Anthropic eyes AGI in 6-12 months

Anthropic's pivot to an aggressive timeline signals the end of the "safety-only" era and the start of an AGI arms race.

• –Dario Amodei's Davos 2026 warning of a "white-collar bloodbath" within 12 months adds high-level credibility to these rumors
• –The "Claude Mythos" model reportedly doubles internal scaling expectations, potentially automating the AI research loop itself
• –Achieving AGI-level software engineering within a year would collapse current developer roadmaps and infrastructure planning
• –Public communications to government bodies (OSTP) still hedge with a "late 2026" date, suggesting internal breakthroughs are outpacing official disclosures