The aim of artificial general intelligence, or AGI—technology that equals or surpasses human cognitive capacities—is what tech companies are racing toward. Elon Musk, the CEO of Tesla, and Dario Amodei, the CEO of Anthropic, have both forecast that human-level AI may be developed as early as this year. Business executives are optimistic about AI, but if it is not managed, experts warn that it might have disastrous effects.
In an interview with host Steven Bartlett on The Diary of a CEO podcast in November, former Google insider and AI specialist Tristan Harris talked about the quest for artificial general intelligence (AGI), which he admits the majority of business executives think might be achieved as early as 2027. The rush to develop AI on par with humans, according to Harris, may lead to detrimental incentives for unbridled growth, which would eventually worsen safety, security, and economic well-being.
Harris described it as a competitive logic that reinforces itself. It incentivizes everyone to take the most shortcuts, to care the least about safety or security, to care less about how many jobs are lost, and to care less about the well-being of ordinary people.
AI businesses operate with little restriction these days. Biden-era AI policies that protected workers facing job disruptions and ensured safe and secure implementation were taken back by President Donald Trump on the first day of his second administration. Additionally, Trump signed an executive order in December that forbade the regulation of the technology, avoiding a patchwork of state regulations that he claimed may “stymie innovation.” According to Harris, the average American does not benefit from unrestricted AI development.
According to him, “the default path is not in [the people’s] interest.” “By default, businesses are racing to release the most potent, mysterious, and uncontrollable technology we’ve ever created with the greatest incentive to compromise on safety.”
A greater labor risk than NAFTA and immigration
Harris’s primary worries regarding the present course of AI development are the impending effects of the technology on the labor market. He stated that the potential for powerful AI to completely replace human labor should be a bigger worry than the loss of jobs due to immigration.
According to Harris, if you’re concerned about immigrants stealing employment, you should be much more concerned about AI. There are millions of new digital immigrants who can operate at superhuman speed, have Nobel Prize-level skills, and will labor for less than the minimum wage.
Early studies demonstrate the growing influence of AI on employment. According to a recent Stanford University research that examined payroll data, AI is resulting in a 13% reduction in employment for people in their early careers. Around 55,000 layoffs occurred in 2025 as a result of the rapid introduction of AI, according to outplacement agency Challenger, Gray & Christmas. Citing a desire to use AI, Microsoft cut 9,000 workers last year. Salesforce also eliminated 4,000 customer service positions due to an AI push.
AI is similar to the North American Free Trade Agreement. It’s like NAFTA 2.0, Harris added, referring to the North American Free Trade Agreement, a trade agreement between the United States, Canada, and Mexico that detractors, including Trump, claim harms the US job market. Except that, instead of China, which will perform the manufacturing labor for cheap, this country of geniuses in an AI-created data center will arrive on the world scene, doing all of the cognitive labor in the economy for less than the minimum wage.
Brewing backlash
According to Harris’ prediction in the interview, growth may be impeded by the present unstructured AI rollout “unless there’s a massive political backlash because people recognize that this issue will dominate every other issue.”
Some states are already experiencing the first rumblings of that fallout. 26 states have passed some form of AI legislation in spite of Trump’s executive order, according to law firm Bryan Cave Leighton Paisner. States like California and New York have proposed strict regulations pertaining to data-use transparency and safety. Quick regulation is necessary, according to Harris, since as AI replaces jobs, human political authority may be diminished as human labor becomes less economically valued.
“Human political power will no longer be significant at this time,” he declared. “Does the state still require people? The AI firms account for nearly all of their GDP. This political class and political power base suddenly turned into a useless class.
