Sam Altman’s escape from the Silicon Valley pack is a sprawling farm on the finish of a highway that snakes by way of the vine-swept hills in Napa Valley. I spot the 40-year-old with barely tousled hair within the open-plan kitchen of the extensive bay-windowed home, and I step proper in. His bemused look tells me that I wasn’t anticipated. I’m, it seems, practically an hour early, however the man behind ChatGPT will end up a gathering and be a part of me within the backyard. I wait beneath a grapevine-shaded pergola that runs alongside the home.
For a Lunch with the FT, Altman supplied to prepare dinner a easy vegetarian meal at his farm as a substitute of assembly me at a restaurant of his selection, the place he’s more likely to be hounded by selfie-seekers. Since OpenAI, the corporate he runs, launched the generative AI mannequin in 2022, Altman has been catapulted to the standing of worldwide celeb. Final 12 months, he married his software program engineer boyfriend they usually’ve just lately had a child boy by way of surrogacy (he consulted ChatGPT on which crib to purchase) so he’s been spending extra time on the Napa farm.
Altman takes only some minutes to affix me outdoors. He has constructed OpenAI into one of many fastest-growing firms ever, with a staggering valuation of greater than $250bn, and accelerated a fierce race for AI supremacy: the pursuit of synthetic common intelligence, when a machine can surpass the cognitive talents of people, not solely absorbing data however reasoning and studying by itself.
Nevertheless it has been a rollercoaster journey wherein Altman has been fired and rehired by his personal firm, his character and dedication to the protected improvement of AI subjected to bruising questions. He’s wrestled with Elon Musk, with whom he co-founded OpenAI, and sparred with Scarlett Johansson, who accused him of utilizing an “eerily related” voice to her personal to coach a chatbot. Having stolen a march on extra established opponents (Google particularly, provided that it has lengthy had the lead in AI analysis), he’s been courted by presidents and prime ministers and he has seduced a number of the world’s largest buyers.
I discover Altman brimming with confidence as our dialog ranges from AI merchandise to the existential query of an AI future {that a} handful of optimistic technologists are steadily main us to, whether or not we prefer it or not. Radiating ambition, he feels like a person satisfied of his personal future. He tells me that he has the “coolest, most necessary job perhaps in historical past” and whereas he used to assume AI was as consequential because the Industrial Revolution, he now reckons the “explosion in creativity” makes the Renaissance a extra apt analogy.
We’re assembly quickly after the discharge of OpenAI’s o3, a extra superior AI mannequin with improved functionality to cause and generate pictures. It’s, he says, an necessary step in direction of the creation of AI brokers that may execute duties on people’ behalf, and which all main AI firms are furiously pursuing. “Persons are saying like, that is . . . genius-level intelligence,” he gushes.
No sooner was the software launched than customers flooded the web with pictures generated within the type of the Japanese animation home Studio Ghibli. That gave Altman and OpenAI a sensational advertising and marketing increase, however renewed questions concerning the liberal use of others’ mental property to coach AI fashions and generate artwork. Altman says compensation for artists could also be required (his firm has carried out licensing offers with publishers, together with the FT), however he prefers to place the instruments out into the world after which discover solutions to questions that come up.
“There are some people who find themselves, like, ‘all AI artwork is horrible’, however then there’s plenty of artists who’re, like, ‘that is the very best software ever, it’s just like the invention of the digicam’,” says Altman. “We agree we’d like a brand new enterprise mannequin for this type of a world, however what it’s, the group remains to be form of feeling their manner by way of. I do know that we’ve obtained to converge on what it ought to be.”
We’re again within the kitchen and I watch Altman season with cumin the yellow and orange carrots grown on the farm, that are then roasted within the oven. With spectacular willpower, he chops an unlimited quantity of garlic, which he tosses right into a pan with crimson chilli peppers, walnuts, parsley and pecorino to make what seems like a Californian tackle aglio e olio spaghetti. The salad leaves, with thinly sliced carrots and radishes, are within the fridge already and wish solely dressing. Altman visibly enjoys cooking and, as I’ll quickly discover out, is fairly good at it.
As we discuss, I seek for clues in his upbringing that trace at his future stardom. He says there are none. “I used to be like a form of nerdy Jewish child within the Midwest . . . So know-how was simply not a factor. Like being into computer systems was form of, like, uncommon. And I definitely by no means might have imagined that I’d have ended up engaged on this know-how in such a manner. I nonetheless really feel form of surreal that that occurred.”
The eldest of the 4 youngsters of a dermatologist mom and a father who labored in actual property, Altman learn plenty of science-fiction books, watched Star Trek and appreciated computer systems. In 2005, he dropped out of Stanford College earlier than graduating to launch a social networking start-up. In these days, AI was nonetheless in its infancy: “We might present a system a thousand pictures of cats, and a thousand pictures of canines, after which it [the AI] might accurately classify them, and that was, like, you have been dwelling the excessive life.”
Quickly, Altman was operating Y Combinator, a start-up accelerator that had backed his first enterprise. He was nonetheless there when he began OpenAI in 2015 as a non-profit with a mission that synthetic common intelligence ought to profit all of humanity. Musk was a co-founder who financed it with tens of hundreds of thousands of {dollars} however then fell out with Altman and left its board in 2018 in one among Silicon Valley’s most carefully watched feuds. Musk, who has been rising his personal AI rival firm, xAI, alleges in a lawsuit that Altman had strayed from the corporate’s founding mission. Altman has countersued: “It appears clear to me why he’s doing all of this,” he says of Musk. “As a result of he’s attempting to decelerate a competitor and he doesn’t like that he’s like not profitable in AI.”
Menu
Sam Altman’s farm
Napa Valley, California
Pasta with chillies, garlic, walnuts, parsley and pecorino
Roasted carrots
Salad leaves with carrots and radishes
Espresso
Tea
Musk could have his private motives however the debate over learn how to “win” in AI — which requires large computing energy and funding — with out loosening dedication to security, has lengthy divided OpenAI. The splits exploded into the open in November 2023, when the non-profit board immediately fired Altman, with one board member later accusing him of misrepresenting data and in some circumstances mendacity to the board.
Altman prevailed. He was reinstated inside days, after practically all workers threatened to resign and Microsoft, then the corporate’s largest monetary backer, supplied to rent Altman and his staff. “It was very painful and really embarrassing that this entire factor occurred, and nobody, together with me, actually knew all of it on the time. What I needed was to go sit on a seashore and get better however I needed to simply maintain operating the corporate and now clear up a huge mess.” That mess, he says, included prospects and buyers asking whether or not they might depend upon the corporate.
The storm led to the departure of a few of OpenAI’s greatest researchers and left lingering questions on Altman. Nevertheless it additionally cemented his standing as the corporate’s undisputed chief, with a brand new board that backs him.
His ambitions too have expanded: early this 12 months, he starred in one other headline-grabbing occasion, showing on the White Home with Donald Trump to announce a three way partnership with Japan’s SoftBank that can elevate a whole lot of billions of {dollars} to develop AI infrastructure, together with information centres.
However whereas he has shifted the narrative round OpenAI’s mission as its business potential has turn out to be extra obvious, makes an attempt to vary the corporate construction to a extra conventional for-profit enterprise have met with resistance, from Musk in addition to from AI consultants who insist the corporate should stay beneath a non-profit board to fulfil its mission of creating know-how to profit all humanity.
I ask Altman whether or not he learnt from the tried coup. One criticism of him is that he tells folks what they wish to hear, relying on what’s expedient. All he’ll admit to is that he does choose to keep away from battle and he’s needed to be taught shortly learn how to run such a posh firm. “Within the final two years we have now gone by way of a decade and half of a traditional firm’s progress.”
It’s a busy day on the farm. Altman’s mom is visiting, as are his in-laws. His husband and son, in addition to a colleague, are in the home too. They arrive out and in of the kitchen. Altman lays the bowls of meals on a desk, we serve ourselves and return to take a seat within the backyard. The pasta is scrumptious, with simply the correct amount of spice, the carrots crunchy, and the salad tastes mild and juicy.
As a chief disrupter who’s hyperactive on social media and, by his personal description, OpenAI’s “advertising and marketing division”, he generates a continuing stream of reports. Just lately that has included a dispute with an estranged sister who has accused him of sexually abusing her when she was a toddler (the remainder of the household has backed his denials and he says he feels each compassion and upset in direction of a sister who’s “had a tough time of it for a very long time”).
Altman claims that being aggressive just isn’t one among his defining options — “Am I not an outlier by way of competitiveness,” he asks aloud, “in comparison with different tech CEOs?” — but he relishes discuss of profitable. He admits that he has entertained operating for governor of California (not a presidential bid although, as some have claimed); his favorite manner of describing the attain of ChatGPT just isn’t 800mn customers however “10 per cent of the world”.
However how a lot does the aggressive race finally matter? Whereas Silicon Valley has been sinking large funding into AI, DeepSeek, a Chinese language start-up, launched a mannequin this 12 months developed on a restricted finances. That instructed that AI fashions have been turning into commoditised and the US technological edge over China was diminishing. Altman says there may be an “asterisk” to the commoditisation narrative: “Most of those fashions will probably be commoditised. The frontier fashions I don’t assume will probably be.” He’s, after all, anticipating to prevail in frontier fashions but additionally win within the commoditisation sport, given what number of customers are already connected to ChatGPT.
And the way will OpenAI ship the returns on large funding? Altman hints at his final purpose, however describes it as only one compelling thought: when a subscription to ChatGPT turns into a private AI, by way of which customers log into different providers. “You might simply take your AI, which is gonna get to know you higher over the course of your life, have your information in and be extra personalised and you possibly can use it wherever. That will be a really cool platform to supply.”
The solar is simply too sturdy so we return inside. Altman is sipping his tea, curled up on the lounge couch, his arms wrapped round his legs. We discuss concerning the AI-dominated future his son will inherit. Ailments could also be extra quickly cured by AI, and sectors from schooling to banking reworked. However it’s a world that additionally raises existential questions on the way in which we dwell. Why ought to society belief a handful of AI males to resolve on the form of the longer term? In a response unlikely to persuade, he says these creating the know-how are “dedicated to assembly the gravity of the second with accountable know-how”.
AI progress is shifting at such breathtaking velocity that some consultants favour slowing down till internationally agreed norms and laws are put in place. Two years in the past, Altman himself signed a assertion with others within the subject cautioning that “mitigating the chance of extinction from AI ought to be a world precedence alongside different societal-scale dangers reminiscent of pandemics and nuclear struggle”.
Altman insists he hasn’t modified his thoughts and there will probably be moments when holding again could also be required. For now, he appears happy with the rollout of instruments for folks to experiment with and assess the dangers. “The world must find out about it [AI], the world must weigh in on it, very closely. By having our customers assist us resolve what the boundaries ought to be, like studying this collective worth, perform and preferences of humanity,” he says.
Some advances do scare him. After releasing a reminiscence characteristic that permits the AI to register previous behaviour, he has heard of circumstances the place customers turn out to be too emotionally depending on the AI. “Persons are like, that is my new greatest good friend, you possibly can by no means delete this model, I would like this factor . . . I’ve little question that we, society, will work out learn how to navigate this, however that’s a brand new factor that’s simply occurred and you may think about all types of ways in which it goes actually flawed.”
Extra alarming, I notice, is a future wherein AI brokers talk with one another with out instruction by people. Altman explains that maybe it’s not an agent that creates different brokers however an AI system that’s so good, so trusted, that it in impact controls what people do. “It [the AI model] turns into simply higher than we form of have a conception for.” This sounds so spooky that even he appears alarmed by his phrases.
Altman doesn’t strike me as a person who entertains doubt however I ask how his grand plan of constructing an AI large might go flawed. Is he assured that OpenAI will exist in 10 years? “Fixing fences and taking good care of cows” could be his plan B, he jokes. Extra critically, he says: “We might make a flawed analysis wager, you recognize, we might fall behind on product to anyone else. It’s like we’re doing a really difficult factor.”
We’ve been speaking for greater than two hours and his husband, holding the child, joins us in the lounge, and after some fussing, the tiny toddler is falling asleep. I ask whether or not Altman finds his courageous new world, wherein people usually are not probably the most clever factor on the planet, threatening — if not for him, then for his son?
He’s, predictably, too enthralled by his AI creation to really feel menace. “Do you assume you’re smarter than o3 proper now? I don’t . . . and I really feel fully unbothered, and I wager you do too,” he says. “I’m hugging my child, having fun with my tea. I’m gonna go do very thrilling work all afternoon . . . I’ll be utilizing o3 to do higher work than I used to be in a position to do a month in the past. I’ll go for a stroll tonight. I believe it’s nice. I’m extra succesful. He [his son] will probably be extra succesful than any of us can think about.”
Roula Khalaf is editor of the Monetary Instances
Discover out about our newest tales first — comply with FT Weekend on Instagram and X, and enroll to obtain the FT Weekend publication each Saturday morning
