24th July - AI News Daily - Silicon Valley vs. Beijing: The High-Stakes Race Between US Labs and China's Open-Weight Champions

Send a text

AI News Summaries

https://s.server489.com/AI-2025-07-24

AI Tweet Summaries

https://s.server489.com/XAI-2025-07-24

Major Developments: OpenAI secured a $30 billion Oracle partnership, significantly expanding AI infrastructure capacity. Google DeepMind began recruiting in Asia Pacific to build more globally representative models. Major government AI action plans were released, with Trump's agenda focusing on deregulation and unified policy. OpenAI lost Slack integration access due to data protection concerns. Diode raised $11.4M for next-gen robotics hardware. China Telecom launched AI Flow to integrate LLMs with telecom networks. AI is now helping conservationists fight invasive plants. 

New Tools: Factory 1.4 brings AI Droids to Slack for workflow automation. DeepSite offers AI-based website building via natural language. Goedel Prover V2 enables automated mathematical proof verification. Music Arena lets users judge AI-generated music. Elicit enhanced research automation capabilities. Local Chat expanded model support. Higgsfield Steal enables web image recreation and remixing. 

LLM Advancements: US labs lead in frontier models while China excels in open-weight alternatives. Google's Gemini 2.5 Pro solved IMO math problems. Qwen3 and QuotientAI's models outperformed rivals. Alibaba's Qwen3-Coder surpassed Kimi K2. Anthropic discovered "subliminal learning" between models. Meta's Llama 3.1 reproduced copyrighted text verbatim. Grok 4 showed strong benchmarks but practical limitations. 

Feature Enhancements: YouTube Shorts launched an AI-powered media creation platform. Google integrated Gemini's video generation features. wandb added colorblind accessibility features. KerasHub released ultra-efficient HGNetV2. Cerebras Systems now offers free, fast inference. 

Showcases: Aeneas AI decodes ancient inscriptions. RALU technology speeds up diffusion models 7x. TimeScope created a rigorous video comprehension benchmark. 

Key Discussions: Demis Hassabis suggests AI could unlock fundamental scientific mysteries. Safety debates intensify around "safetyism" vs rapid AI expansion. International AI development shows US-China divide. Anthropic's "subliminal learning" findings challenge model interaction assumptions. Copyright concerns follow Llama 3.1's text replication. 

Industry Trends: Governments worldwide accelerate AI adoption. AI talent wars intensify with packages up to $300M. Big tech integrates AI across product lines. Enterprise AI adoption surges. AI transforms content creation. Security and ethical risks increase with AI advancement. Legal sectors struggle to keep pace. Societal impact concerns grow as ChatGPT reaches 500M weekly users. Energy demands prompt sustainability efforts. AI delivers breakthroughs across sectors despite occasional setbacks. 

Support the show

🌍 INAI • The Open AI Hub

The Intelligence Atlas → the world’s most comprehensive, open hub of AI knowledge. 2 Million+ tools, models, agents, tutorials & daily news—free for all, updated every day.

https://github.com/inai-sandy/inAI-wiki