AI Doesn't Have to Fail for the Bubble to Burst — Distillation and Cheap Models Could Do It
budputnal · reddit · 2026-09-22
A Reddit discussion argues the AI bubble doesn't require AI to fail — the business model could erode first.
Key points:
- Companies are spending billions, but if cheaper models do what most people need, why keep paying extra for slightly better ones?
- Distillation lets competitors train on your model's answers, and Chinese companies can cheaply buy American training data to close the gap — so how lasting is the lead those billions buy?
- Massive spending bets on future profits while competition keeps pushing prices down. What stops cheaper models from eating those profits?
More from AGI Musings
- Rishi Bommasani co-founds nonprofit STABL with SayashK and Arvind Narayanan — RishiBommasani · 2026-11-12
- Investor: AI diffusion will be slowest in consumer software — Oura, Strava, Labcorp still have no real AI strategy — MartinGTobias · 2026-11-12
- On AI Progress: When Models Lag on Your Load-Bearing Skill, You Can't See Them Improving — dfrsrchtwts · 2026-11-12
- Slack's David Hoang: agents beat people on coordination, proposes 'boundary agents' — davidhoang · 2026-11-12
- CS student mourns coding flow states he'll never experience in the AI era — chrissyykat · 2026-11-12
- Ford's 370% Turnover Led to $5 Days — But ASI Never Quits — krishnan · 2026-11-12