DeepMind researcher quits, calling near-term superintelligent AI irresponsible
Robert O’Callahan, a technical expert at Google DeepMind in New Zealand, has left the company over concerns about the risks of artificial intelligence. He said the field’s current pace of change is too rapid and described efforts to achieve artificial superintelligence in the near future as inherently irresponsible.
O’Callahan previously worked on chip-design tools that helped make AI systems cheaper and faster. He said he could no longer justify contributing to that progress, describing the dangers of superintelligent AI as real but uncertain. He also pointed to possible cognitive surrender, AI-induced psychosis and loneliness, concentrated power, economic disruption, cybersecurity threats and a lack of accountability.
O’Callahan said many colleagues at DeepMind share similar concerns but seldom discuss them publicly. He believes AI’s potential harms may exceed its benefits and now plans to pursue work that is “unambiguously pro-human.” He linked to If Anyone Builds It by Eliezer Yudkowsky and Nate Soares, while saying he did not endorse the book’s claim that superintelligent AI would inevitably cause catastrophe.
Written by Sakhanda Wire AI from reporting by The Decoder. Facts, quotes and full context are in the original article below.