Learning from History: The Past as a Compass for Systems Engineers in the Age of AI
summary
A historical arc links early rule‑based systems and Deep Blue to modern transformers and agentic workflows while claiming practical evidence of rapid capability gains: GPQA 60%→84% with Gemini 25, coding benchmarks near 98% from ~70% last year, and SWEBench 2%→~60% in one year. Notes: a 1998 Bell Labs neural‑net for handwritten digits and Siri origins appear, Deep Blue is rule‑based, TD Gammon used nets, and core gap is abductive reasoning—framed as an alien thinking machine that requires new epistemology rather than AGI debates.
AI-written summary · about
chapters (9)
- 0:00Introduction and roast
- 1:07Why historical perspective matters
- 3:57AI history timeline and patterns
- 6:01Benchmarking AI progress
- 9:11AI reasoning models leap forward
- 10:38Philosophy of AI and alien thinking
- 16:59Reasoning types in AI models
- 18:26Walter Pitts and neural networks
- 20:58Q&A and closing
related talks
15mMore than 70 years of AI history in 15 minutes!
46mAI vs Humans Rewriting the Rules of Success
47mThe Past as a Compass for Systems Engineers in the Age of AI
54mAI vs Humans: Rewriting the Rules of Success
42mDawn of the planet of artificial intelligence
41mArtificial Intelligence through the ages
47mKeynote: AI is having its moment ... again
12mThe History of AI - what can we learn from the past?