UN experts warn against exaggerating the dangers of artificial intelligence

Experts on a United Nations committee dealing with artificial intelligence affairs warned yesterday against the “exaggeration” by some proponents of this technology regarding its risks, calling for a “scientific” discussion to understand the issue more rationally.

Joel Paral, executive officer at Google DeepMind and member of the International Scientific Body for Artificial Intelligence, told the French Press Agency (AFP) while participating in an event held on the sidelines of the United Nations General Assembly in New York, “As scientists, it is important that we rely on science to distinguish between what we know and what we do not know. Spreading fear is not a good thing.”

Investigative journalist Maria Rees, Nobel Peace Prize laureate, said during the same event, “We need to change our approach to the risks of artificial intelligence, as the conversation has long been about existential risks and that this technology will cause the destruction of everything, that robots are coming to kill us, or that the matter is treated as if it has no importance at all.”

She added, “Reality is more complex and nuanced,” warning that “the discussion cannot be turned into a kind of ideology.”

Earlier this month in September, former artificial intelligence researcher at Anthropic, Jacob Coxon, sparked a wave of concern when he said that those developing this technology “seriously believe it may kill us all by the end of the decade.”