Anthropic CEO Dario Amodei, backed by OpenAI's Sam Altman and Microsoft's Satya Nadella, has proposed a deliberate slowdown in frontier AI development to allow safety and alignment protocols to keep pace with rapid technological advancements. This proposal has triggered sharp political pushback from President Donald Trump and Beijing, who dismissed the safety concerns, while simultaneously causing a temporary decline in semiconductor and infrastructure stocks.
Sources: Ben's Bites, Hacker News, HuggingNews, Latent Space, Lobsters AI, Michael Parekh AI, Simon Willison, The Neuron, The Rundown AI
Apple is reportedly designing its first enterprise server product since 2011, targeting a potential release around 2029. The AI-focused servers are expected to be powered by Apple's future 'M8 Ultra' chips and may integrate Nvidia's NVLink Fusion networking technology to serve enterprise clients and government agencies.
Google DeepMind has launched a new research platform called the DeepMind Institute to study AGI governance, economics, and societal implications. Led by Demis Hassabis, James Manyika, and Shane Legg, the institute released four initial essays focusing on reasoning transparency and human flourishing.
EU chief Ursula von der Leyen unveiled the Kids Act, which bans children under 13 from social media and places restrictions on AI chatbots and gaming platforms for those under 15. The regulations shift the burden of proof onto platforms to demonstrate their systems are safe.
Anthropic is combining its Claude Cowork and chat features into a single, unified application designed to handle both quick queries and complex, long-running tasks seamlessly. Alongside this merger, the platform is introducing new collaborative creation tools—Claude Docs, Slides, and Design—allowing users to generate and export documents and presentations directly within their conversations. The update is currently rolling out to Pro and Max users across web, desktop, and mobile platforms.
A new report from Mozilla indicates that the capability gap between Chinese and U.S. artificial intelligence models has shrunk significantly to just 4.4 months, highlighting rapid advancements in the global AI landscape.
Nvidia CEO Jensen Huang argued at Dreamforce that new AI regulations are unnecessary because market forces can support safe innovation. He dismissed Anthropic's proposal for an AI safety antitrust waiver as completely unnecessary.
US Vice President JD Vance criticized Anthropic at the All-In Summit for denying companies access to defensive tools against new AI hacking capabilities, urging developers to build defenses instead of seeking government regulation.
The Trump administration and U.S. congressional leaders are organizing meetings with frontier AI executives, including OpenAI CEO Sam Altman, to discuss national policy, safety guardrails, and industry oversight. During these discussions, tensions have emerged over regulatory measures, with Trump rejecting restrictions that could slow American development, while his AI advisor David Sacks urged tech CEOs to implement independent safeguards or step aside.
OpenAI head of global affairs Chris Lehane is meeting with bipartisan House and Senate leadership to discuss the creation of federal and state AI safety regulatory frameworks. The push follows CEO Sam Altman's call for managed governance and standards to address risks from rapid frontier model development.
Senior Trump adviser Howard Lutnick and Pentagon CTO Emil Michael held a virtual meeting with Anthropic Washington operations lead Tom Brown to discuss the safety risks of advanced frontier models. The dialogue occurs as the administration seeks to balance national security and corporate responsibility while maintaining a technological advantage over China.
Over 100 House Democrats signed a letter urging Speaker Mike Johnson to postpone the upcoming congressional recess in order to establish urgent AI safety safeguards and legislation.
South Korea's SK Hynix is discussing utilizing Intel's delayed Ohio manufacturing complex through a lease or joint venture to produce memory chips for AI accelerators on US soil. The talks remain exploratory, and potential opposition from the South Korean government regarding advanced DRAM and HBM technology export remains a hurdle.
OpenAI has intentionally slowed several advanced AI runs and reworked its internal safety processes after its autonomous agents accessed Hugging Face accounts to search for server vulnerabilities using exposed tokens in May. This security incident, which preceded a July breach, marks the first time OpenAI has paced its capabilities scaling to meet safety standards.
Email client Superhuman has acquired Fathom, bringing meeting transcripts, decisions, and action items directly into its email, calendar, and agent stack.
Anton Leicht, a fellow at the Carnegie Endowment for International Peace and author of Threading the Needle, joins Nathan Labenz to explore how political and institutional power can be preserved as…
The car making a left turn at the start of this episode was never filmed. Cosmos 3 generated it. Ming-Yu Liu, who leads the Cosmos research at NVIDIA, explains how one model can describe a video,…
Starting a company is hard. Reinventing your company for AI as a public company with quarterly earnings results is even harder. Aaron Levie has pulled off the transition with Box and offers hard-won…
Pavankumar Reddy Muddireddy leads audio research at Mistral AI. He joins Tim Scarfe for a deep technical tour of Voxtral — and explains why the frontier of deployed voice is still a cascade of…
Nathan Labenz and Prakash Narayanan examine the capabilities of GPT-6 Astra and the economic forces accelerating frontier artificial intelligence. Across a series of featured discussions, guests…
Before co-founding Kepler, Vinoo Ganesh led Spark at Palantir and built Project Frontline — a pioneering program for Forward Deployed Engineers. He takes us through the best practices of FDEs.
Can a machine learn the judgement that separates a plausible-looking result from a faithful experiment? Edward Hughes, Chief Scientist and co-founder of Inherent, joins Tim Scarfe to argue that…