After Moonshot’s Kimi K3, U.S. AI rivals launch new models and upgrades
Google and Meta both released new models this week, even as U.S. users wait for an open-weight model breakthrough
In the weeks following the launch of Chinese AI lab Moonshot AI's Kimi K3 model, U.S. AI firms have taken the initiative to counteract the situation by releasing their own open-weight models and upgrades. Meta CEO Mark Zuckerberg published a 6,500-word manifesto outlining his vision of a "positive AI future," where every user would have a personalized tutor and coach with PhD-level expertise in every subject, aiming to enhance various aspects of their lives.
Meta also announced the release of new models, such as Muse Spark 1.2 and Muse Glimmer. Muse Glimmer is described as a 30-billion-parameter open agentic model optimized for always-on local agent workflows, which can run on a Mac or PC with a single consumer GPU. This enables various use cases, including local agents, function calling, local coding, and LLM-as-a-judge evaluation.
Google, on the other hand, saw changes in its senior leadership and AI models. Jeff Dean, Google's chief scientist and 30th employee, announced his departure to co-found an AI startup called Discovery Loop, while Demis Hassabis stepped down as Google DeepMind CEO to become the unit's chairman and Alphabet's chief scientist. Google is under pressure to develop a model that can outperform China in the AI race, but the progress of its anticipated Gemini 3.5 Pro model remains uncertain.
The week also witnessed Google launching a new line-up of Pixel phones, equipped with Gemini AI integrations. These devices showcased upgraded hardware but lacked radical innovations. Additionally, Google announced the launch of Gemini 3.7 Flash, which delivers an improved developer experience, enhanced adaptability, and better instruction following.
Anthropic, the maker of Claude, attracted attention for its announcement regarding AI labeling. In compliance with the European Union's AI Act, Claude-generated text will carry embedded watermarks, while generated files will include digitally signed provenance metadata. However, Anthropic did not provide technical details about the watermarking process, raising concerns among users about potential false positives and negatives in AI detection tools.
Written by urgent.news from The Hindu - Sci-Tech's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
