Famed AI researcher launches controversial startup to replace all human workers everywhere

Every now and then a startup of Silicon Valley is launched with such a “absurd” mission that it is difficult to distinguish if the startup is for real or just satire.

That’s the case with MechanizeA startup whose founder and the non-profit AI research organization he has founded, Epoch sets Spies on X after he announced it.

Complaints include both the mission of the startup and the implication that it mitigates the reputation of his respected research institute. (Even director of the research institute posted On X: “Yay exactly what I wanted for my birthday: a comms crisis.”)

Mechanize was launched on Thursday via a Post on X By his founder, the famous AI researcher Tamay Besiroglu. The goal of the startup, Besiroglu wrote, is “the complete automation of all the work” and “the complete automation of the economy”.

Does this mean that Mechanize works to replace every human employee with a bone from an AI agent? In essence, yes. The startup wants to provide the data, evaluations and digital environments to make employee automation of any possible task.

Besiroglu has even calculated the total addressable market of Mechanize by collecting all the wages that people are currently being paid. “The market potential here is absurdly large: employees in the US are paid in total around $ 18 trillion per year. For the whole world, the number is more than three times larger, around $ 60 trillion per year,” he wrote.

Besiroglu, however, clarified to WAN that “our immediate focus is indeed on white borders” instead of manual workers who require robotics.

The reaction to the startup was often brutal. X user Anthony Aguirre replied: “Huge respect for the work of the founders in Epoch, but sad to see this. The automation of most human work is indeed a huge price for companies, and therefore strive for many of the largest companies on earth. I think it will be a huge loss for most people.”

But the controversial part is not only the mission of this startup. The AI ​​Research Institute of Besiroglu, Epoch, analyzes the economic impact of AI and produces benchmarks for AI performance. It was believed that it was an impartial way to check the performance of SATA Frontier model makers and others.

This is not the first time that Epoch is in controversy. In December, Epoch revealed that OpenAi supported the establishment of one of its AI benchmarks, who subsequently used the Chatgpt maker to reveal his new O3 model. Social media users found the era that should be more in the front about the relationship.

When Besiroglu announced Mechanize, X User Oliver Habryka replied“Unfortunately, this seems to confirm an approach that Epoch Research was directly working on the frontier, although I hoped it would not be literally from you.”

Besiroglu says that Mechanize is supported by a WHOs WHO: Nat Friedman and Daniel Gross, Patrick Collison, Dwarkesh Patel, Jeff Dean, Sholto Douglas and Marcus Abramovitch. Friedman, Gross and Dean have not returned WAN’s request to comment.

Marcus Abramovitch confirmed that he has invested. Abramovitch is a managing partner at Crypto Hedge Fund Altx, and Self -written “Effective altruist.”

He told WAN that he invested because “the team is exceptional in many dimensions and has thought deeper at AI than someone I know.”

Also good for people?

Nevertheless, Besiroglu argues to the no -sayers that doing agents all the work will actually enrich people, not impoverishment, due to ‘explosive economic growth’. He points to one Paper that he published About the subject.

“Complete automation of work can generate enormous abundance, much higher standard of living and new goods and services that we cannot even propose today,” he told WAN.

This can be true for those who own the agents. That is, if employers pay for them instead of developing them in -house (presumably by other agents?).

On the other hand, these optimistic outlook looks over a basic fact: if people have no jobs, they do not have the income to buy all the things that the AI ​​agents produce.

Yet Besiroglu says that human wages in such ai-automated world should actually increase because such employees are more “more valuable in complementary roles that AI cannot fulfill.”

But don’t forget that the goal is that the agents do all the work. When asked about this, he explained: “Even in scenarios where wages could fall, economic well-being is not only determined by wages. People usually receive income from other sources such as rental prices, dividends and welfare of the government.”

So maybe we all earn our living from shares or real estate. If that is not, there is always well -being – if the AI ​​agents pay taxes.

Although Besiroglu Vision and Mission are clearly extreme, the technical issue he wants to solve is legitimate. If every human employee has a personal crew of agents who helps them produce more work, economic abundance could follow. And Besiroglu is undoubtedly at least one thing: a year in the age of AI agents they don’t work very well.

He notes that they are unreliable, do not store information, struggle to complete independent tasks as asked: “and cannot perform long -term plans without going off the rails.”

However, he is hardly alone to work on fixes. Gigantic companies such as Salesforce and Microsoft are buildings agent platforms. OpenAi is too. And startups from Agent are abundant: from tasks specialists (outgoing sales, financial analysis); To those who work on training data. Others work on Agent Price determination Economy.

In the meantime, Besiroglu wants you to know: mechanize is hiring.