GENEVA (AP) — OpenAI CEO Sam Altman was the star speaker at the annual AI for Good conference on Thursday, speaking at the U.N. telecommunications agency's annual gathering about how to harness the social potential of artificial intelligence technology.
But Altman spent part of his virtual appearance dodging tricky questions about governance, the AI voice controversy and criticism from ousted directors.
Altman's appearance to talk about the benefits of AI comes as the company is grappling with growing concerns about its business practices and its handling of AI safety.
He was among the technology leaders at the Geneva gathering, which featured two-day speeches and talks on the applications of artificial intelligence to robotics, healthcare, education, sustainable development and more.
The latest wave of complaints against OpenAI coincided with the company's product launch in early May, which drew the ire of Hollywood star Scarlett Johansson, who said she was shocked by how ChatGPT's voice sounded “eerie” to her own, even though she had previously declined Altman's request to lend her voice to the system.
In an interview with The Atlantic editor Nicholas Thompson, Altman spoke at length about AI-related subjects, but declined to answer questions about some sensitive topics, such as explaining why Chat GPT's creators use an actor's voice that resembles Johansson.
“It's not her voice. It can't be. Sorry for the confusion. Apparently you think so,” Altman said. “People will disagree about how similar the voices are, but we don't. It's not her voice. Yeah, we don't think so… I don't know what else to say.”
Thompson's magazine had signed a product and content deal with OpenAI the previous day, but has not responded since.
He then asked Altman about OpenAI's governance, including the idea of a governance committee.
“We are continuing to discuss how we will implement governance and I don't think I should say much more at this time,” Altman said.
“Tell me more,” Thompson replied, drawing laughter from the audience.
“Sorry, but I'll have to decline,” Altman replied.
Oversight of big AI companies like OpenAI has become an increasingly heated topic. Shortly after the Johansson controversy, retired researcher Jan Rijke said the company was “putting safety second to flashy products” and that disagreements with its top leaders had reached a “breaking point.”
Reike's departure comes days after the departure of co-founder Ilya Sutskever. The two co-led OpenAI's “Superalignment” team, which was centered around the organization's founding mission of safely developing better-than-human AI, or so-called artificial general intelligence. The team has now been disbanded and replaced by a separate safety committee.
Thompson asked about comments by Georgetown University researcher Helen Toner, who was one of the OpenAI board members who was ousted last year after a messy power struggle with Altman. Toner, part of a group concerned about AI safety risks, accused Altman on the podcast of withholding information, misrepresenting or “outright lying” about what's going on at OpenAI.
As an example, when ChatGPT was released in November 2022, “the board was not informed of it in advance. We found out about ChatGPT on Twitter,” she said.
“I don't agree with her recollection of events,” Altman said. Toner added, “I sincerely hope that AGI will come out well. I commend her for that and wish her the best. I'm not going to go into a point-by-point rebuttal here.”
OpenAI is riding a new wave of generative AI technology that garnered public attention 18 months ago with the launch of ChatGPT, an early standout AI system that churns out novel text, images and videos based on a vast database of online writing, books and other media. Other large companies and AI startups have since released their own versions of the technology.
While the UN's “AI for Good” initiative predates the recent craze for generative AI, the rapid advancement and commercialization of the technology has attracted global attention.
“Artificial intelligence is transforming our world and our lives,” U.N. Secretary-General Antonio Guterres said in a video, noting the technology's promise for improving education and healthcare in remote areas, improving crop yields and early warning systems for natural disasters.
Political leaders around the world are particularly concerned about AI tools that could accelerate the spread of disinformation online: With just a few commands and requests, computer-generated text and images can be spread across social media and the internet, blurring the line between fake news and reality.
“To realise that potential, we need AI that reduces bias, misinformation and security threats, rather than exacerbating them,” Guterres said, arguing that developing countries also needed to “use AI themselves” and help “connect the unconnected” around the world.
——
Chang reported from London. AP Technology Writer Matt O'Brien in Providence, Rhode Island, contributed to this report.
—–
The Associated Press and OpenAI have a licensing and technology agreement that gives OpenAI access to portions of the AP's text archive.
