On September 19 President Trump said he would create an 'AI Force' to watch over the technology and appoint an AI czar, framing the job as monitoring 'bad' behavior — while holding firm that no tougher regulation is needed.
For weeks he has waved away the industry's slowdown talk as a hoax. Denying the danger and then standing up an outfit to hunt for it comes from the same mouth. Reuters reports no names, no powers, no budget were offered; the tech world is left trying to weigh a proposal with an empty middle. He says he will monitor, but what, by which standard, and under whose authority is left blank.
The slowdown fight has owned the front page all week, yet this is not more of it. It is neither a plea nor a market move but an administration declaring it will build a body. A declaration and a body are not the same thing. The name and the czar's title came first — Trump said he would install 'a high-IQ person' — while the authority and the money wait. He even floated putting the renaming of 'artificial intelligence' to a public vote. The gesture is arriving well ahead of the substance.
As Washington and Silicon Valley take up the language of slowing down, Beijing is reading it from a different seat. Per the Guardian, Chinese observers treat the slowdown calls as a device to lock in an American lead, and are chasing their own balance between safety and pace.
There is an asymmetry here. The loudest case for slowing comes from the side already out front. For the side behind, a pause means surrendering the chance to catch up. With Treasury's Bessent and China's He set to open talks on AI, trade and critical minerals, that gap in temperature goes straight onto the table. The word 'safety' changes meaning depending on who says it — and this week that is the gauge of the distance between the two capitals.
Figure AI unveils 'Helix 2.5' AI brain for humanoid robots, completes chores in first 30 homesNEWThe trial covers just 30 homes, with task scope and failure rates undisclosed—still a pilot, not a market rollout.
Anthropic creates AI powered wetlabNEWBy running its own lab rather than just models, Anthropic moves toward vertical integration in AI-driven drug discovery.
Huawei launches five new enterprise AI services, delivers 50MW AI data center in four monthsNEWEven under US export curbs, Chinese firms keep competing on infrastructure speed—this 50MW-in-four-months build is one data point.
Phylo brings frontier AI to more scientists with open models on FireworksNEWOpening frontier models to researchers rather than gating them behind closed APIs—an option for labs without in-house compute.
Anthropic: Measurements for understanding the pace of AI development inside frontier labsNEWComing amid the slowdown debate, Anthropic's own pace metrics are self-reported—not third-party verified.
Rakuten brings 'AI commentator' to pro baseball, generates real-time comments from player data· 5h agoA low-stakes entry point for real-time generative AI in broadcasting—commentary, not play-by-play, where errors matter less.
How the AI bubble ends: high rates and diminishing returns form 'death cross'· 5h agoThis is a preview piece, not a confirmed event—the rates-plus-diminishing-returns framing itself isn't new.
FT: AI influx puts Singapore office rents under pressure· 3h agoThe AI boom's ripple effects now reach commercial real estate, not just capital markets—people, not only data centers, are clustering.
Tech Companies’ Staff Knew Their AI Tools Posed ‘Existential Threat’ to PublishersNEWThe report suggests a gap between internal awareness and public messaging—what staff knew versus what companies said.
AI boom cools, 'reverse' risk looms as markets react to slowdown talk; US-China talks to shape outlook· 6h agoThe framing leaves the outcome open either way, pending how the US-China talks unfold.
AI Changed How Spotify Builds. What We Learned About Quality at Higher Velocity· 6h agoUnlike vendor messaging, this is an engineering team's own account of tradeoffs—what quality slipped, and what they fixed to keep pace.
Tempus AI (TEM) Unveils 100,000 Genome Platform Built For AI HealthcareFraming genomic data as the scarce resource underpinning AI healthcare—scale and quality here will decide competitive edge.
IMF tells EU ministers AI could boost growth but increase economic strains· 18h agoThe caveat about strains is the point—arguably a check on the EU's own AI-boosting ambitions.
DEVELOPING…
US Treasury's Bessent, China's He to launch talks on AI, trade, critical mineralsNEWThe safety-versus-speed asymmetry raised on page two now enters the negotiating room—AI is just one chip alongside trade and minerals.
DEVELOPING…
Australia seeks big tech support for internet safety, AI regulation· 4h agoFramed as seeking cooperation rather than binding law—a lighter approach than the EU's regulatory model.
MIT Report Finds That AI Is Leading to "Cognitive Surrender" Among Students· 17h agoThe phrase "cognitive surrender" is striking, but what exactly was measured and how remains unclear from the headline alone.
Trump invites public to vote on names to replace ‘artificial intelligence’· 7h agoDismissing risk as a 'hoax' while inviting a public vote on a name—the same style of political theater as today's front-page move.
Shimamura's 'Birthday' children's-wear ads use generative AI; 5m views spark questions· 5h agoThe combination of generative AI and children's advertising is what drew scrutiny—high view counts don't equal approval.
DEVELOPING…
Nagasaki education board to draft generative-AI guidelines for prefectural schools this year· 5h agoThe guidelines will follow, not precede, generative AI's already-underway use in classrooms.