The Old Debate Over 'Too Dangerous' AI Models
kimmonismus · x · 2026-07-18
The author revisits past debates over whether GPT-2, GPT-3.5, and GPT-4 were "too dangerous to release," pointing out that two years later, the world hasn't collapsed despite the emergence of more powerful models.
Using this to poke fun at the exaggerated narratives surrounding model release risks, the author notes that a new generation of models has since arrived, suggesting that the "too dangerous to release" arguments should be viewed with more skepticism in the face of reality.
Related event: Industry Reflects on AI Release Risks and Doomsday Predictions(4 posts)→
More from AGI Musings
- Gary Marcus says LLMs still cannot really do math on their own — GaryMarcus · 2026-07-22
- Gary Marcus says LLM math skills are like knowing only a car’s engine size — GaryMarcus · 2026-07-22
- AI may make digital work infinitely leveraged while offline life gets more human — illscience · 2026-07-22
- Better AI math could save researchers time by killing false conjectures earlier — prateekj · 2026-07-22
- AI’s economic forecasts are split by nearly a quadrillion dollars by 2035 — bittingthembits · 2026-07-22
- Open source is becoming tech’s soft power, says Kevin Xu — kevinsxu · 2026-07-22