The probability of human extinction due to the development of artificial intelligence (AI) could be at least 95 percent. This risk was assessed by the head of the non-profit organization Machine Intelligence Research Institute, former engineer at Google and Microsoft Nate Soares, as quoted by The Times.
"You take a huge amount of computing power and throw it at a truly gigantic amount of data so that it somehow changes the computers. Nobody knows what is happening inside these things," said the expert.
Soares' organization is engaged in developing safe AI and identifying potential dangers of technological progress. According to him, the current situation can be compared to a car speeding towards a cliff. "I’m not saying we can’t stop the car, but we are just barreling towards the edge," he claims.
The article states that artificial intelligence is currently at the beginning of its development, but according to various experts, within 12 months AI could evolve to the point where it can think on par with humans. Following that, artificial superintelligence may emerge, capable of solving problems beyond human capability. For example, it could help create a cure for cancer or organize intergalactic routes. As a result, there is a risk that at some point AI will perceive humanity as an obstacle.
However, the publication notes that Soares presents an extremely negative scenario. Another expert, executive director of the PauseAI movement Holly Elmore, believes that the risk of human extinction due to AI is only 15-20 percent. Tesla CEO Elon Musk agrees with a 20 percent probability, while Google CEO Sundar Pichai assesses the risks at 10 percent.
Leave a comment