OpenAI is talking about a discovery that could threaten humanity

Before the decision to fire the CEO of OpenAI, Sam Altman, the board of directors received a letter from the employees, which spoke about a “powerful discovery” that could threaten humanity. Reuters writes about this with reference to sources within the company.

According to the interlocutors of the agency, this is a breakthrough in the creation of general artificial intelligence (AGI), which can understand the environment like a human and perform any intellectual tasks. Accordingly, by learning, AGI can surpass a human in solving most of these tasks.

OpenAI and the authors of the letter did not respond to questions from Reuters. However, one of the interlocutors of the agency said that in an internal message the management acknowledged the existence of the letter and work on the Q* (Q-Star) project.

According to the second interlocutor, Q* is able to solve mathematical problems, although so far his abilities are at the level of a junior high school student. However, this instills in the researchers “great optimism about the future success of the project.”

Now calculation tools use algorithms and can perform a limited number of operations. Mastering mathematics implies that the AI ​​will have the ability to reason, that is, it will be able to generalize, learn and understand. This can help in scientific activity.

The authors of the letter did not specify exactly what threats Q* carries. Scientists have long argued about the danger posed by highly intelligent machines: for example, if they decide that the destruction of humanity is in their interests.

The letter was only one of the reasons for Altman’s dismissal, which the board of directors motivated by the fact that Altman was not always frank with them, writes Reuters. Also, in the announcement about the dismissal of the general director, there was the following reason:

“OpenAI was created to achieve our mission: to ensure that shared artificial intelligence will benefit all of humanity. The Council remains fully committed to this mission.”

A few days later, after the threat of 700 out of 770 OpenAI employees resigning, Altman managed to “reach an agreement in principle” on his return to the post of CEO.