Core Materials

• Harvard Business Publishing Cases – 18 Cases in Course Pack (~ $100)

  February 3, 2025 Case selection is finalized. Go to this link to purchase from Harvard Business Publishing.

• Open Source Software (Free!)

  • Transformers
  • LangChain
  • Gradio

• Other readings, if applicable, will be provided on an as needed basis through canvas

Optional Readings

1. Economics of AI
   • Prediction Machines (Agarwal, Goldfarb, and Gans 2018) & Power and Prediction (Agarwal, Goldfarb, and Gans 2020) — (Now) classic books on how AI affects business at a fundamental economic and strategic level. Good introduction to the Economics of AI.
   • Competing in the Age of AI (Iansiti and Lakhani 2020) — B-School profs discuss how AI-driven processes are reshaping business operations, removing traditional constraints on scale, scope, and learning, and necessitating a fundamental rethinking of business and operating models to remain competitive.
   • Human + Machine (Dougherty and Wilson 2024) — Consultants perspective on how artificial intelligence (AI) is transforming business processes by fostering collaboration between humans and machines
   • Co-Intelligence (Mollick 2024) — Ethan’s a Wharton Prof and leading thinker on AI applications in the world. See also Ethan’s substack One Useful Thing and HBR Column Inspiring Minds
2. Recent AI Industry War Stories
   • Genius Makers: The Mavericks Who Brought AI to Google, Facebook, and the World (Metz 2021) — Stories behind Hinton, Le Cun, Bengio, and other thinkers that created the technologies that power AI today.
   • The Worlds I See: Curiosity, Exploration, and Discovery at the Dawn of AI (Li 2023) — Autobiography of Fei-Fei, creator of image net and luminary in the field.
   • The Nvidia Way: Jensen Huang and the Making of a Tech Giant (Kim 2024)
     • DNY January 6, 2025: I’m just starting to read this! Will update if I recommend.
3. Artificial Intelligence Textbooks
   1. Artificial Intelligence: A Modern Approach, 4th US ed. (Russell and Norvig 2020) — available online
      • Good for historical framing on AI Technologies, not as machine-learning focused.
   2. Deep Learning (Goodfellow et al 2016) — available online
      • a bit outdated, but good primer on the foundations of ML that led to Deep Learning
   3. Machine Learning: A Probabilistic Perspective (Murphy 2012) — google books
      • Introduces a Bayesian perspective on statistical learning, helpful for framing neural nets in the context of other approaches. See also Pattern Recognition and Machine Learning (Bishop 2006, amazon) for a similar perspective. Classic textbook, very influential, probably not worth reading anymore
      • There’s a more recent re-written version: Probabilistic Machine Learning: An Introduction (Murphy 2022) — seems to do the same thing but updated based on advances in the 2010s. I haven’t read it though.
4. Open Source Software
   • **Working in Public: The Making and Maintenance of Open Source Software** (Eghbal 2020)
5. Relevant Blogs
   • OpenAI / Anthropic / Google AI / DeepMind / Hugging Face, among others — leading AI groups
   • Epoch AI — research group focused on AI Trends
   • Ethan Mollick’s One Useful Thing — Wharton Prof, AI Scholar
   • Jack Clark’s Import AI — blog by a co-founder of Anthropic
   • Simon Willison — former square engineer, turned open source developer.
   • Latent Space — newsletter targeting “AI Engineers”, seem to do a good job covering recent developments in AI literature. More technical.

Computing Infrastructure

AI is compute-intensive, and we will build AI applications in this course that will benefit from better compute. Access to a computer equipped modern browser technology (i.e., the latest version of Chrome, Safari, or Firefox) is a must.

Nevertheless, this course is accessible to those without top-end computers. In particular, this course will take three steps to ensure equitable access to technology for all students:

• All technical assignments will be distributed via Google Colab, which can be executed for free by students.

• The course will focus on teaching AI through the lens of open-source software, which is freely accessible.

• The one exception is the use of OpenAI’s API. Instead, the course instructor will provide a limited budget for OpenAI token credits. Each student will be provided an OpenAI API key upon request.

  Details on shared API keys will be released before January 16, 2025, in time for the first technical assignment.

More technically inclined students are free to explore alternative ways to implement AI technologies, not limited to:

• Using Managed API Services from other providers like Google or Anthropic
• Locally running an Open Source LLMs on your own machine or on GPUs provided through Colab.