Using Generative AI to Strengthen & Accelerate Learning
summary
Shows transformers as the mechanism behind generative AI, demonstrates a Sunno demo that turns a paper title into musiclike output, and contrasts engines by example: a Greg Brockman notebook image yields working website code from ChatGPT while Claude initially refuses then concedes. Notes token limits — Claude handling ~200k tokens allowed analysis of a 100k‑word book and produced a usable forward in ~20 seconds. Covers hallucination risk, tools like WriteSonic, Semantic Scholar, MathGPT/Wolfram links, and when AI accelerates retrieval versus when deliberate practice is still required.
AI-written summary · about
chapters (10)
- 0:00Intro
- 3:03What does ChatGPT do?
- 12:32The value of metaphor
- 21:357 int. companies with their foundational LLMs
- 26:25Large language models
- 34:33Cheating or "learning lightly"
- 36:11The Flynn Effect
- 39:08How to think about the new complex world
- 46:45Good books on GenAI
- 48:30Outro
related talks
17mUnlock the power of Generative AI: From understanding its principles to utilizing its power
32mHow generative AI is Changing the way people interact with computers
37mGenerative AI: Beyond (Just) Hype
15mHarnessing the Power of Generative AI for Intelligent Applications
35mGenerative AI in the real world
59mKeynote: The past, present, and future of AI for application developers
45mBringing the Power of AI to Your Application
41mThe Unreliable Computer Revolution - Embracing the Double-Edged Sword of GenAI