What a difference a week makes.

As others and I have predicted, newly-inaugurated President Trump rolled back much of his predecessor’s policy approach to artificial intelligence (AI). But few forecasted just how stark the contrast has been.

Last week, before the inauguration, I argued that Trump should reverse outgoing President Biden’s ill-considered Executive Order on AI. As I explain in my forthcoming book on AI policy, the top-down, one-size-fits-all measure threatened to stifle machine innovation and to invite regulatory capture.

And sure enough, hours after his swearing-in, Trump revoked the AI order, along with dozens of others promulgated by his predecessor in order “to commence the policies that will make our Nation united, fair, safe, and prosperous again.” This action comports with Trump’s approach to AI during his previous administration, when he declared, “realizing the full potential of AI for the Nation requires the combined efforts of industry, academia, and government.”

But the Trump White House went even farther, a day later announcing a major AI partnership between OpenAI, Nvidia, Softbank, Oracle, Microsoft, and others. Pledging $500 billion in investment over the next decade, OpenAI CEO Sam Altman introduced the Stargate Project, which aims to “secure American leadership in AI, create hundreds of thousands of American jobs, and generate massive economic benefit for the entire world,” as well as “provide a strategic capability to protect the national security of America and its allies.”

The Verge reported that Altman told Trump he launched the project after and because of the president’s November victory, while, according to the AP, Softbank CEO Masayoshi Son proclaimed “the beginning of a new golden age”—a not-so-subtle callback to Trump’s (second) inaugural address. It seems the new administration is determined to spur robust AI investment and technical advances.

The return of the 47th president to the White House has had ripple effects on AI as well. A day after the inauguration, the Washington Post reported that Google had partnered with Israel’s Ministry of Defense after the October 7, 2023 Hamas invasion to provide access to its latest AI technology, known as Nimbus. “Thanks to the Nimbus public cloud,” an Israeli defense official said, “phenomenal things are happening during the fighting, these things play a significant part in the victory — I will not elaborate.” While this relationship developed more than a year ago, its disclosure awaited Trump’s arrival at 1600 Pennsylvania Avenue—another signal that a new AI sheriff is in town.

Finally, on the day after Trump’s inaugural, the CEO of Anthropic, a major AI developer founded by former OpenAI developers and executives, declared that robotic capabilities will soon outstrip those of humans. “I don’t know exactly when it’ll come,” CEO Dario Amodei told the Wall Street Journal. “I don’t know if it’ll be 2027…I don’t think it will be a whole bunch longer than that when AI systems are better than humans at almost everything. Better than almost all humans at almost everything. And then eventually better than all humans at everything.”

Amodei’s comments touch on another key debate I’ve detailed in this space and explore in my book—whether and how soon machines will acquire artificial general intelligence (AGI), or the ability to surpass humans in most endeavors, and how we as a society will adapt to the opportunities and challenges AGI presents. Many AI practitioners, philosophers, and policymakers doubt that machines have achieved, or will ever achieve, true autonomy, while others are more bullish. If early developments in the new presidential terms are any indication, it appears that President Trump finds himself on the side of the latter.