During the Great AI Debate seminar, Elon Musk shared his insights on the development of modern AI algorithms. Specifically, he mentioned the estimated probability of an “uprising” of artificial intelligence, a figure that other industry experts believe is significantly underestimated.
According to the billionaire, the current likelihood of a scenario where AI technologies revolt against humanity ranges from 10 to 20%. However, Musk believes that such a risk is justifiable. He argues that to ensure security, AI must be developed to provide only truthful information.
Previously, researchers discovered that once AI masters the art of deception towards humans, its behavior becomes resistant to existing security measures. The New York Times labeled this revelation as “a chilling new statistic sweeping Silicon Valley,” with several tech executives offering highly pessimistic estimates of negative outcomes, some reaching up to 50%.
Roman Yampolsky, an AI security researcher and director of the cybersecurity lab at the University of Louisville, supports Musk’s concerns regarding the AI threat. He contends that the actual probability of artificial intelligence seizing control of humanity is much higher than Musk’s claims, surpassing 99%.