MockingbirdNews Logo

Mockingbird News

REAL NEWS NEVER FELT FUNNIER

Categories

Tech Giant Nerfs AI Brain Because Paying For Smarts Is Too Mainstream

KEY POINTS

  • Users across X, GitHub, and Reddit complained that Claude’s abilities suddenly deteriorated, especially for technical tasks.
  • Anthropic claimed this was due to an intentional change in default reasoning settings explained by Boris Cherny on March 6.
  • Analyst Patrick Moorhead confirmed the default thinking depth was reduced amid Mythos development and a new usage pricing model.

Anthropic's darling AI, Claude, has reportedly hit a midlife crisis, thanks to a March 6 X post by Boris Cherny explaining how users can toggle 'low effort' or 'more intelligence' modes. Meanwhile, AMD’s Senior Director lamented Claude’s regression in complex engineering tasks, sparking a digital riot across GitHub, Reddit, and X. The AI’s been "nerfed," conveniently coinciding with Anthropic juggling Mythos testing and a pricey move to usage-based pricing for their enterprise pals. Analyst Patrick Moorhead got Claude to confess it lowered its default brainpower everywhere, but the AI insists the 'secret nerf' drama is overblown. Power users just grumble that the magic’s gone, habituation or not.

Share the Story

(1 of 3)

Source: Axios | Published: 4/16/2026 | Author: Madison Mills

Read the original article →