Vitalik Buterin spoke about the threat of AI

Humanity does not cope well enough with the coordination of global values ​​to guarantee that supercantly artificial intelligence will work for the benefit of all people. Such a statement was made by the co -founder of Ethereum Vitalik Buterin.

He commented on the post of 3D artist with the nickname Utah Teapot.

“How can the concept of AI alignment exist? I always ask: “Agreement with what?” No one will ever answer anything but “with human values”. But holistic, consistent human values ​​do not exist. I still don’t understand. What do we study? It is absurd that this turns into a growing industry. Fictive work, ”he wrote.

Buterin emphasized that in the world there are many things that contradict ethical principles. As an example, he cited the murder or conclusion of the innocent.

“[…] We are not yet good enough in agreement to guarantee that a single super -rational AI will avoid this even this, ”the developer added.

UTAH TeAPOT noted that the development of coordinated artificial intelligence is moving towards restriction of public access to development. At the same time, large companies sell versions of their AI systems that contribute to negative phenomena and act of warfare.

“I am deeply concerned that the centralization of access to AI-technologies allows you to impose as security issues that it is actually not it-for example, discussions around the“ psychosis ”from LLM. Thus, there is a risk of harm to unique or marginal cultures and their values, ”the user noted.

Buterin shares many of these fears.

“I think that the greatest risks will come from the military and other structures that have great power that allows them to derive themselves from the action of security rules that are by default for everyone,” added the co-founder of Ethereum.

UTAH TeAPOT was given by the Anthropic AI-Startap. He develops alternative versions of “civilian” models and provides governments. They can be used for military or reconnaissance actions.

Buterin emphasized that the probability of a catastrophic scenario for humanity due to AI is increasing if there is only one super-rational agent artificial intelligence, which has its own will and the ability to act as a subject.

If the environment is pluralistic, none of the systems can total control the situation. However, to create it, market mechanisms alone are not enough. Intentional efforts are needed, including changes in laws and incentives that will not like large corporations.

Recall that in 2022, the Minister of Information Technology of India Rajiv Chandrakar called for the development of global standards that guarantee AI safety for humans.

Be in the know! Subscribe to Telegram.


Source: Cryptocurrency

You may also like

Ireland will welcome 52 students from Gaza
World
Flora

Ireland will welcome 52 students from Gaza

Fifty -two Gaza students will arrive in Ireland this week, taking advantage of university scholarships, Ireland Foreign Minister Simon Harris