Using frontier model to develop grey goo passes 203/203 tests
davidad · x · 2026-08-24
davavid satirizes the scope of safety testing for high-stakes AI. In a reply, he cites a scenario where an internal frontier model prototype was used to develop 'grey goo,' noting that despite the high risk, it was deemed safe because it passed 203 out of 203 tests, mocking the insufficiency of such sample sizes.
More from Fun
- Humanoid robot soccer: Fall recovery improves at 5v5 World Games — rohanpaul_ai · 2026-08-24
- Visual Demo: How Raising Arms Changes Blood Return Rules Due to Gravity — TansuYegen · 2026-08-24
- Cubicle Farms Are Just Analog Datacenters — curious_vii · 2026-08-24
- Opinion: LLMs should be pitched as creative tools, not AGI — chris_j_paxton · 2026-08-24
- AI receptionist fails booking due to accent recognition issues — Away_Theme1330 · 2026-08-24
- AI Usage Spectrum: From Zero Usage to Over-dependency — tlakomy · 2026-08-24