TL;DRAI history runs in waves — booms of hype followed by "AI winters" of cut funding.
The field has cycled: 1950s-60s optimism, the first winter when promises outran results, 1980s expert-systems boom and bust, then a long quiet until data + GPUs revived neural networks (~2012, AlexNet). Each winter followed over-promising. Knowing this tempers today's hype — and explains why practitioners distinguish genuine capability from marketing.
Key points
Repeated boom/bust cycles ("AI winters")
Winters followed over-promising
2012 deep-learning revival (data + GPUs)
History counsels measured expectations
Common mistakes
Assuming progress is monotonic
Taking every hype cycle at face value
Forgetting compute/data, not just ideas, drove revival
Try it: Name what triggered the AI winters and what ended the last one.
Example code
<!doctype html><html><head><meta charset="utf-8"></head>
<body style="background:#06040d;color:#e6e0ff;font-family:monospace;padding:20px"><pre>hype → over-promise → winter (funding cut)
2012: data + GPUs → deep learning revival</pre></body></html>