Burkov skew AI hype: 'deterministic LLMs' and 'first agents' are old tricks rebranded
burkov · x · 2026-09-23
Andriy Burkov satirizes AI product hype: removing sampling randomness becomes a 'deterministic LLM for regulated industries', a Playwright-driven browser LLM becomes the 'world's first agent', and majority-voting across LLMs is repackaged as 'The Ensemble' — with his proof that 51%-accurate uncorrelated models push accuracy toward 100%.
More from Fun
- Economist blasts academic fraud culture: defending obvious cheats like Ariely and Gino hurts honest researchers — Afinetheorem · 2026-09-23
- Two-year-old AI video aesthetics are now impossible to recreate with latest models — joshgonsalves_ · 2026-09-23
- The 'Claude Factory' meme: AI circles' absurdist joke about model personification — nabla_theta · 2026-09-23
- 'Remember tab complete?' How fast AI coding paradigms have moved — yoobinray · 2026-09-23
- Opus 4.7 vanishes from Claude's model selector, sparking attachment drama — repligate · 2026-09-23
- tszzl: glossy 3D animation is now the single most important skill for launching a new AI model — tszzl · 2026-09-23