Claude Opus 5 backlash signals potential end of AI scaling laws

gerardsans · x · 2026-08-20

Users report that despite higher benchmark scores, Anthropic's Claude Opus 5 performs poorly in daily use, exhibiting issues like ignoring instructions, stopping mid-task, and degraded reasoning. This has led to a wave of subscription cancellations and sparked doubts about the effectiveness of pure scaling strategies.

Related event: Claude Users Report Sharp Quality Drop, Anthropic Scrambles to Respond(2 posts)→

Original post →

More from AGI Musings

AGI Musings channel →