Trump pledged to gut Biden’s AI rules, as OpenAI eyes landmark infusionA policy proposal presented by the AI start-up warns that if the United States doesn’t invest in the technology it will lose its technological edge to China.
https://www.washingtonpost.com/technology/2024/11/13/openai-nuclear-subsidies-trump-ai-china/OpenAI outlined a proposal on Wednesday calling on the United States to support the artificial intelligence industry with a landmark infusion of funding and resources, previewing a potential lobbying and policy battle emerging for the incoming Trump administration.
he proposal, presented by OpenAI’s head of global affairs, Chris Lehane, at a think tank event in Washington, calls for special economic zones with fewer regulations to incentivize new AI projects, a fleet of small nuclear reactors to power data centers aided by the U.S. Navy and a “North American Compact” allowing U.S. allies to collaborate to bolster the field, according to a document provided by OpenAI.
The proposal comes as President-elect Donald Trump has announced plans to repeal President Joe Biden’s AI executive order, arguing it “hinders” innovation. Yet Trump has emphasized some of the same issues raised in the proposal, including the increased need for power for the U.S. to lead on advanced AI technology, signaling potential common ground. Trump allies earlier this year drafted a plan to cut tech regulations and invest in military tech.
Without government investment, the United States will lose its technological edge over China and other competitors, OpenAI leaders warn in the document.
“Given the stakes, we need to think big, act big, and build big,” the OpenAI proposal reads. “These decisions determine whether a nation leads or lags in technological innovation, often with far-reaching consequences for economic competitiveness and national security.”
OpenAI spokeswoman Liz Bourgeois declined to comment beyond the proposal.
The policy recommendations signal an escalation of OpenAI’s relationship with the U.S. government, pushing the federal branch to invest in, subsidize and boost the AI industry. OpenAI CEO Sam Altman has lobbied governments in the United States and the Middle East to support a plan to raise billions of dollars to build an infrastructure company that would build new chips.
“I think what you will see is … this focus on the importance of American leadership in AI innovation,” said Jennifer Huddleston, a senior fellow in technology policy at the Cato Institute. “That’s certainly something that we can expect to continue, and aligns with some of the rhetoric we’ve seen from the Trump White House.”
AI and Big Tech executives have said the race for better AI is essential to America’s future and have warned that if the United States does not maintain its technological edge, then China will use AI to constrain American military and commercial power. Politicians have emphasized that the push for the government to boost the AI industry has bipartisan backing.
China’s high-tech ambitions began raising alarms in Washington in 2015, when Chinese leader Xi Jinping’s Made in China 2025 plan disclosed the nation’s aim to rival the United States in key technological fields. In 2017, soon after Trump began his first presidency, Beijing policymakers added a target to become the world’s leader in AI by 2030.
While the United States is still well ahead in artificial intelligence, China has narrowed the gap and has certain advantages such as lower labor costs for time-consuming tasks such as training AI models, an asset that worries U.S. officials.
Biden tried to boost the U.S. tech industry while slowing down China’s AI progress with billions of dollars of investments in new computer chip factories and restrictions on exporting high-end AI tech to China. Trump has already signaled an aggressive military approach to China, with the appointments of fierce China hawks such as Rep. Michael Waltz (R-Florida) and Sen. Marco Rubio (R-Florida) to important foreign policy positions.
Training and running AI software is extremely expensive and energy-intensive. Companies need to buy thousands of state-of-the-art computer chips and run computations through them constantly, using electricity to run them at full speed and water to keep them cool. Google, Microsoft and Facebook have spent tens of billions of dollars this year building new data centers and filling them with computer chips.
Despite Big Tech companies’ pledges to lower their carbon emissions, many have ramped up their energy use as they build data centers to train new AI models and run AI applications for their customers. The boom in new data centers for AI has, in some places, increased energy prices for everyday customers.
During the campaign, Trump met with supporters in Silicon Valley and said on the popular All-In tech podcast that he recognized how much the AI industry needed energy production.
“They need electricity at levels that nobody’s ever experienced before to be successful, to be a leader in AI,” Trump said. “The amount of electricity — that needs double what we have right now and even triple what we have right now.”
Some tech leaders have said improvements in AI software and computer chip efficiency will bring down the technology’s soaring energy costs. But Altman has argued that a massive increase in computing power is necessary even with these improvements.
“The history of the U.S. is one of iconic infrastructure projects that moved the country forward: the auto industry, the Tennessee Valley Authority, the Manhattan Project, the interstate highway system,” OpenAI said in the proposal.
The document suggests a “National Transmission Highway Act” that would invest in improvements to the electrical grid, high-speed internet and new energy sources for training AI. The proposal calls for the government to “unblock the planning, permitting and payment” for new projects.
OpenAI also called on the government to commit to buying energy from new electricity-generating projects to make them less risky for private companies to build; the government should also fund new jobs programs to train workers for data centers. A “North American Compact” for AI would create an economic bloc of countries in the Western Hemisphere that could allow money, people with expertise and crucial supplies to smoothly cross borders, giving the United States even more economic heft to win the AI arms race.
AI proponents often lament the United States’ lagging approach to nuclear power generation, and some AI companies have struck deals to get shuttered nuclear plants back online or invested in start-ups trying to build cheaper, smaller reactors. The OpenAI proposal suggests tapping the U.S. Navy’s expertise in running the small nuclear reactors that power some of its submarines and surface ships. A renaissance in nuclear technology will help drive reindustrialization in the United States, OpenAI said.