Path
AI Foundations
Mental models through labeling, prompting, Claude & Claude Code, cloud frontier models, Bedrock, and personal vs agentic workflows. 22 lessons · about 320 minutes total.
- 01What “AI” usually means today
- 02What generative AI is
- 03How modern AI systems feel from the outside
- 04Large language models in plain language
- 05Why vectors and embeddings
- 06Transformers in plain language
- 07How transformers work
- 08How a model is built (high level)
- 09GPUs vs CPUs — why GPUs for model compute
- 10How data is labeled
- 11Evaluating AI outputs
- 12Prompting as clear communication
- 13Prompt engineering in practice
- 14Limits, risks, and responsible use
- 15A simple personal AI workflow
- 16The tool landscape (map, not tour)
- 17Agentic workflows (the loop)
- 18How clouds serve frontier models
- 19Claude basics
- 20Claude Code basics
- 21AWS Bedrock basics
- 22Where to go next