Welcome to Introduction to AI Applications!

Lecture 01: Introduction

Danilo Freire
danilo.freire@emory.edu

Department of Data and Decision Sciences
Emory University

Welcome to DATASCI 185! 🎉

Lecture overview

Today’s agenda

  • Hello and welcome!
  • Instructor and TAs
  • Motivation and course overview
  • Assignments and grading
  • Course logistics

Course materials

Course repository: https://github.com/danilofreire/datasci185

Course website: https://danilofreire.github.io/datasci185

All course materials, including lectures, code, assignments, and project guidelines, are available on our GitHub repository and course website

We will use Canvas for course administration, including submitting assignments, accessing grades, and receiving announcements

Please take some time to get to know both platforms, and reach out if you have any questions

Note

Please remember to check the course repository regularly for updates and announcements!

Nice to meet you! 😊

Instructor

A bit about me

Danilo Freire

Visiting Assistant Professor in the Department of Data and Decision Sciences

MA from the Graduate Institute Geneva, PhD from King’s College London, Postdoc at Brown University, Senior Lecturer at the University of Lincoln, UK

Research interests: computational social science, experimental methods, policy evaluation, political violence, organised crime

What about you? (time permitting!)

  • Now it’s your turn! 😉

  • Please introduce yourself! 👋

  • Tell us your name, your major, and what would you like to learn in this course!

My teaching philosophy

  • I love teaching and aim to make learning fun
  • Classes where students participate are the best!
  • Hands-on activities help you learn better
  • I am always available to help and answer questions. And I mean it!
  • Your feedback helps me improve my teaching. Please let me know what is working and what is not 😉

Teaching assistants

  • We have four amazing TAs with us this semester!

  • Tom Suo and Sissi Li will be answering questions during our lectures and holding office hours. You can reach out to them at and

  • Philip Wang and Anita Osuri will also be grading your assignments and quizzes (with my oversight). Their email addresses are and

  • We are all here to help you! So feel free to ask questions during class, office hours, or via email 😃

Office hours

What for and what not for

  • What office hours are meant for:
    • Applying tools in practice
    • Discussion of issues related to the assignments
    • Boosting your knowledge of AI and data science more generally
  • What these sessions are not meant for:
    • Summarising lecture content
    • Solving the assignments for you

Class etiquette

  • Learning a new topic can be tough and push you out of your comfort zone. If the course pace is too fast, let us know. I expect your commitment, but I do not want anyone to fail
  • You are all keen on AI, but your backgrounds may vary. That is great! Some sessions might be more engaging than others. If you are bored, help others or explore additional resources (we have plenty!)
  • Please be respectful to each other
  • Ask questions whenever you need to!

Learning objectives

By the end of this course, you will be able to:

  • Explain the main ideas behind contemporary AI systems in plain language

  • Identify common failure modes of AI systems and the data issues that cause them

  • Read and assess claims about AI in news articles, product pages and policy documents

  • Design a small, realistic plan for an AI application, including data needs, evaluation, and a basic harm-mitigation strategy

  • Reflect critically on ethical, legal and social questions raised by AI deployment

Course overview

Module Topic Key Questions
0 Orientation What is AI? What will we learn?
1 AI Design How are AI systems built?
2 Perception How do machines read, write, see and hear?
3 RAG & Pipelines How do we make AI reliable?
4 Ethics & Bias What can go wrong?
5 Policy & Impact How is AI regulated?
6 Applications Real-world uses and limits

What makes this course unique

  • We’re starting a new AI track in the Data Science curriculum! 🚀

  • This is the first offering of DATASCI 185 — help us make it great! 🙌

  • Why is this course different from others?

    • Non-technical approach: No heavy maths or programming required
    • Critical perspective: Learn to ask the right questions about AI
    • Real-world focus: Learn how AI is applied in practice
    • For all backgrounds: Whether you’re a humanities or STEM student, this course is for you!

Logistics

Course information

  • Syllabus: Available on our course repository and website. The course is designed to be self-contained. The syllabus includes links to the slides we will use in class, along with recommended readings, and problem sets. I will upload slides throughout the term as we progress

  • Schedule: Lectures are on Mondays and Wednesdays from 4:00 to 4:50 pm

  • Office Hours: We’re always happy to help. Please reach out to the TAs or me via email to schedule a time

  • Materials (recap): All course materials will be available on:

Assignments

How you will be graded

  • Problem sets: Ten of them, due on Wednesdays at 11:59 pm (50%)
  • In-class quizzes: Five of them (30%)
  • Final project: Due on the last day of class (20%)
  • Late policy: 10% off per day late
  • Collaboration: You can discuss assignments with your classmates, but you must write your own code and submit your own work. AI is allowed, but you must disclose its use and ensure you understand the output
  • Academic integrity: Please refer to the syllabus for the university’s policy on academic integrity

Motivation:
What the heck is AI anyway? 🤖

What is AI?

  • A simple definition: Artificial Intelligence (AI) is a branch of computer science focused on creating systems capable of performing tasks that typically require human intelligence
  • These tasks include learning, reasoning, problem-solving, perception, language understanding, and many more!
  • AI can be classified into three main categories:
    • Narrow AI: Designed for specific tasks (e.g., virtual assistants, recommendation systems)
    • General AI: Hypothetical systems that possess human-like intelligence across a wide range of tasks
    • Superintelligent AI: Surpasses human intelligence in all aspects (still theoretical)
  • AI is powered by various techniques, including machine learning, deep learning, and natural language processing

AI taxonomy

Source: McKinsey & Company (2024)

Why should you care about AI?

  • AI already impacts many aspects of our daily lives:
    • Virtual assistants (e.g., Siri, Alexa)
    • Recommendation systems (e.g., Netflix, Amazon)
    • Dating apps (e.g., Tinder, Bumble)
    • Social media algorithms (e.g., Facebook, Instagram)
  • AI is transforming industries:
    • Healthcare (e.g., diagnostics, personalised medicine)
    • Finance (e.g., fraud detection, algorithmic trading)
    • Transportation (e.g., autonomous vehicles, route optimisation)
    • Education (e.g., personalised learning, 24/7 tutoring)
  • If the predictions are correct, AI will be one of, if not the most, important technological advancements of our time

Key questions we’ll explore

  • What can AI actually do vs. what’s hype?
    • Can AI truly think or is it just pattern matching?
    • Why does ChatGPT sometimes give wrong answers?
    • What can and cannot be automated?
  • Why do AI systems fail?
    • What happens when training data doesn’t represent everyone?
  • How does data affect AI behaviour?
    • Garbage in, garbage out
    • Who decides what counts as “correct” data?
  • What are the ethical implications?
    • Should AI make life-or-death decisions?
    • Who’s responsible when AI makes a mistake?
  • How should we regulate AI?
    • Innovation vs. safety trade-offs
    • What role should civil society play in governing AI?

AI benchmarks and human performance

Source: Artificial Intelligence Index Report (2025)

True or false?

  • AI understands language like humans do
    • False! AI predicts likely next tokens (what are they?)
  • AI will replace all jobs
    • Probably not. AI transforms jobs more than it eliminates them
  • AI hallucinations are easy to spot
    • False! AI sometimes confidently states incorrect information, and it’s not always obvious
  • We can test AI systems by deliberately trying to trick them
    • True! This is called adversarial testing or red-teaming
  • AI systems that learn without any human labels exist
    • True! This is called unsupervised learning, and it finds hidden patterns automatically
  • AI systems can explain why they made a particular decision
    • False! Many AI systems are “black boxes”, that is, we don’t know for sure how they came up with their answers
  • The order of words in a sentence affects how AI processes it
    • True! Attention mechanisms let models weigh the importance of different words
  • The same AI model can be fine-tuned for many different tasks
    • True! This is one of the most powerful aspects of modern AI

Next class

  • Topic: A brief history of AI and the recent shift
  • What we’ll cover:
    • From symbolic approaches to data-driven learning
    • The transformer architecture explained in plain language
    • How we got from early AI to ChatGPT
  • Main takeaway: Understanding how AI evolved helps us understand where it’s going
  • Preparation: Read the assigned articles on AI history and transformers

Suggested video: The Thinking Game

… and that’s all for today! 🎉

Thank you very much for your attention! 🙏🏻

Have a great day! 😊