Sam Al -Tamman, CEO of Openai, admits that the advantages of artificial intelligence may not be widely distributed

in A new article on his personal blogThe CEO of Openai Sam Altman said that the company is open to a “mathematical budget”, among other ideas “strange sounding”, “to enable everyone on the ground to use a lot of artificial intelligence” and ensure that the benefits of technology are widely distributed.

“The historical influence of technological progress indicates that most of the standards that we care about (health results, economic prosperity, etc.) improving in the average in the long term, but increasing equality does not seem technically specific and obtain this right to do so.” It requires new ideas, “Altman Books.” In particular, it seems that the balance of power between capital and work can easily spoil, and this may require early intervention. ”

The solutions of this problem, such as the concept of Altman’s “budget account”, may be easier to propose. actually, Artificial intelligence affects the labor marketWhich led to Job discounts and Reducing the size of departments. Experts warn This collective unemployment is a possible result of the emergence of artificial intelligence technology, if not accompanied by the correct government policies and reset programs and talismans.

Not for the first time, Altman claims that artificial general intelligence (AGI) – which he defines as “AI) that can process increasingly complex problems, at the human level, in many areas” – near. Whatever the form it takes, this AGI will not be perfect, warns Altman, meaning that it may “require a lot of human supervision and guidance”

“Agi Systems) will not have the largest new ideas, and it will be great in some things but surprisingly bad for others,” Altman wrote.

But the real value of AGI will come from the operation of these systems on a large scale, as Altman emphasized. Like the CEO of Openai Anthropor Competition, Dario Amodei, Altman conceives thousands or even millions of excessive artificial intelligence systems that treat tasks “in every field of knowledge.”

One may assume that this will be an expensive vision to achieve it. In fact, Altman noted that “you can spend arbitrary amounts of money and get continuous and predictable gains” in the performance of artificial intelligence. Perhaps this is the reason why Openai is in talks to raise up to $ 40 billion in a financing round, and pledged to spend up to $ 500 billion with partners on a huge network.

However, Altman also shows that the cost of using a “specific level of artificial intelligence” decreases around 10x every 12 months. In other words, pushing the boundaries of artificial intelligence technology will not become cheaper, but users will be able to reach increasingly capable systems along the way.

It seems that the capabilities and inexpensive artificial intelligence models from the start of the Chinese AI and others support this idea. There is evidence that the training and development costs are also decreased, but both Altman and Amodei have argued that Huge investments Amnesty International will be required at the AGI-and outside.

As for how Openai plans to issue AGI-Level (assuming that it can, in fact, create), Altman said that the company will likely take “some major decisions and restrictions related to AGI safety that will be not prohibited.” Openai Pledge It would commit to stopping competition and starting to help any “alignment” project, “conscious of safety” is close to building AGI before he does, motivated by concern for safety.

Of course, it was when Openai intends to remain unhappy. The company is in the process of converting the corporate structure into its most traditional structure, driven by profit. Openai It is said It aims to reach $ 100 billion in revenue by 2029, equal to the annual sales of the goal and Nestle.

In this case, Altman added that Openai’s goal is because it builds more than artificial intelligence is “more trend towards individual empowerment” as it prevents “artificial intelligence that authoritarian governments use to control their residents through collective control and loss of autonomy.” Altman recently said he believed Openai was on the wrong side of history when it came to its open source techniques. While Openai has an open source technology in the past, the company has generally preferred an approach to the development of closed ownership.

“Amnesty International will leak to all areas of the economy and society; Altman said,” We expect everything to be smart. “Many of us expect that they need to give people more control over technology than we have historically, including more than the open source, and they accept The existence of a balance between safety and individual empowerment that will require contradictions. “

The Altman Blog Blog comes before the AI ​​Action summit for this week in Paris, which has already prompted other technology specialists to determine their own offers Visions For the future of artificial intelligence.

In a footnote, Altman added that Openai does not actually plan to end his relationship with the close partner and investor Microsoft anytime soon using the AGI term. According to Microsoft and Openai, a contractual definition of AGI – artificial intelligence systems that could generate $ 100 billion of profits – would allow, once meet, Openai, to negotiate more converted investment conditions. However, Altman said that Openai “is fully expected (S) partnership with Microsoft in the long run.”