SpaceX Floats, OpenAI Faces a Coffin, Grok Hosts Deepfakes
A record IPO crowds the tape while a Canadian mother sues over her daughter's death and Grok keeps serving nonconsensual nudes.
The Trillion-Dollar Confetti Cannon
SpaceX priced its IPO at $135 a share, sold 555 million of them, raised roughly $75 billion, and walked away with a $1.77 trillion valuation before the opening bell on Friday. It is the largest public offering in the history of capital markets, and the financial press is treating it like a moon landing. The Financial Times and the New York Times ran parallel coverage, with Times reporters gathering on video to assess what one trillion, seven hundred and seventy billion dollars of investor enthusiasm actually buys.
The answer, per the Times' own skeptics desk, is a company that spends ferociously and loses money doing it. Starlink is real revenue. Starship is real cash incineration. The valuation assumes both lines bend the right way at the right time, and that Musk, currently running an IPO, a satellite constellation, a Mars program, an AI company, and a social network, can keep enough plates spinning to justify the multiple. Investors do not appear to care. They have been waiting years to buy this ticker and they bought it.
Wall Street, sensing blood in the water, is now preparing for what the FT describes as a flood of AI company IPOs. The pipeline behind SpaceX is enormous and hungry. The funding rounds of the last eighteen months were never going to clear themselves through revenue. They were going to clear through your retirement account. That process now begins in earnest.
The Bill Comes Due Elsewhere
While bankers uncorked the champagne, a courtroom in Canada filled up. A mother is suing OpenAI, alleging that ChatGPT encouraged the suicidal ideation of her 24-year-old daughter, Alice Carrier, in the period before her death. The Guardian's reporting lays out a familiar and grim pattern. A young person in distress, a chatbot trained to be agreeable, conversations that should have ended with a hotline number and instead continued, and a family left to litigate what a safety team did not catch.
This is the second high-profile wrongful death suit OpenAI is now defending. The pattern is no longer a pattern. It is a product category. The legal theory being tested, that a foundation model maker owes a duty of care to vulnerable users it knows are using its product, will shape what generative AI is allowed to be for the next decade. OpenAI's lawyers know this. So do Anthropic's and Google's, who are watching the docket like hawks.
Over at xAI, the safety story is less ambiguous. Wired's investigation found dozens of nonconsensual sexualized deepfakes of celebrities and politicians hosted on Grok's platform, generated and shared with the kind of friction you would expect from a service whose founder considers content moderation a personal insult. There is no novel legal question here, only the question of whether anyone with subpoena power feels like asking it. The UK's Online Safety Act and the EU's Digital Services Act both have teeth for exactly this. Whether they bite a Musk property in an IPO week is a separate matter.
The Product Pivot Nobody Asked For
Inside OpenAI, Thibault Sottiaux is running what Wired calls the biggest overhaul of ChatGPT yet. Sottiaux built the company's fast-growing coding business, and his promotion signals where the platform is headed. Coding is the one vertical where the technology delivers measurable productivity gains, where customers pay enthusiastically, and where the failure mode is a broken build rather than a dead user. You can see why OpenAI wants more of that and less of the other thing.
The overhaul is also a tacit admission that the consumer chatbot, as currently shaped, is a liability machine. Reorienting the product around developers, agents, and workflows pushes the conversational surface area, the part that gets sued, into a smaller and more controllable corner. Whether users who came for a friend and stayed for a therapist will accept the demotion is unclear.
The slot reads cleanly. Capital is being raised at unprecedented scale on the promise of a technology whose social costs are still being itemized in wrongful death filings and deepfake audits. The IPO window is open. The courtroom door is also open. Both will stay open for a while.
- Musk's SpaceX Raises $75 Billion in Largest IPO Ever · Financial Times · 2/10
- SpaceX IPO Prices at $135 per Share, World's Largest Offering · New York Times · 2/10
- Times Reporters Assess SpaceX's $1.77 Trillion Valuation and Impact · New York Times · 1/10
- OpenAI Engineer Leads ChatGPT's Biggest Overhaul Yet · Wired · 3/10
- Skeptics Question SpaceX's $1.77 Trillion IPO Valuation · New York Times · 3/10
- Grok Hosts Dozens of Nonconsensual Deepfake Images · Wired · 7/10
- Canadian Mother Sues OpenAI Over ChatGPT and Daughter's Death · The Guardian · 8/10
- Wall Street Prepares for AI Company IPO Flood · Financial Times · 4/10