Getting everything running

TL;DR#

We got through what this course is for, why an agent is good at writing code, and what a prompt actually costs. We ran out of room in the setup block, so most people leave with a repository and a hello.py rather than a notebook. Every step of that setup is now written down at /guides/getting-started/, and working through it at your own pace before next Tuesday is the whole of what you owe.

What we covered#

A chatbot answers, an agent finishes. Both definitions came out of the room. The piece worth keeping is the loop: a model that calls a tool, observes the result, and reasons again until the goal is met. Observation is the last step, and it is why agents are good at code specifically. Code can be verified. You can run it and find out. The same reason explains why mathematics is moving fast right now and why work whose results are hard to check is harder to hand to an agent.

A prompt is bigger than what you typed. Four words went into a live agent and over a thousand tokens went out. The rest is the system prompt, plus everything already said in that session. Two consequences. The Babson API budget is spent faster by re-running a failing cell twenty times than by asking twenty good questions. And a long session carries its whole history with every message, so starting a fresh one when you change topic is free and worth doing.

SQL is not a query in Excel. Excel stops a little over a million rows and is poor at relating separate things to each other. A database does both, and SQL is the language for asking it questions. The part that matters here is that SQL is text. A model trained on language works well with a language you type and poorly with an interface you drag.

What the course grades. You may use an agent for the analysis. You are accountable for what you hand in, at the level of a block or a function rather than a line: what does this do, and how do the pieces fit. Every assignment states which mode applies. There are no exams.

Git. A repository is a folder for one project. A commit saves a set of changes with a message describing them. A push sends those commits to GitHub, which is the copy other people can see.

The setup steps#

The setup block moved faster than it should have. Everything it covered is now a numbered page you can walk alone: /guides/getting-started/.

These stopped people, and none of them means anything is wrong with your machine.

• A terminal asking for your password shows nothing at all while you type it. No dots, no moving cursor. Your keyboard is not broken. Type it and press Enter. The installer should not need a password at all, because it installs into your own home folder. If it asked, sudo was in the command. Take it out, paste the line exactly as the guide has it, and it will not ask again.
• uv is not found right after installing it. A terminal reads the list of programs when it opens, so one that was already open has an old list. Close it and open a new one.
• macOS may warn that .local/bin is not on your PATH. Same fix, a new terminal.
• Coloured text in a terminal is not an error. PowerShell in particular greys out text it is predicting for you. What matters is whether the words Error or Traceback appear.

If something still fails after that, send the error itself rather than a description of it. Select the text, copy it, and paste it into an email. A screenshot works too.

What we did not get to#

The Northwind example was named and then set aside, along with the notebook. We pick both up next session. Nothing there is assumed knowledge.

Public links#

• The MIT report on AI in teaching, learning and research
• What employers actually asked for, the job-posting analysis from the session, which gets updated through the term
• uv installation
• Getting Started, where the setup steps now live

Before next session#

1. Work through /guides/getting-started/ to the end. Stop wherever it breaks and bring the error.
2. You need two things in your oim7510 repository: a README.md with something of yours in it, and a hello.py that runs. Both committed and pushed.
3. Finish the pre-course survey if you have not, and register your GitHub account with your Babson address.
4. Watch for your Babson AI key. Updated 9/14: there is nothing to apply for. Babson's AI team emails you the key, and Your Babson AI Key has the setup.

Checkpoint 1 closes Sunday 9/20 and covers the first two sessions. Each of its items is simply present or absent. Tonight gets you two of them: a repository that exists, and more than one commit in it. The notebook and the agent setup come from next Tuesday, so there is nothing to chase on those yet.