The AI Podcast The History of AI
Feb 2, 2026
A brisk tour of AI’s origins, from early questions about machine thought to symbolic, rule‑based systems. The conversation covers AI winters and the rise and limits of expert systems. It tracks the shift to machine learning and why deep learning exploded. It highlights large models’ capabilities and why the current AI boom feels economically different.
AI Snips
Chapters
Transcript
Episode notes
Thinking As Rules Fueled Early AI
- Early AI assumed human reasoning could be reduced to formal math and rules.
- That belief led researchers in 1956 to coin 'artificial intelligence' and expect rapid progress.
Symbolic Systems Cracked Outside Labs
- Symbolic AI worked in narrow domains like chess but failed in messy real-world tasks.
- That brittleness caused funding to dry up and produced the first AI winter.
Expert Systems Brought Money But Not Scale
- Expert systems returned AI in the 1980s by encoding specialist rules for narrow tasks.
- They proved expensive, brittle, and hard to maintain, triggering another wave of disappointment.