AITLDR
Trump Eyes OpenAI Stake, Google Pays SpaceX $11B/Year for Compute

Trump Eyes OpenAI Stake, Google Pays SpaceX $11B/Year for Compute

The US government wants a piece of OpenAI, Google is writing nine-figure monthly checks to SpaceX, and Apple is about to reintroduce Siri. Again.

Subscribe free →
The US Government Wants Equity in OpenAI

The US Government Wants Equity in OpenAI

President Trump confirmed he's in active discussions about deals that would give the American public a financial stake in AI companies — with OpenAI apparently at the top of the list. This comes as OpenAI completes its long-running restructuring from nonprofit to for-profit and cements its status as one of the most valuable private companies on Earth. The White House framing is pure populism: "the American people should benefit from AI's success."

Why it matters: A government equity stake in a frontier AI lab would be unprecedented — and would raise serious questions about regulatory capture, research independence, and what "oversight" actually means when Uncle Sam is a shareholder.


Google Is Paying SpaceX $920 Million Per Month for Compute

Google Is Paying SpaceX $920 Million Per Month for Compute

Google has inked a deal to pay SpaceX $920M *per month* — that's $11 billion annually — for computing resources. A Google spokesperson attributed the deal to "unexpected demand" for its recently launched AI products, suggesting its own infrastructure simply can't keep up. This is a staggering number even by hyperscaler standards, and it's going to a company not traditionally known as a cloud provider.

Why it matters: When Google — which operates some of the world's largest data centers — has to go outside for compute at this scale, it tells you everything about how severely AI demand is outpacing infrastructure buildout across the entire industry.


Reid Hoffman Bails on Microsoft's Board to Go All-In on AI Drug Discovery

Reid Hoffman Bails on Microsoft's Board to Go All-In on AI Drug Discovery

LinkedIn co-founder and longtime Microsoft board member Reid Hoffman is stepping down after what the article calls "a very profitable decade" to go full "founder mode" at Manus, his AI drug discovery startup. Hoffman has been one of Silicon Valley's most prominent AI boosters, but this move signals he's done advising from the boardroom and wants skin in the game.

Why it matters: When one of AI's loudest evangelists cashes out a cushy board seat to build a startup himself, it's a data point on where the real upside — and the real action — is perceived to be right now.


Apple Is About to Reintroduce Siri. Again.

Apple Is About to Reintroduce Siri. Again.

Ahead of WWDC on Monday, The Verge lays out Apple's awkward AI situation: the company first "launched" a new AI-powered Siri at WWDC 2024, then it largely failed to ship, and now Apple is gearing up to reintroduce it — this time reportedly with Gemini integration under the hood. The piece frames Apple's years-long lag not as pure embarrassment but as a potential advantage: watching competitors stumble before shipping something polished.

Why it matters: If Apple ships a genuinely capable Siri to 2+ billion devices this fall, the AI assistant landscape changes overnight — but the company has cried wolf on this before, and credibility is running thin.


Meta Built an AI-Generated Clickbait News Feed — And It's Already a Mess

Meta Built an AI-Generated Clickbait News Feed — And It's Already a Mess

Meta's standalone AI app now has a "For You" section that surfaces AI-generated articles complete with AI-generated images and AI-generated text. The results are predictably chaotic — The Verge highlights an AI image of the royal family featuring two Queen Elizabeths. This isn't a bug report; it's how the feature shipped.

Why it matters: Meta is essentially industrializing misinformation-adjacent content and pumping it directly into a feed, which should terrify anyone who watched what happened when humans were writing Facebook's clickbait — this is that problem on steroids.


An Indiana Town's Data Center Fight Gets Ugly

An Indiana Town's Data Center Fight Gets Ugly

A proposed $2 billion data center in Shelbyville, Indiana has become a full-blown political controversy — and the mayor made it worse. Shelbyville Mayor Scott Furgeson was caught on camera saying that "No Data Center" signs only appear in "shitty houses" and that "most of them are rentals." The clip went wide and crystallized a growing national tension between communities bearing the costs of AI infrastructure (power, water, noise, property disruption) and the economic promises made to get them to accept it.

Why it matters: Data center siting is becoming one of the most contentious local political issues in America, and dismissing residents as renters in bad houses is exactly the kind of response that turns opposition movements into organized ones.

Quick Hits

  • A USB-connected Creative Sound Blaster speaker can be hacked over-the-air to infect connected PCs — and Creative doesn't consider it a vulnerability. Ars Technica
  • A former IBM cybersecurity exec turned whistleblower alleges the company covered up multiple data breaches from the mid-2010s and never disclosed them. TechCrunch
  • The anti-phone startup wave is real: founders are raising money for in-person social experiences and DIY computers explicitly designed to keep you offline. TechCrunch

AI TLDR

Get AI Daily in your inbox every morning — free

Free forever·Unsubscribe anytime·5 min read
Share this edition: