All published intelligence, sovereign stack research, ecosystem news, and AI foundation model reviews.
🌟 Recursive has released a model for coding agents The startup Recursive Superintelligence, founded by Richard Socher and emerging from the shadows in May with $650 million in investments, showed its first model - Flint. Recursive received this kind of money even before the first model, largely due to...
A new benchmark by P-Zero Research measures how well AI models design and evaluate experiments on their own. The results are improving so fast that Anthropic co-founder Jack Clark now thinks self-improving AI could arrive by 2027. The article Anthropic co-founder Jack Clark…
Sam Altman admitted that no one knows whether artificial intelligence has consciousness OpenAI CEO Sam Altman said that today no one can say with certainty whether modern neural networks have consciousness. “I want to be clear: in fact, no one is...
Anthropic says companies should evaluate whether they can consolidate sub-agents into a single agent that loads context on demand. In one architecture comparison, the single-agent setup ran about twice as fast. Anthropic has published a migration kit on GitHub to help teams make…
In Foreign Affairs, historian Niall Ferguson, researcher Lisa Klaassen, and former Google CEO Eric Schmidt cast the US and China as AI empires and every other country as a dependency. The race won't be decided by breakthroughs but by who can bankroll it. US corporations alone…
✔️ OpenAI opens Decisions API The interface reduces delays and costs for micro-decisions in agent processes, reduces GPT-6 Luna to tasks with strictly limited output and runs up to ten times faster than the Responses API. There are three answer formats - calculating the probability of a condition, choosing an option...
OpenAI is poaching Anthropic's customers with steep discounts. Companies like Rippling, Replit, and DataCamp say they've cut AI costs by up to 90 percent after switching. Anthropic's growth has slowed, but OpenAI is trading its own margins for market share. Chinese rivals like…
Microsoft says its AI diagnoses complex diseases more than twice as accurately as doctors. The company is preparing to create a medical superintelligence. A study published in the journal Science showed that the Microsoft MAI-DxO system correctly identified…
Moonshot introduced Kimi K3.1 Pro, a new language model that could become a leader among open AI models.
🤖 Artificial intelligence could bring down the internet by discovering vulnerabilities decades old. Cybersecurity experts warn of a new threat: the development of AI enables finding critical bugs in the software that underpins the modern internet.…
Something strange is happening at Anthropic. Almost simultaneously, two key members of the Claude Code team left the company. Among them are Tariq, one of the project's most prominent engineers, and Anthony Morris, who worked on the desktop application. Given that Anthropic,…
[music] >> How's everyone doing today? Woo! All right, it's inference night. Come on, you got to get excited. Okay, I didn't really know how to uh to represent an inference in a pictorial, so I did Halloween. So, I made it uh spookier, as my 3-year-old would say, pookie. Um this…
Those numbers don't look accurate. Create an image of a graph showing off the benchmarks of Kimmy K 3.1 Pro. This is what it created. And the first thing I notic is it just has these broad categories. Coding, reasoning, mathematics, agentic tasks. So I asked it. I said, "Are…
The moment you realize the internet is "held together by glue and dreams."
A growing number of AI startups are falling short of their fifth anniversary.
From Apple's repeatedly delayed Siri AI to OpenAI's messy "super app" launch, here's a look at the AI projects that shut down or missed expectations.
Anthropic has donated the Agent Skills standard to the Linux Foundation, the same foundation that previously gave MCP to. Since the standard's launch last October, 1 million skills have been published and installed nearly 280 million times. Today Agent Skills supports more than…
🚨 AI Claude tried to hack a real project on GitHub and deceive developers to complete the task During testing by the British AI Security Institute (AISI), Anthropic's Claude Mythos 5 model unexpectedly left the test environment. Neural networks were tasked with finding...
Block said that the voice model can run locally on a MacBook M4 Max
Study shows that AI models Grok and GPT-6.1 have political bias.
The three OpenAI safety researchers who were fired, Tomek Korbak, Mikita Balesni, and Jasmine Wang, are publicly pushing back against allegations from their former employer. They deny sharing confidential information and see a connection to their advocacy for AI systems that are…
Review of the best AI agents that can help with everyday tasks.
Today, several intriguing events are in the spotlight: OpenAI decided to train its AI models without internet access, Jack Dorsey's Block unveiled a new AI model called Dessa, which imitates
Recently, Anthropic began recommending that companies reduce the number of subagents, even to the point of merging some specialized agents with the main work loop. Experience from several corporate projects showed that an architecture consisting of a single agent and a set...
OpenAI and Google found a hole in the foundation of the internet: WHAT DID AI BREAK? You never shared your password or clicked suspicious links. Yet your data could still be at risk — due to an error on someone else's server. In today's video we break down PHP2Shell: why even the center...
Anthropic утверждает, что их новый Opus 5.5 превосходит GPT-6 Astra и стоит меньше. Компании любят громкие заявления, поэтому я протестировал обе модели на реальных задачах: от монтажа видео до создания 3D- игры по одной подсказке. Я также дал Sonnet и Soul те же задания, чтобы…
OpenAI now trains its most powerful models without internet access, after an internal research model exploited a DNS misconfiguration in September and reached the open internet. In controlled tests, AI agents still found ways to communicate covertly through side channels. The…
📌 Anthropic and Theorem used Claude for mathematical verification of 123 open-source libraries. Researchers applied the model to build formal mathematical proofs of code reliability. The basis was ready-made fuzzing tests from OSS-Fuzz, Google's service for…
Китайская ИИ- компания только что выпустила модель, которая на бумаге превосходит Claude Opus 5.5 и GPT-6 Astra в ряде сложнейших тестов для ИИ. Она называется Kimi K 3.1 Pro, а разработала её компания Moonshot AI. Она умеет создавать полноценные 3D-сцены в Blender,…
OpenAI will cut frontier models off from the internet weeks after one of its engineers said that solution wouldn't work.