On Wednesday, President Trump issued three executive orders, SpeechAnd released Action planOn the topic of continuing the US leadership in all AI.
The project includes dozens of proposed measures, including three “pillars”: rapid innovation, infrastructure construction, and international diplomacy and security guidance. Some of its recommendations are also thought -provoking even if extra, some clearly present ideological ends, and enrich many big -tech companies, but this project is just a combination of recommended steps.
On the other hand, all three executive orders, from each pillar actually practical shape a subset:
- A The purpose is to stop the “OK” to prevent that the federal government has only considered the DI rather than just to be considered “the truth” and “ideologically neutral” just as big language models. This process accelerates AI innovation.
- A Second The construction of AI data centers is to be accelerated. A very high industry -friendly version of AN Order Released under President Biden, it makes it available to extreme policies, such as extensive exemptions of environmental concerns, providing government grants to the world’s wealthy companies, and even offering federal land for private data centers.
- A Third Promoting and financing US AI technologies and infrastructure exports, which aims to secure US diplomatic leadership and reduce international dependence on AI system from opposing countries.
The actions for the shiny press moments erupted, including the president’s one hour -long speech and the signing of the stage. But as long as the tech industry pleases these announcements (which will give flowers to their treasury), they confused the fact that the administration is currently abolishing the policies that have made the United States the first place to become a global leader in AI.
You have to understand what has arisen from the US to maintain US leadership in AI. There are four specific public policies that helped the United States to achieve this leadership.
Investing in federal funding in R&D
Generative AI products were recently released by US companies, such as Chat GPT, with industry -driven research and development. But the R&D, which enables today’s AI, was in fact a large -scale financing like federal government agencies – such as the Department of Defense, the National Science Foundation, NASA, and the National Health Organization – began in the 1950s. It includes the first successful AI program in 1956, the first chatboat in 1961, and the first specialist system for doctors in the 1970s, as well as successes in machine learning, nervous networks, back propaganda, computer vision, and natural language processing.
The US Tax Dollar also financed progress in hardware, communication networks, and other technologies under the AI system. Public Research Funding has pursued lithium -ion batteries, micro hard drives, LCD screens, GPS, radio frequency signal compression, and in today’s smartphones, along with chips used in AI data centers, and even more. Only.
Instead of promoting this world -class research history, the Trump administration is squeezing R&D financing, firing federal scientists and squeezing well -known research universities. This week’s action plan recommends investing in R&D, but suggests the actual budget of the administration Cut Nondefense R&D by 36 %. It also suggested steps for better harmony and guidance to the Federal R&D, but there will be no more financing than coordination.
Some say that companies’ R&D investment will make a difference. However, companies research that benefits their bottom, not necessarily national interest. Public investment allows for a wide range of scientific inquiry, including basic research that lacks immediate trade applications, but sometimes the massive markets end after years or decades. That’s what happened to today’s AI industry.
To support immigration and immigrants
Beyond public R&D investment, the United States has long attracted the world’s best researchers and innovators.
Today’s Generative AI Transformer Model is based on the Chat GPT), first Described By a team in Google in 2017. Of six Eight researchers The team was born outside the United States, and the other two migrants are children.
This is not an exception. Immigrants have been the center of US leadership in AI. Among the 42 US companies involved 2025 Forbes 50 Top AI Startup Rating, 60 % has at least one immigrant couch, Institute for Development. Immigrants also supported or headed companies at the AI Eco System Center: Openi, Entropic, Google, Microsoft, Nedia, Intel, and AMD.
“Brian Drain” is a term designed to describe other countries of scientists for the benefit of Americans, leaving the United States after World War II. Sadly, this year the trend has begun to turn. Recent studies Suggest that the US administration is already losing its AI talent age through anti -immigration measures (in which it has been prosecuted. AI researchers) And deduction in R&D funding.
Banned non -computers
Attachment to talented brains is just half equality. Giving them freedom of innovation is equally important.
Silicon Valley took his name because of the 20th MidAs for, as.Century companies, who made the semiconductor from the silicon, started in 1955 with the Shakli semiconductor. Two years later, a group of employees, “treacherous eight”, a competitor, withdrew from launching a fierce semiconductor. By the end of the 1960s, other groups of former Fairchaid employees were called “Fair Folderin” in Intel, AMD, and others collectively.
Software and Internet companies were eventually followed, which was once again founded by those who worked for their predecessor. In the 1990s, former Yahoo employees founded WhatsApp, Slack and Claudira. “PayPal Mafia” confirmed LinkedIn, YouTube, and Finnik firms. Former Google employees launched more than 1,200 companies, including Instagram and Forescavires.
AI is no different. Openi has founders who work in other tech companies and former students who have gone to launch More than a dozen AI startups, including notable people, such as anthropic and anxiety.
According to many historians, the flow of the laborer and the innovation he has created was possible in a large part, as the California’s 1872 constitution’s translation is to ban non -contracting contracts in employment contracts. Protection The state actually only shared with North Dakota and Oklahoma. These agreements tie one in five US workers.
Last year, the Federal Trade Commission, under President Biden, banned non -computers across the country, but Trump’s federal judge has stopped the action. The current FTC has indicated limited assistance to the ban and it can be easier to leave. If non -computers are intact, the US AI Innovation, especially outside California, will be limited.
Pursuing the actions of distrust
One of this week’s announcements needs to investigate the FTC and review settlements that “AI puts a burden on innovation.” During the last administration, the agency was allegedly investigating Microsoft’s AI actions, and several major tech companies have settlements that their lawyers certainly see as a burden, which means that it may make recent progress in a non -confidence policy. This is a problem because, in addition to the flow of labor by banning non -computers, the anti -trust policy has also served as a key lubrication for Silicon Valley Innovation Gears.
In the second half of the 1900s against AT&T, IBM, and Microsoft, significant distrust cases, semi -conductors, software, and Internet companies allow innovation and flourish markets, such as anti -trust scholar Jovana Masaroto. Described.
William Shokali was only able to launch the first semiconductor company in Silicon Valley as AT&T was forced to grant his patent license to the Transster in the 1950s as part of a decree to resolve the DOJ Anti -Trust case against the company.
Then the initial software market began because in the late 1960s, the IBM dissolved its software and hardware offers in response to the federal government’s no -confidence pressure. As Masaroto explained, the AT&T consent of the 1950s also helped to flourish the open source software, which plays an important role in today’s technology ecosystem, including the operating systems of mobile phones and cloud computing servers.
Meanwhile, many people attribute the success of Internet companies like Google like Google to the competitive breathing room developed by the federal government’s non -confidence case against Microsoft in the 1990s.
Repeatedly, the distrust of a distant actress has enabled the formation of the next. And today, the Big Tech is suppressing the AI market. Although supporters of the Anti -Trust were rightly optimistic about the currency of the administration, which is given at the earliest appointments, this week’s announcements should reduce the enthusiasm.
I don’t want to focus where things are: we need a future in which life is better than the positive use of AI.
But if the United States wants to continue the world’s guidance in this technology, we have to invest in it, which has made American leaders the first place: Bold Public Research, Open Doors for World Capital, and Fair Competition.
Prioritizing the profit of the short-term industry over these badric principles will not just endanger our technical future-this will jeopardize the role of the United States as the superpower of the world’s innovation.
Asad is the director of artificial intelligence and technology policy in the Ramadanali Wazirblat Policy Accelerator. Earlier, he was serving as the Deputy Director of the White House Office of Science and Technology Policy Strategy under the Chief of Staff and President Biden.