Benchmarks are meaningless, the differences between models are just vibes, dev argues

gnukeith · x · 2026-09-22

Developer gnukeith rants that benchmarks are "absolutely meaningless," that the scale-of-intelligence narrative is pure hype and marketing, and that the only real difference between models is "vibe" — while joking that nobody, except perhaps Google itself, knows what Google is doing. The take captures growing fatigue with leaderboard-driven model evaluation.

Related event: Dev claims benchmarks are meaningless, only 'vibes' separate models(2 posts)→

Original post →

More from Models

Models channel →