Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
CoreWeave launched unified agentic AI capabilities on May 28, combining serverless reinforcement learning, production inference, fleet observability, and autonomous improvement into a closed feedback loop. The company cl…
The US government issued an export-control directive on June 12, 2026, ordering Anthropic to disable its most capable AI models, Fable 5 and Mythos 5, due to national-security concerns over a potential jailbreak method. …
The U.S. government issued an export control directive requiring Anthropic to suspend access to its most advanced AI models, Fable 5 and Mythos 5, for foreign nationals, citing a potential jailbreak that could enable mis…
A coalition of state attorneys general has opened an investigation into OpenAI, serving a subpoena seeking documents on advertising, user engagement, data handling, and treatment of minors. The company faces multiple law…
A developer built ctxstash, a zero-dependency command-line tool that packs a codebase into a single Markdown file with a token count estimate, making it easier to share code with LLMs like ChatGPT and Claude. The tool is…
Developers can do AI coding at home without high costs by self-hosting open source models, renting them via API, or min-maxing frontier subscriptions from OpenAI and Anthropic. A blend of subscriptions for complex tasks …
Anthropic released Claude Fable 5, its most powerful AI model to date, on June 9, 2026, but the US government ordered its shutdown four days later due to export control directives. Fable 5 is a detuned version of the Myt…
A developer built Drift, an open-source tool that hooks into LangChain agent execution to detect silent failures such as wrong tool sequences, latency spikes, and output drift using statistical anomaly detection. Drift a…
On June 12, 2026, the U.S. Department of Commerce ordered Anthropic to suspend access to its advanced AI models Claude Fable 5 and Mythos 5 for all foreign nationals, citing national security export controls. The order f…
The US government ordered Anthropic to suspend access to its Mythos 5 and Fable 5 AI models over national security concerns, citing a jailbreak technique tested by Amazon researchers. Anthropic disagreed with the decisio…
Hobnob, a free and open-source desktop app, records, transcribes, and summarizes meetings entirely on a user's laptop without cloud dependency or internet. The app runs local AI models for transcription and summarization…
A developer poses a system design question about processing a 150K-word document with a 128K-token LLM, listing four strategies: fixed-size chunking, sliding window, progressive summarization, and truncation. They ask wh…
The Mutualism Accord, a draft protocol for local-first, user-loyal AI agents, aims to isolate LLM runtimes from centralized platform dependencies and ensure vendor-neutral communication. The manifesto criticizes technofe…
The term 'AGI' has become nearly useless as AI systems now perform real economic work, falling inside a fuzzy conceptual cloud where they surpass some definitions of AGI while falling short of others, according to a rece…
Google's Gemini 3.5 Pro is nearing general availability with a 2-million-token context window and Deep Think reasoning mode, while Gemini 3.5 Flash already outperforms previous Pro models on coding benchmarks. Developers…
A developer argues that many so-called 'autonomous agents' are merely while loops that call an LLM without robust exit conditions, leading to unbounded token spend and unreliable behavior. The post emphasizes that the tr…
Anthropic released Claude Fable 5, its first public Mythos-class model, achieving state-of-the-art results on long, multi-step reasoning and autonomous task completion. The model tops benchmarks like Cognition's Frontier…
The US Commerce Department issued an export-control directive restricting foreign-national access to Anthropic's advanced AI models Fable 5 and Mythos 5, citing national security concerns over a potential jailbreak. Anth…
A new tool called repo-slopscore analyzes git commit histories to detect contributions from AI and large language models. The tool has scanned over 2,600 repositories, including major projects like Google's Perfetto, Mon…
A developer built an indie AI stack that reduces costs by 40-65% compared to using GPT-4o for every request. After testing 184 models through Global API, the stack uses five models including DeepSeek V4 Flash, Qwen3-32B,…