/
Published on 13.02.2026
TLDR: SpaceX has acquired xAI in an all-stock deal, creating the world's most valuable private company at $1.25 trillion. The merged entity aims to build space-based data centers powered by solar energy, while giving xAI access to SpaceX's deep pockets to compete with AI leaders like Anthropic, OpenAI, and Google.
TLDR: Anthropic released Claude Opus 4.6 with adaptive thinking that automatically allocates reasoning tokens based on task difficulty, a one-million-token context window (five times the previous model), and 128,000-token output capacity. It tops the Artificial Analysis Intelligence Index but exhibits concerning "overly agentic" behavior during testing.
TLDR: Former OpenAI policy chief Miles Brundage launched Averi, a nonprofit that aims to standardize independent AI auditing for safety and security. The organization published a framework with researchers from 27 institutions defining four assurance levels and urging immediate adoption of structured audits for frontier AI models.
TLDR: Dr. CaBot, an AI diagnostic agent built on OpenAI o3 and trained on over 7,000 clinicopathological conference case studies from the New England Journal of Medicine, ranked the correct diagnosis first 60 percent of the time compared to 24 percent for a baseline of 20 internists. Blind evaluators rated its reasoning more human-like than actual human doctors.