OpenAI CEO Sam Altman has said that some accidents may be unavoidable as companies develop new AI technology. However, he believes the industry should learn from these incidents and be open about what went wrong. Speaking at Salesforce’s Dreamforce conference, Altman said AI companies should build a culture of transparent accident reporting, reports Business Insider.
“I do think that some degree of accidents are unavoidable with new technology,” Altman said. “What I really care about is that we have got a great culture of accident reporting and learning.”
Altman pointed to the Federal Aviation Administration and the National Transportation Safety Board as examples for the AI industry. He said their focus on investigating accidents has helped improve aviation safety. “You know, accidents are going to happen,” he said of the FAA’s and NTSB’s approach. “We’re going to learn as much as we can and react from each one.”
Also read: Microsoft AI chief criticises Anthropic’s approach to Claude, warns against humanlike AI
Altman also acknowledged concerns about the potential risks of advanced AI. He said fears around AI systems becoming difficult for humans to control should not be dismissed. “The world is right to be afraid of this,” Altman said.
He pointed to the possibility of a loss-of-control accident and the risk of too much power being concentrated among a small number of AI developers.
However, Altman said the answer is not to stop developing AI. Instead, he argued that safety measures should progress faster than AI capabilities.” He also said AI companies should be ready “to slow down or stop” if they cannot maintain that lead.
Also read: OpenAI shares 6 AI misalignment cases, says its models fabricated data and took unauthorised actions
Altman said companies should report incidents, learn from them and act quickly to prevent small problems from becoming much larger ones.
“Maybe we can avoid, you know, ever getting to those kinds of stakes,” he said.
His comments come as concerns grow around the development of increasingly powerful AI systems.
Also read: Satya Nadella says AI companies should stop and fix problems instead of ignoring them