Sam Altman's Empire Continues to Expand: Capital, Connections, and the AGI Roadmap

Last December, Altman and Iger announced a partnership that shocked both Silicon Valley and Hollywood. According to the agreement, OpenAI would gain authorization to use many of Disney’s classic characters—including Mickey Mouse, Darth Vader, Cinderella, and others—for its video generation application Sora. This AI tool can generate highly realistic videos with the simplest text prompts.

This was a stunning alliance, as Disney has long been renowned for its strict protection of intellectual property, and Hollywood as a whole views AI as an existential threat. The negotiations took over a year. According to the agreement, Disney can use videos generated by Sora on its Disney+ streaming platform. Altman also successfully persuaded the entertainment giant to make a $1 billion equity investment in OpenAI, securing Hollywood’s most significant endorsement for the AI giant.

“Sam hopes this investment is both a symbol of confidence and essentially a way to cement the partnership,” Iger stated. “It gives Disney more direct interests in this collaboration.”

This also speaks to Altman’s influence—as OpenAI’s scale has rapidly ballooned. On the first day of Trump’s second term, Altman appeared at the White House alongside Trump, Oracle co-founder Larry Ellison, and SoftBank billionaire investor Masayoshi Son to announce the “Stargate Project”: a $500 billion U.S. AI infrastructure investment commitment. This was a bold move, fitting the style of both an extremely expansionist president and risk-loving investors like Masayoshi Son. But the one pushing for even greater scale was Altman himself.

“We discussed it, and he said ‘the more, the better,’” Masayoshi Son said. “The more, the better.”

Altman stated that Trump is easy to work with in the AI field, even though the current administration’s nationalist policies don’t completely align with his or OpenAI’s philosophies.

“His duty is to ensure America wins. I see our mission as serving all of humanity,” Altman said. “There is some conflict between those two.”

That said, as OpenAI embarks on large-scale plans for the future, there is some synergy in both parties’ expansionist tendencies. In addition to ChatGPT, Sora, and the secret hardware project led by Jony Ive, the company is also developing custom AI chips, building a social media app to compete with X, and even considering humanoid factory robots. In January this year, OpenAI released a suite of software tools for healthcare organizations and launched a freemium, ad-supported business model for ChatGPT. OpenAI’s Chief Research Officer Mark Chen said the company aims to develop an AI researcher “intern” within the next year to help the team accelerate ideas.

“We are moving toward a system capable of autonomous innovation,” Altman said. “I think most people in the world haven’t really grasped what that means.”

Critics observe all this and believe Altman just wants to make OpenAI “too big to fail,” but his allies disagree.

“I don’t think there’s any secret plan,” said OpenAI Chairman Bret Taylor. “Everyone is just very excited about the impact AI will have on humanity.”

Graham believes this is simply Altman’s nature.

“If he sees an opportunity no one else is seizing, he finds it hard not to act,” he said, noting that the former protégé has a particular weakness for underestimated things. “I bet he even finds it hard to resist buying commercial real estate in San Francisco.”

Altman holds shares in over 400 companies, which might give the impression he lacks focus. Several OpenAI employees have said they worry the company is trying to do too much in too short a time. They are concerned whether the company can maintain its lead in the model race, especially after the generally underwhelming performance of GPT-5. And when Apple chose Google’s AI model to power the next-generation Siri, employees were shaken—this deal was almost in OpenAI’s pocket, since OpenAI was already supporting Apple Intelligence.

“Yeah, that wasn’t great,” said one engineer. “A lot of us thought it was a done deal.”

Altman himself says he focuses “110%” of his energy on OpenAI and its core mission of artificial general intelligence (AGI). The definition of AGI itself is vague—it could be achieved in three years, thirty years, or even longer. At one point, he even declared victory outright:

“We’ve basically built AGI, or we’re very close.”

When asked about this statement, Microsoft CEO Satya Nadella poured cold water on it. “I think we’re still far from AGI,” he said with a smile. “We have a pretty good advancement process. It’s not up to Sam or me to declare it.”

Even as one of OpenAI’s most important partners, Nadella admits that there is an inherent “friction” between the two companies in the AI field.

“There will be gray areas,” he said. “So the term ‘frenemies’—I think it’s a fitting description of our relationship.”

A few days later, Altman retracted his statement. “That was meant in a spiritual sense, not literally,” he said.

He acknowledged that achieving AGI would require “many medium-sized breakthroughs. I don’t think we need one giant leap.”

Altman understands that his motives are hard for some people to comprehend. Even his longtime mentor Graham says, “It’s hard to know what goes on in his head.” The OpenAI CEO’s insistence on immediate, radical expansion has often drawn criticism. For example, he has publicly pledged to invest $1.4 trillion over the next eight years, mainly in AI chips and data centers. In his view, keeping up with the exponential growth in AI usage requires such capital and computing power—it’s “self-evident.”

“Then everyone else in the world will say, ‘You have to face financial reality.’ And I’m not very good at balancing those two opposing perspectives at the same time,” he said.

Altman has set a rather simple succession plan for OpenAI: hand the company over to an AI model. If the goal is to advance artificial intelligence enough to run a company, he asks, why not his own?

“I would never get in the way of that,” he said. “I should be the most willing person to do it.”

And then?

He said that apart from OpenAI, he has no other career ambitions—except for one: in a world after AGI arrives, he might find passion in a new kind of job that doesn’t yet exist.

“Most of the things I really wanted to accomplish are done,” he said. “I feel like now I’m just earning extra credit.”

Editor: Chen Yujia