Anthropic claims 2x R&D speedup, yet stays tied with OpenAI for a year
adam_karvonen · x · 2026-11-21
AI researcher Adam Karvonen questions why Anthropic's reported 2x AI R&D acceleration hasn't produced any visible lead: from November 2025 to November 2026, Opus and GPT models remain roughly equal in capability, with Claude only better at understanding intent. He notes the two frontier labs have been tied for a full year and asks why cracking AI R&D hasn't translated into noticeable advantage.
More from Companies & People
- Intel Fellow publicly calls out Google over employee's threatening AI-anger tweet — JFPuget · 2026-11-21
- From -₹1 crore debt to Google and a ₹2 crore side business: a founder's story — RajaswaPatil · 2026-11-21
- Nvidia down 39% in six months as shareholder slams board and calls to reinstate Jensen Huang — LinusEkenstam · 2026-11-21
- If Anthropic cracked AI R&D with 2x speedup, why is it still tied with OpenAI? — adam_karvonen · 2026-11-21
- VC lists six mistakes by Nvidia's board that contributed to a 40% stock decline — garrytan · 2026-11-21
- Musk says Ashlee Vance's 10,000-word Neuralink feature is 'probably 80% accurate' — elonmusk · 2026-11-21