JOHN JAY COLLEGE OF CRIMINAL JUSTICE

The City University of New York

524 W. 59th Street, New York, NY 10019

MODULE 1 READINGS
HOMEWORK PACKET
Instructor: Jen Page (jepage@[Link])
Module 1 Dates: Jan. 27, Jan. 29, Feb. 3, Feb. 5

Reading Length/ Homework Page # in

(Author + Title) Time Due Date Packet
R. Matthew Shockey, “The Thurs. Jan.
Fact-Value Distinction, 8 pages 29 before 2
Core Values, and Ethics” class
39 minutes (the
Hard Fork Podcast, “We
transcript is for in Thurs. Jan.
Asked Roblox’s C.E.O.
class; you can just 29 before 10
About Child Safety. It Got
watch the linked class
Tense”
video as homework)
R. Matthew Shockey, Tues. Feb. 3
5 pages 24
“Making Claims” before class
R. Matthew Shockey,
Tues. Feb. 3
“Evaluative and Non- 9 pages 29
just before
Evaluative Claims”
Anthropic, "Claude
Thurs. Feb.
Teaches Jen about
18 pages 5 before 38
Philosophical
class
Argumentation”

YOU’RE INVITED! Module 1 Office Hours/Homework Party

Tue. Jan. 27, Thurs. Jan. 29, Tues. Feb. 3
1:40 to 2:55 pm (Community Hour)
New Building, 8th Floor by the windows (just outside the Philosophy Dept)

1 Ethics and Info Tech

John Jay College
 R. Matthew Shockey, “The Fact-Value Distinction, Core
Values, and Ethics”

From:
Shockey, R. Matthew. 2020. “The Fact-Value Distinction, Core Values, and
Ethics.” In Critical Thinking and Applied Ethics.

Link:
Indiana University Pressbooks

How to cite in-text:

(Shockey 2020, Handout, Par. #)*

* “Par. #” means “Paragraph number”

How to cite in Works Cited:

Shockey, R. Matthew. 2020. “The Fact-Value Distinction, Core Values, and
Ethics.” PHI 216 Handout.

What are Values?

1. Values are what determine our views about what is good and bad, right and
wrong, and what should or shouldn’t be done. Our values thus determine how
we think the world ought to be (even if it often isn’t that way).

2. Values are, to put it another way, normative for us. And they are unavoidable.
We live our lives based on what we value; we judge other people and our
institutions based on our values; and we judge ourselves based on whether we
live up to our values. There’s no human life that isn’t enmeshed in values.

3. There are different kinds of value: ethical or moral values reflect our basic
views about how we ought to live and act (more about these below); aesthetic
values reflect our views about what is beautiful and ugly, artistically pleasing
or displeasing (for instance, the music you like to listen to, the movies you like
to watch, the fashion you prefer, etc.); other values shape what we think is
good and bad about everything from sports teams to food to personal hygiene.

2 Ethics and Info Tech

John Jay College
4. Some values may reflect our individual opinions. If I say pizza is good or it
would be awesome if the Cubs never make it to another World Series, those
express my opinions about things I value, but not in a particularly deep or life-
defining way. I can recognize that other people value those same things
differently than I do and not lose any sleep over it. If you dislike pizza or want
to root for the Cubs, so be it.

5. Other values are not particularly fundamental, but they are still socially agreed
upon, and not a matter of individual opinion. What counts as politeness or good
manners are like this. The fact that you should dress nicely for a job interview,
or chew with your mouth closed when you’re eating with others, or not stay
parked at a meter longer than the time allowed, don’t really reflect deep
ethical values (although they are about respecting others, so there is some
ethical element to them). They are social norms about what good behavior
looks like, but nobody really suffers if you violate them.

6. On the other hand, if I say that murder is wrong, or that it is good to help those
who are in need, I am expressing ethical values that are nearly universal. It’s
not just that we happen to agree that murder is bad behavior, or that charity is
a virtue. They reflect some sort of ethical truth that is rooted in something other
than mere agreement or opinion. It’s a deep philosophical question as to what
this sort of ethical truth really is and how we know it. You can take more
philosophy classes to explore that question. Here the point is just for you to
see that there are differences among the kinds of values we hold. This leads
to the following:

CRUCIAL POINT: Values are not (at least much of the time) a matter
of mere opinion. You have probably been taught a fact-opinion
distinction at some point, but that is not the distinction we are making
here.

7. Most of us also hold values that we believe ought to be universally shared,

even if in fact they aren’t. These define our ethical outlook — our view about
how people, including but not only us, ought to live and act.

8. And it’s for this reason, as Socrates recognized, that differences over values
often lead to conflict. If we didn’t think our values correct and someone else’s
incorrect, we would just always live and let live. But sometimes we can’t do
this. And so we argue and fight and sometimes end up in violent conflict with
one another as a result.

3 Ethics and Info Tech

John Jay College
9. Typically such conflicts arise when there is an issue about what we will call
core values. We’ll use this term to refer to those things which are taken to be
of fundamental value to a person or group of people, which typically permeate
and shape their entire lives, and which define who they are.

10. Some examples:

• One of the core values that Americans have taken to be central since the
country was founded is individual liberty. People have argued about what
exactly this is, how to achieve it, who should have it, etc. But even amidst
such ambiguity and disagreement, the idea that individual liberty is central
to America has remained for many a guiding principle. As such, it provides
a standard by which laws, policies, actions, and politicians are judged: do
they promote or inhibit individual liberty?
• In many societies, social stability has been taken to be among the highest
values. This has led in many cases to systems of law and social hierarchies
that are quite restrictive. In such societies, individual liberty may be
perceived as of comparatively little value.
• For many religious believers, love of and obedience to God defines what is
of the highest value. To live according to this value means endeavoring to
regulate one’s life according to the demands of one’s faith. This value is
supposed to trump other values, for instance, those associated with wealth
and power.
• For scientists, truth and knowledge are regarded as of utmost value, and
the practices and institutions of science are designed to promote the value
of truth and knowledge in the face of all the obstacles we face in achieving
them. Individual scientists are supposed to look for reasoned evidence for
any claims they make, and to be open to revising their views when new
evidence or theoretical advances require they do so.
• For many artists, beauty has been taken to be a core value. What beauty
is, how to create or represent it in a particular medium, etc., have defined
a great deal of artistic effort. (Modern art often does not take beauty to be
a core value. As I recently heard one artist put it, “there are already enough
pretty pictures in the world.” So, such artists must be driven by some core
value other than beauty.)

11. The list could go on: fairness, power, kindness, charity, wealth, honor, familial
connection, ecological stability, etc. — lots of things figure as core values for
people. Sometimes they (we!) have multiple core values, and sometimes even
multiple incompatible ones.

4 Ethics and Info Tech

John Jay College
12. Whichever ones we have, they will shape our lives in various ways. So, for
instance, if you think that a core value is kindness, and that throughout your
life, you should strive to be kind to others, then there are going to be lots of
more particular values you’ll also hold. You’ll think it’s good to compliment
people, to smile and be friendly, to give them help when they need it, etc. If
you think that beauty is of the highest value, you may be willing to earn less
money by pursuing art than you could in a different career. Or, if you think
wealth or power is what’s most important, you may act immorally in pursuit of
them and believe you are justified in doing so.

13. How do you identify core values? Maybe the easiest way is to pretend you
are a four-year old who’s at the stage where they keep asking “why?” to
everything. If someone says, “you should recycle your plastic water bottle” or
“I’m buying a car,” or whatever, you can start asking “why?” or “so what?” and
keep asking it until there’s no further answer that can be given. For example:

• You should recycle your plastic water bottle.

• Why?
• Because it’s better to recycle than just to throw things away.
• Why?
• Because it takes energy and resources to produce things, and we shouldn’t
use more than we need.
• Why?
• Because resources are limited and we need to make sure they don’t run
out.
• Why?
• Because human life depends on these resources to survive.
• So what?
• Well, we need to make sure humans can continue to survive.
• Why?

14. At this point, there’s not much more to say. Being more conservation-minded
is grounded in the value of sustaining human life, and there’s not really an
argument for that value that most who hold it would be able to give. Not that
you couldn’t keep pressing the issue (Socrates surely would), but for most
who hold human life to be of value, it’s just self-evident. They don’t really have
reasons for holding it. For those that do, it’s that core value that drives their
concern about recycling and similar things. Of course, one discarded water
bottle isn’t going to end all life on the planet. But the totality of our wasteful
actions has an immensely negative impact on our continued ability to survive,
or at least to thrive. So, someone who has as a core value our ability to live

5 Ethics and Info Tech

John Jay College
 well will look at each action we take through the lens of that value, even
seemingly insignificant actions.

15. Note that this conversation could have taken a different direction and ended
in a different core value:

• You should recycle your plastic water bottle.

• Why?
• Because a lot of plastic we throw away ends up as litter or even washed
into the ocean.
• So what?
• As plastic breaks down, it can harm other species.
• So what?
• Well, we shouldn’t cause harm to other species if we can avoid it.
• Why?

16. Here we’ve arrived at the idea that the lives of other species are of value, not
just our own. If you hold that value, you will be concerned about a lot of the
same things that those who only value human life are concerned with. But, as
with valuing human life, for most who are really driven by the value of the lives
of other species, that value is basic, it’s not something they will necessarily be
able to give further reasons for.

17. Here’s the other example from above:

• I’m buying a car.

• Why?
• I can’t get to work unless I can drive.
• So what?
• I need to be able to get to work.
• Why?
• Because I need the money the job provides.
• Why?
• Because without it I can’t afford college.
• So what?
• If I don’t go to college, I won’t be able to have a career in the health
sciences.
• So what?
• It’s important for me to be able to work in a field that helps others.
• Why?
• Because helping others is one of the most important things we can do.
• Why?

6 Ethics and Info Tech

John Jay College
18. Again, at this point answers run out. For most people who think helping others
is important, it’s a basic value that steers them, but one for which they
probably can’t give reasons for.

19. If you think these examples stop too soon, try to push them a little further. See
where reasons run out — or where they would realistically run out for most
people whose values we are trying to understand. If you’re clever enough, you
can probably keep them going past that point, but the main idea here is that,
for most of us most of the time, we have values that guide us that are basic,
that we don’t or can’t justify by appeal to anything else.

What are Facts?

20. Our values reflect how we think things ought to be, even when they are not
actually like that. When we talk about facts, however, we are talking about
how the world is, regardless of what we think about it or hope it to be or believe
it ought to be. Facts are those things that are true of the world independently
of what we happen to think or wish were true.

21. There are facts about the past, facts about the present, and facts about the
future (though this gets a little tricky, as we will see); there are facts about
particular things, about classes or categories of things, and facts about all that
is; there are facts about the natural world and human society; and there are
facts about fictional beings, numbers and other mathematical objects, and
(perhaps) supernatural beings, none of which are part of the empirical world
in any straightforward way.

22. Many facts may be unknowable, many can only be known with some degree
of probability rather than certainty, but, still, they remain facts. Science aims
to know facts, but so do we in lots of areas of our lives where science doesn’t
reach. Faith makes claims about fact, but typically thinks those claims cannot
be established by ordinary empirical or scientific means.

23. Sometimes it’s hard to focus on the facts because of our values, or because
of our pre-existing beliefs. We are all subject to psychological biases that
shape how we process information we are presented with, what sorts of
questions we think are legitimate, etc.

24. Perhaps the most significant bias is what’s called confirmation bias. This is
a tendency we all have to be more willing to accept information if it confirms
what we already think is true. It leads us to seek out such information rather

7 Ethics and Info Tech

John Jay College
 than evidence that might require us to change our minds. For example, think
about how people tend to seek out news from sources that reflect their political
perspective and to refuse to take seriously any news from competing sources.
That’s a common example, but even the most rigorous scientists are not free
from confirmation bias.

25. You can see the danger, though: if what you already believe is false, then
confirmation bias will prevent you from learning what’s true.

26. Still, knowing the limits we have and the biases that shape how we think allows
us to work to correct for them, and to seek knowledge of facts that is objective,
that doesn’t reflect merely our distorted individual perspective on the world.

27. That we ought to seek objective knowledge is itself an ethical claim, one that
depends on the value of such knowledge. So, even as we seek knowledge,
we cannot escape the demands our values place on us.

What is Ethics?
28. I referred above to an “ethical outlook” as one that reflects a view about how
people, including but not only us, ought to live and act. Our ethical outlooks
are tied to what sorts of lives we think are worth living, and what it takes to live
such lives. They are usually rooted in some identifiable core value or values,
but they also include more particular values as well that are connected to
them.

29. “Ethics” refers to this domain of values and what we believe about it — our
general and specific beliefs about how we ought to live and act and treat one
another. We all have an “ethic” we live by, even if we don’t spend much time
thinking about it or trying to articulate or defend it. So, we can speak
interchangeably about our ethics and our ethical outlook.

30. Now, notice that in this chapter we’ve been talking about ethics without
making specific claims about what we ought to do or how we ought to live.
Instead, we’ve making a number of distinctions and defining a bunch of terms
that are involved in describing ethics (and other evaluative activity). This is
what philosophical ethics does: it involves taking a step back from our lives
in order to reflect on what ethical evaluation involves, in order to understand
it better. The goal is to help us avoid unethical action that results from
confusion (e.g., confusing values with facts). It may or may not lead to claims
about what we concretely ought to do or how we should live…

8 Ethics and Info Tech

John Jay College
31. Wrapping things up for this chapter, then: one of the concrete goals in this
course is to start thinking about how to identify our core values and see just
how much they shape our lives and affect how we engage in our day-to-day
reasoning. We can hardly begin to understand ourselves or each other without
recognizing the different core values that structure our lives.

32. Recognition isn’t enough, however, for we can’t be good learners and critical
thinkers if we aren’t aware of our own core values and how they shape how
we engage with the world and the people around us. That awareness is central
to the ability to reason about our values and to listen to others doing the same
thing, i.e., to engage in ethical reasoning. To do this is to take values seriously,
to regard them as not just mere opinion.

33. Being able to reason about our values won’t mean we can resolve every
conflict, of course, but it may let us ratchet down the emotions that lead to
unproductive or violent exchanges, and it may help us to change our own
minds, even if not others’.

34. To reason about values, as with reasoning about other matters (which we will
also talk about), we need to be able to identify the claims we make that involve
or express what we take to be true about what is of value. These are what
we’re going to call evaluative (or sometimes normative) claims. The next
module will look at them, along with the contrast group, non-evaluative claims,
which, as its name suggests, covers everything that doesn’t involve a
judgment about what’s of value.

9 Ethics and Info Tech

John Jay College
 Hard Fork Podcast, “We Asked Roblox’s C.E.O. About
Child Safety. It Got Tense”

From:
Roose, Kevin, and Casey Newton. 2025. “We Asked Roblox’s C.E.O. About
Child Safety. It Got Tense.” Hard Fork Podcast, The New York Times, November
21. [Link]
[Link].

How to cite in-text:

(Roose and Newton 2025, Handout, page #)

How to cite in Works Cited:

Roose, Kevin, and Casey Newton. 2025. “We Asked Roblox’s C.E.O. About
Child Safety. It Got Tense.” PHI 216 Handout.

Instructor’s Note: If you’re an auditory learner,

please do not read the transcript! Instead, watch the
YouTube video of the interview using the QR Code
provided or do a YouTube search for “roblox ceo hard
fork.” However, you’ll still need the transcript in class.
If you’re a visual learner, you’re welcome to read the
transcript instead of watching the video. Hard Fork interview

Also, if you’re unfamiliar with Roblox, it may be helpful

to watch the trailer for the gaming site; use the other
QR Code.

Roblox Trailer

This summer, the gaming platform Roblox hit a new milestone: More than 25
million users, at the same time, played Steal a Brainrot, one of the hit games on
the site.

Though the company has been around for nearly 20 years, its popularity has
exploded among children since the Covid-19 pandemic. Today Roblox has more
than 150 million daily active users and functions as one of the primary online
gathering places for preteens. (Unlike Instagram, TikTok and Discord which

10 Ethics and Info Tech

John Jay College
require users to be at least 13 to make an account, Roblox allows users as young
as 5 to join.)

But precisely because Roblox is so popular with children, it has also attracted the
attention of adult predators. Three states — Louisiana, Kentucky and Texas —
have filed lawsuits against Roblox over child safety issues. And more than 20
lawsuits accusing Roblox of enabling sexual exploitation have been filed in federal
court this year.
Roblox, for its part, continues to ship new safety features, including a new facial
age estimation technology that the company announced this week. But the
question remains: Are these safeguards enough?

To talk about the company’s approach to safety, we spoke with Roblox’s co-
founder and chief executive, David Baszucki. Here’s our conversation, lightly
edited for clarity and length.

Casey Newton: Dave Baszucki, welcome to “Hard Fork.”

David Baszucki: It is great to be here. Thank you.

Newton: You have joined us today to talk about this new age-gating policy
that Roblox is rolling out to protect kids. And I think we should start by just
talking about the scope of the problem here. What has led you to this point?
And how do you think of the problem of predators on Roblox?

Baszucki: We think of it not necessarily just as a problem, but an opportunity as

well. How do we allow young people to build, communicate and hang out together?
How do we build the future of communication at the same time? So we, you know,
we’ve been, I think in a good way, working on this ever since we started. And when
we were — this was almost 18 or 19 years ago — when we first launched the
company and we had just four of us sitting in a room, we were literally the
moderators, like we would rotate all the time. And so fast-forward to where we are
today, it’s just like every week, what is the latest tech? At the scale we’re at, 150
million daily actives, 11 billion hours a month, like what is the best way to keep
pushing this forward? And as you correctly note, we’ve just started adding that
we’re going to be using facial age estimation with A.I. to complement that.

Newton: Right, so tell us a little bit about how this works. All users of Roblox
now, if they want to communicate with others, they’re going to have to kind
of submit to a face scan?

11 Ethics and Info Tech

John Jay College
Baszucki: That’s correct. And I think long-term face scan isn’t the only signal. Like
mathematically, the more signals that are convoluted, the better the accuracy and
the better we can get. So we already have signals: We have behavioral signals,
we have photo ID upload signals, we have user suggested age. But we’re
complementing that with a facial age estimation signal that literally, you know, in
an ephemeral way, use your phone, use your PC, scan an image, do a good A.I.
estimation — the image isn’t stored — and start using that to complement all the
other things we do. And we already do a pile of stuff: We have very, very good text
filters; there’s no image sharing; we’re doing a bunch of monitoring. But adding
one more signal to that can really help us make cool decisions.

Kevin Roose: I want to learn more about this face scanning technology, but
first, I’m just struggling to understand exactly how many of your users are
which ages. My impression from the outside, as a nondaily active user of
Roblox, is that a lot of kids who are 9, 10, 11, 12 are on the app. What is sort
of the most popular age band of active users on Roblox?

Baszucki: This is really interesting. And I think this goes to what we’ve been
saying for a while in that maybe for the last five to 10 years, people look at Roblox
and say, well, These are kind of these primitive games and everyone on the
platform is pretty young. We’ve seen that migrate over the last five years from
predominantly under 13 to — now I believe we recently said — self-reported almost
two-thirds over 13. And we can see those numbers growing and growing, like
fastest growing segments on Roblox are 21 and up right now.

Roose: Or at least people who say they’re 21 and up.

Baszucki: Yeah. Now I think those numbers have a fair amount of current efficacy.
We’re obviously balancing those with the age estimation because we want a tighter
estimation, but just looking at the size of the cohorts and the growth rates, there’s
a fair amount of accuracy to that.

Newton: You know, when the E.U. rolled out some age assurance technology
earlier this year, we saw a lot of kids that just took a photo of a video game
character and used that to pass the age checks. How confident are you that
your systems can avoid that kind of tinkering?

Baszucki: I think we have a lot of opportunity here in that we already have other
signals that are maybe not as reliable, but when combined are very reliable. We
also have the opportunity — this isn’t just a one-time thing. Roblox experiences
are — you’re holding a phone. In many cases we can use signals from your phone
or the camera to animate your face, for example. So, this isn’t kind of a one-stop-

12 Ethics and Info Tech

John Jay College
shop thing. This is possibly checking as we go. If we see weird signals, do another
age check as we go. So I’m pretty optimistic that kind of the more naïve, like “take
a picture of that” — we’re gonna go way beyond that. And it’s gonna get harder
and harder to fingerprint in that way.

Newton: So after you roll out these changes, you’re going to sort users into
bands and, depending on if you’re like 13, you’re only going to be able to talk
to people within a few years of your age. For most of Roblox’s history
though, that has not been the case. Any anonymous adult could talk to any
anonymous minor. Why was that the case?

Baszucki: I think we’ve always from Day 1 really focused very heavily on quality
of text filtering. And if anything, we’ve had very, very historically restrictive text
filtering over time. One of the biggest complaints on Roblox — and arguably trading
in a good way safety for growth — has been complaints of communication on the
platform. Like, there’s a lot of Roblox memes and themes. If you go back and dig
into them, everyone is just saying “hashtag, hashtag, hashtag” and there’s been a
lot of …

Newton: Because that’s how you blur out words that aren’t supposed to be
used.

Baszucki: That’s correct, so there’s a lot of historical memes about historical

“hashtag, hashtag, hashtag.” With A.I., these filters have gotten better and better.
The state of where we are now is primarily getting into the blocking of adversarial
attempts to broach the filter, and adversarial attempts to try to share PII [personally
identifying information], which we’re getting better and better at as well. On the age
estimation, it’s just — I would say — recently come of age to the point where we
would complement it. But I would say our strategy is always complement with
whatever technology is out there to put that together.

Newton: Yeah, I mean, I don’t know. When I read these lawsuits and these
investigations into the company, it does not seem like predators are having
that hard of a time getting around your filters. Like, you know, you read the
allegations and it’s, you know, people kind of change a few keywords, a few
letters here. They’re able to get what they need. So I’m curious what’s made
you so confident that things are working?

Baszucki: I don’t want to comment on it. We do see the chat logs of those. And
we can see interesting and, many times, ways of people trying to — I’d say, many
times people who are fairly sophisticated, and I’m not going to say all of them,
many times kids who are over 13, who actually in any other app are not running in

13 Ethics and Info Tech

John Jay College
a filtered situation, unfortunately, figuring out how to jump to some other platform
where they can communicate unfiltered, where they can share images, all of it. It’s
one of the primary things we’re doing is trying to keep people on our platform. It’s
an interesting industry situation. I would say, we’re not waiting for the rest of the
industry. We’re like, we’re always trying to stay ahead of it. On the California age-
appropriate design code, for example, we’re like, we are already doing all of that
stuff. And the same thing with age estimation: We’re not doing this because of any
laws that are coming, we think it’s the right thing to do.

Newton: Yeah, I mean, you mentioned sort of kids leaving the platform, and
it does become clear in those lawsuits that Roblox is kind of, you know,
where predators go to find kids, but then they do try to —

Baszucki: (interjecting) Well, I would say in the lawsuit or the press, yes. I would
categorically reject the actual description of what you’re saying. But yes, what you
might read in the news would portray it that way, but I —

Newton: (interjecting) You don’t think you have a problem with predators on
the platform.

Baszucki: I think we’re doing an incredible job at innovating relative to the number
of people on our platform and the hours, in really leaning into the future of how this
is going to work.

Newton: So, let me ask the question about the other platforms, though,
because I mean, I agree with you that this is not a story about one company.
Like, what we see is predators will come to a place like Roblox, then they try
to get a child somewhere else. Like, Discord is a platform that kind of comes
up a lot. It seems to me like you could probably stop a lot of abuse if you
prevented those platform shifts from happening. So like, what can you do in
that regard? Like, you know, in my mind, if you see somebody chatting like,
“Hey, what’s your Discord username?” You know, and it’s an adult talking
to a minor — that seems like it’s a red flag. Maybe you’re doing some of that
kind of detection already —

Baszucki: (interjecting) I would actually say that is a very simple red flag. Like,
that sounds like text filter many prior generations. So I would say the techniques
are much more sophisticated than that. We’re constantly getting better than that
and sometimes it’s adversarial. Like, we’re getting into, you know, if we
cryptographically were going to try to have an exchange of how to share your Snap
handle or whatever handle. We see some of that starting, like things we’ll have to
essentially prevent in the future.

14 Ethics and Info Tech

John Jay College
Newton: I don’t understand. Like, what do you mean like a cryptographic
exchange to share a Snap handle?

Baszucki: Imagine you and I were trying to share the name of an airport
somewhere in the world and we were getting blocked by all of the filters. We might
actually start getting really creative, like use the third letter of every fifth word to
spell out the airport. And those are the kind of things we get into right now.

Roose: I have sort of a basic question, and forgive me if this betrays my

naïveté about Roblox and how the platform works, but like, why have a sort
of open-ended text communication feature at all on a game that is very
popular with young people? Like, I have a son, he’s too young to play Roblox
or any other game, but I can imagine saying as a parent of a slightly older
child, like, “I just don’t want my kid playing games that allow them to chat
with other people in this way.” There are lots of games out there that don’t
have that kind of feature. Why is this so core to the way that Roblox works?

Baszucki: Well, I would highlight a couple things. First, there’s enormous benefits
for having people be able to communicate, and we think we see those benefits.
There’s a lot of kids who find their peeps on Roblox. There are a lot of kids when
they’re lonely or isolated actually — I’m not going to use it as an excuse or anything
— but I get a lot of calls from famous parents who said, “Look, my kid would be
dead if it were not for finding a community on a platform like Roblox to stay
connected with.” And I would also say we do take it as an enormous responsibility
to design for not the sophisticated parent who goes in and says, “Look, I’m going
to turn off who you can chat to and all of that.” We have to design for all parents
and think through that. But we do believe we have enormously great technology
up and down the stack. We think it’s part of building an experience like this. You
know, if you go out there in the gaming space, I think it is very common to support
communication. I’m not going to list any other games. And typically, we are way
beyond in, I’d say, our innovation level, responsibility level, on both the filtering, on
the image sharing, on all of these things.

Roose: Is it fair to say that you think that any platform of Roblox’s size, with
the number of young users Roblox has, would be experiencing similar safety
challenges?

Baszucki: I wouldn’t categorize any of that. I would say we’re at a scale that is
somewhat mind-boggling. So I’m not going to run any, you know, excuse numbers
of Roblox relative to any other platforms. But I would say a scale of 11 billion hours
is very different than a scale of 10 million hours. It’s a thousand times the size of

15 Ethics and Info Tech

John Jay College
it. And I would say, the scale we’re at is quite enormous: It’s 150 million people
every day; it’s many more than that every month; and relative to an individual
game, the scale’s enormous. We have experiences on our platform right now
which have 25 million people playing at the same time. And, traditionally in the
gaming market, the record prior to that, I think, was 11 or 12 for all of Fortnite. And
prior to that, big numbers are considered a million for a single game. Like, I
remember when Minecraft hit a million users. So, think about that: a whole platform
at some point in time, like Minecraft hitting 1 million users. A single game on Roblox
hitting 25 million users? So this ——

Newton: Right. But you guys want that scale, right? You court it, you send
push notifications, you give people daily login rewards. Like, you want the
millions and the millions to show up.

Baszucki: I think we actually see there being an incredible opportunity. If that

scale is going through platforms with good systems, it’s actually a big thing. And
the scale already exists. Like, the gaming space, in a way, is $190 billion. So, we
now have about 3 or 4 percent of that — I guess 3 percent — coming through
Roblox. So I would say we like the scale: It creates an opportunity for individual
game creators who might be making their game by themselves, without part of
systems like that, to make it as part of a more overall system.

Newton: Here’s the dynamic that I think that people find frustrating. We have
many large tech platforms right now. They seek this enormous scale. They
reach previously unseen heights. And then they notice they have a problem,
and they go in and they try to solve it to some degree. And then, they keep
growing further, and then there’s another problem. And I think, to some
degree, that dynamic is probably inevitable. But I think what a lot of people
want, what a lot of parents want, is some sense that people have a plan for
what is gonna happen when the 25 million people show up, right? And I think
they’re looking to people like you to see around these corners and to not
wait ’til 20 years in existence to say: Maybe we should sort these people into
age bands before they start communicating with each other.

Baszucki: Yeah, I would push back on that a bit. And I can hear what you’re
saying. But I would actually say, for people inside of our company, the way we
think about it and the way we do look ahead of it — we’re constantly looking ahead.
And you know, it goes back to the very first thing we did probably a month into the
company. So a month into the company, fun little story, there’s about 200 people
on Roblox. And what should we do? Should we keep growing it or should we go
build a safety system? And so 200 people in, we said: Let’s stop everything, let’s
go build a safety system, and let’s let the four of us go and become moderators.

16 Ethics and Info Tech

John Jay College
So, behind the scenes, I do agree with you, enormous responsibility, enormously
important to pre-empt this. It’s why a lot of the architecture and decisions we’ve
been making have been made, but we’re gonna keep doing them, right? And you
could imagine Roblox at 20-X this scale, having 50 percent of the gaming market
space all in a single place. Like, we’ll keep innovating on top of that.

Roose: Last year, activist fund and short seller Hindenburg Research came
out with a report arguing, among other things, that Roblox was,
“compromising child safety in order to report growth to investors.” One of
the things they pointed to was that there had been a decline in your spending
on trust and safety. You have denied that. Why? What did they have wrong?

Baszucki: Fun. Let’s keep going down this. And so, first off, Hindenburg is no
longer in existence, correct? So, you should report on that. They went out of
business for some reason. The other thing is, have you researched that
information?

Roose: Um ——

Baszucki: Because you’re asking me about this, I’m curious if you’ve done your
own research.

Roose: I have, I have done some research about ——

Baszucki: And what’s your take on that?

Roose: Well, I’m asking you.

Baszucki: So it’s really interesting, because I think we’re diving into a situation
where we’re getting better, better, better. But would you ask the same situation of
someone who converted from maybe hypermanual labor making cars by hand to
an assembly line? The assembly line might be infinitely better and at the same
time not have so many people. So, I think what they’re referring to is moving more
and more and more and more to A.I. Arguably, is it better to have A.I. systems
reviewing every image, every piece of text, all of that, or to keep doing that
manually? So I think that was a misnomer. And I think if you interview anyone in
our safety team or safety systems, that would be just part of making our systems
better.

Roose: So, just so I have the clear response, the argument is that the shift
has been from a more manual safety team and system to a more automated

17 Ethics and Info Tech

John Jay College
review process that may do a better job even if it costs less to operate
because ——

Baszucki: Yeah, I would actually lean in harder. The shift would be constantly
moving to higher quality, more accurate and more diligent systems. Can we review
more? Can we do more better? And say you and I, we’re a system, what is the
best way to review every single image going into Roblox at higher and higher
quality? And say we figured out the best way to do that is to do it with better and
better A.I. models. Would you pick that?

Roose: Yeah.

Baszucki: Good, so you’re aligning with what we did. High-five.

Roose: But I’ve also been covering social networks ——

Baszucki: Thank you for supporting our Roblox decision matrix.

Roose: But I also remember 15 years ago when ——

Baszucki: I’m so glad you guys are aligned with the way we run Roblox. High-five.

Roose: Right, I also remember 15 years ago when Mark Zuckerberg was
saying ——

Baszucki: Is this a stealth interview where actually you love everything we’re
doing and you’re here to stealthily support it?

Newton: We were told you wanted to come in and talk about your new age-
gating technology, so you know, we thought we’d get into it.

Roose: Let me just finish this up.

Newton: Yup.

Roose: I totally buy what you’re saying about using the best technology on
the market to keep people safe.

Baszucki: And I want to highlight, I came here because I love your podcast and
came to talk about everything. So if our PR people said, “Let’s talk about age-
gating for an hour,” I’m up for it, but I love your pod. I thought I came here to talk
about everything.

18 Ethics and Info Tech

John Jay College
Roose: Oh, yeah, we’re gonna talk about Steal a Brainrot too.

Newton: Yeah, that’s a fun one.

Roose: That’s part of my plan here too, but just to finish the thought: You
know, Casey and I both covered social networks back in the 2010s, when it
was very popular for people like Mark Zuckerberg to say, “Oh, we were gonna
use A.I. to solve all of our content moderation. It’s gonna take down
everything that’s scammy or false and we won’t need these people
anymore.” And investors loved that. And then they tried to do that and they
found that actually, oh, the systems weren’t that reliable yet and maybe they
were taking down too much content or maybe they were missing stuff that
was bad.

Baszucki: Well, look at all these social networking companies. Like, go to 2016,
for example, which is a reliance on automation and whatever. We’ve seen a lot of
things where not being on top of what’s going on in your company and these
algorithms leads to a lot of weird dystopianess as well. So I think we saw in 2016
those algorithms got a little, kind of into an optimization loop on weird things as
well. So I think, anyways, yeah, I love your questions. I’m glad you guys are
supporting how we’re running Roblox on our A.I. technology.

Roose: But you’re confident that the technology has gotten good enough
that you can use it to do moderation at a platform of Roblox’s scale.

Baszucki: The scale we’re at is absolutely mind-boggling right now. Like

absolutely mind-boggling. And every system: image review, text review, all of that,
constantly getting better and better.

Newton: I mean, look, Dave, I can tell you’re getting a little bit frustrated here.

Baszucki: I’m actually not frustrated. I just, I’m so excited to talk — I thought we
were going to be talking about fun, funny things in the industry and all of that. So
I’m not frustrated, I’m just kind of figuring out how much fun time I have with you
guys and how much fun time we have versus kind of this superfocus. But I’m happy
to talk about whatever you guys want.

Newton: Well, let me just, maybe just try one last thing here. Because this is
what I really want to understand. It is very difficult to do content moderation
at scale. We have learned that from every single company. And I think that
platforms provide various benefits to society. I’m on a lot of them every day,

19 Ethics and Info Tech

John Jay College
OK? So I feel like we’re kind of on the same page there. And I think being a
C.E.O. of one of these companies can sort of feel like being the head of the
Department of Homeland Security, where it’s like the best thing that happens
during your tenure is that nothing happens, right? Nothing explodes,
nothing blows up ——

Baszucki: I want to push back on that. And the reason I want to push back on that
is, that implies kind of a looking-over-your-shoulder defensive, oh-shit posture.
That’s actually not the optimal posture to build an amazing company. The optimal
posture to build an amazing company is: Oh my gosh, we are uniquely responsible
for a lot of people 12 and under. There’s so much constant learning, constant
innovation that if we can do this right, like, the future continues to be big and bright.
And I feel we’re uniquely in an opportunity to do excellent work in this 6-through-
12 zone where many, many other platforms, you know, bounce off of it because
they just can’t do it. They’re just not doing it right. So that is not how we run the
company. We view it as a high-responsibility situation, but an incredible opportunity
if we can do it right.

Newton: So then, my question is, How do you measure success? Like, how
do you as the C.E.O., as you look at everything you’re doing, feel like we’re
applying the right number of resources and they have the right error rates
and we’re at a place where I’m comfortable telling parents, “Bring your kids
to Roblox”?

Baszucki: We see all of those metrics internally all the time. They’re not publicly
shared. We’re watching filter rates. We’re watching this rate. We’re watching all of
that rate. Those metrics constantly get better, as we add more and more A.I. on
top of it. As you said, we are constantly running at a higher, higher scale as well.
My kids are on Roblox. We have thousands of people in the company. Their kids
are all on Roblox. But I would say, and I’ve said it on other things, a parent is the
ultimate arbiter of responsibility. And if a parent out there, for any reason, is not
comfortable with their kid going to the park, not comfortable with their kid playing
with a certain toy, not comfortable being on Roblox, who am I to say, right? And
so, I do leave that to the parent's responsibility. I just operate as an example. Like,
what would I do with my own kids?

Roose: Do you think that screen time for children is getting a bad rap? I
mean, this is a constant thread of conversation among the parents I know.
Everyone’s constantly worrying, oh, am I giving my kid too much screen
time? Are they doing enough offline activities with their friends? Are they
getting sucked into these like addictive online experiences with these A.I.
companions? What is your view on screen time for children?

20 Ethics and Info Tech

John Jay College
Baszucki: So when my son was really young, we had pretty radical screen time
limits. And you can imagine as you're a parent, you have your second, third and
fourth kid, you loosen up on that. I’m actually kind of supportive of it in a way. And
I do think there’s a bunch of contributing factors, like what is the ratio of blue light
to nonblue light, how much of the screen time is solo consumption versus group
consumption, what time of the day is the screen time. So I would say 2 a.m. heavy
blue light, solo consumption, short-form video, that’s arguably a different type of
screen time than, you know, noon with your friends for a while playing hide-and-
go-seek. But I do think this is something worth thinking about.

Newton: Well, anything that we did not get to today that you wanna make
sure that we talk about, any project that you’re working on that is exciting
for you at the moment?

Baszucki: Oh, wow. I thought we were going to talk about Polymarket and, like,
dropping, like, things into Roblox — like, earnings calls.

Newton: Wait, where are you on Polymarket?

Baszucki: So I assume you are both big fans.

Newton: We’ve talked a lot about prediction markets on the show.

Roose: Yup.

Baszucki: So, I don’t know if you had the experience, independent of political party
or anything, just a real-time science experiment. On election night, I had four
screens, like journalists of all journalists, and like, how do I quickly go from
journalist to journalist? And then with my family, we were watching Polymarket as
well, and we were comparing Polymarket to the comments of all the journalists.
And I do think Polymarket was coming through, like, 30 to 45 minutes ahead of
them. And, like, it was an interesting thought experiment to see, like, OK, we’re at
86 percent here, whether you like or dislike the way it was going, and then see
various journalists, like, say different things.

Newton: Yeah, so you think this is becoming a really interesting signal to

understanding what’s happening in the world right now?

Baszucki: I think, mathematically, there’s the notion that prediction markets, if

they’re liquid and transparent, capture, in a way, all of the world’s information and
every signal. And if there is one election counter way out in the corner in some

21 Ethics and Info Tech

John Jay College
weird county that knows it’s going to go this way, if they could go bet $10,000, like,
that’s gonna nudge the thing and make the information more perfect. So, I’m a
huge fan of that.

Roose: Would you ever put a prediction market inside Roblox — like, let kids
bet with their Robux and say, “I bet he’s gonna steal a Tung Tung Tung
Sahur?”

Newton: Or, “I bet he’s not gonna Dress to Impress.”

Baszucki: We would — I think we would have to do that — once again, I’ll share
some of the complexity. Every single country in the world has different legislation
around loot boxes and kid gambling. And so we would have to be — it sounds very
fun and obvious. Like, I love that —

Roose: Oh, to be clear, I think this is a horrible idea.

Newton and Baszucki: (laughter)

Roose: I was just interested if you were thinking about it.

Baszucki: Well, I actually think it’s a brilliant idea if it can be done in an educational
way that’s legal. And so, imagine no free Robux, no free prizes, just a game called
the Dress to Impress Predictor, where it’s not like trying to get kids’ money or
anything like that. I would be a big fan of it.

Newton: Interesting.

Roose: Yes, start them young. That’s what I always say. When it comes to
gambling, you’re never too young.

Newton: When I was in elementary school we did do a straw poll on who

would win the 1988 presidential election.

Roose: Oh yeah?

Newton: Yeah. So, some of these old ideas come back eventually.

Baszucki: But the straw poll was not a prediction market. And if you gave every
kid 10 cents of lunch money and, like, some of them got to buy an extra snack and
some lost it, you could get a completely different result than the straw pool you
would get in any fifth-grade classroom.

22 Ethics and Info Tech

John Jay College
Newton: Interesting to think about.

Roose: OK, we’re out of time, but Dave, can I just make one selfish request
of you?

Baszucki: Yes.

Roose: So my kid right now is about 3-and-a-half years old. Presumably

sometime, maybe four or five or six years from now, he’s gonna ask me,
“Can I play Roblox?”

Newton: Oh, it’s gonna be about a year from now.

Roose: (laughs) Probably a year from now. Will you fix the problems with
safety on your platform by the time that my kid is old enough to ask me that
question?

Baszucki: I’m going to go more out of the box than that. Roblox is an amazing
platform right now for your kid. And I’m happy to go deep with you, possibly
nondisclose you, show you metrics, all of that, to do whatever it takes to get you
over the hump.

Roose: OK.

Newton: All right. There it is.

Roose: There it is.

Newton: All right. Thank you, Dave.

Baszucki: Thank you guys.

23 Ethics and Info Tech

John Jay College
 R. Matthew Shockey, “Making Claims”

From:
Shockey, R. Matthew. 2020. “Making Claims.” In Critical Thinking and Applied
Ethics.

Link:
Indiana University Pressbooks

How to cite in-text:

(Shockey 2020, Handout, Par. #)*

* “Par. #” means “Paragraph number”

How to cite in Works Cited:

Shockey, R. Matthew. 2020. “Making Claims.” PHI 216 Handout.

Claims: What They Are and What They Are Not

1. There are lots of things we do with language:
• We ask questions (“why did you order the French fries?”)
• We give orders and make threats (“order the fries, or else!”).
• We express feelings and sensations (“yum, fries”, “ouch, too hot!”).

2. But one of the most important things we do is to state what we take to be true.
When we do this, we make a claim.

3. A claim is a statement, expressible as a declarative sentence, that is either

true or false. (We’ll use the terms “claim” and “statement,” and sometimes
“judgment,” interchangeably.)

4. Here are some examples of claims/statements:

• You ordered the French fries.
• I told you to order the French fries.
• French fries are delicious.
• French fries are made from potatoes.

24 Ethics and Info Tech

John Jay College
 • French fries are made from tomatoes.
• If you eat too many French fries, then you’ll get a stomach ache.
• French fries are bad for you.
• French fries are good for you.
• French fries are not good for you.
• Either French fries are good for you, or they are not.
• I like French fries.
• It is immoral to eat French fries.
• Hobbits are especially fond of French fries.
• The largest French fry ever made was over 2 pounds.

5. Each of these is either true or it isn’t (and obviously not all of them can be true,
since some state opposite and so incompatible things). But:

CRUCIAL POINT: Whether or not a claim is actually true is a separate

issue from whether it is a claim.

6. This is most clear when you look at claims that are obviously false, as in the
case of “French fries are made from tomatoes.”

7. (For you food nerds out there: fries used to be called French fried potatoes,
which implies there is a technique of French frying that could, in principle, be
used to cook other things, possibly even tomatoes (though the water content,
at least for ripe ones, would make that a challenge. But the term “French fries”
is pretty much exclusively used to refer to potatoes. This points to the
importance of definitions in knowing what claims are being made, which we’ll
discuss below.)

8. In many cases you may not know whether a given claim is true, but you can
still see that a claim is being made. I have no idea, for instance, whether the
claim “the largest French fry ever made was over two pounds” is true. I doubt
it is, but who knows? Still, I can see that there’s a claim there, and it’s either
true or it’s not, regardless of whether I know it to be.

9. In other cases the truth of a claim may depend on some context or other, such
as who is saying it. This is true in the case of, e.g., “I like French fries.”
When I, Prof. Shockey, say it, it is true. But if someone who didn’t like fries said
it, they’d be saying something false. See below for more on context.

10. So, to make a claim is to make an assertion, which can then be evaluated for
its truth.

25 Ethics and Info Tech

John Jay College
11. By contrast, a question or a command or an exclamation or an
expression or exclamation of feeling isn’t like that. If I say, “Do you like French
fries?” And you respond, “false!”, that doesn’t really make any sense.
Likewise, if I say “Hand over the fries!”, or “Gross, French fries!,” and you
respond, “true!”, that shows you don’t really understand what a command or
an exclamation is.

12. We’re focusing on claims because our big-picture aim in becoming better
critical thinkers is to get better at tracking the truth. For that, we need to
know what form our truth-tracking takes, and claims are at the heart of this.
They are also the building blocks of inferential reasoning (inferring some
things from others). When presented in a certain way, inferences are what
we’ll call arguments, which are chains of claims where some are used to
support or prove another. We’ll talk more about arguments in later modules.

13. First, there are three kinds of claims we are going to talk about. Evaluative
claims, non-evaluative claims, and definitional claims. We’ll finish off this
chapter looking at the last of these. You’ll learn about evaluative and non-
evaluative claims in the next chapters.

Defining Terms and Considering Context

14. There are two ways in which defining terms is important. First, some claims
state what is true by definition. Second, we need to be clear on what our words
mean so that we know what claims we are making. If there is ambiguity or
confusion, it tends to lead to lots of pointless discussion and argument…

Definitional Claims
15. Here’s how a not infrequent and always aggravating conversation I used to
have with my then nine-year-old son would go: he misuses a word, I tell him
what the word really means, and he responds: “that’s not what it means for
me.” I then say, “well, you can use the word in that way, but no one else will
understand you.” He’s often okay with that, hence the aggravation on my part.
(He’s now 11, and he’s gotten a little better on this, though he now delights in
finding other ways to aggravate me.)

16. The point here is that words typically have fairly fixed meanings. It’s not up to
us as individuals to decide what these meanings are. Of course, these
meanings change over time; many words have multiple meanings; casual

26 Ethics and Info Tech

John Jay College
 usage can differ from formal; etc. But, as any dictionary will show you, words
have relatively fixed definitions — and they need to if we are to be able to
communicate with one another.

17. When we state definitions of words or ideas, we make definitional claims,

that is, claims about the meanings of words or concepts.

18. If, for instance, I say “a caboodle is a collection of things,” you may not know
whether that is true. So you open your dictionary app (or blow the dust off your
print copy) and check, and you find that the claim is true. (My son got a kitten
and named him Kaboodle, based on the expression “the whole kit and
caboodle” — Kitten Kaboodle, get it? So I myself just looked up the word
“caboodle” in order to make my claim about what it means.)

Kitten Kaboodle, in a rare moment of rest.

19. If, on the other hand, I say, “a biped is a four-footed animal,” you know, or can
look up and find out, that my claim is false. Bipeds have two feet, not four.

20. Being clear on definitions is necessary in many cases to avoid mistaking one
claim for another. Suppose I report to the police, having just witnessed a heist
in South Bend, “the robbers are on the bank.” If what I mean is that they have
climbed to the top of the building they robbed (the bank), but the police think
I mean they ran to the edge of the river (the bank), that’s a problem. The claim
they think I am making is false, whereas the one I am actually making is
true. The problem is that “bank” is ambiguous, and the same sentence, “the
robbers are on the bank,” can mean two different things. This is a somewhat
silly example, but it serves to illustrate a common danger: unless we are sure
we are interpreting our words in the same way, we can easily make mistakes
in understanding each other and so in assessing the truth of what we say to

27 Ethics and Info Tech

John Jay College
 each other. We often waste a lot of time arguing about things because of such
misunderstandings. (These are what are called “verbal disputes.”)

Context
21. Issues connected to definitions bleed into issues of context. Sometimes words
that have one usage in common parlance are given a more restricted or
specific technical definition in another… [T]ake the word “dry.” If someone tells
you “the wine is dry,” you might think they are talking nonsense, for wine is a
liquid, and liquids are wet, not dry. But once you know that “dry” is used by
people in the wine world to refer to wine with low levels of residual sugar
(sugar that hasn’t been converted to alcohol during fermentation), then the
claim “the wine is dry” makes sense. Dry wine is just wine that is not sweet.

22. Here’s another example where context matters. Suppose I say, “Earl is tall.”
In order to assess whether that is true, it’s not enough to know Earl’s actual
height. Suppose Earl is 6′ 1″. Is he tall? Well, he’s taller than the average
American man. If he were a 1st-grade teacher and you saw him next to his
students, he’d look really tall. But if you saw him next to an NBA team, he’d
look short. In order to know whether the claim is true, then, you have to know
what class of people I’m judging Earl in relation to.

23. Lots of words work this way: to say it’s hot outside, or that someone is smart,
or that the music is loud, etc., is to assume a context where heat, intelligence,
or loudness are measurable in some reasonably stable way. Relative to the
context, the claim may or may not be true. And even if it is true in one context,
it might not be in another.

24. There’s a great deal more that could be said about how words vary according
to context. All meaning is contextual in some way or other. As competent
language users, we are constantly taking context into account. If you
remember this, then you can be careful not to make mistakes based on
placing a claim in the wrong context, and you can recognize that you can’t tell
what kind of claim you’re dealing with in some cases without being told the
context.

28 Ethics and Info Tech

John Jay College
 R. Matthew Shockey, “Evaluative and Non-Evaluative
Claims”

From:
Shockey, R. Matthew. 2020. “Evaluative Claims.” In Critical Thinking and Applied
Ethics; Shockey, R. Matthew. 2020. “Facts, Non-Evaluative Claims, and
Opinion.” In Critical Thinking and Applied Ethics.

Links:
Indiana Pressbooks

How to cite in-text:

(Shockey 2020, Handout, Par. #)*

* “Par. #” means “Paragraph number”

How to cite in Works Cited:

Shockey, R. Matthew. 2020. “Evaluative and Non-Evaluative Claims.” PHI 216
Handout.

Evaluative Claims
1. Many of the claims we make directly express our judgments about what we
value, as when we say that something is good or bad (or evil), or is right or
wrong. The values in question are often our ethical values, but other sorts of
values (aesthetic, etc.) are also expressed in such judgments.

2. Other claims make judgments that express our values more indirectly, but still
rely on them. Chief among these are claims about what should or shouldn’t be
done. To say that something should be done is to say that it would be good to
do it.

3. Together these kinds of claims are what we are going to call evaluative claims.

4. An evaluative claim is one that makes a statement about what is good or bad,
or right or wrong, or what ought or ought not to be done.

29 Ethics and Info Tech

John Jay College
5. This idea of an evaluative claim and what it expresses is absolutely crucial to
the course. It also is the concept that causes the most confusion for students.
You’ll get plenty of practice in the homework and on the quizzes with it, but
start now by memorizing it and really carefully thinking about the examples that
follow.

6. To help remember this, say the word “evaluative” out loud, and hear that
“value” is part of it.

7. Note that evaluative claims can also be comparative claims about what is better
or worse, or superlative claims about what is the best or worst.

8. Here are some examples of evaluative claims:

• Star Wars is the greatest movie of all time.

• George Lucas should not have made any more Star Wars movies after
Return of the Jedi.
• Health is one of the most important goods for a happy human life.
• Doctors should respect their patients’ autonomy.
• A good doctor is respectful towards their patients.
• Everyone should have the right to free health care.
• People should each have to pay for their own health care.
• Barack Obama is a great person.
• Donald Trump is a great person.
• You should treat others as you would have them treat you.
• It is better to be hated than loved.
• Panda Express has better food than Taco Bell.
• Panda Express should expand their menu to include more vegetarian
options.
• We should prevent the earth from getting any warmer.
• One ought to recognize and admit when one lacks knowledge.
• Abortion is wrong.
• Every woman should have the right to have an abortion.
• Eating animals is morally wrong.
• Eating animals is morally permissible.
• Brussels sprouts are delicious.
• I like donuts. [But see the discussion in the next section for how this can
also be a non-evaluative claim.]

9. There is a lot of diversity in the claims here. Some of these are about what the
right actions or policies are, some are about personal taste, some are about

30 Ethics and Info Tech

John Jay College
 moral/ethical matters, and some are about aesthetic matters. For some of the
claims, it’s unlikely that dispute over them would lead to violence. Others have
actually been disputed violently.

10. For many of these, you probably have a view about whether you think it’s
true or not.

CRUCIAL POINT #1: whether or not a claim is true is a separate issue

from whether it is evaluative or non-evaluative.

CRUCIAL POINT #2: Evaluative claims are not the same as opinion
claims. See the next chapter on non-evaluative claims for more about
what opinion is and isn’t. (I know I’m repeating myself here, but that’s
because this is really important.)

11. For any of the claims on this list, you can ask, “why should I believe that?” To
answer that question means providing reasons for why the claim should be
believed to be true. To engage in this kind of reason-giving is to engage in
argument — providing reasons (what we’ll call “premises”) for a claim (what
we’ll call a “conclusion”).

12. Depending on the kind of value in question, the reasons will be quite different.
If you ask why I think Brussels sprouts are delicious, all I can really say is that
they taste good to me. If you ask the person who says Star Wars is the
greatest movie of all time why they think that, they might answer by telling you
that it’s got compelling characters, innovative special effects, a classic story
line, etc. There’s room for difference of opinion here, but there’s also room for
rational discussion based in careful analysis of this film and others. If you ask
the person who says that eating animals is morally wrong why it is, they might
say that we shouldn’t cause unnecessary pain and suffering, and eating
animals requires we cause them pain and suffering. Here the appeal is to
values that the person making the claim thinks should be universally shared
and recognized, without there being room for differences of opinion. And so
on for any of these claims.

13. One of the key points, though, is that evaluative claims that are about a
particular case or example or person or situation almost always rests on some
sort of general value (a core value in the case of ethical claims). So, the
person who thinks that eating animals is wrong has, among their core values,
the belief that it is good to alleviate or prevent pain and suffering. The belief
about eating animals applies that core value to a particular case. The person
who thinks Star Wars is great has general views about the importance of

31 Ethics and Info Tech

John Jay College
 characters, story line, and special effects. Those general views are used to
assess Star Wars in relation to other films…

14. That’s partly how our values work: they are general, and we apply them in
particular instances to evaluate what’s good and bad and so what we should
and shouldn’t do.

15. Ultimately these are the things we’re interested in: how we reason based on
our values (applying general ideas to particular cases), and the related
process of giving reasons to support the claims we take to be true. Good
critical thinking requires recognizing claims, distinguishing them for the
reasons that support them, and being able to provide those reasons in a
cogent way….

Non-Evaluative Claims
16. Where values reflect how we think things ought to be, facts are those things
that are true of the world independently of what we happen to believe about
it.

17. There are facts about the past, facts about the present, and facts about the
future. There are facts about particular things, about classes or categories of
things, and facts about all that is. There are facts about the natural world and
human society. And, there are facts about fictional beings, numbers and other
mathematical objects, and (perhaps) supernatural beings, none of which are
part of the empirical world in any straightforward way.

18. Many facts may be unknowable, many can only be known with some degree
of probability rather than certainty, but, still, they remain facts.

19. Science aims to know facts, but so do we in lots of areas of our lives where
science doesn’t reach…

20. [A]t least in many cases, isolate the facts from our evaluative perspectives,
and we should always try to do so. When we do try to do this, and we think
we know what the facts are, we can then state what we believe to be true.

21. To state a belief about the facts is to make a particular kind of claim, what we
will call a non-evaluative claim.

32 Ethics and Info Tech

John Jay College
22. A non-evaluative claim is a claim that does not express the values of the
person making it (so, it says nothing about what is good or bad, right or wrong,
or should or shouldn’t be done)…

23. Here are some examples of non-evaluative claims:

• Water is composed of hydrogen and oxygen.

• The earth revolves around the sun.
• The sun revolves around the earth.
• The earth’s average temperature is increasing.
• It will rain tomorrow.
• The Pittsburgh Pirates will win the next three World Series.
• Dinosaurs roamed the earth millions of years ago.
• Sugar tastes sweet.
• Drinking a lot of soda will harm your body.
• Medical doctors are required to receive extensive education and training.
• Many people do not like their jobs.
• God exists.
• God does not exist.
• Either God exists or God does not exist.
• The soul is immortal.
• There is no such thing as a soul.
• Frodo is a hobbit.
• Frodo is an orc.
• If Frodo is a hobbit, then he is smaller than an orc.
• 2+3=5.
• 2+3=6.
• The sum of the squares of the sides of a right triangle is equal to the square
of the hypotenuse.
• I like donuts.

24. Observe that, as with evaluative claims, there is actually a lot of diversity here.
Some of these are obviously false (which ones?). Thus, as emphasized
above, to say that a claim is non-evaluative is not to say that it is true, only
that it purports to state a fact, i.e., an objective truth (without voicing a view
about what’s good/bad/right/wrong/should/shouldn’t be done).

25. As with our list of evaluative claims, for each of these you can ask, “why should
I believe that?” And whoever is asserting the claim can then try to give reasons
why.

33 Ethics and Info Tech

John Jay College
26. In some cases, the “why?” question will be answered by science. This requires
giving empirical evidence in conjunction with whatever theory is relevant. So,
if you ask “why should I believe that the earth’s average temperature is
increasing?,” a climate scientist will be able to give the data they have
collected and an explanation of how the earth’s climate mechanisms work in
order to answer your question. Similarly for the claims about water, the earth
and sun, dinosaurs, and the weather.

27. Other claims require objective information to support them, but there is more
room for non-experts to have access to that information. If you ask why you
should believe the Pirates will win the next three World Series, for instance, I
can answer by giving you my analysis of their players, coaching, etc., based
on my following the team carefully.

28. Observe that, in many of these cases, including the scientific ones, no matter
how good the evidence provided for the claim is, there’s always some degree
of uncertainty. Those who assert the claims are thus making a judgment that
reflects their individual assessment of the evidence. In that sense, they are
expressing an opinion. But note two things: (1) these are opinions backed by
reasons, and (2) there are no values involved in these opinions. So, here we
can see that it is possible to have opinions about facts, but, to the extent that
they are based on reasons and evidence, they are not merely subjective. It is
possible to assess the evidence provided by objective standards. (And, sadly,
when one does so, the earth is warming, and the Pirates are unlikely to win
the next three World Series.)

29. Other non-evaluative claims need to be supported by reasons in other ways.

Take the claim about triangles. It expresses the Pythagorean Theorem. If
someone asks, “why should I believe that?”, it is possible to provide
a proof that it is true, on the basis of certain other claims that define terms
such as “straight line” and “angle.” It is not a generalization from data about
triangles, it is a necessary fact about them. No one who can understand the
proof can deny the truth of the Theorem, so there is no room for individuals to
have legitimate differences in the judgments they make about it. Mathematics
in general works this way: its claims (most of which are far more abstract than
the Pythagorean Theorem, and refer to objects which are in principle
unobservable) are established with logical certainty, not empirical data.
(Empirical scientists rely on mathematics to do their work; but there is a vast
difference in how truth is established for empirical and mathematical claims.)

30. Or take the claims about God’s existence. Assuming we can agree on what
“God” means (as at least some have been able to do), a being matching the

34 Ethics and Info Tech

John Jay College
 definition either exists or doesn’t. So one of the listed claims is true and the
other one isn’t. Like mathematical claims, which one of these is true isn’t a
question for empirical science (at least on most understandings of what “God”
means and how empirical science works). There’s a long history of arguments
for and against God’s existence (which you can learn about in other
philosophy classes if that interests you!). Some of these arguments are
supposed to work like mathematical proofs; others work more like scientific
claims that admit of some uncertainty. The assumption in both cases,
however, is that it is not a matter of subjective opinion which claim is true.
And, in neither case, does the truth depend on the values of the one making
the argument. Your own values may be rooted in your belief in God (or in a
rejection of that belief), but that does not mean that your
values determine which claim about God’s existence is true.

31. Finally, take the first claim about Frodo. If you’ve either seen the Lord of the
Rings movies or read the books by Tolkien that they are based on, you’ll know
that this claim is true. But Hobbits don’t really exist! That means that you must
determine its truth by finding evidence within the fictional world in which Frodo
lives. As in the previous cases, nothing about what you value is relevant in
determining whether this claim is true. The fact that we’re dealing with a
fictional world doesn’t change that.

32. So, some Crucial Points to make about this:

CRUCIAL POINT #1: claims about the past and future are typically
non-evaluative claims. Just because there is uncertainty in such claims
(as when you say what the weather will be, whether a particular team will
win the World Series next year, or that a species of animal existed along
time ago) doesn’t mean you aren’t making a claim that is or is not
objectively true – you just can’t know its truth with 100% certainty.

CRUCIAL POINT #2: building on the previous point, claims that express
uncertainty or probability (as most claims about the future do)
are not, just because of that, evaluative claims. (They may
involve individual judgment, but that does not mean that the values of the
individual who is making the judgment are involved.)

Thus, CRUCIAL POINT #3: Non-evaluative claims are not to be

contrasted with opinions. You can express an opinion about the facts,
when there is some uncertainty and room for individual judgment about it.
(In the case of doctors, we even call their assessments of a patient’s
condition “opinions,” and it’s often good advice to get a “second opinion.”

35 Ethics and Info Tech

John Jay College
 But these “opinions” are, in fact, expert judgments. They reflect the best
assessment that the doctors can make about the facts; they do not express
the doctor’s values.)

Telling the Difference between Claims

33. So, how do you know whether a given claim is non-
evaluative, evaluative, or definitional?

34. The first thing to do is to look and see if it has language that implies it’s
evaluative, i.e., whether it clearly says that something is
good/bad/right/wrong/should be done/shouldn’t be done. If it does, it’s almost
certainly evaluative. (I say “almost certainly” because language always has
exceptions to its rules, and this rule is no exception.)

35. Once you are sure a claim is not making an evaluative judgment, you then
need to ask whether the claim is non-evaluative or definitional. Definitional
claims are really a kind of non-evaluative claim — they state the facts about
the meanings of words or ideas — but they are special enough that we are
giving them their own category. So, check to see if the claim is just about what
a word or concept/idea means. To do this, ask if you could determine its truth
just by looking at a dictionary. If you could, it’s definitional.

36. If you’ve ruled out evaluative and definitional, then non-evaluative is what’s
left. But remember, there is variety here. Be particularly careful to recognize
that among these are claims that involve individual judgment about empirical
matters (doctor’s “opinions,” predictions about the weather or the performance
of a sports team, etc.). It’s tempting to call these evaluative, but they are not.

Context, again
37. Now, about the claim “I like donuts.” In the previous chapter I gave a context
where this is essentially equivalent to me saying “Donuts are good,” which
expresses my (positive) evaluation of them. But suppose someone is coming
around the offices in Wiekamp and taking a poll of people in order to decide
what treats to buy, and they ask me “Do you like donuts?” If I say in response
“I like donuts,” then what I am doing is reporting a fact about my food
preferences. In that context, the same string of words is used to express a
different, non-evaluative claim. I might actually think it’s bad that I like donuts
(since I know they are unhealthy), even as I report the fact of my liking them.

36 Ethics and Info Tech

John Jay College
38. In fact, a great deal of social science, as well as the work of product
development and marketing departments, involves tracking the
facts about people’s views of what’s good and bad. So, for instance, a
researcher reporting on the results of a survey might make a claim of the form
“X% of Americans think abortion is wrong.” But that claim, though about what
people’s values are, is not itself an evaluative claim.

39. Thus, Crucial Point #4: In general, claims about what people believe is
good and bad are non-evaluative. They describe people’s values
without endorsing them.

40. The fact/value distinction and the corresponding non-evaluative/evaluative

claim distinction are helpful ones to make. Lots of our reasoning can be
clarified as to whether we are assuming or trying to establish what we think is
good/bad/right/wrong or trying to figure out what is true independent of our
values. Not every case fits neatly into the distinctions or is 100% clear as to
which side it falls on, but when we learn to recognize the ones that do, we can
make better sense of the murky ones.

37 Ethics and Info Tech

John Jay College
 Anthropic, "Claude Teaches Jen about Philosophical
Argumentation”

From:
Anthropic, "Claude Teaches Jen about Philosophical Argumentation." Claude,
Sonnet 4.5 version, 20 Jan. 2026, [Link]
8d9c-0e677e1bd64b.

Link:
Claude

How to cite in-text:

(Anthropic 2026, Handout, Par. #)*

* “Par. #” means “Paragraph number”

How to cite in Works Cited:

Anthropic. 2026. “Claude Teaches Jen about Philosophical Argumentation.” PHI
216 Handout.

Instructor’s Note: Below is a lightly edited copy-paste of conversation with a

GenAI chatbot, Claude by Anthropic, where I pretend to be a philosophy student
taking Ethics and Info Tech and ask it to teach me about what philosophical
argumentation is. The resulting reading is a twofer! You get to learn about what
philosophical argumentation is, plus you get to have meta-level philosophical
thoughts about the uses of AI chatbots as tutors, whether Claude is flattering me
to keep me engaged, and so on! Looking forward to talking more in class!

JEN:

I am a 200-level college student taking my first philosophy course, which is called

Ethics and Information Technology. Please tutor me on what a philosophical
argument is. Please make connections to the distinction between evaluative and
non-evaluative claims, as well as what it means to give reasons for something.
Please give me information in small chunks, check my understanding as we go
along, and provide sources when relevant.

38 Ethics and Info Tech

John Jay College
 CLAUDE:

Great! I’m excited to help you learn about philosophical arguments. Let’s start with
the foundation and build from there.

What is a claim?
Before we can talk about arguments, we need to understand claims. A claim is
simply a statement that can be true or false. For example:
• “It’s raining outside”
• “Privacy is important”
• “This app collects user data”
All of these are claims because they’re asserting something about the world that
could be evaluated as true or false.

Two types of claims: Evaluative vs. Non-evaluative

Here’s where it gets interesting for ethics:

Non-evaluative claims are descriptive - they describe facts about the world
without making value judgments. Example: “Facebook has 2 billion users.”

Evaluative claims make value judgments - they say something is good/bad,

right/wrong, should/shouldn’t be done. Example: “Facebook should not sell
user data to third parties.”

Notice the difference? One just states a fact, the other makes a moral judgment.

Quick check: Can you tell me which of these is evaluative and which is non-
evaluative?
1. “AI systems can process large amounts of data quickly”
2. “Companies ought to obtain informed consent before collecting personal
information”

Take a shot at categorizing these, and let me know if you’d like me to explain
anything further!

JEN:

1 is non-evaluative. 2 is evaluative.

39 Ethics and Info Tech

John Jay College
 CLAUDE:

Excellent! You’ve got it exactly right.

Let me reinforce why:

1. “AI systems can process large amounts of data quickly” - This is just
describing what AI can do. It’s a factual claim about capability.
2. “Companies ought to obtain informed consent...” - The word “ought” is a
dead giveaway that this is evaluative. It’s making a moral claim about what
companies should do.

Helpful tip: Look for words like “ought,” “should,” “good,” “bad,” “right,” “wrong,”
“just,” “unjust” - these often signal evaluative claims.

Now let’s talk about arguments:

A philosophical argument isn’t a disagreement or a fight. Instead, it’s a structured
way of giving reasons for a claim.

An argument has two main parts:

• Premises: These are the reasons you’re offering (they’re also claims)
• Conclusion: This is what you’re trying to convince someone of (also a
claim)

The premises are supposed to support or justify the conclusion.

Here’s a simple example:

• Premise 1: Autonomous vehicles will reduce traffic accidents.
• Premise 2: Reducing traffic accidents is morally good.
• Conclusion: Therefore, we should develop autonomous vehicles.

See how the premises give reasons for accepting the conclusion?

Question for you: In this argument, which claims are evaluative and which are
non-evaluative?

JEN:

It sounds like Premise 1 is non-evaluative, and Premise 2 is evaluative. I think the

Conclusion is also evaluative because it’s a claim about what we should do, but
I’m not sure and would like an explanation from you.

40 Ethics and Info Tech

John Jay College
 CLAUDE:

You’re absolutely correct on all three, and your reasoning is spot-on!

• Premise 1 is non-evaluative - it’s a factual prediction about what

autonomous vehicles will do.
• Premise 2 is evaluative - it’s making a moral judgment that reducing
accidents is “morally good.”
• The Conclusion is also evaluative, and you identified it perfectly. The
word “should” signals a value judgment about what we ought to do. It’s
prescribing action based on moral reasoning.

Here’s the important pattern you’ve discovered:

In ethics, we almost always need both types of claims to make a compelling
argument:
• Non-evaluative premises tell us facts about the world (what is)
• Evaluative premises provide our moral principles (what we value)
• Together, they support an evaluative conclusion about what we should
do

This is sometimes called bridging the “is-ought gap” - we can’t get from pure facts
to moral conclusions without some evaluative principle connecting them.

Let’s deepen your understanding of “giving reasons”:

When we give reasons in philosophy, we’re trying to show that if the premises are
true, then the conclusion should be accepted. We’re building a logical connection.

Quick exercise: Can you think of an argument related to information technology

or AI? Try to identify one non-evaluative premise, one evaluative premise, and an
evaluative conclusion. Don’t worry about making it perfect - just give it a try!

JEN:

I’ll try.

Non-evaluative premise: Many GenAI companies have free versions of chatbots

that users can access just by signing up.

Evaluative premise: GenAI companies want to hook more and more people into
using their chatbots, which giving out free versions allows them to do.

41 Ethics and Info Tech

John Jay College
Evaluative conclusion: People should remember that “if something’s free, you are
the product” when using free GenAI chatbots.

CLAUDE:

Great effort! You’re thinking about a real and important issue in tech ethics. Let me
give you some feedback to help refine your understanding:

Your non-evaluative premise - Perfect! This is a straightforward factual claim

about what GenAI companies do.

Your second premise - This is actually trickier than you might think. Let me break
it down:
• “GenAI companies want to hook more and more people” could be non-
evaluative (describing their business goal)
• BUT the word “hook” has negative connotations, which sneaks in some
evaluation

More importantly, this premise doesn’t yet give us a moral principle - it’s still mostly
describing what companies want to do.

Your conclusion - You’re right that this is evaluative (it says what people “should”
do), but notice it’s pretty far from your premises. We’d need connecting steps to
get there.

Let me show you how we might restructure this:

• Premise 1 (non-evaluative): Free GenAI chatbots collect user data for

profit.
• Premise 2 (evaluative): People have a right to understand how their data
is being used.
• Conclusion (evaluative): Therefore, people should be informed that they
are the product when using free chatbots.

See how Premise 2 now provides a moral principle (about rights) that connects the
fact to the conclusion?

Does this distinction make sense? What makes a premise evaluative vs. just
descriptive?

42 Ethics and Info Tech

John Jay College
 JEN:

Yes, the distinction makes sense, but when you say “People have a right to
understand how their data is being used,” where does something like a “right”
come from?

CLAUDE:

Wow, excellent question! You’ve just hit on one of the deepest issues in ethics and
philosophy. This is exactly the kind of critical thinking your philosophy course wants
you to develop.

The short answer: Different ethical theories give different answers about where
rights come from, and philosophers disagree about this!

Let me sketch a few major approaches:

1. Natural rights theory: Rights are inherent to being human - they exist
independently of governments or laws. (Think: Thomas Hobbes, John
Locke, or the language in the Declaration of Independence about
“unalienable rights”)
2. Social contract theory: Rights come from agreements (explicit or
implicit) that people make with each other to live together in society.
(Think: Rousseau, Rawls)
3. Consequentialist view: We should treat people as having certain rights
because doing so leads to better outcomes for everyone. (Think: John
Stuart Mill)
4. Legal positivism: Rights are whatever laws and institutions say they are
- they’re social constructs, not natural features of the universe.

Here’s the key insight for making arguments:

When you use evaluative premises like “people have a right to X,”
you’re assuming some moral foundation. In a complete philosophical argument,
you might need to defend that assumption too! This can lead to what philosophers
call “infinite regress” - every premise needs justification, which needs justification,
which needs...

Question for you: Does this make philosophical arguments seem impossible, or
does it just mean we have to be clear about our starting assumptions? What do
you think?

43 Ethics and Info Tech

John Jay College
 JEN:

When you’re talking about an infinite regress, it does make philosophical

argumentation seem impossible. Is philosophical argumentation basically just our
opinions, but disguised as being deeper than that by using pretentious
terminology?

CLAUDE:

That’s a fantastic question, and I really respect your skepticism here. You’re
wrestling with something philosophers themselves debate intensely. Let me
address this head-on.

No, philosophical arguments are NOT just dressed-up opinions. Here’s why:

1. Arguments are publicly evaluable

Unlike mere opinions (“I just feel privacy is important”), arguments
give reasons that others can examine, challenge, and test. We can ask:
• Are the premises true?
• Does the conclusion actually follow from the premises?
• Are there counter-examples?
• Are the assumptions consistent?

This makes philosophical thinking much more rigorous than opinion-sharing.

2. The infinite regress problem has solutions

Philosophers handle foundational premises in several ways:
• Coherentism: Beliefs justify each other in a web of mutual support
(not a linear chain)
• Foundationalism: Some basic premises are self-evident or don’t
need further justification
• Reflective equilibrium: We adjust our principles and judgments until
they fit together coherently

3. Progress happens through critical dialogue

Even if we start from different foundations, argumentation helps us:
• Identify where we disagree (factual vs. moral premises)
• Test internal consistency
• Discover implications we hadn’t considered
• Refine our thinking

44 Ethics and Info Tech

John Jay College
Real example from your field:
Two people might disagree about AI regulation. Through argument, they might
discover they actually share the same values (human autonomy) but disagree
about facts (whether AI threatens it). That’s progress!

Here’s an analogy: Science also rests on unprovable assumptions (that the

universe is orderly, that our senses are somewhat reliable). Does that make
science “just opinions”? No - because the methods of testing, evidence, and
reasoning matter.

What do you think? Does this distinction between “mere opinion” and
“reasoned argument” seem meaningful, or are you still skeptical?

JEN:

OK, your explanation makes sense to me. So, if I feel like I have an opinion about
something, can you walk me through the process of testing it and making it a
reasoned argument?

CLAUDE:

Absolutely! I’d love to help you work through that process. This is one of the most
practical skills you’ll learn in philosophy.

Let’s do this step-by-step:

Step 1: State your opinion clearly

First, just tell me what you think about some issue in information technology or AI
ethics. It can be something you feel strongly about, something you’re uncertain
about, or even just a hunch you have.

Don’t worry about making it sound philosophical yet - just tell me what you believe
in your own words.

What issue would you like to work with?

(And just so you know where we’re headed: Once you share your opinion, I’ll guide
you through identifying what factual claims it depends on, what values it assumes,
possible objections, and how to structure it as a proper argument. But let’s start
with what YOU actually think!)

45 Ethics and Info Tech

John Jay College
 JEN:

I worry about GenAI isolating humans from real, meaningful interactions other even
more than previous technologies - like when you Google something instead of
asking a friend who’d know the answer, but on steroids, because GenAI gives
bespoke answers to specific questions even better than Google

CLAUDE:

This is a really thoughtful concern, and it touches on something many people are
worried about. Let’s work through turning this into a philosophical argument!

Step 2: Let’s identify the different types of claims in your worry

I’m hearing several things in what you said. Let me break them down, and you tell
me if I’m understanding you correctly:

Possible factual/non-evaluative claims:

• GenAI provides more personalized answers than Google does
• GenAI can substitute for asking friends questions
• Previous technologies (like Google) already reduced some human
interactions

Possible evaluative claims:

• Real, meaningful human interactions are valuable/important
• Reducing human interactions is bad or concerning
• There’s something worse about GenAI replacing human interaction than
Google replacing it

Step 3: Clarify what you mean

Before we build the argument, I want to make sure I understand your core worry:
1. Are you concerned that using GenAI is morally wrong for individuals, or
that it will have bad societal consequences, or both?
2. When you say “meaningful interactions” - what makes an interaction
meaningful to you? Is it the human connection itself, the quality of
answer, something else?
3. What exactly is “on steroids” about GenAI compared to Google - is it just
that the answers are better, or something else?

Take your time with these - your answers will help us build a stronger argument!

46 Ethics and Info Tech

John Jay College
 JEN:

Here are my answers:

1. I worry that it will have bad societal consequences.

2. I think the human connection itself is what’s meaningful. By making

ourselves vulnerable to each other, like admitting we don’t know
something and asking another human to help, it deepens a relationship.
Deep human relationships seem like a very valuable part of the human
experience and what it means to be human.

3. What I meant is that GenAI offers really customized answers to

questions, so people might feel silly/stupid/annoying asking another
human for an answer and taking up their time when they could just
“Claude” their question.

CLAUDE:

Excellent answers! You’re being really thoughtful about this. Now we have much
clearer material to work with.

Step 4: Identify your core values and factual assumptions

Based on what you’ve said, here’s what I’m seeing:

Your core values (evaluative premises):

• Deep human relationships are valuable and important to human
flourishing
• Vulnerability and interdependence strengthen relationships
• We should be concerned about societal changes that weaken human
connection

Your factual assumptions (non-evaluative premises):

• GenAI provides sufficiently good answers that people will use it instead of
asking humans
• When people use GenAI instead of asking friends/colleagues, they lose
opportunities for vulnerable moments
• These lost opportunities will accumulate into weaker relationships at a
societal level

Step 5: Test the factual assumptions

This is where critical thinking comes in. Let’s examine these factual claims:

47 Ethics and Info Tech

John Jay College
Question 1: Is it true that people will replace human asking with GenAI, or might
they use GenAI for different types of questions? (For example: asking Claude
about philosophy homework vs. asking a friend for restaurant recommendations)

Question 2: Are there any ways GenAI might actually enable deeper human
conversations? (For instance: learning enough background to have a more
meaningful discussion with someone?)

I’m not saying your worry is wrong - I’m asking you to think about what evidence
would support or undermine your factual claims. What do you think about
these questions?

JEN:

Question 1 - I think people already ask Claude for both philosophy homework help
and restaurant recommendations, as well as many other things.

Question 2 - Maybe you’re right that GenAI could give more background info, or
help people who are neurodivergent get useful strategies for
communicating/interacting with other humans.

CLAUDE:

Great! You’re doing exactly what philosophers do - considering evidence and

counter-arguments honestly, even when they complicate your initial view.

Step 6: Consider objections and refine your position

You’ve identified some important complications:

1. GenAI might have mixed effects - reducing some interactions while

enabling others
2. Different people might be affected differently - your neurodivergent
example is excellent
3. The type of question matters - some substitutions might be more
concerning than others

This is really important: A good philosophical argument doesn’t have to be

black-and-white. You can still argue that GenAI is concerning for human
connection on balance or in certain ways, even while acknowledging benefits.

48 Ethics and Info Tech

John Jay College
Step 7: Structure your argument
Let me try drafting an argument based on our conversation. Tell me if this captures
your concern:

• Premise 1 (non-evaluative): GenAI can answer many questions that

people previously would have asked friends, family, or colleagues.
• Premise 2 (non-evaluative): Asking others for help creates moments of
vulnerability that can deepen relationships.
• Premise 3 (evaluative): Deep human relationships are essential to
human flourishing.
• Premise 4 (non-evaluative): If GenAI significantly reduces these
vulnerable interactions, it will weaken human relationships at a societal
level.
• Conclusion (evaluative): Therefore, widespread GenAI use poses a
significant risk to human flourishing that we should take seriously.

What do you think? Does this capture your worry? Would you change any
premises? Do you see any weak points someone might challenge?

JEN:

I think it sounds mostly good, but I’m wondering why you’re saying that “Asking
others for help creates moments of vulnerability that can deepen relationships” is
non-evaluative. It sounds pretty evaluative to me. Or do philosophers think there
can be claims that are a mix between evaluative and non-evaluative?

CLAUDE:

Wow, excellent catch! You’re absolutely right to question this - your philosophical
instincts are getting sharper.

Let me unpack this carefully:

“Asking others for help creates moments of vulnerability that can deepen
relationships”

You’re right that this feels partly evaluative. Here’s what’s going on:
• “Asking others for help creates moments of vulnerability” - This part
is non-evaluative (just describing what happens psychologically/socially)
• “deepen relationships” - This is where it gets tricky! “Deepen” sounds
positive, like it’s saying something good happens

49 Ethics and Info Tech

John Jay College
Can claims be mixed? Yes, actually!
Philosophers recognize that some claims blend factual and evaluative elements.
“Deepen” could mean:
• Descriptively: The relationship becomes more intimate, trusting,
interdependent (non-evaluative - just describing what happens)
• Evaluatively: The relationship becomes better (evaluative - making a
value judgment)

How to handle this:

We have two options:

Option A: Rewrite Premise 2 to be purely non-evaluative: “Asking others for help

creates vulnerable interactions that increase intimacy and interdependence”
Then add the evaluative judgment separately: “Relationships with greater intimacy
and interdependence are more valuable”

Option B: Accept that Premise 2 contains both elements and be explicit about it

Which approach seems clearer to you? And do you see how being precise
about evaluative vs. non-evaluative helps us identify exactly where someone might
disagree with the argument?

JEN:

I think both Options A and B sound good, and what it maybe helps me realize is
that two people constructing similar philosophical arguments based on their initial
similar opinions can have different specific premises based on how they formulate
things, but reach the same conclusion.

Does that mean that philosophical arguments are subjective, like one’s subjective
perspective on whether a movie or song is good or bad? Can philosophical
arguments ever be wrong?

CLAUDE:

Another excellent question! You’re really digging into the heart of philosophical
methodology here.

Short answer: No, philosophical arguments are NOT subjective like taste in
movies, and yes, they can definitely be wrong - but let me explain the nuances.

50 Ethics and Info Tech

John Jay College
Why multiple formulations of “the same” argument doesn’t mean
subjectivity:

You’re right that we can formulate arguments differently and reach similar
conclusions. But notice what stays constant:
• The logical relationships between premises and conclusion
• The truth or falsity of the factual claims
• The coherence of the moral principles

Think of it like math: I can prove the Pythagorean theorem multiple ways, but that
doesn’t make math subjective. The different proofs are all constrained by logic and
mathematical truth.

How philosophical arguments can be WRONG:

1. Factually false premises: ”GenAI never makes mistakes” - empirically
false
2. Logical invalidity: The conclusion doesn’t follow from the premises
3. Internal inconsistency: Your premises contradict each other
4. Hidden assumptions: Relying on premises you haven’t stated

Example of a bad argument:

Premise 1: Social media connects people globally. Conclusion: Therefore, social
media always improves relationships.
This is invalid - the conclusion doesn’t follow. We’d need additional premises about
whether global connection necessarily improves relationships.

But here’s the tricky part about evaluative premises:

You can’t prove “deep relationships are valuable” the same way you prove
“2+2=4.” BUT:
• You can show it’s consistent or inconsistent with other values someone
holds
• You can examine its implications and see if someone accepts them
• You can appeal to shared intuitions or common human experiences

Key difference from movie taste:

If I say “The Godfather is bad” and you say “The Godfather is good,” we’re just
expressing preferences.

But if I argue “Surveillance capitalism is wrong because it treats people as mere

data sources, and treating people as mere means is morally wrong,” you can:
• Challenge whether surveillance capitalism really treats people as mere
means (factual)

51 Ethics and Info Tech

John Jay College
 • Question whether my Kantian principle (people as ends) is justified
(evaluative, but arguable)
• Point out inconsistencies (do I accept other implications of this principle?)

This is rational discourse, not just preference-sharing.

Question for you: Think about your GenAI argument. Can you imagine ways
someone could show it’s wrong that go beyond just having different preferences?
What would make your argument stronger or weaker?

JEN:

Tell me what surveillance capitalism is?

Also to answer your question, I think some research studies showing that GenAI
helps or hurts human connections could make my argument stronger or weaker
(depending on what these studies say).

CLAUDE:

Great question! Let me explain surveillance capitalism, and then I’ll respond to
your excellent point about research.

Surveillance capitalism:
This is a term coined by scholar Shoshana Zuboff (in her 2019 book The Age of
Surveillance Capitalism). It refers to a business model where companies:

1. Collect massive amounts of data about users’ behavior, preferences,

location, social connections, etc.
2. Analyze this data to predict and influence user behavior
3. Sell these predictions to advertisers or use them to manipulate users
toward profitable actions

The key insight is that you are not the customer - you’re the raw material. Your
behavior is extracted, processed, and sold. Companies like Google, Facebook,
and many others profit not from selling you products, but from selling predictions
about your behavior to others.

Zuboff argues this is a new form of capitalism that treats human experience as free
raw material for commercial exploitation.

52 Ethics and Info Tech

John Jay College
Now, about your answer regarding research studies - this is EXCELLENT
thinking!

You’ve identified exactly what philosophers look for: empirical evidence that
could support or undermine factual premises.

Your argument rests on factual claims like:

• GenAI reduces certain human interactions
• These interactions are important for relationship formation
• The net effect is negative for human connection

Research could test these! For example:

• Longitudinal studies on GenAI users’ relationship quality
• Surveys about what types of questions people ask AI vs. humans
• Psychological studies on vulnerability and relationship formation

Here’s the sophisticated part you’ve grasped:

Even if we can’t prove evaluative premises empirically, we can test
the factual premises empirically. If the facts turn out differently than you assumed,
you might need to revise your conclusion even if your values stay the same.

Does this connection between philosophical arguments and empirical

research make sense?

JEN:

Yes it does, but my follow-up question is about what the point of philosophy is. If
research studies an important part of making philosophical argument stronger or
weaker, isn’t empirical research much more important than philosophy if we want
to answer important questions about the world?

CLAUDE:

This is an amazing question - honestly one of the most important questions about
philosophy itself. You’re thinking like a real philosopher now!

Short answer: Philosophy and empirical research need each other. Neither can
do the job alone. Let me show you why.

What empirical research CAN’T tell you:

Imagine a perfect research study shows: “GenAI users have 30% fewer face-to-
face conversations with friends.”

53 Ethics and Info Tech

John Jay College
That’s valuable data! But it doesn’t answer:
• Is that bad? (evaluative question)
• What counts as a “meaningful” relationship anyway? (conceptual
question)
• Do people have a right to use GenAI even if it harms
relationships? (normative question)
• Should we regulate GenAI based on this? (policy question combining
facts and values)
•
Science can tell us what is, but not what ought to be. This is sometimes called
the is-ought problem (famously identified by philosopher David Hume).

What philosophy does that research can’t:

1. Clarifies concepts: What do we mean by “privacy,” “autonomy,”
“fairness,” “harm”? Research can’t proceed without clear definitions.
2. Identifies values: What matters morally? Why? Research assumes
values but doesn’t justify them.
3. Examines logical consistency: Can you coherently hold beliefs X and Y
together? This is logic, not empirical study.
4. Asks foundational questions: Why trust scientific method? What makes
evidence valid? (You can’t use science to justify science without circular
reasoning!)

A concrete example from your field:

Suppose research shows: “AI facial recognition is 95% accurate.”
• Science tells us the accuracy rate
• Philosophy asks: Is 95% accurate enough when the 5% errors
disproportionately affect people of color? What does “fairness” mean in
this context? Do people consent to being surveilled even if it’s accurate?

Here’s the key insight:

Every empirical research project makes philosophical assumptions:
• What questions are worth studying? (values)
• What counts as good evidence? (epistemology)
• How should we treat research subjects? (ethics)

And every philosophical argument benefits from empirical evidence about how the
world actually works.

They’re partners, not competitors.

54 Ethics and Info Tech

John Jay College
Think of it this way: Research is like having really accurate maps and
measurements of the terrain. Philosophy is like deciding where you’re trying to go
and why. You need both to take a good journey.

Does this help clarify why philosophy remains important even in an age of
powerful empirical research? Do you see how your GenAI argument actually
needs BOTH philosophical reflection (about what relationships matter and
why) AND empirical data (about how GenAI affects behavior)?

JEN:

Yes it does. OK, I have to run. Can you wrap up what we talked about?

CLAUDE:

Absolutely! Here’s a quick summary of what we covered:

Main concepts:
• Philosophical arguments have premises (reasons) that support a
conclusion
• Evaluative claims make value judgments (good/bad,
should/shouldn’t); non-evaluative claims state facts
• Good arguments typically need both types of claims - facts alone can’t
get you to moral conclusions

Key skills you practiced:

• Distinguishing evaluative from non-evaluative claims
• Turning an opinion into a structured argument with clear premises
• Identifying your factual assumptions vs. your values
• Considering objections and counter-evidence
• Recognizing that empirical research and philosophy work together

Your argument about GenAI and human connection: You developed a

thoughtful concern about how GenAI might reduce vulnerable interactions that
deepen relationships, structured it as a proper philosophical argument, and
identified what kind of evidence could strengthen or weaken it.

Big philosophical insights you had:

• Arguments aren’t just opinions - they’re publicly testable through logic and
evidence
• The “infinite regress” problem doesn’t make philosophy pointless
• Philosophy and empirical research need each other

55 Ethics and Info Tech

John Jay College
You asked really sharp questions today. Good luck with your Ethics and
Information Technology course - you’re going to do great!

56 Ethics and Info Tech

John Jay College