OpenAI CEO Sam Altman (Altman; Altman) believes that despite growing public concerns about the various risks of the development of artificial intelligence (AI; artificial intelligence), people should have more confidence in his company and other peer companies that they will make the right choices in AI development.
"The world should trust us to do the right thing because it is the right thing and we feel deeply how important it is," Altman said at a conference in San Francisco on Tuesday (September 15).
Altman noted, however, that concerns about AI are valid because the capabilities of these tools are advancing rapidly.
"We don't have to think as hard now as we did before to imagine how this could go wrong," he said. "I think the world is right to be afraid of this."
Altman's remarks came as he attended an annual conference hosted by software company Salesforce.
This is the first time he has spoken publicly since a post sparked a heated discussion online last week. The post was written by a retired researcher at artificial intelligence company Anthropic. The researcher claims that if left unchecked, artificial intelligence could wipe out all of humanity by the late 2020s.
Several other AI company executives and experts responded that they agreed with the idea, but did not explain how they came to their conclusions or how exactly AI could destroy all of humanity.
The heated discussions triggered by these statements have recently prompted the outside world to intensify scrutiny of artificial intelligence development. In response, Anthropic CEO Dario Amodei called for a slowdown in all AI development and urged governments to regulate the industry.
The article drew praise from Altman, but also from Demis Hassabis, co-founder of Google-owned DeepMind, and Elon Musk, owner of social media platform X and artificial intelligence assistant Grok.
By Tuesday, a growing number of AI industry leaders were expressing support for industry self-regulation rather than government intervention.
Altman said he believes artificial intelligence companies like his own are largely capable of regulating themselves.
"We're going to get this right, and I'm very confident in our company and the industry's ability to do this safely," Altman said. He also said he's convinced the companies will "let alignment and safety be well ahead of capability development" and if they can't, they will "slow down or stop."
After Altman’s speech, Meta CEO Mark Zuckerberg wrote on
"Any lab that doesn't take alignment seriously will fall behind," Zuckerberg wrote. "Labs will face significant legal liability if their models cause harm, so they also have strong incentives to prevent this from happening."
Jensen Huang, head of Nvidia, also said at the same conference on Tuesday that whether to release a new version of related technologies should be decided by artificial intelligence companies and should not be managed by external forces.
"We don't need new laws or regulations," Huang said, adding that "false trade-offs" should not be made between the speed of innovation and the safety of artificial intelligence products.
Nvidia, the world's most valuable company, has seen its profits surge due to strong demand for the artificial intelligence computing chips it produces.
"Safety is paramount," Huang said. "However, safety is essentially an engineering issue."
He also said that if at any time the head of an AI company lacks confidence in its product, he should choose not to release it.
"It's an obvious thing to do... go as fast as you can, but whenever you feel that the organization is out of control, you should pause," Huang said.
Some industry insiders believe that the recent panic about artificial intelligence has been exaggerated and is being used to create hype for the industry.
During the conference, Altman addressed hundreds of business people, telling them they should use AI tools to aid their work, but also need the tools to protect their businesses from potential AI cyberattacks.
The idea of leaving the artificial intelligence industry entirely to the self-regulation of its executives is undesirable to some in the industry.
Anthropic executive and co-founder Jack Clark told the BBC on Monday (14th) that allowing artificial intelligence to become a "completely unregulated industry" is "rolling the dice with huge risks."
Patrick Hillman, chief operating officer of Logical Intelligence, noted on Tuesday that there is a lack of confidence in tech companies to actually do anything for the public good. The chairman of this company is Yann LeCun, who is regarded as a leading figure in the artificial intelligence industry.
Hillman said: "Today, the institution that Americans distrust the most is probably Silicon Valley, followed by Washington. I have worked and lived in both Silicon Valley and Washington. I can assure you that they have brought this distrust on themselves."
He added: "If you yourself believe that what you are building is dangerous, then take concrete actions to demonstrate what you are prepared to stop doing."
Leaders at OpenAI and Anthropic say they have recently begun working toward some kind of industry-wide security agreement.
In a brief appearance at Tuesday's conference, Amodei said Anthropic is currently "in dialogue" with other companies in the industry about working to improve safety standards and inspections for AI tools and development.
