On Curiosity
summary
Argues curiosity is the engine that pulls future ideas into the present and demonstrates it with concrete demos: a 2020 GPT‑3 prototype constrained by a 2,000‑token context and ~$75 per million tokens feeding a JSX‑in‑browser compiler; a 2021 experiment that builds a custom HTML parser to compress pages so a model can attempt to buy AirPods (it sporadically fails, confusing terms of service); and a recent Gemini 2.5 Pro video feedback demo that grades basketball and running form and produces coach‑level advice.
AI-written summary · about
chapters (8)
- 0:00The power of curiosity
- 1:34Early explorations in AI
- 4:35Testing AI agency
- 5:35AI as an excavation tool
- 6:33Learning through unconventional paths
- 8:18Navigating the current AI landscape
- 10:14A moral obligation to innovate
- 11:41Live demos and archival examples
related talks
23mIntelligence as a Material
16m#BuildPropulsionLab at #DevoxxUK: @jbaruch with @codepo8
42mUX+AI: Mastering the UX of AI-Powered Products
44mBuilding experiences for the superhuman future
48mIgnite Sessions
24mHow Agentic AI Empowers Creative Thinking
45mFrom Click to Conversation: Let's Replace Buttons and Forms with an LLM!
45mContent First Prototyping