Geoffrey Hinton, a prominent figure in the field of artificial intelligence (AI), has resigned from his position at Google and expressed concerns about the potential risks of AI development. The 75-year-old Hinton stated that he regrets his past work in the field and warned about the dangers of AI chatbots, which he finds “quite scary.” While he believes that chatbots are not currently more intelligent than humans, he expressed concern that they could soon surpass us in intelligence. Hinton’s departure from Google marks a significant moment in the ongoing conversation about the ethics and risks of AI.
Dr. Hinton, a British-Canadian cognitive psychologist and computer scientist, has retired from his job at Google, citing concerns about the dangers of developments in AI. Although his pioneering research on deep learning and neural networks paved the way for current AI systems, he acknowledged that his age also played a role in his decision to retire at the age of 75. Dr. Hinton warned that AI chatbots could soon surpass the level of information held by the human brain, with GPT-4 already eclipsing human general knowledge by a considerable margin. He stressed the need to address these developments’ rapid progress and potential consequences.
Dr. Hinton expressed concerns about the potential misuse of AI by “bad actors” in his New York Times article.
When asked by the BBC to expand on this, he stated that it was a worst-case scenario and gave an example of a possible situation where Russian President Vladimir Putin grants robots the power to create their own sub-goals. Dr. Hinton warned that this could eventually lead to the sub-goal of “needing more power”.
He highlighted that digital systems have many copies of the same set of weights, which can learn separately but share knowledge instantly, unlike biological systems. This is why chatbots can have significantly more knowledge than any individual.
Dr. Hinton stated that he did not intend to criticize Google and that they had been responsible. In a statement, Google’s chief scientist Jeff Dean also emphasized their commitment to responsible AI and their efforts to understand emerging risks while innovating boldly.
