While the normies were resting, I was mainlining tech discourse all weekend to bring you the latest trends, rumors, fights, and innovations from the sweatiest corners of the internet. This week, OpenClaw joins OpenAI, Anthropic really wants you to think it’s ushering in God, and the AI bros discover it’s good to have personality.
This is Manic Monday Twisted Tuesday.
An Open relationship began on Valentine’s Day…
Peter Steinberger, the Austrian founder of the viral AI agent OpenClaw (fka MoltBot, fka Clawdbot), announced on Saturday that he is joining OpenAI to do ... stuff with agents. Sam Altman did not add many details in his statement on the acquisition, saying of Steinberger, “He is a genius with a lot of amazing ideas about the future of very smart agents interacting with each other to do very useful things for people.” Which is Altman’s extremely specific way of saying that he is not quite sure what OpenAI is going to do with OpenClaw or Steinberger insofar as generating revenue, which is currently OpenAI’s top concern.
Despite the deal’s many headlines, the OpenClaw craze remains largely contained to AI evangelists with social media addictions. Altman said he expects “this will quickly become core to our product offerings,” but even if we assume “this” means an OpenClaw-like, personal agent product (OpenClaw will move to a foundation and remain open-source, which OpenAI has had a lot of success with), I imagine mass adoption will take longer than Altman’s balance sheet is hoping for.
…and Meta, uninvited to the love-fest, buys itself flowers
And by flowers, I mean a different AI agent company called Manus. It’s a classic Zuck move (see: Stories eating Snapchat’s lunch, Reels eating TikTok’s), to which I say, good for you, buddy, good for you.
Anthropic is beefing with the Pentagon…
After months of heated negotiations about how the military might use Anthropic’s government-specific LLM Claude Gov, Defense Secretary Pete Hegseth said he is “close” to cutting ties with the AI lab, Axios reported yesterday. Reps from the Department of War have said they should be able to use Claude for “all lawful purposes,” including uses Anthropic CEO Dario Amodei has publicly opposed, including mass surveillance of civilians and autonomous deployment of weapons.
Last year, Anthropic won a two-year contract with the defense department, but this frisson opens the door to the next contract to be handed to other labs like OpenAI, Google, or (god forbid) xAI.
…meanwhile, their cyber-elf philosopher queen is on the PR circuit
Elsewhere in Anthropica, the Wall Street Journal published a profile of the lab’s “resident philosopher” Amanda Askell. The profile leans into Askell’s anthropomorphization of Claude — by “developing Claude’s understanding of itself,” she is like “a parent raising a child” — and paints Askell as a white-haired, blue-eyed, elfin philosopher (Ph.D. from NYU) from the High North (Scotland), who’s otherworldly, wise, and ageless (she’s 37).
The discourse surrounding the profile mainly involves people wondering whether it’s wise to hand over Claude’s moral compass to a singular person with biases and quirks and an unclear political worldview. But there’s a more important point to raise here: Anthropic’s publicity team has been on a run lately, inserting its lab into high-minded features about AI and God and the balance between good and evil. These articles provide little insight into the lab’s business; more importantly, they advance Anthropic’s brand story as an intellectual and truth-seeking hero among a swarm of greedy, lowbrow competitors. If you can’t build a profit, you might as well build a cult.
AI people have decided that being cool and artsy is valuable after all
“Taste is a new core skill” and “startups in the future will get acquired merely for the founder’s taste profile” is what thenerds are saying, which is about the stiffest and tech-poisoned way to say it’s good to have interests and know yourself. Awesome!
The week ahead: Will MrBeast join the a16z New Media cabal on its recent hiring blitz?
