Openai CEO Sam Altman acknowledges that the benefits of AI cannot be widely distributed ncvrs.com

THE New essay on your personal blogSam Altman, Openai CEO, said the company is open to the “calculation budget”, including “Strange-Hangos” ideas to “allow everyone on the ground to use a lot of AIs” and secure to make the benefits of technology widespread.

“The historical impact of technological development suggests that the majority of caring indicators (health results, economic prosperity, etc.) improve on average and in the long term, but increasing equality does not seem to be a technological point of view and this right may It can be a right. They require new ideas, ”Altman wrote. “In particular, the power of power between capital and labor may be easily confused and may require early intervention.”

The solutions to the problem, such as the “Compute Bettye” concept of Altman, can be more easily recommended than implemented. Already, AI affects the labor marketresulting work -cuts and class reductionOr Experts warned This mass unemployment is a possible result of AI Tech’s growth, if not with proper government policies, with the contact and further development programs.

Altman is not the first to claim that artificial general intelligence (AGI), which defines, is close to “(an AI) system that deals with increasingly complex problems in many areas, human level, human level”. Whatever the form it is, this Agi will not be perfect – warns Altman in the sense that “you can require a lot of human supervision and direction”

“(Agi Systems) won’t be the biggest new ideas,” Altman wrote, “and some things will be great, but surprisingly bad in others. “

But the true value of AGI comes from the enormous run of these systems, Altman said. Similarly to Opena rival Antropic CEO Dario Amodei, Altman is planning thousands or millions of hyper-capable AI systems that manage tasks in “all knowledge work”.

It is assumed that this will be an expensive idea. In fact, Altman observed that “can spend any amounts and receive continuous and predictable profits” in AI performance. This is probably why Openai allegedly raised $ 40 billion in negotiations in a funding round and promised to spend up to $ 500 billion on a huge data network with partners.

However, Altman also claims that the cost of using the “one -given AI level” is about 10 times a month. In other words, pressing the boundary of AI technology will not be cheaper, but users will have access to increasingly capable systems along the way.

Capable, cheap AI models from China’s start -up from Deepseek and others support this idea. Evidence suggests that training and development costs are also reduced, but Altman and Amodei have claimed huge investments You will need AGI-level AI and beyond that.

As for how Openai plans to publish AGI-level systems (assuming it can actually create them), Altman said that the company is likely to “make some major decisions and restrictions on AGI security, which will be unpopular.” Opena After promising To commit to stopping competition and helping to help the “security conscious” project that is close to the construction of Agi before it does, because of the worry about security.

Of course, when the Openai nonprofit organization remained. The company is ongoing to convert its corporate structure into a more traditional, profit -oriented organization. Opena allegedly Its purpose is to reach $ 100 billion in revenue by 2029, the same as Target and Nestle’s current annual sales.

In this situation, Altman added that Openai’s goal, when he builds a stronger AI, is to “be directed toward individual authorization” while preventing the “AI used by authoritarian governments to control their population in mass observation and by loss of autonomy. ” Altman recently said that he said Openai was on the wrong side of history when he was boiling open source technologies. While Openai has an open source technology in the past, the company generally preferred a patented, closed source development approach.

“AI leaks into all areas of the economy and society; We expect everything to be smart, ”Altman said. “Many people expect people to have more control over technology than historically, including open sources, and accept that there is a balance between security and individual authorization that requires compromises.”

Altman’s blog post arrives before the weekly AI action summit that has already made the other technological names to outline their own sight For the future of AI.

In a footnote, Altman added that Openai does not actually plan to end his close partner and investor Microsoft, using the term AGI. Microsoft and Openai were said to have been a contractual definition of AGI systems, which could result in $ 100 billion in profits, would once allow Openai to negotiate more favorable investment conditions. However, Altman said that Openai “fully expects to be partnership with Microsoft in the long run.”