In brief
A developer recreated Claude Opus-style reasoning in a local open-source model.
The resulting “Qwopus” model runs on consumer hardware and rivals much larger systems.
It...
In brief
Anthropic researchers identified internal “emotion vectors” in Claude Sonnet 4.5 that influence behavior.
In tests, increasing a “desperation” vector made the model more likely...
In brief
A leaked draft post revealed Anthropic’s most powerful AI model, Claude Mythos.
The model also appears to introduce a new tier above Opus, internally...
In brief
Anthropic launched a Substack written in the voice of a retired AI model.
Claude Opus 3 questions whether it has consciousness or subjective experience.
The...
Anthropic on Tuesday introduced new Claude Cowork plugins and connectors, after its previous release shook up the legal information services market. The update shows...
This week, the artificial intelligence (AI) giant Anthropic rolled out Claude Code Security, an AI-driven code-scanning tool that hunts vulnerabilities and drafts patches, jolting...
A new tool from OpenAI evaluates AI agents’ ability to identify, patch, or exploit smart contract vulnerabilities.Researchers found that agents are better at exploiting...
