1.4 — The two AI winters 🧊

🧊 The two AI winters. When promises outrun results, the funding freezes.

• 1st winter (1974-1980): after the scathing Lighthill Report (1973), which judged AI unable to scale. Mostly academic.
• Upturn (1980-1987): expert systems become a business — the market exceeds 1 billion dollars in 1986.
• 2nd winter (1987-1993): specialized hardware (the “LISP machines”, computers dedicated to the AI of that era) collapses against mainstream computers; the expert-system companies sink with it.

👉 The lesson still holds today: beware the “hype”. AI advances in waves of enthusiasm… and disillusionment.

🟢 In plain words — an “AI winter” is a moment when promises so far outrun results that the money (the funding) stops flowing and research slows down.

✅ In short

• AI went through two winters (1974-1980 and 1987-1993) when funding froze.
• Common cause: an excess of unkept promises.
• A reflex to keep: tell the “hype” apart from real results.