The CEO of Openai, Sam Altman, admits that the advantages of AI may not be widely distributed
In a New test on his personal blogThe CEO of Openai, Sam Altman, said that the company is open to a “calculation budget”, among other ideas “with strange consonance”, to “allow everyone on earth to use a lot of AI” And to ensure that the advantages of technology are widely distributed.
“The historical impact of technological progress suggests that most of the metrics we care for (results for health, economic prosperity, etc.) are improving on average and long -term, but the increase in equality seems Not technologically determined and realizing the right one may require new ideas, ”wrote Altman. “In particular, it seems that the balance of powers between capital and labor could easily spoil, and this may require early intervention.”
Solutions to this problem, such as the concept of Altman’s “calculation budget”, can be easier to offer than to execute. Already, AI has an impact on the job marketresulting in job cuts And Department reduction. Experts have warned This mass unemployment is a possible result of the rise of AI technology if it is not accompanied by good government policies and reskilling and reduction programs.
Not for the first timeAltman claims that general artificial intelligence (AG) – which he defines as “[an AI] The system that can tackle the increasingly complex problems, in human level, in many areas ”- is close. Whatever form is needed, this act will not be perfect, warns Altman, in the sense that it can “require a lot of supervision and management of human”
“”[AGI systems] Will not have the greatest new ideas, “wrote Altman,” and it will be great in certain things but surprisingly bad to others. “”
But AG’s real value will come from the execution of these systems on a massive scale, said Altman. Similar to the CEO of Openai Rival Anthropic, Dario AmodeiAltman is considering thousands or even millions of hyper capable AI systems that attack tasks “in all areas of knowledge”.
We could assume that it will be an expensive vision to achieve. Indeed, Altman observed that “you can spend arbitrary sums and obtain continuous and predictable gains” in the performance of the AI. This may be why Openai is would have been in talks collect up to $ 40 billion in a financing round and has promised to spend up to $ 500 billion With partners on a huge network of data.
However, Altman also argues that the cost of using “a given level of AI” falls by about 10 times every 12 months. In other words, postponing the limit of AI technology will not become less expensive, but users will have access to increasingly capable systems along the way.
AI models capable and inexpensive of The startup of Chinese AI Deepseek and others seem To support this notion. There is evidence suggesting These training and development costs also drop, but Altman and Amodei argued that massive investments will be necessary to reach AG level AI – and beyond.
As for the way Openai plans to publish AG level systems (assuming that this can, in fact, create them), Altman has declared that the company would probably take “certain major decisions and limitations related to AC security which will be unpopular “. OPENAI Once promot That he would undertake to stop competing and start to help any project “aligned on the value”, “concerned with security” which approaches the act before doing so, for the sake of security.
Of course, it was at this time that Openai intended to remain a non -profit organization. The company is in the conversion process of its corporate structure into that of a more traditional organization focused on profit. OPENAI would have aims to reach $ 100 billion in income by 2029, equal to Target and the current annual sales of Nestlé.
This being the case, Altman added that the objective of Openai, because it strengthens a more powerful AI will be “tend more to individual empowerment” while preventing “AI being used by authoritarian governments to control their population by mass monitoring and loss of autonomy ”. Altman recently said he thought that Optai was on the wrong side of history with regard to the opening of his technologies. While Optai has open source technology in the past, the company has generally favored an approach to the proprietary source development.
“The AI will infiltrate in all areas of the economy and society; We expect everything to be intelligent, ”said Altman. “Many of us expect to need people more control over technology than we have done historically, including more open source, and accept that there is a balance between security and individual empowerment which will require compromises. “
Altman’s blog post was before the summit of this week’s action in Paris, which has already prompted other technological notables to describe their visions For the future of AI.
In a footnote, Altman added that Optai in fact does not plan to end his relationship with a close partner and Microsoft investor as soon as possible using the term AG. Microsoft and Openai would have had a contractual definition of the act – AI systems that can generate $ 100 billion in profits – which, once encountered, allowed OpenAI to negotiate more favorable investment conditions. Altman, however, said that Optai “is expecting fully[s] Be associated with long -term Microsoft. »»
