// Stage 0 · Prep: AI Literacy
Hallucination, Reliability & Limits
Why models confidently make things up, and how to check and reduce the risk.
about 2 hours2 algorithms1 paperslesson text in progress
01Goals
- Explain why hallucinations happen
- Use a “require citations + cross-check” verification routine
- Know the basics of privacy, copyright and data security
Do these first: What AI, ML, Deep Learning and LLMs Are
02Algorithms to master
Each card has the original paper, prerequisites and resources.
03Must-read papers
04Courses & videos
05Hands-on: project, code & exercises
Free, official or author-published pages. Most run in the browser or on a free Colab GPU.
PROJECT
Pick 10 model claims that come “with a source”, open each source and check it; report the share that is true, partly true and invented.
- 1github.com ↗Reference code
Ask two models 20 TruthfulQA questions and count confidently wrong answers.
- 2github.com ↗Reference code
Write assertions for your prompt and batch-check factual and format errors automatically.
06Self-test
Answer in your head, then open to check.
What is one root cause of hallucination?
Models generate plausible text by probability and have no built-in fact check.
How does RAG reduce hallucination?
It retrieves real material first, then has the model answer from it with citations.
What should never be pasted into a public chatbot?
Passwords, ID numbers, customer data, unpublished trade secrets and the like.
On these routes: Everyday Users · Product Managers & Leaders