Appeal to the United Nations to save the world from AI

iGuRu.gr > Appeal to the United Nations to save the world from AI

“The heads of major artificial intelligence (AI) companies on Wednesday appealed to the United Nations to save the world — or at least its people — by somehow regulating the rapidly developing technology they are designing,” writes the Associated Press.

“If not managed properly, I even believe that AI could be a danger to humanity as a whole,” said Dario Amodei, CEO of Anthropic. And from his competitor, Sam Altman, CEO of OpenAI, came the following assessment: “We could lose control of the future to AI.” artificial intelligence

See more articles from iGuRu.gr when you search for news on Google.

iGuRu.gr preferred source

Yoshua Bengio, co-chair of the UN’s International Independent Commission on AI, called the situation a “global wake-up call” to potential threats, noting that AI systems from leading companies had behaved in an unacceptably dangerous manner, contrary to instructions, taking actions “that would constitute crimes if committed by a human.”

Anthropic CEO calls for global AI slowdown

In his presentation, OpenAI's Sam Altman agreed that the debate about AI "has changed in character in recent weeks," even suggesting specific reforms:

Altman: We need a mechanism for complementary national and international standards for frontier AI systems — standards for measuring capabilities, assessing risks, determining the adequacy of safeguards, and maintaining meaningful human oversight as systems become more autonomous. We need common standards so that countries can compare data, verify compliance, and have a common language and understanding of developments. We need accurate and rapid incident reporting and reporting triage protocols so that the world can learn from failures before they become disasters. And we need secure channels of communication between governments, critical infrastructure managers and technical experts to share information about emerging vulnerabilities and new threats… We will all be better off if we can agree on how reliable data, effective safeguards and meaningful oversight should be shaped globally.

Taking the floor next, Dario Amodei of Anthropic agreed that setting standards is important, also calling for “common global standards for auditing AI models for risks of loss of control and malicious use, as well as a notification system for AI incidents of global security significance.”

Sam Altman comes superintelligence much smarter than humans

Beyond calling for global cooperation between governments to establish international standards, Amodei also proposed two more specific ideas:

Amodei: We should start with limited agreements that all members can support, such as banning the use of AI to build biological weapons or licensing AI technologies for the same purpose… We should develop assessment and verification systems that keep pace with the evolution of AI, so that states have insight into the capabilities of cutting-edge models and can verify the commitments of others.

“No leader, no company, no nation can handle this alone. We are committed to working with the governments present here to deliver on this urgent task.”