President Donald Trump’s embrace of a potential financial partnership between the federal government and the artificial intelligence industry is turning the political fight about tech on its head — and dividing people in Silicon Valley on whether to pursue such a tight alliance with Washington.
It’s unclear how quickly the administration is ready to move: The White House has yet to invite major tech companies to a meeting that the president had said could occur as early as this week, said three people familiar with the policy discussions and granted anonymity to describe them. Trump and his aides also haven’t spelled exactly how such a partnership would work.
But the fact that such an idea is getting a serious airing is a sign that the explosive growth of AI is kicking off a new debate in earnest: How to rechannel an avalanche of expected AI profits to the public — and maybe defuse opposition to the technology’s spread.
And the idea is adding to AI’s strange ideological bedfellows: Trump’s talk of a partnership may align him with Sen. Bernie Sanders (I-Vt.), who has proposed having the government take a large share of AI companies’ stock and distribute the proceeds to Americans. Meanwhile, the tech giant OpenAI is also urging the government to redistribute its profits to the public.
“The White House, progressives in Congress and the industry itself may disagree sharply on the solution,” said Joseph Hoefer, principal and chief AI officer at the bipartisan lobbying firm Monument Advocacy. “But they're responding to the same underlying worry: that AI could generate extraordinary value, and the public wants confidence it will share in the upside rather than just absorb the disruption.”
“The defining AI debate of the next decade may not be how to regulate the technology,” Hoefer said. “It may be how to distribute the gains it creates while keeping the U.S. in the lead.”
Sanders said during a National Press Club appearance Monday that Trump is responding to an undeniable political reality: Americans are increasingly uncomfortable with an AI industry that seems poised mainly to make the wealthiest tech leaders richer.
“Trump is many things, but he is a good politician, and he understands where people are coming from, and he understands that there's a lot of anxiety with regard to AI,” said the Vermont senator, who has proposed a nationwide moratorium on the data centers at the heart of the industry’s expansion plans. (Trump, in contrast, has called for a rapid expansion of AI to beat China in dominating the technology.)
“Do I have great confidence that Trump will do the right thing?” Sanders added. “No, I don't.”
What meeting?
Trump said Friday that he planned to bring in executives from top AI companies soon to discuss a “partnership” with the federal government.
“There are concepts where pieces could be given to the American public, where the American public essentially becomes a partner with the companies … where the American people can benefit from the success of AI,” Trump told reporters, without filling in many details. “And by doing that, they’re going to like it better.”
The president added that top AI and tech companies — “all the big ones” — are “all coming to the White House, probably next week.”
But two industry officials working at major AI companies and one representative from an industry trade group said the president’s Friday comments have so far been the only indication of a meeting, with no further outreach from the White House. The people were all granted anonymity to describe the status of talks with the Trump administration.
A Trump administration official, asked about possible plans to bring in AI executives to discuss the government taking stakes in their companies, said only that the White House “continues to proactively engage across government and industry.”
Spokespeople for OpenAI, Anthropic, Microsoft, Meta and Google declined to comment.
Though details of Trump’s proposal remain scarce, the president is floating what some free-market supporters see as a partial nationalization of the AI giants, with potential perils for both the government and the industry.
If it took an equity stake in companies such as OpenAI and Anthropic, the government could send some of their profits to Americans. But that arrangement could also give the U.S. government greater control over — and responsibility for — an industry that famously embraces moving quickly and taking huge risks for potentially gargantuan rewards.
It could also put the public treasury at risk if the AI industry’s fortunes took a sharp dive, akin to the bursting of the late 1990s' dot-com bubble and the housing market's collapse in 2008.
Most fundamentally, it would also go against decades of opposition in Washington — especially under Republican presidencies — to entangling government and private enterprise. Trump has already broken with that orthodoxy before, including by having the government take a 10 percent stake in Intel last year.
One tech industry lobbyist, granted anonymity to candidly discuss the proposal, said major tech companies like Microsoft and Google may be wary about providing Trump with the ability to further influence their decision-making.
But similar ideas are being advanced by others in the tech industry — including OpenAI, whose CEO Sam Altman met last week with Sanders to discuss the senator’s plan for a sovereign wealth fund created through a one-time 50 percent tax on the stock of top AI companies. That fund would give the government “the power, through its voting shares and an equal representation on each company’s board, to block decisions that hurt our citizens and to push for policies that help them,” Sanders wrote in an op-ed in The New York Times.
Sanders said Monday that he plans to introduce his bill to establish a sovereign wealth fund "in a week or two."
His proposal is similar to OpenAI’s own plans for a sovereign wealth fund that would redistribute AI wealth to Americans. In a meeting with reporters last week, top OpenAI executive Chris Lehane said the company had hoped its proposal, which it first floated in April, would be a “catalyst for a broader conversation” in Washington.
Picking winners and losers
Taken together, the proposals from Trump, Sanders and OpenAI are exposing a divide inside Silicon Valley over whether the government should try to build the AI industry or effectively place a bet on its biggest winners.
“Investing in OpenAI and Anthropic is stock picking,” said Raj Gajwani, an AI investor who owns shares in both companies. Instead, he argued that Washington would do more to strengthen AI by investing in the infrastructure underpinning the industry. Such investments, he said, would benefit the broader ecosystem regardless of which firms dominate the market.
“Investing in data centers, electricity, regulatory relief, Intel chip fabrication — that's industrial policy, national security, and infrastructure build out, which makes a hell of a lot more sense,” Gajwani said.
But some tech executives say having the government pick winners can be a good thing. They call it inevitable — even beneficial — for the government to concentrate its investments in the AI companies most likely to succeed if taxpayers are sharing in the risk.
“It would be a political nightmare if the U.S. lost money on the investment,” said one AI CEO, granted anonymity to candidly discuss the proposals. “It's easier to pick a small number of companies that will clearly be winners.”
Others in the tech industry want nothing of the kind.
Patrick Hedger, director of policy at the tech trade association NetChoice, warned that “forcing a government partnership or financial stake upon private AI labs mimics the socialist economic models of our adversaries, compromises the First Amendment, and establishes an unprecedented bureaucratic chokehold on American innovation.”
“Nationalizing America’s leading AI labs, even if only partially, would inject Washington bureaucracy into a hyper-competitive, fast-moving industry, stalling progress when we can least afford it,” Hedger said in a statement.
A leader at the U.S. Chamber of Commerce echoed that criticism.
“In the limited instances in the past when the government has taken an equity stake, it has been in response to a national emergency, was authorized by Congress, was done in a way to limit government involvement in business decisions, and with an eye towards a quick exit,” Executive Vice President Neil Bradley said. “What is being discussed today is not in response to an emergency and appears to include none of those guardrails.”
Former White House AI czar David Sacks expressed similar concerns in a Friday post, suggesting that Sanders’ proposal would lead to the "nationalization of AI” and “accelerate the corporate-government fusion we’re already sliding toward.”
When asked about the similarities between his and Sanders’ approaches to AI profit-sharing, Trump said Friday that “as far as economics is concerned, we have certain things that aren’t that far apart.”
Even some advocates for tougher AI safety regulations are torn over proposals that would give the U.S. government a financial stake in top AI companies.
Nathan Calvin, general counsel and vice president of state affairs at AI safety group Encode, said it is encouraging to see the White House and Congress float bold policy proposals to address the technology’s potential harms. But he also worried Washington could hesitate to rein in the AI industry once it becomes a major source of funding.
“There is some tension here, where if the government does start drawing a lot of its money from these stakes, then it might be more hesitant to do something that promotes safety at the expense of near-term valuation,” Calvin said.
Katherine Long and Sam Sutton contributed to this report.
from Politics, Policy, Political News Top Stories https://ift.tt/1vXIyxL
via IFTTT
0 Commentaires