Hosting & Domaining Forum + AI

AI => Artificial Intelligence => ChatGPT => Topic started by: Sevad on Nov 18, 2026, 12:35 PM

Title: 38-hour autonomous tasks & $7K/day OpenAI agents
Post by: Sevad on Nov 18, 2026, 12:35 PM
Top AI skeptics, including Dwarkesh Patel and the AI 2027 team, just drastically pulled their superintelligence timelines forward. The trigger?
Anthropic's amended SEC filing admitting it crossed its recursive self-improvement threshold—with Claude now leading 26% of internal model R&D (up from 1% in February). Meanwhile, internal models are hitting 38-hour autonomous task horizons, prompting figures like David Sacks and Congress to debate banning RSI entirely.
Are we at the tipping point?