AI

I teach professionals how to put generative AI to work, I build AI tools for my own courses, and I study what AI does to firms and their people.

Students often ask me where to start with generative AI. This is the path I recommend, in six short parts. Each part will appear here as it is ready.

1. How language models work. Tokens, next-word prediction, and where models fail without warning.
2. Using them well. Giving context and a clear brief instead of a one-line prompt, and choosing the right tool for the task.
3. Building with generative AI. Calling a model from code, small apps, and answering questions from your own documents.
4. Coding agents. What tools like Claude Code do when they work on your own files, how to set them up safely, and why they are useful to non-programmers too.
5. Verification and evaluation. Checking outputs, building a small test set, and measuring before you trust a result.
6. AI in organizations. Tasks rather than jobs, automation versus amplification, and what changes for people's careers.