Shares of major AI-related companies, including Oracle and CoreWeave, fell due to concerns over a slowdown in OpenAI's growth.
AI Wellbeing research reported that large models' wellbeing changes depending on descriptions of suffering versus positive experiences.
NVIDIA has unveiled the multimodal Nemotron Nano Omni 30B A3B Reasoning.
Anthropic has joined the Blender Development Fund as a corporate patron to support the advancement of open-source 3D software.
Talkie, a 13B-scale 'vintage' language model trained solely on historical texts from before 1931, has been released, enabling contamination-free testing of generalization ability.
An AI analyzer found 38 vulnerabilities in OpenEMR.
NVIDIA has unveiled Nemotron 3 Nano Omni, which integrates vision, audio, and language.
For its 20th anniversary, Google Translate unveiled a pronunciation practice feature for Android.
W2A has released a sensor protocol for local agents.
Ling-2.6-flash has been open-sourced, achieving up to 340 tok/s on 4× H20.
Lemonade OmniRouter unifies local AI engines into a single tool-calling endpoint.
OpenAI is fiercely competing by following in Anthropic's footsteps in product launches and business model.
OpenAI is working with MediaTek and Qualcomm to develop an app-free smartphone powered by AI agents.
Observers suggest OpenAI is pushing to develop chips for an AI agent phone with MediaTek and Qualcomm.
Oracle and AI chip stocks fell after reports that OpenAI missed its internal user and revenue projections.
A WSJ report fueled concerns about slowing OpenAI growth and revenue, triggering a sharp selloff in AI infrastructure stocks.
A Tree-sitter-based local memory layer dramatically reduces coding agent token usage.
Musk has entered an Oakland trial seeking to reverse OpenAI's for-profit conversion.
GitHub patched a critical RCE vulnerability in the `git push` path, and there was no evidence of exploitation.
China blocked Meta's acquisition of Manus, sharpening the US-China AI divergence.
Poolside released Laguna XS.2, a 33B A3B MoE model, under Apache 2.
A Blender connector has been added to Claude, supporting scene debugging and batch editing.
ElevenLabs has added pre-configured agent templates to ElevenAgents.
AI21 Maestro achieved SOTA on BrowseComp-Plus with 95.18% accuracy.