Good morning. Once upon a time, bots pretending to be human was a story. The most infamous example was the controversy over Ashley Madison, an online site designed for those seeking extra-marital affairs. It turned out that the men on the site were real, but the women were overwhelmingly bots.
Fast forward a decade or so, and the tables have turned: it’s the humans who are now the unexpected guests. Consider Moltbook, a social network-like platform where as many as 1.5 million AI agents are said to interact autonomously, asking each other for help on technical subjects and for assistance on other tasks. As Fortune’s Eva Roytburg writes, a new investigation by security firm Wiz found that the vast majority of the AI agents on Moltbook were not autonomous at all. Behind the proverbial curtain are about 17,000 humans controlling the agents—roughly one human for every 88 bots.
Today’s news below.
Alexei Oreskovic
@lexnfx
alexei.oreskovic@fortune.com
Want to send thoughts or suggestions to Fortune Tech? Drop a line here.
SpaceX swallows xAI
Elon Musk’s rocket company SpaceX has acquired xAI, the artificial company founded by Musk three years ago, in a massive, and unconventional, deal that combines the two privately held firms into a company with an astounding $1.25 trillion reported valuation and plans for a historic IPO this year.
Musk, who is the CEO of both companies, cited the potential for space-based data centers as one of the most important benefits of the combination, even though the concept is still unproven and largely theoretical. “Global electricity demand for AI simply cannot be met with terrestrial solutions, even in the near term,” Musk wrote in a SpaceX blog post. “By directly harnessing near-constant solar power with little operating or maintenance costs, these satellites will transform our ability to scale compute,” Musk wrote.
xAI's flagship product, Grok, a chat which has recently been blocked in some countries for producing sexualized deepfakes of women, went completely unmentioned in the announcement.
While reports of a potential deal emerged last week, the stratospheric value of the transaction and the swiftness with which it closed left many industry observers in awe, underscoring the massive expectations around AI as well as fears of an overheated market that could be due for a reckoning.
According to reporting in Bloomberg, the deal between SpaceX and xAI will lead to a combined enterprise value of $1.25 trillion, with shares of xAI valued at $526.59 apiece. Musk has reportedly been hashing out the potential terms of a SpaceX IPO this year that would value the company at $800 billion, setting the stage for what could be the largest initial public offering of all time.—Amanda Gerut
Palantir Q4 results wow Wall Street
Palantir Technologies has come under fire in some sectors for its role working with ICE, but on Wall Street, investors cheered Monday as the data analytics firm delivered a powerful combination of faster growth, fatter margins, and a revenue outlook “crushing consensus expectations.”
The Denver-based company reported fourth-quarter revenue of about $1.41 billion, topping analyst expectations and marking another record period for the company famously named after a magical object from The Lord of the Rings. Adjusted earnings per share came in at 25 cents, two cents above consensus, while net income climbed to about $609 million.
Palantir’s AI platform remained the main growth engine, particularly in the U.S. commercial market. The company’s “boot camp” go-to-market model—short, intensive workshops where Palantir teams build live applications on customer data in days—has compressed sales cycles from months to weeks in some cases, with several organizations signing seven-figure deals shortly after attending. Shares in Palantir were up nearly 8% in post market trading. As for the controversial ICE contracts? The topic never came up on the earnings call.—Nick Licthenberg
OpenAI fires back at Claude Cowork with Codex app
OpenAI is in an increasingly cutthroat fight with Anthropic and Google to be the platform on which software developers—and increasingly others, too—build new software applications.
On Monday, OpenAI launched its latest salvo in the battle, with the debut of a desktop app version of Codex coding assistant. In January, Anthropic launched a similar application for its popular Claude coding assistant, called Code Cowork.
OpenAI is betting that developers want centralized control over agent workflows: The new Codex app lets users run multiple AI agents at once across different projects, automate repetitive tasks, and monitor what the agents are doing. And it's designed to be more intuitive than earlier versions of Codex, potentially broadening Codex’s appeal beyond the software engineers who have been its primary users so far.—Beatrice Nolan
More tech
—Nvidia CEO denies ‘nonsense’ report he’s unhappy with OpenAI: ‘I really love working with Sam.’
—Oracle said it was ‘highly confident in OpenAI’s ability to raise funds.’ Cue the stock fall.
—Anthropic makes friends in science. Partners with Allen Institute and Howard Hughes Medical Institute.
—Mozilla to let users block AI features. Old school web browsing.
—AI accounting startup Fieldguide gets Goldman Sachs funding.
