Openai CEO Sam Altman is a new essay for his personal blog, and the company is widely distributed in the other “strange sound” ideas and the advantages of “many AI” and the advantages of technology are widely distributed.
“The historical impact of technical progress is that most of the indicators we are interested in (health performance, economic prosperity, etc.) are better in the long term, but equality is not technically determined and does not gain this rights. Altman needs a new idea. “Especially, it seems that the balance of power between capital and labor can be easily messed up, which may require early intervention.”
The solution to this problem, such as Altman’s “Compute Budget” concept, is easier to suggest than running. Already, AI affects the labor market, causing recruitment cuts and reductions by department. Experts warned that mass unemployment is the result of an increase in AI TECH unless it is accompanied by the right government policy and re -investigation and skilled programs.
Altman first claims that the artificial general information (AGI) (AGI) (he is a system that can solve more complex problems at the person level ”, which is not perfect, and this AGI will not be perfect. “Many human directors and directions may be needed.”
Altman said, “(AGI Systems) will not have the biggest new idea.” Some are great but surprisingly not good for others. “
But the actual value of AGI will come to run these systems on a huge scale, Altman argued. Altman, like DARIO AMODEI, CEO of Openai Rivals Anthropic, conceives thousands or millions of hyper -available AI systems dealing with “in all knowledge work.”
It may be assumed that it will be an expensive vision to realize it. In fact, Altman observed that in AI achievements, “you can consume any amount and get sustainable and predictable profits.” For this reason, Openai reported that it was in the talks to raise up to $ 40 billion in the financing round, and promised to spend up to $ 500 billion in partners in a huge data network.
However, Altman also costs about 10 times every 12 months of using “AI -level AI”. In other words, it is not cheaper to promote the boundaries of AI technology, but users can gradually access more competent systems.
China’s AI Start Up DeepSeek and other competent and inexpensive AI models seem to support the concept. There is evidence that education and development costs are also decreasing, but Altman and Amodei argued that huge investment is needed to achieve AGI Level AI.
Altman said that the company plans to plan that Openai wants to launch the AGI Level system (assuming that it can create a real), and the company will create some major decisions and restrictions related to unpopular AGI safety. Openai promised to start supporting the “value alignment” and “safety consciousness” project, which is close to building AGI before it once stops competition and is concerned about safety.
Of course, it was when Openai tried to remain a non -profit organization. The company is in the process of converting the corporate structure into the structure of a more traditional interest -based organization. Openai aims to reach $ 100 billion in sales, the same as the current annual sales of Target and Nestle by 2029.
In this case, Altman added that the goal of OpenAI is to build a stronger AI, which is “more trend for personal authority.” I hold the forest in the forest in the forest. ” Altman recently said that when Openai is in the wrong aspect of history when the technology is open. Openai has open source technology in the past, but the company generally prefers to develop monopoly closed source.
“AI will penetrate all areas of economy and society. We will expect everything to be smart, ”Altman said. “Many of us need to be more historically controlled than us, including open source, and we must acknowledge that there is a balance between safety that requires compromise and individual authority.”
Altman’s blog post is ahead of the AI Action Summit in Paris this week, and briefly explained his vision for the future of other technologies.
In a footnote, Altman added that Openai actually uses the term AGI to end the relationship with nearby partners and investors Microsoft. Microsoft and Openai reported that they have a contractual definition of AGI (AGI). But Altman said, “We expect to be a partner in the long run with Microsoft.”
