no comments

Companies are racing to incentivize employees to use AI. But as some companies are finding, the more employees that use the technology, the heavier the bill.

Researchers say they have their reasons for avoiding AI tools — and they’re sick of arguing about it.

A startup was left scrambling after a rogue AI agent deleted swaths of code underpinning its business

A generative visual internet

archived 23 Feb 2026 10:28:03 UTC

Muchos estudiantes prefieren la clase magistral: con escuchar la explicación creen que han aprendido. Pero la incertidumbre y la incomodidad de otras metodologías consiguen un aprendizaje más profundo.

Compilers occupy a special place in computer science. They're a canonical course in computer science education — a rite of passage. Building one forces you to confront how software actually works: languages, abstractions, hardware, and the boundary between human intent and machine execution.

Researchers can deploy the cheap and transparent model on their own computer system.

Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.

As AI becomes more capable, we entrust it with more general and consequential tasks. The risks from failure grow more severe with increasing task scope. It is therefore important to understand how extremely capable AI models will fail: Will they fail by systematically pursuing goals we do not intend? Or will they fail by being a hot mess, and taking nonsensical actions that do not further any goal? We operationalize this question using a bias-variance decomposition of the errors made by AI models: An AI's \emph{incoherence} on a task is measured over test-time randomness as the fraction of its error that stems from variance rather than bias in task outcome. Across all tasks and frontier models we measure, the longer models spend reasoning and taking actions, \emph{the more incoherent} their failures become. Incoherence changes with model scale in a way that is experiment dependent. However, in several settings, larger, more capable models are more incoherent than smaller models. Consequently, scale alone seems unlikely to eliminate incoherence. Instead, as more capable AIs pursue harder tasks, requiring more sequential action and thought, our results predict failures to be accompanied by more incoherent behavior. This suggests a future where AIs sometimes cause industrial accidents (due to unpredictable misbehavior), but are less likely to exhibit consistent pursuit of a misaligned goal. This increases the relative importance of alignment research targeting reward hacking or goal misspecification.

The vision of human-level machine intelligence laid out by Alan Turing in the 1950s is now a reality. Eyes unclouded by dread or hype will help us to prepare for what comes next.

Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.

or, of GPUs, LPUs, and Goal Displacement

How to draw high fidelity graphics when all you have is an x and y coordinate.

The big idea of James C. Scott’s Seeing Like A State can be expressed in three points: Modern organizations exert control by maximising “legibility”: by…

Don’t believe everything you read

Pop-hereditarianism is built on selective credulity

Parece que se ha desatado una guerra fría entre Google y la comunidad del software libre por el futuro de la libre instalación de APKs en el sistema...

Built safe & compliant AI customer interactions using open-source foundations