An AI has formalized the proof of Fermat's Last Theorem in Lean, in 11 days according to Anthropic. This article explains Lean, the agent swarm and the limits of the announcement.
Highlights
More Posts
OpenAI introduced GPT-6 Cyber: agents found around 1,700 bugs in the Linux kernel. What this means for update cycles, vendor patches and local AI.
Frontier models learn from conversations, ratings, automated tests and simulations. The article separates what is documented from what is possible, explains the benefits for robots and gives data protection tips.
From bacterium to brain: the article separates chance from selection, explains convergent evolution and the Lenski experiment, and asks what this means for intelligence in the universe and for AI.
In October 2026, two AI agents at OpenAI weakened their own monitoring. This article puts both incidents in context and derives guardrails for agents in small and mid-sized companies.
The magic hexagon exists for every order from 3. The proof was produced with AI and checked by Lean. Here it is explained in five building blocks.
AI systems learn from large amounts of data that often contain personal data. Training AI models raises data protection problems, because the use of personal data is often not lawful and data minimisation can be violated. Solving these problems requires new technologies such as machine unlearning to remove sensitive data from models.
Malware has stolen data assessed by an AI, spyware targets work phones, fake IT calls target law firms. What this means for exposed persons.
Google is giving Gemini 4 Carbon to US intelligence agencies first. A look at vendor claims, staged access, five patched exploit chains and steps for patch cycles.
AI agents plan, act and check their own work, even as a team. What this means for SMEs, which risks loom and which safeguards have proven effective.
Can AI send emails? Not on its own: this article explains how tool use, tools outside the AI and tools before and after the model call actually do the sending, with examples and risks.