Google DeepMind employee leaves, fearing accelerated AI development
- Robert O'Callahan left Google DeepMind due to the rapid pace of AI
- He believes the development of AI is too fast and irresponsible
- The departure adds to the list of specialists who are wary of AI risks
- O'Callahan emphasizes the real but uncertain risk of ASI
Robert O'Callahan, an engineer at Google DeepMind, announced his departure from the company through a post on social media X. According to him, the team was working on developing specialized chips that would accelerate the performance of AI models and reduce their cost, but he believes the pace of technology development exceeds reasonable limits.
In a personal blog, O'Callahan published a letter sent to colleagues on Friday, emphasizing that he cannot 'turn a blind eye' to the potential consequences of his work. He stated that his Christian beliefs do not allow him to support projects that may harm people, and called such an approach 'inherently irresponsible'.
The author of the letter noted that he loves his colleagues and appreciates the opportunity to solve interesting problems, but believes that the goal of accelerating and cheapening AI could be harmful to society. According to him, the development of AI is happening 'too rapidly', and the direction in which the industry is moving raises serious doubts.
Business Insider was unable to clarify how long O'Callahan worked at Google, and company representatives declined to comment. The employee himself stated that he lives in New Zealand and suggested that journalists review the full text of the blog for additional details.
Trend of industry departures
O'Callahan's departure adds to the list of specialists leaving major AI companies due to concerns over technology safety. Earlier this month, Jacob Coxon left Anthropic, citing the inability to control future AI systems. Last month, former DeepMind researcher Bilal Chughtai also left the company, fearing that humanity would not have enough time to prevent potential harm from AI.
Unlike some colleagues, O'Callahan does not claim that a catastrophic scenario with artificial superintelligence (ASI) is inevitable. He considers the risk to be real but uncertain, and emphasizes the moral obligation of the industry to minimize it. In his opinion, the probability of a total catastrophe does not reach 100%, however, the pursuit of creating ASI in the near future is already considered irresponsible.
O'Callahan also rejected the hypothesis that warnings about AI dangers are fueled to boost stock prices or eliminate competitors through regulation. He expressed confidence in the sincerity of those raising safety issues and called for a more serious public discussion about the future of artificial intelligence.
Source: 3DNews
