OpenAI leaders comment on the need for a regulatory body for AI.

Artificial intelligence is the tool of the moment. With its rapid evolution, records were broken in a short time with the phenomenal ChatGPT. However, its leaders believe that an international regulatory body is needed, similar to the one that governs nuclear energy. There is strong concern about the malicious use of the tool to create misinformation and deepfakes.

In a post on the company's blog, OpenAI Founder Sam Altman, President Greg Brockman and Chief Scientist Ilya Sutskever came together to provide more details on what can be done.

What does the OpenAI blog post say?

The people leading OpenAI believe that now is the best time to reflect on so-called “superintelligence governance”. The company strongly believes that superintelligence — AI systems in the next 10 years — will be much more powerful than other technologies that humanity has dealt with.

It is said that we can have a dramatically more prosperous future, but that we need to know how to manage the risks to get there. The authors comment that, “given the possibility of existential risk, we cannot just be reactive. A

nuclear energyis a commonly used historical example of a technology with this property; synthetic biology is another example.”

It ends its first sections by stating that we must mitigate the risks of today's AI technology, but that superintelligence will require special treatment and coordination.

He goes on to talk about this relationship of the need for a regulatory body, as well as the already existing International Atomic Energy Agency, of the United Nations, the UN.

“We need some degree of coordination between major development efforts to ensure that the development of the superintelligence occurs in a way that allows us to maintain security and assist in the smooth integration of these systems with the society."

“It is likely that eventually we will need something like an [International Atomic Energy Agency] for superintelligence efforts; any effort above a certain threshold of capacity (or resources such as computing) will need to be subject to an international authority that can inspect systems, require audits, test compliance with security standards, impose restrictions on degrees of deployment and security levels, etc".

Company sees public contribution as important in the process

It is said that the governance of powerful systems must accompany public oversight. I believe that people around the world should be empowered to democratically decide what the boundaries and standards of AI systems The company points out that they still do not know how to design such a mechanism, but that its development is in progress. planning.

Finally, it is emphasized that it is risky to prevent the creation of a superintelligence. It highlights their advantages and that the number of actors building them is increasing rapidly and is inherently part of the technological path. The company asks that the measures have to be accurate.

To access the original publication, click here.