When and how can AI destroy humanity?

Along with the development of the technological segment in the world's leading countries, artificial intelligence (AI) may pose a threat to humanity at some point, says an article published in The New York Times.

Experts say too much autonomy for AI systems and their connection to vital infrastructure, such as power grids, stock markets and weapons, could lead to unintended consequences.

According to Professor Joshua Bengio, a researcher at the University of Montreal, today's AI systems do not pose an existential threat, but it is difficult to predict which moment will become decisive and lead to disaster.

“Today’s systems are not anywhere close to posing an existential risk,” said Yoshua Bengio, a professor and A.I. researcher at the University of Montreal. “But in one, two, five years? There is too much uncertainty. That is the issue. We are not sure this won’t pass some point where things get catastrophic.”

“At some point, it would become clear that the big machine that is running society and the economy is not really under human control, nor can it be turned off, any more than the S&P 500 could be shut down,” said Anthony Aguirre, a cosmologist at the University of California, Santa Cruz.

According to specialists, the AI systems tend to get stuck in endless loops. Researchers gave one system all the resources it needed to replicate itself. It couldn’t do it. In time, those limitations could be fixed.

“As AI systems learn from vastly larger volumes of data than even their creators can understand, these systems are already exhibiting unexpected behavior. Some experts fear that as researchers make these systems more powerful by training them with more and more data, they could develop more harmful and even dangerous habits,” the author of the article adds.