Anthropic News - Page 4 of 42
825 articles • Page 4 of 42
AMD's Instella-MoE-16B Hits 12.7% Speedup With New FarSkip Training Technique
AMD put out Instella-MoE-16B-A3B this week, a Mixture-of-Experts language model built from the ground up on its own Instinct MI300X and MI325X GPUs.
OpenCode adds plugins for citations, source lists and multi‑search support
OpenCode is putting in the plumbing. Three new plugins are turning it from a flashy chatbot into a tool with receipts.
AI Researchers Reveal Token Warehousing Strategy to Cut GPU Computational Waste
The math is brutal. When every GPU inference cycle kicks off with a redundant prefill, recalculating the same tokens from scratch, the waste...
How Ollama’s Context Length Setting Impacts Local Model Memory
Picture this: You load an entire codebase into a local model, expecting an insightful analysis.
AI Researchers Resign, Bots Hire Humans, Anthropic Targeted, Evie Party
The AI industry is eating itself, and the menu is full of irony. Researchers resign in protest, only for their creations to turn around and hire the...
Anthropic embeds Claude AI in Excel, eyes finance market against Copilot
Every AI company wants to sell to banks. Anthropic, tired of waiting, just parked its truck in Microsoft’s driveway.
OpenAI Slows AI Model Releases, Details Safety Shortcomings
OpenAI has spent the past several weeks in crisis mode. According to current and former employees who spoke to WIRED on condition of anonymity, the...
Nonprofit Finds No AI Labs Fully Apply Basic Safety Controls
Anthropic got a C+. So did OpenAI. Nobody else came close. That's the grading from Guidelight, a nonprofit that just published its first scorecard on...
Anthropic's Sonnet 4.6 hits 79.6% on SWE-bench, costs one‑fifth of Opus
The math has shifted. Anthropic quietly released a new model that doesn’t just close the gap, it erases it in key categories.
Anthropic CEO Dario Amodei returns to Pentagon talks to salvage deal
The air in the Pentagon’s negotiating rooms has rarely been this charged, or this fragile.
Chatbots can draft a book by lunch; a third of authors now use AI
Every book you read this year might have been written by a machine before its author finished their coffee.
OpenAI yields to Pentagon, bans bulk U.S. data; Amodei says law not yet
The Pentagon wants OpenAI’s power, but not the moral headache that comes with it.
M-GRPO Boosts Coordination in Multi-Agent Training Over Single-Agent GRPO
Training one AI is straightforward enough. Training five of them to work together is a mess.
Editor Urges Writers to 'Be More Human' as Anti-AI Literary Push Grows
Laura Brooke Robson started writing her novel "Love Is an Algorithm" last year with a strange new worry: sounding too much like a chatbot.
Elgato adds MCP support in Stream Deck 7.4 update, enabling new trigger method
The Stream Deck just learned to listen. Elgato’s 7.4 update introduces MCP support, a quiet but profound shift in how you can talk to your hardware.
Pentagon launches GenAI.mil with Google, says multi-vendor strategy
The Pentagon's new AI platform, GenAI.mil, is now live. Its first model is Google's Gemini. A simple, safe pick? Look closer.
Anthropic Overtakes OpenAI in Quarterly Revenue as Costs Hit Margins
Anthropic closed the second quarter with more revenue than OpenAI, according to figures the two companies gave investors and reported by the Wall...
AI Jailbreak Findings Challenge Industry Self-Regulation
FAR.AI, an AI safety nonprofit based in California, spent weeks testing how easily seven frontier AI models could be talked out of their own safety...
Anthropic launches think tank as it battles Pentagon blacklist
Tension is mounting inside Anthropic, but the AI company isn’t slowing down. As it fights to get off the Pentagon’s blacklist, it’s spinning up a new...
Claude gains shared context in Excel, PowerPoint; Microsoft adds Copilot Cowork
Your spreadsheet just whispered to your slide deck. On Monday, Anthropic announced its Claude AI can now maintain shared context between Microsoft...