There is an ideology at the center of every empire, a belief system that leads to the system and justifies the expansion – even if the cost of that expansion directly defines the declared mission of the ideology.
For European colonial powers, it was a promise to save souls while removing Christianity and resources. For today’s AI Empire, it is artificial general intelligence for “benefiting all humanity”. And Openai is its main campaigner, spreading enthusiasm throughout the industry, which has created AI.
“I was interviewing those whose voices were shaking with the enthusiasm of his beliefs in AGI,” Karen Hao, Journalist and “Bestsailing Writer of”AI Empire“On a recent episode of Equity told Techcrunch.
In his book, Hao compares the AI industry in general, and especially openi, to an empire.
“The only way to really understand the scope and scale of openi’s behavior … really to recognize that they have already become much more powerful than any nation state in the world, and they have consolidated not only economic power, but also the extraordinary amount of political power,” Hao said. “They are tearforming the Earth. They are rebuilding all the geopolitics of all our lives. And so you can only describe it as an empire.”
Openai has “described AGI as a highly autonomous system that improves humans at the most economically valuable work,” one that will increase humanity by increasing abundance “in any way, turbocorizing the economy, and assisting in the discovery of new scientific knowledge that changes the limits of possibility.”
These nebulace promises have promoted the exponential growth of the industry – its massive resource demands, the desire to release the oceans, stressful energy grids, and unused systems in the world. Many experts say in all a future service that can never come.
Techcrunch event
San francisco
,
27-29 October, 2025
Hao says that this path was not unavoidable, and this scaling is not the only way to achieve more advances in AI.
“You can also develop new techniques in the algorithms,” he said. “You can improve the existing algorithms to reduce and calculate the amount of data they need to use.”
But that strategy must renounce speed.
Hao said, “When you define the discovery of the construction of a beneficial AGI, where Victor takes everyone – which was done by Openi – so the most important thing is that there is speed on anything,” said Hao. “Speed on efficiency, speed on safety, speed on searching.”
For Openai, he said, the best way to guarantee speed was to take existing techniques and “simply talk intellectually cheap, which is to pump more data, more supercomputer, in those current techniques.”
Openai set the stage, and instead of falling back, other technical companies decided to fall into the line.
“And because the AI industry has successfully captured most of the top AI researchers in the world, and those researchers are no longer present in academics, you have a complete discipline that is now being shaped by the agenda of these companies, rather than real scientific exploration, instead of real scientific exploration,” Hao said.
Spending, and will be astronomical. Last week, Openai said it expects to burn $ 115 billion cash By 2029. Meta said in July that it will spend $ 72 billion on the construction of AI infrastructure this year. Google hopes he expects to hit $ 85 billion In capital expenditure for 2025, most of which will be spent on the expansion of AI and cloud infrastructure.
Meanwhile, the goals keep running posts, and “benefits to humanity” have not yet become physical, even horms mount. Loss of job loss, concentration of money, and disadvantages such as fuel confusion and psychosis. In his book, developed countries such as Hao Kenya and Venezuela also have a document of workers, which were exposed to material harassing materials including child sexual abuse materials, and very little wages were paid – about $ 1 to $ 2 per hour – in roles such as material moderation and data labeling.
Hao said that this is a wrong trade for the progress of AI against current loss, especially when other forms of AI provide real benefits.
He pointed to the Nobel Prize winner of Google Deepmind, Amino acid sequence and complex protein folded structures, and can now make accurate prediction of the 3D structure of the protein from their amino acids-the discovery and understanding of the 3D structure of the protein can make the 3D structure of the protein from their amino acids.
“They are types of AI systems that we need,” Hao said. “Alphafold does not cause mental health crisis in people. Alphafold does not suffer environmental damage … because it is trained on a very low infrastructure. This material does not manufacture the loss of moderation because (not near dataset) all toxic nonsense that you were scrapping on the Internet.”
There has been a legend about the importance of racing to defeat China in the AI race along with semi-religious commitment to AGI, to have a liberal impact on the Silicon Valley world.
“Really, the opposite is,” Hao said. “The interval has continued to shut down between the US and China, and the Silicon Valley has had an unqualified impact on the world … and the only actor who has come out of it, you can argue, Silicon Valley is the only one.”
Of course, many people would argue that Openai and other AI companies have benefited humanity by releasing chatgpt and other large language models, which promise huge benefits in productivity by automating tasks such as coding, writing, research, customer support and other knowledge-work works.
But the way Openai is structured-part non-profit, part for part-it complicates how it defines and measures its effects on humanity. And this week it is complex from this news that Openai reached an agreement with Microsoft that eventually brings it closer to public.
Two pre-Openai security researchers told Techcrunch that they feared AI Lab began to confuse their profit-profit and non-profit missions-because people enjoy using Chatgpt and other products built on LLMS, it ticks the box to benefit humanity.
Hao resonated these concerns, describing the dangers of consumed so much by the mission that the reality was ignored.
Hao said, “Even when evidence collects that the construction they are doing is actually damaging people in significant quantities, the mission has continued the paper on all that,” said Hao. “There is really dangerous and dark about it, (being) so wrapped in a belief system that you have made that you lose touch with reality.”
