The world should take artificial intelligence as seriously as it takes the climate crisis, said Demis Hassabis, the UK chief executive of Google’s artificial intelligence unit. warning that we cannot afford to delay her answer which must be given.

As the UK government prepares to host an AI safety summit, Demis Hassabis said industry oversight could start with a body similar to the Intergovernmental Panel on Climate Change (IPCC).

Hassabis sounds the alarm calling on the world to act immediately to address the dangers of the technology, which included helping to create bioweapons and the existential threat posed by super-intelligent systems.

“We need to take the risks of artificial intelligence as seriously as we do other major global challenges, such as climate change,” he said. “The international community has been slow to coordinate an effective global response to climate change and we are seeing the consequences now. We cannot afford the same delay with artificial intelligence.”

At the same time, Hassabis, whose unit created the revolutionary AlphaFold program that depicts protein structures, said artificial intelligence could be “one of the most important and beneficial technologies ever invented.”

However, he told the Guardian that an oversight regime was needed and governments should take inspiration from international structures such as the IPCC.

“I think we should start with something like the IPCC, where it’s a science and research agreement with references, and then build on that.”

He added that “what I’d like to see eventually is an equivalent of a Cern for AI security doing international research around it. And then maybe there will be some kind of equivalent of the IAEA one day, which will actually control these things».

The International Atomic Energy Agency (IAEA) is a UN agency that promotes the safe and peaceful use of nuclear technology in an effort to prevent the proliferation of nuclear weapons, including through inspections.

However, Hassabis said none of the regulatory analogies used for artificial intelligence were “directly applicable” to the technology, although we could learn “valuable lessons” from existing institutions.

Last week, Eric Schmidt, the former CEO of Google, and Mustafa Suleyman, the co-founder of DeepMind, called for the creation of an IPCC-like panel on artificial intelligence. Although the British are positive about such a development, they nevertheless believe that its creation should be carried out under the auspices of the UN.