Former Google CEO makes shocking case against AI proposal | Official Website of Louis Velazquez, entrepreneur, finance guy and tech innovator

Financial markets worldwide are currently shrouded in uncertainty, but the world of artificial intelligence (AI) isn’t slowing down.

Even as U.S. President Donald Trump’s tariffs complicate the global supply chain, many experts seem confident that AI adoption and development will continue as companies advance toward the next frontier of the revolutionary technology.

💰💸 Don’t miss the move: SIGN UP for TheStreet’s FREE Daily newsletter 💰💸

Since Chinese AI startup DeepSeek released its R1 model in January 2025, experts have been rethinking their approaches to AI. The fact that DeepSeek trained its model for only $5.6 million and built it on less advanced AI chips has prompted speculation as to the best ways to approach AI development.

From business leaders to government officials, many people seem to have opinions on this pressing topic. But one industry expert recently revealed that he sees some major problems with a recent AI proposal.

Eric Schmidt, former Google CEO and chairman has some blunt words regarding the future of artificial general intelligence (AGI).(Photo by Michael Kovac/Getty Images)

Michael Kovac/Getty Images

Business and government split on newly proposed AI initiative

AI has recently become a popular topic for government officials, as politicians on both sides of the aisle deliberate over how it should be handled.

In November 2024, a congressional commission proposed that the U.S. put together an initiative to develop artificial general intelligence (AGI) similar to the iconic Manhattan Project, created during World War II to produce nuclear weapons. However, former Google (GOOGL) chairman and CEO Eric Schmidt doesn’t believe this is a good idea.

Considered one of the tech sector’s most influential leaders, Schmidt is known for his predictions on the future of technology. Now, he and two other AI experts have published a paper in which they break down the case against a Manhattan Project for AGI.

Artificial general intelligence refers to an AI system that can outperform humans at virtually any task and demonstrate a higher level of proficiency. While reaching AGI would undoubtedly be an important step forward, developing it through a Manhattan Project-style initiative may not be the best way.

At least that’s according to the case made by Schmidt and his co-authors, ScaleAI CEO Alexander Wang and machine learning researcher Dan Hendrycks. In the paper titled “Superintelligence Strategy,” they highlight potential problems with the proposed initiative.

Much of the paper centers around the concept of “Mutual Assured AI Malfunction (MAIM): a deterrence regime resembling nuclear mutual assured destruction (MAD) where any state’s aggressive bid for unilateral AI dominance is met with preventive sabotage by rivals.”

As the authors state:

“The Manhattan Project assumes that rivals will acquiesce to an enduring imbalance or omnicide rather than move to prevent it. What begins as a push for a superweapon and global control risks prompting hostile countermeasures and escalating tensions, thereby undermining the very stability the strategy purports to secure.”

The authors seem to fear that if the U.S. moves forward with the Manhattan Project-style program for AGI development, it could spark retaliation from China. It makes sense that Schmidt would approach this area with caution. After the rise of DeepSeek, he argued that the U.S. would need to take action to remain ahead of China in the AI race.

Schmidt’s case against the AI Manhattan Project may have more layers

But is that Schmidt’s only motivation for arguing against the Manhattan Project for AGI? China has used cyberwarfare to attack the U.S. many times, so the type of retaliation highlighted in the paper likely wouldn’t be anything too new.

Kevin Korte, a member of the board of advisors at Targa.AI, sees a personal incentive for Schmidt to argue against the U.S. government taking the reins in the push to develop AGI supremacy.

“A Manhattan Project for AGI would restrict Google’s ability to monetize its developments ahead of its rivals,” he states. “It might give the government a decisive say in what can and cannot be done with AI.”

He also highlights the possibility of increased regulation on AGI if it is developed through a government initiative, which likely wouldn’t be a factor in work being done by private companies. As Korte sees it, “The opposition to an AGI Manhattan Project is motivated by trying to avoid government interference and regulation and less through a genuine worry about attacks by China.

Dev Nag, CEO of QueryPal, echoes this, stating that Schmidt fears government interference and regulation more than any threats from China.

“This clearly serves the interests of tech giants who want access to government resources and protection, but Schmidt would likely argue that in the global AI race, what’s good for the leading AI companies is good for the country,” he notes.