#LLM
Curated coverage on «LLM»: releases, research, benchmarks, open-source.
Musk v. Altman: $150 Billion Trial Over Control of OpenAI Opens in the US
A high-profile US trial began on April 27: Musk is seeking Altman's removal, up to $150 billion in compensation and an end to OpenAI's conversion into a commercial…
AlphaGo creator launches unicorn startup to build AI "superlearners"
David Silver, whose AlphaGo beat the world Go champion, calls LLMs a dead end and is launching a billion-dollar company to build a new kind of AI.
OpenAI releases GPT-5.5, its smartest model yet for code and multistep tasks
OpenAI has unveiled GPT-5.5, a model that can plan and carry out complex multistep tasks on its own: debugging code, researching data and working across several…
$1 Billion for 12 People: Why Yann LeCun Is Rejecting Language Models
Why DeepSeek Is Delaying V4: A Forced Switch to Chinese Chips
OpenAI unveils Privacy Filter, an open model for stripping personal data from text
OpenAI Offers $25,000 to Break GPT-5.5's Biological Safeguards
OpenAI Turns ChatGPT Into an Operating System With AI Agents
OpenAI has announced autonomous workspace agents in ChatGPT. They can link tools, automate routine work and reshape how enterprise teams operate.
How OpenAI Is Speeding Up AI Agents with WebSockets
OpenAI has updated its Responses API with WebSockets and connection-level caching. The change sharply cuts latency and paves the way for fast, autonomous…
Elastic Memory for AI: How kvcached Tackles the GPU Shortage
xAI's voice model beats GPT Realtime on business tasks
GitNexus: A New Tool Teaches AI Agents to Understand Code Structure
DeepSeek-V4: How New Compression Algorithms Made a One-Million-Token Context Real
Why SEO Won't Die — and What's Really Behind the GEO Hype
Marketers are panicking: LLMs are eating traffic and nobody clicks links. Here's why SEO isn't going anywhere and what every GEO article gets wrong.
LangChain drops reasoning content from CoT models: how to fix the LLM provider bug
LangChain's chat classes don't keep the reasoning block from CoT models, so users wait in silence and developers have to fix it by hand.