Chapter 1: Introduction to Artificial Intelligence (AI)

1.1 What is Artificial Intelligence (AI)?

AI is the field of computer science that focuses on creating machines that can perform tasks
that require human intelligence. AI can be defined in four ways:

1. Thinking Humanly – Understanding how humans think and trying to replicate it in

machines. (Example: Cognitive Science)

2. Acting Humanly – Making a machine behave like a human, passing the Turing Test
(Alan Turing, 1950).

3. Thinking Rationally – Using logic and reasoning to make decisions (based on

mathematics).

4. Acting Rationally – Creating an agent that makes the best decisions to achieve its
goals.

👉 The goal of AI is to create intelligent systems that can reason, learn, and adapt.

1.2 History of AI

AI has evolved in several stages:

Early AI (1950s – 1970s)

 1950 – Alan Turing introduces the Turing Test.

 1956 – The term “Artificial Intelligence” was coined at the Dartmouth Conference.

 1960s – First AI programs using logic and problem-solving were developed.

AI Winter (1970s – 1980s)

 Progress slowed due to lack of computing power and high expectations.

 Governments reduced funding for AI research.

AI Revival (1990s – 2000s)

 1997 – IBM’s Deep Blue beats chess champion Garry Kasparov.

 2000s – AI starts improving with better hardware and machine learning algorithms.

Modern AI (2010s – Present)

 2011 – IBM’s Watson wins Jeopardy! against human champions.

 2016 – Google’s AlphaGo defeats world Go champion.

 Today – AI is used in self-driving cars, chatbots, healthcare, and more.

👉 AI is growing fast due to big data, powerful computers, and deep learning.

1.3 Applications of AI

AI is used in many real-world applications across different fields.

1.3.1 Robotics and Legged Locomotion

 BigDog (2008) – A quadruped robot that moves like an animal and can recover from
falls.

 Atlas – A humanoid robot that can walk, jump, and even do backflips.

1.3.2 Autonomous Planning & Scheduling

 NASA’s Remote Agent – The first onboard AI system to control spacecraft scheduling.

 Mars Rovers – AI-based planning systems help NASA operate rovers on Mars.

 Uber & Google Maps – AI predicts traffic and suggests optimal routes for millions of
users.

 DARPA’s DART – Used for military logistics, planning movement of vehicles, cargo,
and troops.

1.3.3 Machine Translation

 Google Translate – Supports over 100 languages and translates billions of words
daily.

 AI translation is improving, especially for languages with a lot of training data (e.g.,
English & French).

1.3.4 Speech

Recognition

 Microsoft’s AI (2017) – Achieved a 5.1% word error rate, similar to human

transcription.

 Voice Assistants – Alexa, Siri, Google Assistant help users with tasks like setting
reminders and answering questions.

 Google Duplex – Can make phone calls and book appointments using natural-
sounding AI.

1.3.5 Recommendation Systems

 Netflix, YouTube, and Spotify – Recommend content based on user preferences.

 Amazon & Walmart – Suggest products based on purchase history.

  Spam Filtering – AI detects and blocks 99.9% of spam emails.

1.3.6 AI in Games

 Deep Blue (1997) – Defeated world chess champion Garry Kasparov.

 AlphaGo (2016) – Beat human Go champions using deep learning.

 AlphaZero – Learned to play chess, Go, and shogi without human input.

 AI in Video Games – AI has beaten human players in Dota 2, StarCraft II, and Quake
III.

1.3.7 Image Understanding

 AI can analyze and caption images, e.g., “A person riding a motorcycle on a dirt
road.”

 AI is improving but still makes mistakes (e.g., confusing a "no parking" sign for a

 refrigerator).

1.3.8 AI in Medicine

 AI is as good as or better than human doctors in diagnosing some diseases.

 AI applications in medicine:

o Detecting Alzheimer’s disease (Ding et al., 2018).

o Diagnosing cancer from medical images (Liu et al., 2017).

o LYNA system detects metastatic breast cancer with 99.6% accuracy.

 AI adoption is limited by trust, privacy, and ethical concerns.

1.3.9 AI in Climate Science

 AI helps analyze extreme weather events (Gordon Bell Prize, 2018).

 AI is used to fight climate change (Rolnick et al., 2019).

1.4 The Future of AI

 AI is not science fiction but is based on science, engineering, and mathematics.

 AI is growing in fields like robotics, automation, healthcare, and space exploration.

 Challenges include ethics, bias, data privacy, and human-AI collaboration.

Key Takeaways
✅ AI can be defined by how it thinks and acts (human-like or rational).
✅ AI has grown from simple problem-solving to deep learning and robotics.
✅ AI is used in robots, planning, translation, speech, games, medicine, and more.
✅ The future of AI involves better decision-making, automation, and ethical concerns.

Here are some important questions and answers from the given text that will help you
prepare for your exam:

1. Legged Locomotion

Q1: What is BigDog, and why is it significant in robotics?

A: BigDog is a quadruped robot developed by Raibert et al. (2008). It changed the
perception of robot movement by mimicking animal-like locomotion and could recover from
disturbances like slipping on ice.

Q2: What are some key features of the Atlas humanoid robot?
A: Atlas can walk on uneven terrain, jump onto boxes, and even perform backflips,
showcasing advanced mobility and balance in robotics.

2. Autonomous Planning & Scheduling

Q3: What was NASA’s Remote Agent program, and why was it important?
A: The Remote Agent program was the first autonomous planning system for spacecraft. It
could generate plans, monitor execution, detect, diagnose, and recover from problems,
reducing human intervention.

Q4: How does AI help in transportation and logistics?

A: AI-powered planning systems, such as DART (used during the Persian Gulf crisis), handled
massive logistical challenges by optimizing transport routes and scheduling operations. Uber
and Google Maps also use AI for real-time route planning.

3. Machine Translation

Q5: How has AI impacted machine translation?

A: AI-based translation tools, such as Google Translate, now support over 100 languages and
translate billions of words daily. For closely related languages with ample data, AI translation
approaches human-level accuracy.
4. Speech Recognition

Q6: What breakthrough did Microsoft achieve in speech recognition?

A: In 2017, Microsoft’s Conversational Speech Recognition System achieved a 5.1% word
error rate, matching human performance on transcribing telephone conversations.

Q7: How is AI used in voice assistants?

A: AI powers assistants like Alexa, Siri, and Google Assistant, enabling them to answer
questions, perform tasks, and even make phone calls, such as Google Duplex making
restaurant reservations.

5. Recommendation Systems

Q8: How do recommendation systems work, and where are they used?
A: AI-based recommendation systems analyze user behavior to suggest content (e.g.,
Amazon, Netflix, YouTube). AI is also used in spam filtering and auto-suggestions for email
replies.

6. Game Playing

Q9: How did AI defeat human champions in board games?

A: AI programs like Deep Blue (chess) and AlphaGo (Go) learned from human games and
improved through deep learning and reinforcement learning. AlphaZero, without human
input, mastered Go, chess, and shogi through self-play.

Q10: What other games has AI mastered?

A: AI has defeated human champions in Jeopardy, poker, Dota 2, StarCraft II, and Quake III,
demonstrating strategic thinking and adaptability.

7. Image Understanding

Q11: How does AI generate image descriptions?

A: AI models analyze images and generate captions like “A person riding a motorcycle on a
dirt road.” While impressive, some mistakes occur due to misinterpretation.

8. AI in Medicine

Q12: How is AI revolutionizing medical diagnosis?

A: AI matches or surpasses doctors in diagnosing diseases such as cancer, Alzheimer’s, and
skin diseases through image analysis. AI-human collaboration enhances accuracy.
Q13: What are some challenges in adopting AI in medicine?
A: AI adoption faces hurdles like ensuring transparency, reducing bias, and maintaining data
privacy. However, FDA approvals for medical AI applications are increasing.

9. AI in Climate Science

Q14: How is AI used in climate science?

A: AI models detect extreme weather patterns and analyze climate data, improving
prediction accuracy. AI-driven research has even exceeded exaop (10¹⁸ operations per
second) processing speeds.

These questions cover key topics and are formatted for easy recall during exams. Let me
know if you want additional questions or a more detailed explanation! 😊