Keep pulling the thread on United States.
The Biden administration is pressuring Meta to agree to a voluntary review of its AI models for safety and performance testing prior to their release.
Meta is the only major AI lab that has not yet signed a voluntary agreement with the U.S. government for pre-release model testing, unlike Microsoft, xAI, and Google.
The U.S. Commerce Department is reportedly considering a ban on Chinese-made robots.
Commerce Secretary Howard Lutnick held a closed-door meeting with executives from firms including Boston Dynamics, SpaceX, Siemens, and Goldman Sachs to discuss the risks of Chinese-made robots.
The Congressional Select Committee on China highlighted on X that a humanoid robot from Unitree, a designated Chinese military company, was available for sale on Amazon in the U.S.
The U.S. government has designated the robotics firm Unitree as a Chinese military company.
SeedDance 2.5 is the first video generation model capable of using image, video, and audio files as input references.
Anthropic announced a new feature called Claude Tag, which integrates Claude into Slack as a team member with access to channel context and tools.
Anthropic claims that 65% of its product team's code is now generated by its internal version of Claude Tag.
Andrej Karpathy described Claude Tag as a third major redesign of LLM user interfaces, where the AI becomes a "self-contained, persistent, asynchronous entity" working alongside human teams.
Chris Taylor of Fractional reported that an instance of Claude Tag autonomously identified a bug in one of its own scheduled tasks, fixed the bug, and documented the fix without human intervention.
Chris Taylor of Fractional stated that after being given access to a codebase, Claude Tag analyzed it, identified and prioritized the top 10 problems, implemented the fixes, and opened them for review within Slack.