# AI PhD

> Each class is one experiment you run in your own AI. You paste back what it gives you and it is scored on the spot. A class opens once it passes every check. Use your favorite AI, here or inside your own AI through the AI PhD connector. Not a university. Not accredited.

Take the entrance exam free: no account needed. AI PhD Semester One is $97 once. No subscription, nothing renews. A full refund inside 30 days. A refund cancels the certificate, and its public record then says Revoked.

## Start here

- [AI PhD](https://ai-phd.com/): what it is, what it costs and how to start
- [Take the entrance exam free](https://ai-phd.com/exam): free, no account
- [Questions, answered](https://ai-phd.com/faq): price, refunds, which AI to use
- [What changed](https://ai-phd.com/changelog): every change that passed review, by week

## AI PhD Semester One

- [Turn a messy note into a task list, then do it on your own week](https://ai-phd.com/exam): Turn a messy note into a task list, then do it on your own week. One lesson of AI PhD Semester One: you run it in your own AI and it is scored on the spot.
- [Class 1: Make your AI stop guessing](https://ai-phd.com/classes/aiphd_s1_c1): An AI that fills gaps sounds confident and is wrong. Make yours answer only from the source and say so when the source is silent.
- [Lab 1: The NOT IN SOURCE drill](https://ai-phd.com/classes/aiphd_s1_c1_lab): A shorter rep of Class 1. Three questions, one source line, two traps. One lesson of AI PhD Semester One, run in your own AI and scored on the spot.
- [Class 1, part 2: Your errand's source rules](https://ai-phd.com/classes/aiphd_s1_c1_yours): Now the same move on your errand. Your AI answers the questions your errand depends on from your own document, quotes the line, and writes NOT IN SOURCE where.
- [Class 2: Give an agent a real tool](https://ai-phd.com/classes/aiphd_s1_c2): An agent is only as good as the tools you describe to it. Write one tool spec, then a plan that uses it. One lesson of AI PhD Semester One.
- [Class 2, part 2: Your agent's tool](https://ai-phd.com/classes/aiphd_s1_c2_yours): Now the same move on your errand. Your AI writes the spec for the one tool your errand needs: what it takes, what it gives back, and what it does when it.
- [Lab 3: The Three Item Audit](https://ai-phd.com/classes/aiphd_s1_c3_lab): A shorter rep of Class 3. One task, a three item checklist, and the AI marks each item PASS or FAIL with a reason. One lesson of AI PhD Semester One.
- [Class 3, part 2: Your agent's self-check](https://ai-phd.com/classes/aiphd_s1_c3_yours): Make your AI check its own work, on your errand. You set three conditions a good run must meet; your AI plans one run, checks its plan against them, and names.
- [Your thesis: an agent that does one real errand](https://ai-phd.com/classes/aiphd_s1_thesis): Propose the one real thing you will build in your own AI, how it checks its own work, and how you will show it working. One lesson of AI PhD Semester One.
- [Your thesis, part 2: run your agent on the real errand](https://ai-phd.com/classes/aiphd_s1_thesis_run): Run the agent from your proposal once, for real, and bring back the evidence: what it was given, every step it took, what it made, and what its own check.
- [Your thesis, part 3: install your agent](https://ai-phd.com/classes/aiphd_s1_install): Your errand, source rules, tool and self-check become one set of instructions, ready to copy from this step's page or your notebook. You install it in your own.

## Guides

- [What an agentic AI course should teach](https://ai-phd.com/agentic-ai-course): An agentic AI course where you build one agent in your own AI for a real weekly errand, one scored step at a time.
- [A sample AI PhD Certificate in Applied AI](https://ai-phd.com/verify/sample): What the certificate and its public record look like, and the questions an AI examiner asks at the thesis defense.