended on top of it.

898
02:36:02.440 --> 02:36:09.349
Uthappa Outskill: We have given you access on the registered email ID that we have with.
That's where we have given access to.

899
02:36:09.350 --> 02:36:18.480
Uthappa Outskill: Any questions in accessing this, or let's say, apart from SOP, if you have
certain questions that you have here on which email you to use and all those things, please,
there is…

900
02:36:18.480 --> 02:36:32.559
Uthappa Outskill: The document that Karthik shared will be shared in the, the same section
where I showed you pre-read materials. Just below that, we will add a section for the tools
access, where we'll add this document. Go over it after the session is done, or tomorrow.

901
02:36:32.730 --> 02:36:47.020
Uthappa Outskill: And then there is a clear documentation. Links for each of the tool is also
provided. We will also be adding or sharing the Google form, where you have to fill for any
issues that you have in terms of redeeming it, and we will get to it, alright?

902
02:36:48.650 --> 02:37:12.140
Uthappa Outskill: Where do we use these tools? That's fairly written in that document. You can
use MLE for multiple purposes. You can use it to summarize your YouTube videos, upload a
data and get the analysis of it, or even transcribe the audios. And at the same time, there's a
prompt library also built in. There's a bunch of things available under MLE which you can
access.

903
02:37:17.700 --> 02:37:26.439
Uthappa Outskill: where can I find the coupons? [Link]. You can find it on the
document that Karthip will be sharing. The URL will be available there.

904
02:37:27.950 --> 02:37:41.589
Uthappa Outskill: Yeah, alright. So, folks, now, let's jump into the Q&A part, and then if there is a
few more questions towards the end, I can still see if we can take it up. So, we'll quickly get to
the Q&A. Let me present. Dilip, you're here.

905
02:37:42.760 --> 02:37:44.129
Dileep Karri: Yeah, I'm here, I'm here.

906
02:37:55.020 --> 02:37:57.579
Uthappa Outskill: Okay, let me present this slide.

907
02:38:02.190 --> 02:38:07.100
Uthappa Outskill: So, I'm gonna stop taking new questions here right now until we finish this.

908
02:38:20.730 --> 02:38:21.490
Uthappa Outskill: Okay.

909
02:38:22.370 --> 02:38:28.729
Uthappa Outskill: Anything operational, I'll probably take it up, and if Deep has anything to add
to it, please do.

910
02:38:29.900 --> 02:38:31.149
Uthappa Outskill: Will we get…

911
02:38:32.060 --> 02:38:37.569
Uthappa Outskill: perspective of, okay, if this is something that I've already answered, we will
give you the deck.

912
02:38:37.720 --> 02:38:39.810
Uthappa Outskill: As well as the post-read material.

913
02:38:40.850 --> 02:38:44.049
Uthappa Outskill: POCs is something that we have already kind of addressed.

914
02:38:46.350 --> 02:38:47.790
Dileep Karri: Yeah, I'll dig this up.

915
02:38:48.400 --> 02:38:58.540
Dileep Karri: Aside from my own domain, does the AI generalist job currently exist, or should we
apply problem-solving thinking with AI in our daily lives or job roles?

916
02:38:58.540 --> 02:39:11.190
Dileep Karri: Very important questions. I think you should start applying it right away, because
the AI generalist role that we are talking about is futuristic in nature, so some people have
already started, doing and,

917
02:39:11.200 --> 02:39:34.760
Dileep Karri: hiring for these roles. They might not be calling it an AI generalist, but for example,
there is one role called AI Operations Expert, which I know that some companies are hiring for,
which are a person reporting to the founder, and the job of this person is to look at all the
activities in the company and see how they can be accelerated with or without AI.

918
02:39:34.760 --> 02:39:38.350
Dileep Karri: So that is this AI ops person that people are hiring for.

919
02:39:38.350 --> 02:39:40.920
Dileep Karri: So, again.

920
02:39:41.370 --> 02:39:52.150
Dileep Karri: I think we'll need to wait a little bit more time to see that maturity coming, and as I
said, we are preparing for the future, and therefore, that's how this is going to be there. Okay?

921
02:39:53.430 --> 02:40:07.079
Dileep Karri: There can be more than 4 to 5 ways to solve a problem. Can we leverage AI to
explore alternatives? The same goes for AI tools, how to smartly, explore alternatives. See,
for…
922
02:40:07.100 --> 02:40:15.979
Uthappa Outskill: If there are 4 to 5 ways to do that, you can nudge AI and say, okay, tell me
more, tell me more, tell me more. See, the way to solve a problem.

923
02:40:15.980 --> 02:40:16.780
Dileep Karri: is…

924
02:40:17.000 --> 02:40:33.189
Dileep Karri: to solve it, okay? I would mention that 95% of the problems, the, yes, you might
have to solve it more efficiently, and if you have to solve it efficiently, you have to put it in the
problem framing itself, saying that I want this problem to be efficiently solved.

925
02:40:33.770 --> 02:40:38.910
Dileep Karri: So that if you put a constraint, your answer will get even more tighter. So that's…

926
02:40:39.120 --> 02:40:47.179
Dileep Karri: what we can do. You can explore a lot of solution trees, as I would call it, but one
thing I want to mention is don't

927
02:40:47.760 --> 02:41:09.389
Dileep Karri: get stuck in exploring solutions over and over and over again, and not really go to
the actual solution and, you know, kind of, you know, implement that, okay? So, you cannot…
don't get into stuck at this AI analysis phase, where you're just solving questions over, you know,
exploring various solutions, but you're not exactly solving the problem.

928
02:41:09.390 --> 02:41:12.329
Dileep Karri: How to smartly explore alternatives.

929
02:41:12.330 --> 02:41:18.919
Dileep Karri: So, the best way is, there is this, site called… there is an AI tool for that.

930
02:41:19.060 --> 02:41:29.799
Dileep Karri: Okay? Or, yeah, something like that. There's an AI tool for that. You get very good
examples of other tools, which you can look at it over there. Test it out.
931
02:41:29.800 --> 02:41:40.259
Dileep Karri: I would say, if you're starting in your journey, don't try to test multiple tools. First,
get comfortable with few tools, and then keep expanding and go for multiple tools. That's what…

932
02:41:40.260 --> 02:41:55.359
Dileep Karri: You should be able to do that. How to integrate [Link] with other
applications? Actually, no integration is required. Wherever your cursor is, it will be, co… the text
you dictated will be copied there, so there is

933
02:41:55.360 --> 02:42:01.109
Dileep Karri: No direct, thing you need to do. It's a native thing. You can do it directly.

934
02:42:01.700 --> 02:42:13.130
Dileep Karri: provide realistic examples of when it is advised not to use AI, what the actual
solution would be in instances where AI is not required to create value.

935
02:42:13.990 --> 02:42:20.770
Dileep Karri: I'll tell you one great example of where you should not use AI. You should not use
AI in nuclear reactors.

936
02:42:20.980 --> 02:42:27.580
Dileep Karri: Why? Because the effect of any error in a nuclear reactor is going to be
catastrophic.

937
02:42:27.830 --> 02:42:37.070
Dileep Karri: So then, if your nuclear reactor meltdown is going to be catastrophic, and even
though you know AI is smarter than you to manage AI, you will not do that.

938
02:42:37.430 --> 02:42:52.440
Dileep Karri: Okay? So you won't put AI in places where the effect of something's going wrong
could be catastrophic, that is number one. Second thing is, defense, right? If you are doing
defense, you will have local models, etc. You won't…

939
02:42:52.440 --> 02:43:04.170
Dileep Karri: go for ChatGPT and say, you know what, I want to go with war with that country,
how will I do this, do that, so that, again, you should advise not to use AI over there.

940
02:43:04.610 --> 02:43:08.100
Dileep Karri: You can use AI, but it has to be a very secure environment.

941
02:43:08.220 --> 02:43:17.289
Dileep Karri: So, very, very important to think through in that manner. Other cases is, if you feel
a human has to be in the loop.

942
02:43:17.640 --> 02:43:23.199
Dileep Karri: And a human decision is very critical, you won't use AI there.

943
02:43:23.620 --> 02:43:36.790
Dileep Karri: For some cases, you will not be able to use AI also. For example, if the application
doesn't have an API, again, what is API, etc, you'll learn a little ahead, but there also, you will
not be able to use AI.

944
02:43:37.300 --> 02:43:42.279
Dileep Karri: How do we assess criticality in the matrix? What lens should I use?

945
02:43:42.440 --> 02:43:53.319
Dileep Karri: Give it a little bit time, you will understand the wisdom as you kind of move ahead
and interact with multiple mentors, but criticality means if a mistake happens.

946
02:43:53.450 --> 02:43:56.419
Dileep Karri: will you survive? How will you survive?

947
02:43:56.610 --> 02:44:01.370
Dileep Karri: Okay? For example, if I showed the ad to the wrong person.

948
02:44:01.530 --> 02:44:04.139
Dileep Karri: That is a mistake I can survive.

949
02:44:04.500 --> 02:44:22.460
Dileep Karri: But if I am using, again, if I am kind of doing a financial transaction on AI, and I
debited money from the wrong account, or debit, like, instead of $1,000, I added an extra zero
due to AI error. $10,000 was debited.

950
02:44:22.460 --> 02:44:26.179
Dileep Karri: Now, that is highly critical and catastrophic, so that is the lens

951
02:44:26.180 --> 02:44:33.440
Dileep Karri: you should use, if something is a mistake, how devastating the effect would be.
That is how you should address that.

952
02:44:35.350 --> 02:44:43.929
Dileep Karri: Okay? How is Whisper Flow AI better, etc, and all that. See, whenever you ask me
a question, which tool is better, why, etc.

953
02:44:44.030 --> 02:44:48.519
Dileep Karri: I am not here to defend a particular tool, okay? I am using that tool.

954
02:44:48.700 --> 02:44:54.260
Dileep Karri: Okay? So what that means is, if you want, you can use it. If you don't want, feel
free not to use it.

955
02:44:54.530 --> 02:45:01.379
Dileep Karri: I don't want this because later, I don't want you all to say that, hey, you
recommended this tool, this tool did not work for me.

956
02:45:01.700 --> 02:45:11.860
Dileep Karri: Right? We created, we did this partnership so that we feel these tools will be
helpful. Now, if you say, strongly say that, okay, I don't want to use the tool, it is completely up to
you.

957
02:45:12.390 --> 02:45:16.479
Dileep Karri: Okay? So, I would say, if you're having,

958
02:45:16.480 --> 02:45:32.860
Dileep Karri: this… how… how are they different? If you feel inbuilt voice typing feature will help,
please go ahead and do that. Absolutely not required. And if you want to find the answer to this
question, you will have to take that trial, see how it is working for you, and if it is really giving you
a boost.

959
02:45:32.860 --> 02:45:35.769
Dileep Karri: Go ahead and go for the paid subscription. If not, just drop it.

960
02:45:37.420 --> 02:45:44.859
Dileep Karri: which features of Claude made you prefer it? Again, that is a very personal choice.

961
02:45:44.940 --> 02:45:47.830
Dileep Karri: Okay, I will only say this.

962
02:45:47.830 --> 02:46:05.820
Dileep Karri: I felt that the answers it gave, or the writing it did, is close to my voice. So when I
gave it my directions, it tried to give me exactly close to the voice that I prefer, and that is how I
prefer it. That's the reason I like Claude, nothing big other than that.

963
02:46:08.720 --> 02:46:16.859
Dileep Karri: Okay, many IT organizations don't allow the use of tools, etc, so how can we
leverage this? So, again.

964
02:46:17.320 --> 02:46:20.080
Dileep Karri: coming back to my slides, I said.

965
02:46:20.230 --> 02:46:25.140
Dileep Karri: Don't, get stuck on tools, work with frameworks.

966
02:46:25.140 --> 02:46:46.320
Dileep Karri: Okay? Now, I gave you various frameworks. Prompting frameworks will be taught
from tomorrow onwards. There are multiple things you'll be taught, where the tool does not
matter. If the organization doesn't allow, it is very difficult to do that, so you will have to ask them
to allow it, or say that take the version of Azure or Amazon Bedrock, which is more secure.
967
02:46:46.440 --> 02:46:51.720
Dileep Karri: That is how you can do it. If your organization does not elevate, there is nothing we
can do

968
02:46:51.740 --> 02:47:06.519
Dileep Karri: To circumvent that. The only thing we can ask is, if your organization doesn't allow
it, at least you learn it, so that in a time in future, when the organization allows this, you are at a
level wherein you can contribute there.

969
02:47:07.440 --> 02:47:08.190
Dileep Karri: Okay?

970
02:47:08.750 --> 02:47:28.039
Dileep Karri: Is Spark an outskill IP? Okay, whatever we teach, there is no IP at all, okay? There
is absolutely no intellectual trademark in this thing. Of course, the content is owned by us, that is
there, but it is not something like that. It is just like,

971
02:47:29.610 --> 02:47:33.209
Dileep Karri: You know… How would I say it?

972
02:47:35.670 --> 02:47:51.160
Dileep Karri: It is just an acronym, I thought, of a way to organize information so that you
remember it. Like, it is just like that. Okay? I could have easily told Spark as some, you know,
CRASP also. I could have said K-R-A-S-P, it wouldn't change anything.

973
02:47:51.210 --> 02:47:58.799
Dileep Karri: I just tried… so basically how I did it was, I did some research, I came to know
these principles, I fed it.

974
02:47:58.800 --> 02:48:12.310
Dileep Karri: to an AI, and said, can… how do I make it palatable to first-time learners? And it
told me that, you know what? You can think of… it didn't tell me this, by the way. It gave me 10
different things. Then I used my intelligence to kind of

975
02:48:12.310 --> 02:48:19.620
Dileep Karri: take it forward from there. I used AI as a great sparring partner, and that's how I
kind of came up with the SPAR framework.

976
02:48:19.880 --> 02:48:36.899
Dileep Karri: didn't really get the R of the Spark framework. I think I have showed multiple
examples of R exactly, and I also told you that if you're not able to get it, it's okay. As I said, if
you're not able to find it to anything relatable after even 3 examples, what you can do is, for
now, park it.

977
02:48:36.900 --> 02:48:46.319
Dileep Karri: Maybe as you go along the route, you will kind of figure it, because SPAK, if you
understand, that itself is a great framework for you to understand whether you use AI or not.

978
02:48:48.360 --> 02:48:56.980
Dileep Karri: How do we download WhisperFlow? Can we attach it to Perplexity as well, etc.
Now, here is something I ask you for such questions.

979
02:48:57.180 --> 02:49:06.819
Dileep Karri: Right? How to download Whisper Flow is something that you can even search on
the internet and get it. Okay? So this is a fact-based question.

980
02:49:07.120 --> 02:49:18.180
Dileep Karri: Right? I can put this exact question in perplexity, it will give you a much more
detailed and much more exhaustive and a comprehensive answer than what I would give.

981
02:49:18.500 --> 02:49:29.419
Dileep Karri: I'm not trying to tell you that, you know, don't, you know, this is not a great
question, etc. No, this is a completely judgment-free zone. You can ask every question that you
want.

982
02:49:29.420 --> 02:49:40.049
Dileep Karri: I wanted to give you constructive feedback, because as AIGs, you will also need to
have the art of figuring out how to use AI. Start using AI tools.

983
02:49:40.080 --> 02:49:47.619
Dileep Karri: Okay? Even right now, once the advent of AI has come, what I have done along is,
I stopped remembering facts.

984
02:49:48.070 --> 02:49:59.350
Dileep Karri: Because facts are something I can search, I can Google, I can perplexiate, or
whatever. It is frameworks and thinking where I spend most of my time on, because that is
something that I retain.

985
02:49:59.740 --> 02:50:13.189
Dileep Karri: So, for this question, I would suggest you to kind of check it out, and you can try to
do this. And the other question I'll answer, like, you know, the question was.

986
02:50:13.190 --> 02:50:20.370
Dileep Karri: Perplexity makes a lot of mistakes, unlike ChatGPT. Now, that's a question that I
will answer. See.

987
02:50:20.370 --> 02:50:36.210
Dileep Karri: every model, when they release an update, some things work better and some
things slightly break. So, what I would suggest is, if you feel Perplexity is making mistakes,
always have a fallback LLM, where you are comparing the answers and deciding, okay, hey.

988
02:50:36.690 --> 02:50:42.210
Dileep Karri: this answer looks legit. If I put the same query in one or two LLMs.

989
02:50:42.760 --> 02:50:50.059
Dileep Karri: two, three LLMs, and they give me the same answer, I know that there is a very
small chance that they are making mistakes. But…

990
02:50:50.410 --> 02:50:58.309
Dileep Karri: If it is diverging, I know that there will be some mistakes which are happening, and
that would be, that, okay?

991
02:50:58.420 --> 02:51:03.589
Dileep Karri: So, on this case, if you find a LLM which works well for you.

992
02:51:03.590 --> 02:51:21.500
Dileep Karri: try to use that ChatGPT with citations, etc. Perplexity, in between, it gave some
errors, but it has improved, and one hack I can tell you is always prompt it back and say,
double-check your sources and give me the right answer. If it is doing something wrong, it tries
to recover from that.

993
02:51:23.810 --> 02:51:34.569
Dileep Karri: In the Zoom desktop app, the Notetaker app cannot activate due to permission
issues during the ongoing meeting. Is there an SOP, available,

994
02:51:34.720 --> 02:51:38.889
Dileep Karri: do for this. So again, that is something Zoom does, okay?

995
02:51:38.950 --> 02:51:49.520
Dileep Karri: If you have too many note-takers kind of coming in, Zoom automatically blocks
saying that it's a malicious software, etc. I think Zoom has an AI companion below.

996
02:51:49.520 --> 02:51:59.349
Dileep Karri: Okay, so when you open your below bar, there is a button called AI Companion,
where you can open, and you can chat with it, and you can get this thing, etc. Plus, we provide
you post

997
02:51:59.690 --> 02:52:04.570
Dileep Karri: session notes also. So you don't have to worry about this, aspect at all.

998
02:52:05.010 --> 02:52:17.260
Dileep Karri: Will we get a copy of this would help us understand who we might want to work
with. See, most of the questions are anonymous, so we… you will not even know who has
asked those questions.

999
02:52:17.630 --> 02:52:37.520
Dileep Karri: Other thing is, when we are doing a session directly, that time you will have the
transcript, some of the questions which are already there, but sometimes we might do it in a
different manner, like a breakout room, etc, when we find the topic to be large, and we feel we
need to do a breakout room, etc. In those cases.

1000
02:52:37.520 --> 02:52:42.089
Dileep Karri: I think it will not be feasible to kind of do that.

1001
02:52:42.340 --> 02:52:47.599
Dileep Karri: Are there other such frameworks? If so, what is the mechanism to decide which
framework to use?

1002
02:52:48.130 --> 02:52:54.980
Dileep Karri: Number one, first thing is, I would ask you to first adopt these frameworks and take
them step by step.

1003
02:52:55.140 --> 02:52:58.530
Dileep Karri: Okay? Other frameworks, etc, what I will do is…

1004
02:52:58.800 --> 02:53:10.399
Dileep Karri: you start adopting this, start learning as we go along, because I can tell you 20
frameworks for doing the same thing. What I'll end up happening is you'll get confused.

1005
02:53:10.660 --> 02:53:29.949
Dileep Karri: Okay? So, I would say, at least, beginning, trust and start with one framework.
Then you can go experiment by yourself and even evolve your own framework. You can say,
you know what, Spark is not this thing. If you want, you can, go ahead and

1006
02:53:30.330 --> 02:53:35.769
Dileep Karri: you know, kind of… I built Sparks, which is an even better framework than Spark.
Feel free to share it across.

1007
02:53:37.160 --> 02:53:48.769
Dileep Karri: Is Spark an industry framework or something that you have come up? I think I've
already told you. So, there is a lot of industry wisdom which I need to kind of crystallize with the
help of Spark, so that's what I have done.

1008
02:53:49.170 --> 02:53:58.120
Dileep Karri: Do I need to get all yes for Spark for using AI? Not really. Even if 4 out of 5 are yes,
you can use… go ahead and use it.
1009
02:53:59.950 --> 02:54:03.270
Dileep Karri: How to decide which tool to use for which problem.

1010
02:54:03.560 --> 02:54:07.139
Dileep Karri: It will come with experience.

1011
02:54:07.140 --> 02:54:25.909
Dileep Karri: Okay? It is not something that can come to you with day one. It'll come with
experience, it'll come with experimentation. So, two levels. First level, follow the tools and the
frameworks the mentors teach you, start adopting them. That is step one. Once you have done
some level of adoption, then experiment with other peripheral tools.

1012
02:54:25.910 --> 02:54:29.500
Dileep Karri: That might be doing the same thing, and take it forward from there.

1013
02:54:31.670 --> 02:54:32.400
Dileep Karri: Okay?

1014
02:54:32.740 --> 02:54:43.050
Dileep Karri: Can YouTube video from… can Emily identify and extract distinct ultra-sonography
scan, including diagnosis and pathologies explained by the narrative alternate AI tools?

1015
02:54:43.080 --> 02:54:58.199
Dileep Karri: very specific tool, try, there is an AI for that, and see if it can help, but for a very…
it's a very niche use case, which I don't think, is possible. One thing I can tell you is, if you
upload this particular

1016
02:54:59.080 --> 02:55:06.329
Dileep Karri: photo, and if ChatGPT is giving you a good diagnosis, that means any AI tool can
do that.

1017
02:55:06.610 --> 02:55:11.009
Dileep Karri: Because at the end of the day, this underneath would be the foundational LLMs
only.
1018
02:55:12.510 --> 02:55:34.459
Dileep Karri: difference between thinking model of others, you are trying to solve delivery stream
with 3C. What are reasoning models, etc. I think tomorrow's session, you will understand a little
bit more better. Today was just, like, a overview and a simple thing to get you started. So
tomorrow you'll understand what this thing, because further, you need to know a prompting
mechanism which will make it more clear.

1019
02:55:35.890 --> 02:55:44.739
Dileep Karri: Okay? Why Whisper flow? How is Whisper flow better than dictation? Please
experiment and figure it out for yourself. It makes life easy.

1020
02:55:44.980 --> 02:55:47.460
Dileep Karri: Again, we are not here to…

1021
02:55:47.620 --> 02:56:02.729
Dileep Karri: you know, propagate one tool over another and say that this tool is best, this tool is
not good, etc, and all that. What we will do is, we will show what these tools can do. Up to you
to pick it up as such. And actually, Whisperflow was not in any of the silvers.

1022
02:56:02.980 --> 02:56:12.070
Dileep Karri: as my, this thing, I have used Whisper Flow. You want to use it, you are your best
judge, try to use it or not use it, completely up to you.

1023
02:56:13.730 --> 02:56:25.830
Dileep Karri: how to apply Spark and 3C for some brainstorming, or create something new. I
think what I showed you was exactly that. How do you do that, etc. But see, if you're trying to
create something new.

1024
02:56:26.050 --> 02:56:36.399
Dileep Karri: it is very hard to get AI to really help. It can help in terms of thinking frameworks,
etc, and all that, and that's where you should be able to do that.

1025
02:56:36.740 --> 02:56:46.789
Dileep Karri: Okay? So, I would say starting, first, adopt a few things. I know there's anxiety,
how do I do this for that, how do I do this for this, etc.
1026
02:56:47.190 --> 02:57:00.399
Dileep Karri: pace yourself a little slowly, otherwise you'll get confused. For brainstorming or to
create something new, it is… original thought is required, and that is exactly what the Spark
framework told. Like, if you want to, if you, like.

1027
02:57:00.400 --> 02:57:08.079
Dileep Karri: If you want to create a light bulb, you cannot use AI, because that doesn't exist,
according to… if you're looking at Edison example.

1028
02:57:09.600 --> 02:57:14.229
Dileep Karri: How to log into [Link]? I think it's the same email that you have.

1029
02:57:15.540 --> 02:57:23.419
Uthappa Outskill: So, use your registered email ID that you have access for the dashboard for.
So the same email ID if you log in, it'll be, available for you.

1030
02:57:24.880 --> 02:57:32.610
Uthappa Outskill: And this question, I just kept it so that to tell you I did not get what this,
means, so I'm just marking it as done.

1031
02:57:33.000 --> 02:57:43.919
Dileep Karri: Yes. Every domain has a strategic element and a tactical or operational element. It
would be useful to see AI application to solve both types of problems.

1032
02:57:44.190 --> 02:57:45.110
Dileep Karri: Sin.

1033
02:57:46.260 --> 02:57:47.550
Uthappa Outskill: This…

1034
02:57:47.550 --> 02:58:02.080
Dileep Karri: Tactical and operational element solving is something you'll see when we go to the
automations. Right now, the tactical or operational element we cannot bring to life, because that
is what automation kind of solves it for us. So you'll have to be a little…
1035
02:58:02.970 --> 02:58:08.760
Dileep Karri: patient till we get to that element. So, we, again, we are in day session one.

1036
02:58:08.850 --> 02:58:26.300
Dileep Karri: Okay? So you'll have to trust the process. We have 6 months to go. A lot of this will
get clearer and clearer. But I really like this question. This is the right question that an AIG would
definitely have. So, whoever asked is kudos for this question. These are the kinds of questions
we would like to see. So what,

1037
02:58:26.370 --> 02:58:43.470
Dileep Karri: We will see is… we will see some use cases where we will try to solve some
aspect of it, but please understand that the operational tactical element would come only when
you come to the automations and the build phase, because that is where your,

1038
02:58:43.470 --> 02:58:51.570
Dileep Karri: applications hit the ground, and you actually solve problems. Here, you lay the
groundwork for solving the problem, and that's what you'll see over here.

1039
02:58:52.700 --> 02:59:03.199
Dileep Karri: Why my chat is disabled? So, I kept disabling and enabling the chat overall. I think
you joined a little bit late. This is done so that people don't lose focus when a topic is going on.

1040
02:59:05.140 --> 02:59:06.180
Dileep Karri: Okay?

1041
02:59:06.350 --> 02:59:17.679
Dileep Karri: tools access everything, I think Uthapai had already mentioned. Zoom AI
Companion does not work, etc, and all that. It works most of the times. You can just check it out.
It may be that

1042
02:59:17.680 --> 02:59:28.380
Dileep Karri: your Zoom hasn't been upgraded or something. Again, Zoom AI Companion not
working is not something we can control. If we have enabled it from our end, and you should be

1043
02:59:28.380 --> 02:59:29.900
Dileep Karri: Able to do that.

1044
02:59:30.220 --> 02:59:45.939
Dileep Karri: Spark framework doesn't pattern, repeatable and knowledge cover the same
aspect? So, the way it is framed is slightly different. I think I explained it then also. Pattern
recognition is more in terms of, you know.

1045
02:59:47.160 --> 02:59:49.700
Dileep Karri: Exactly, like, what…

1046
02:59:50.000 --> 02:59:56.009
Dileep Karri: how is it, like, I'll tell you this, okay? Let me give it, like, instead of saying, what is
PRK?

1047
02:59:56.020 --> 03:00:11.519
Dileep Karri: Right? I will tell you what Spark asks you the questions to do. I think that'll be a
better thing than… because what I'm realizing right now is when I'm explaining each letter,
people are getting confused, like, this letter, how to do that letter, how to do, etc. Spark
framework, the key thing we are doing is.

1048
03:00:11.760 --> 03:00:30.459
Dileep Karri: Do I define the proper inputs and the outputs, number one? Do I have this
problem… is this problem something like a absolutely new problem that has never appeared
before, or is it a problem that has appeared before and people have already solved it?

1049
03:00:30.460 --> 03:00:40.350
Dileep Karri: Right? For example, how to, you know, how to get to Jupiter in lesser fuel is a
problem, nobody has solved it.

1050
03:00:40.470 --> 03:00:59.609
Dileep Karri: Plus, how to launch a rocket in space is what people have already solved, so that
is the difference. Repeatable is, is it a routine, repeatable activity that it is being done, or is it an
activity that requires some kind of a human or emotional element? That is repeatable.

1051
03:00:59.700 --> 03:01:11.760
Dileep Karri: Knowledge is more like, is it a knowledge in the public domain that everybody
mostly should know, or is it well documented, or is the knowledge sparse?

1052
03:01:11.790 --> 03:01:25.869
Dileep Karri: Okay? Now, as I go along with it, I understand that people might get a little bit more
confused, etc, as we go along, but if you're able to get the directional understanding that this is
what Spark is for, you should be able to get it.

1053
03:01:28.050 --> 03:01:31.960
Dileep Karri: Do we have the Pro or Max? I think already Utapa has answered.

1054
03:01:32.830 --> 03:01:47.400
Dileep Karri: clarify which jurisdiction, the drug example, etc. So, yeah, in the clarity, that was an
important part of it. Good catch, I should have done that. Again, as I said, I did a cross
oversimplification just to show you how an AIG would.

1055
03:01:47.400 --> 03:01:57.490
Dileep Karri: approach a problem statement. I am not trying to solve that drug example right
away there and then. I was just trying to show you how I would apply this framework to that
example.

1056
03:01:59.400 --> 03:02:12.540
Dileep Karri: Can AI help me discover the price of a product I'm trying to launch, or parameters
of the product discovery process? If you have got identified competitors and you want to do
competitive price analysis, that you can do.

1057
03:02:12.540 --> 03:02:20.040
Dileep Karri: But pricing of a product discovery is something you have to figure out by
experimentation. AI will not magically discover a price that works.

1058
03:02:20.040 --> 03:02:23.000
Dileep Karri: Because pricing is something customers decide.

1059
03:02:23.030 --> 03:02:34.429
Dileep Karri: So, you have a price, you test it out, you have another price, you test it out, etc.
So, the framework of what to do, how to test it out, AI can help, but the actual pricing, you will
have to discover.

1060
03:02:36.140 --> 03:02:49.589
Dileep Karri: A job hunting strategy and interview preparation using AI. Again, day one, my
friend, we will get to some of it. As I said, we will definitely cover some general aspects of AI,

1061
03:02:49.590 --> 03:02:59.919
Dileep Karri: And what exact use cases, we will decide based on what it is. Your feedback will
be taken, but we are too early to kind of get started with such use cases right away.

1062
03:03:02.260 --> 03:03:12.889
Dileep Karri: Spark evaluation requires some prior knowledge. What if that knowledge is
missing? If you don't have any domain knowledge, a domain expert has to be involved, because
without prior knowledge.

1063
03:03:13.750 --> 03:03:23.590
Dileep Karri: Okay, without any kind of prior knowledge, if you try to use AI, it is going to be
disastrous. Okay? So, have a consultation with some person

1064
03:03:23.880 --> 03:03:32.209
Dileep Karri: Or definitely try to get an understanding of that and educate yourself by having an
expert on your side.

1065
03:03:33.110 --> 03:03:35.810
Dileep Karri: Otherwise, it's going to be very difficult to do this.

1066
03:03:37.790 --> 03:03:46.259
Dileep Karri: How can we cater the gray areas if any of the frameworks, pointers is maybe?
Again, majority if it is yes, just go ahead and do that.

1067
03:03:48.210 --> 03:03:51.349
Dileep Karri: Pace is very slow, so today, purposefully,

1068
03:03:51.470 --> 03:04:01.659
Dileep Karri: See, again, I understand your anxiety and your concerns, like, pace is slow. For
some of you, pace will be very slow, some of you it is just right, some of you it is fast.

1069
03:04:01.960 --> 03:04:11.770
Dileep Karri: Allow yourself the time to a little bit adjust, okay? We are just in session one, day
one of the cohort, etc. And this is the fifth cohort.

1070
03:04:11.770 --> 03:04:25.940
Dileep Karri: Okay? So we have learnings from previous cohorts which have implemented, and
that's what we try to do in today's session. Today's session was a crystallization of the questions
people had on the back of their mind when they reached the midpoint of the cohort, where they
said, we are confused.

1071
03:04:25.940 --> 03:04:31.899
Dileep Karri: So we took those feedback and tried to address it before the cohort starts, so that
you have a better journey.

1072
03:04:32.050 --> 03:04:36.149
Dileep Karri: Right? So, trust us, trust the process, and I'm going to tell you this.

1073
03:04:36.150 --> 03:04:56.340
Dileep Karri: It is a prediction which I know has come true for every cohort. You are going to
become very good friends with all of us. We are good people at heart, okay? So, we will become
very good friends. Yes, we will do mistakes. It's not that we are not going to do any kind of
mistakes, but we are open to understand those mistakes, learn from them, and do better.

1074
03:04:56.600 --> 03:04:57.440
Dileep Karri: Okay?

1075
03:04:58.590 --> 03:05:18.259
Dileep Karri: Right. Can I first enter a prompt randomly without the Spark framework, and later
one, as per results, I rethink my strategy and use Spark for better results? Feel free to
experiment, and AIG's core muscle is experimentation. You want to experiment with that?
Please do that, okay? And if you feel it works for you, adopt that process.

1076
03:05:23.060 --> 03:05:32.159
Dileep Karri: Okay, the photo and visually sharing questions can fit into some parts of Spark
Framework. So, okay, if you are,

1077
03:05:32.300 --> 03:05:50.989
Dileep Karri: asking some parts of it doesn't really fit into the SPARC framework. You can think
that, okay, partially it is fitting, you can use it to build it, okay? So, partial fit also is fine. If you're
just… it is basically trying to, okay, what I'm trying to say is, before you start applying AI to solve
a problem, think

1078
03:05:51.110 --> 03:05:59.269
Dileep Karri: And take a step back, and ask yourself, should you use AI to solve the problem?
This framework is just to help you with that.

1079
03:06:00.750 --> 03:06:14.900
Dileep Karri: Okay, so this is a directional thing. Now, if you're asking questions like how many
minimum parameters, how many maximum parameters, what is the threshold score, etc, very
difficult to answer that, because it is a very practical directional kind of stuff.

1080
03:06:15.010 --> 03:06:18.260
Dileep Karri: Okay? What I would say is.

1081
03:06:18.590 --> 03:06:26.130
Dileep Karri: A mandatory thing is clarity. Does the problem statement have clarity, where a
human in the loop

1082
03:06:26.460 --> 03:06:31.820
Dileep Karri: can really help? Or, you know, etc, can it be really helped overall?

1083
03:06:32.530 --> 03:06:35.560
Dileep Karri: Okay? So, that's what I would suggest.

1084
03:06:35.730 --> 03:06:53.870
Dileep Karri: It appears to be that it can be used across most of the problems if you're using AI
as a partner. Yes, absolutely, right? What I try to tell you is, at least there are some problems,
you know, AI cannot help, and that part, you have to drop it. For example, one more example I
can give you. If there is a country that has banned AI,

1085
03:06:53.940 --> 03:07:05.649
Dileep Karri: And, the compliance kind of says that you cannot use AI there. Then, this
framework will easily tell you that you cannot put input and output, you cannot analyze patterns,
all of it will come to know.

1086
03:07:06.430 --> 03:07:07.270
Dileep Karri: Okay?

1087
03:07:07.760 --> 03:07:15.839
Dileep Karri: Why can't AI discuss non-existent topics? Why can't it be used for things that lack
repetition? AI is a good… see.

1088
03:07:15.960 --> 03:07:32.900
Dileep Karri: For that, you need to understand that mastermind session, where we told you how
LLM learns things. That is the large language model. Large language model, underlying text is
all the text that it has to think… it has learned on. It analyzes the pattern, so that means

1089
03:07:33.000 --> 03:07:38.880
Dileep Karri: Topics that were already discussed, it does very good. If it is not discussed, it can
hallucinate.

1090
03:07:39.540 --> 03:07:42.600
Dileep Karri: Okay, that is the underlying answer of this question.

1091
03:07:45.140 --> 03:07:59.690
Dileep Karri: provide relatable examples of the high leverage ratio, like finance, etc, and all that.
I addressed it before, we will not take such types of questions, we'll take generic things. For
those things, you will have to apply for yourself.

1092
03:08:00.270 --> 03:08:03.099
Dileep Karri: Okay? And you will have to take it forward.

1093
03:08:03.300 --> 03:08:07.409
Dileep Karri: When we find a solution or developing something with AI, we might note

1094
03:08:07.410 --> 03:08:26.569
Dileep Karri: have complete structure of idea, but step by step will get clear, so how do we
approach AI? At least some partial clarity you will have. Use that partial clarity to move ahead,
and that whatever AI gives will shine more light that will help you to do it further. It's an iterative
process. It's not a single short thing.

1095
03:08:26.780 --> 03:08:45.349
Dileep Karri: When the output of AI comes, you will get ideas which you can then put it back,
then again, come back and do something, etc. And so it's an iterative process, and, like, how
you have mentioned that I might not have a complete, clear step-by-step, that is the same
iterative step-by-step procedure I have shown.

1096
03:08:45.410 --> 03:08:53.579
Dileep Karri: While I compressed a few steps in a single shot, but… and if you see the spirit of
what I had done, it was a very iterative first thing.

1097
03:08:54.700 --> 03:09:09.270
Dileep Karri: Explain leverage matrix with an example. What do you mean by critical and
impact? I think I have addressed that, okay? If you have… if you have some task where,
where… I'll tell you one easy understanding. Customer service.

1098
03:09:09.270 --> 03:09:18.220
Dileep Karri: critical, and if I'm having AI to do it, my manpower will reduce quite a bit, and
therefore, there is a huge impact on some metric.

1099
03:09:18.230 --> 03:09:25.210
Dileep Karri: So what is the impact on a metric, and what is the criticality that I'm looking at?
That's how we will do this.

1100
03:09:26.170 --> 03:09:37.600
Dileep Karri: Okay, when a user needs support in a fast-paced environment, these frameworks
could be viewed as a roadblock, any workaround in place to teach for such situations.

1101
03:09:37.720 --> 03:09:44.910
Dileep Karri: So, if you feel a framework is becoming, like, a complete roadblock for you.

1102
03:09:45.060 --> 03:09:52.659
Dileep Karri: At least have a heuristic which, you know, kind of summarizes the framework and
helps you take a quick decision.

1103
03:09:52.710 --> 03:10:05.920
Dileep Karri: Like, for example, Amazon's founder, right, he has a very good heuristic called
two-way door decisions, which is saying that if you can reverse a decision, take it.

1104
03:10:06.340 --> 03:10:15.740
Dileep Karri: Now, that is not a right decision framework when you think about what is a SWOT
analysis, etc, etc, etc. But it's a heuristic that can help you move faster.

1105
03:10:15.980 --> 03:10:28.090
Dileep Karri: Okay? So I would say, if you feel you can reverse a decision by using… when
you're using AI, using AI, you're solving a task which can be reversed later, then use AI. There is
no harm in that.

1106
03:10:28.280 --> 03:10:30.209
Dileep Karri: That is how I would, recommend.

1107
03:10:30.800 --> 03:10:35.030
Dileep Karri: Can you explain how to use Spark again for food delivery industry?

1108
03:10:35.150 --> 03:10:44.909
Dileep Karri: If you want a repetition, I had repeated, I told it twice, I think you can check the
recording, I think that would be the right way, because I had explained it by writing stuff also
over there.

1109
03:10:45.560 --> 03:11:10.550
Dileep Karri: I have to renegotiate contracts for your practice. I feel like the knowledge and
output can be a yes, as I can upload the old contracts with clear request. Then, I think that's a
great use case for AI, because you have an old contract, you have the new lines you want to
edit, and you can definitely kind of do that. So, but it'll be… yeah, I think it can be done. I feel
that it's a…

1110
03:11:11.070 --> 03:11:16.679
Dileep Karri: a good thing, because it's repeatable, the knowledge is already there, it's a great
use case for AI.

1111
03:11:17.910 --> 03:11:33.940
Dileep Karri: But I have no… no, okay, this is a second question, okay. But I have no knowledge
how to build… negotiate or build contracts, so it would be some abdication. Okay, so even
though you have no… you don't have any knowledge, but the knowledge is out there in the
world. So that means, definitely, you can use AI.

1112
03:11:34.340 --> 03:11:43.299
Dileep Karri: So, what, like, don't… so, what I mean by abdication versus amplification is, when
you get your

1113
03:11:43.530 --> 03:11:47.440
Dileep Karri: Draft of the contract, at least do one manual overview.

1114
03:11:47.520 --> 03:12:06.329
Dileep Karri: Okay? Just, you also read it once to just be on the safe side. Don't just say, AI,
okay, AI, build the contract, read it for me and tell me if there are any issues. Don't abdicate that
way. Have that amplification where you give it guidance, the negotiated contract comes, you
read through it, give it feedback, and get it done.

1115
03:12:07.230 --> 03:12:10.910
Dileep Karri: Okay? Integrate Whisper flow, I think I've already addressed.

1116
03:12:15.270 --> 03:12:23.319
Dileep Karri: any specific reason to use Perplexity over other AI, like ChatGPT, if you could
quickly share what to use when. Again.

1117
03:12:23.490 --> 03:12:43.360
Dileep Karri: Tools usage is something you will gain with experience. It's a very subjective thing.
For the first cohort, we had given some recommendations, but then the cohort members came
back saying, you know what, you told me this, but for my use case, I felt this was better. Should
I do it or not? So we had these multiple recurrences. From then on, what we see is.

1118
03:12:43.890 --> 03:12:56.770
Dileep Karri: Those tools are pretty good. Perplexity is more like cited resources. ChatGPT is
like a Swiss knife, wherein if you're not sure what to use, you can use ChatGPT freely.

1119
03:12:57.050 --> 03:13:06.660
Dileep Karri: Okay? So, again, exact use case, what to use, when, etc. It is something that will
come with experience and experimentation. It's a very subjective question.

1120
03:13:06.660 --> 03:13:24.299
Dileep Karri: Right? Even if I ask the best AI expert in the world, they will be like, okay, what do
you exactly do? What is your workflow? Okay, for your workflow, for your use case, these tools
should work, these tools will not work. It's a very personalized question. And for something
extremely personal.

1121
03:13:24.800 --> 03:13:28.460
Dileep Karri: are extremely subjective. Experimentation will only work.

1122
03:13:30.680 --> 03:13:44.460
Dileep Karri: Okay? You spent a significant amount of time teaching us how to apply it in one
domain when we work in several domains. If I show you 10 domains also, there will be 11th
domain, which I will not be able to show.

1123
03:13:44.460 --> 03:13:52.420
Dileep Karri: Again, I've mentioned it before, I am… here, we are not trying to solve by AI for X,
AI for Y, AI for multiple domains.

1124
03:13:52.420 --> 03:14:07.720
Dileep Karri: I showed you how you could apply to one domain, where, again, it was not a food
delivery domain, okay? I just showed this was an example, and I tried to show that. And going
forward also, we will try to show examples in that way only.

1125
03:14:07.720 --> 03:14:24.829
Dileep Karri: Where it is not about relatability, it is about understanding. Did you understand how
the framework was applied? If you did not understand even how the framework was applied,
then there's a problem. But if you understood how the framework was applied, now you can go
back and apply it to your domain very easily.

1126
03:14:24.830 --> 03:14:28.620
Dileep Karri: Okay? So for some folks who are like, what is my assignment, what should I do?

1127
03:14:28.670 --> 03:14:37.590
Dileep Karri: Go back, organize your notes, try to follow the same thing that I have done. Some
of you will fail, will fall flat on the face, and that is completely okay.

1128
03:14:37.650 --> 03:14:39.750
Dileep Karri: That is how learning happens.

1129
03:14:40.660 --> 03:14:50.850
Dileep Karri: Okay? Whatsapp group, so we'll run it completely on the Circle community itself,
so we won't do any, WhatsApp group, right now.

1130
03:14:52.350 --> 03:15:16.119
Dileep Karri: Can we have an overview on jargons like LLM, RAG, etc. We will come to that. We
will take some understanding, and we will kind of go through. As we go through multiple
sessions, most of the architecture will do that, because if I go with this approach, where I say,
okay, here is a glossary, let me explain the terms. I can explain 25 terms to 30 terms in

1131
03:15:16.210 --> 03:15:28.320
Dileep Karri: one to two hours, but then you'll be like, okay, I don't know how to apply, I don't
know what is this, you just told me the abbreviation and expansion of that. That's why we will
expand, and we will teach these concepts as we move along.

1132
03:15:30.630 --> 03:15:35.589
Dileep Karri: Okay? Example used with responses, this will be part of the post-reads that we will
share.

1133
03:15:37.140 --> 03:15:56.390
Dileep Karri: Little more brief on AI-dependent and a prompt engineer. See, again, as we move
along, we'll understand more about what prompt engineer does and AI-dependent and
everything, but what you need to understand is don't rely too much on AI that, okay, I'll give you
a very good metric. Suppose tomorrow chat GPT stops working.

1134
03:15:56.830 --> 03:16:04.099
Dileep Karri: Can you do your work without ChatGPT? Are you capable enough to think and
execute without that?

1135
03:16:04.120 --> 03:16:16.609
Dileep Karri: Or are you paralyzed, thinking chat GPT is not there? I need ChatGPT, I'm having
withdrawal symptoms, and I'm not able to solve the problems with that. That is the difference of
an AI-dependent person.

1136
03:16:16.610 --> 03:16:27.729
Dileep Karri: Okay? And AIG is more like, okay, ChatGPT is not there, I am still capable enough
to kind of do this. Maybe with AI, I would have taken more time, I will do it in lesser time.

1137
03:16:32.430 --> 03:16:56.419
Dileep Karri: Okay? Food industry problem, again, I addressed this food industry thing overall.
As I told how much… that's why I told examples like Uber Eats, etc, and all that, so that
everybody relates a little bit to what that example is. How much of our examples I'll take, this will
not work out. We have enough experience there. So, best part is to take a common example,
which everybody can… see.

1138
03:16:56.420 --> 03:17:03.660
Uthappa Outskill: Relatability is not what we are solving for. We are solving for AI application.
Relatability is something.

1139
03:17:03.660 --> 03:17:11.130
Dileep Karri: You will have to figure out by applying AI and see for yourself. Fundamentals of AI
we will teach. This is not AI for X.

1140
03:17:11.210 --> 03:17:18.360
Dileep Karri: It's not AI for marketing, AI for finance, etc. This is AI generalist, which is for
everything, so that's why
1141
03:17:18.360 --> 03:17:38.419
Dileep Karri: how much ever examples I'll show you, I will not be able to satisfy everybody in the
room. It is simply not possible, and that is not the intention also, because today you are in one
domain, tomorrow you will be asked to move to a different domain with the way things are
progressing, so you'll need to be able to be adaptable for all of that.

1142
03:17:38.670 --> 03:17:50.469
Dileep Karri: Okay? What is VIP coding? VIP coding is trying to build applications, without
writing code, and we will be… that is the build phase that will unlock overall.

1143
03:17:52.690 --> 03:17:54.410
Uthappa Outskill: I think I've answered this question already.

1144
03:17:55.550 --> 03:18:03.530
Dileep Karri: is it possible to individually go back to a specific slide in the presentation, etc? So,
one thing is.

1145
03:18:03.750 --> 03:18:23.020
Dileep Karri: We, the mentors will go slow, but cannot be super slow, wherein, for one person,
we cannot do that. So what'll happen is, you'll have post-session notes, you'll have the
recordings, etc, so you can actually go back, and how many times you want, you can go to the
slide and, you know, go back and understand the nuances until you finish that.

1146
03:18:25.980 --> 03:18:37.830
Dileep Karri: See, pre-read document, right? Let me do this. The pre-read document was just
meant to give you, like, okay, hey, you know what? These are some things you need to read
before the session. It is not

1147
03:18:37.830 --> 03:18:47.870
Dileep Karri: Pre-reads is not something that will be taught in the session. Pre-reads is like a
fundamental foundation, so that when you go through that, you will understand what is covered
in the session.

1148
03:18:48.180 --> 03:18:52.629
Dileep Karri: Okay? So, we will be doing the same thing, going ahead.
1149
03:18:53.800 --> 03:18:56.690
Uthappa Outskill: Just to clarify one thing here, Dilip, I think…

1150
03:18:56.800 --> 03:19:02.560
Uthappa Outskill: The pre-reads that we had given were program pre-reads, and also kind of,

1151
03:19:02.560 --> 03:19:21.520
Uthappa Outskill: your recap of your mastermind, so all those things that were discussed in your
mastermind. Plus, there are certain fundamental things that it's not necessary that we kind of,
give all of that information, because it's freely available information out there. These are just for
you to know… good to know stuff before you start off with the program. So.

1152
03:19:21.520 --> 03:19:26.899
Uthappa Outskill: The pre-rate for Session 1 was not available. For Session 2, we have already
added a pre-rate, which will be

1153
03:19:26.900 --> 03:19:35.390
Uthappa Outskill: Like as Dilip mentioned, not something that will be covered completely, but
something that you should know, good to know, so that you will be able to relate to things in
your session.

1154
03:19:38.150 --> 03:19:47.889
Dileep Karri: Okay? Before session, heads up can be given to come prepared with questions.
One thing is, the questions are something you get when you're listening to the lecture, so…

1155
03:19:48.390 --> 03:19:53.219
Dileep Karri: Every session, there'll be a Q&A, so that it at least is going to be there.

1156
03:19:54.300 --> 03:20:07.439
Dileep Karri: Right? Will we have a script of this in all forthcoming sessions? Post-read notes,
post-reads are going to be given, and will be there. Miss the onboarding session, the recording
is there on your dashboard. If you have difficulty, you can reach out to the team.

1157
03:20:08.870 --> 03:20:12.580
Dileep Karri: For a one-year subscription tool with the… I think this is already answered.
1158
03:20:15.050 --> 03:20:24.270
Dileep Karri: Will access to be for 6 months duration? I presume beyond that period, yes.
Beyond that period, it'll be a personal subscription. You'll have to renew by yourself.

1159
03:20:27.030 --> 03:20:40.640
Dileep Karri: So, you will get a WhatsApp DM for this cohort. Basically, you'll get some
WhatsApp DMs. If you're not getting those, just check in and see if you're getting the right
number, etc. Just get in touch with the team for that.

1160
03:20:42.030 --> 03:20:56.879
Dileep Karri: What if you have subscriptions to one or more tools on that list? Can you redeem
the coupon code with another email? I think it is tied to the same email that we have given. It's
not possible for that. See, again, these partnerships, we have

1161
03:20:56.890 --> 03:21:03.480
Dileep Karri: taken a lot of pain to do that. And this is, like, you know,

1162
03:21:04.040 --> 03:21:08.060
Dileep Karri: it is hard, like, we are doing the best we can over here, that's all I can say.

1163
03:21:09.200 --> 03:21:11.949
Dileep Karri: Where is the dock on the LMS?

1164
03:21:13.020 --> 03:21:14.729
Uthappa Outskill: I'll give out at the end of this.

1165
03:21:14.730 --> 03:21:15.700
Dileep Karri: Yeah, sure, sure.

1166
03:21:16.730 --> 03:21:25.699
Dileep Karri: Chrome extensions, unsafe. So, Chrome extensions built by people whom you
don't trust are definitely unsafe, but I hope you trust them.

1167
03:21:26.000 --> 03:21:31.629
Dileep Karri: Emily is a trusted partner of Outskill, so you can assume that it's going to be safe.

1168
03:21:32.810 --> 03:21:41.209
Dileep Karri: Are the tools to be used for the course execution, or are these merely
complementary and discretionary, not directly connected? It's a mix of both.

1169
03:21:41.520 --> 03:21:53.330
Dileep Karri: What we have done is, the partnerships we have done are tools which we feel can
be helpful to you as an AIG, and some tools are useful in the course content also.

1170
03:21:54.810 --> 03:21:56.780
Dileep Karri: Okay, it's a mix of both of that.

1171
03:21:59.390 --> 03:22:00.180
Dileep Karri: Right?

1172
03:22:00.180 --> 03:22:00.790
Uthappa Outskill: On board.

1173
03:22:00.790 --> 03:22:01.480
Dileep Karri: Absolutely.

1174
03:22:01.910 --> 03:22:05.590
Uthappa Outskill: Thanks, Eliph, thanks for taking up all those questions.

1175
03:22:05.950 --> 03:22:17.260
Uthappa Outskill: Now, for the folks, if, let's say, any of your questions, you could not take it up
here, I mean, there's also a good amount of time that was spent here in the session right now. If
there are any unanswered questions.

1176
03:22:17.490 --> 03:22:23.350
Uthappa Outskill: As we discussed in our onboarding call, three steps. First, you ask the
community.
1177
03:22:23.760 --> 03:22:32.820
Uthappa Outskill: Sorry, just before that, use AI, use, go back to the recording play reads,
post-its, everything, figure out yourself first, as a generalist.

1178
03:22:32.910 --> 03:22:49.709
Uthappa Outskill: not getting solved, ask the community. So, on Circle, you can put it on your
group chat, or we will be unlocking an Ask Mentor Wala channel, in the next couple of… either
today, sorry, tomorrow, or latest by tomorrow, end of the day, or Monday first off, we'll be
unlocking that group.

1179
03:22:49.800 --> 03:23:00.070
Uthappa Outskill: ask your questions in that channel. A mentor will get to it. We will try our best
to answer it within 24 hours. They're also working professionals, so expect some delays if

1180
03:23:00.250 --> 03:23:04.260
Uthappa Outskill: On some of the days, it doesn't get answered within 24 hours, but we'll try to
get it done.

1181
03:23:04.320 --> 03:23:29.140
Uthappa Outskill: Not getting solved in Circle as well, then come to office hours, ask your
question, get your doubts clarified. So, every question of yours, doubts of yours, we want to get
it addressed, right? Some of the things we will be able to take it up in these core sessions, some
of the things we'll have to take it up in our official channels that we have on Circle, or our office
hours. So, this is how… the same structure is how we will be applying it for the entire program.
So, good that we started on this

1182
03:23:29.140 --> 03:23:31.910
Uthappa Outskill: Node, where we kind of figured all these parts out.

1183
03:23:33.430 --> 03:23:36.119
Dileep Karri: Uzup, just one final thing, before I leave.

1184
03:23:36.300 --> 03:23:49.590
Dileep Karri: One thing, folks, I know some people have rated us 1 and 2, really respect that
candid nature of yours. Only request is, please tell us why you rated us low, so that we take
those corrective actions.
1185
03:23:49.740 --> 03:24:04.740
Dileep Karri: Okay? And, basis that, we should be able to, solve that. So, Utapa has already
shared his email, so please, email there overall, so that even I can take a look, and we can
make corrective actions going forward.

1186
03:24:06.500 --> 03:24:09.660
Uthappa Outskill: Yes, I appreciate… sorry. Let me reshape it again.

1187
03:24:13.580 --> 03:24:18.670
Dileep Karri: Is it? Some… I think somebody said some countries follow one as the first grade.

1188
03:24:23.060 --> 03:24:25.199
Uthappa Outskill: That you may be.

1189
03:24:25.200 --> 03:24:30.329
Dileep Karri: Okay, so, okay, let me put it this way. Some of you genuinely did not like the
session.

1190
03:24:30.420 --> 03:24:45.739
Dileep Karri: Okay? Just reach out to Utapa, so that, you know, for next time onwards, we'll
recalibrate it and tell it more clearly, if in case you felt that way. But if you genuinely did not like
the session, you felt it was… it had no value at all.

1191
03:24:45.740 --> 03:24:52.710
Dileep Karri: We respect your opinion, just reach out to us so that we can pick it up from there
and see what we can do to help.

1192
03:24:53.410 --> 03:25:06.620
Dileep Karri: not just for this session, for every session, we would, you know, expect such things
so that if you don't tell us what is wrong, we cannot correct it, okay? It's a community, it's a place
where you'll have to give us that info.

1193
03:25:07.820 --> 03:25:08.430
Uthappa Outskill: Awesome.
1194
03:25:08.940 --> 03:25:10.080
Uthappa Outskill: So…

1195
03:25:10.080 --> 03:25:15.049
Dileep Karri: before I leave, sorry Uptapa, last thing before I leave. I'll just share this and I'll
leave. Folks.

1196
03:25:15.750 --> 03:25:23.339
Dileep Karri: Arvind, do you have… better trust now that we have started this journey, okay? Do
you trust outskilled folks?

1197
03:25:24.120 --> 03:25:32.650
Dileep Karri: Yes, we have started. Yes, there are few gaps we'll cover. Do you trust? Awesome.
Love it, love it, everyone. Yeah, Uda power to you.

1198
03:25:33.180 --> 03:25:41.299
Uthappa Outskill: Thank you. Thank you so much, Dilip. I think that was a good start to the code
overall. I also want to take a pulse check on your day one right now.

1199
03:25:42.150 --> 03:25:50.439
Uthappa Outskill: If I have to kind of ask you to describe your day one experience in a, in a
word, right? I mean, whatever it might be.

1200
03:25:50.740 --> 03:25:56.599
Uthappa Outskill: what is that word that's in your mind right now? How would you describe your
day-one experience as?

1201
03:26:01.190 --> 03:26:05.260
Uthappa Outskill: Powerful Spark 3C. Perfect. Eye-opening, awesome.

1202
03:26:05.790 --> 03:26:06.770
Uthappa Outskill: Boom.

1203
03:26:10.880 --> 03:26:14.709
Uthappa Outskill: Thank you, thank you, Abhini. We'll do our best to make it better.

1204
03:26:18.030 --> 03:26:33.349
Uthappa Outskill: Got it. Got it. I mean, mostly I see that, you guys have a positive, start. I hope,
and we all, as partners, we will try to ensure that we continue, having a similar experience
throughout.

1205
03:26:34.090 --> 03:26:46.780
Uthappa Outskill: two things before I wrap up. One, anybody who has not, yet filled the
onboarding form, Karthik, can we share the onboarding form here? Now, to clarify what this is…

1206
03:26:46.890 --> 03:26:58.340
Uthappa Outskill: people who have filled this onboarding form, you're given us an email ID on
which you need to be contacted, or you need to be given access to, which we have done, and
that is the email ID which you will be able to access your tools,

1207
03:26:58.400 --> 03:27:05.040
Uthappa Outskill: the coupon platform as well. For those people who did not fill the form, and
until yesterday we were trying to follow.

1208
03:27:05.040 --> 03:27:19.749
Uthappa Outskill: whoever has not filled the form, or we're not able to get through, we have
given the access on your payment email. It's basically the email that you used to make the
payment for. We have given you access on that. You need any corrections in that, or if this is not
correct, Karthik, if you could share the onboarding form again.

1209
03:27:19.750 --> 03:27:31.259
Uthappa Outskill: If there's someone here who has not filled that form yet, please try to get that
filled, and we'll get it addressed. Second thing, on, the tools, platform, links, and other things.

1210
03:27:31.260 --> 03:27:45.939
Uthappa Outskill: allow us time, at least until tonight or in the morning, we will be adding a doc
in the same section where all the content, everything has been listed. All the information will be
listed down there. You'll also be having a Google Form to fill if you have any issues,

1211
03:27:45.940 --> 03:27:51.720
Uthappa Outskill: accessing it. We will share this announcement also in the group, as in when
we do it.

1212
03:27:51.730 --> 03:27:55.150
Uthappa Outskill: Or, reach out to Karthik. Simple, if you have issue, reach out to Karthik as
well.

1213
03:27:55.670 --> 03:28:10.170
Uthappa Outskill: Cool. I hope that answers most of the question. Again, in the interest of time
and the people available here, any further questions that you want to get answered, please use
our LMS, post your question, and we'll get to it.

1214
03:28:10.650 --> 03:28:33.930
Uthappa Outskill: And don't forget, if you like the session, also please drop a gratitude note to
the link, or the session in the channel as well, which will help further motivate our mentors. So,
we'll meet again tomorrow, 7.30pm IST, the same time again, where we will be starting to
explore the prompting techniques, the art of talking to LLMs. So, until then, see you, good night,
good morning, good afternoon, whatever is your time zone.

1215
03:28:33.930 --> 03:28:35.619
Uthappa Outskill: We'll see you tomorrow again.