New Delhi | OpenAI CEO Sam Altman on Thursday said he shares concerns about the concentration of AI power in the hands of a country or a company, and believes that the only path forward is to "heavily democratise" the technology and put tools in the hands of people, even if it comes with challenges and downsides.
Responding to a question from PTI, Altman said that all my study of history suggests that concentrating all of the AI power in the hands of one company or country, even if it was in the name of safety, would be a "disastrously bad" thing to do.
"I do share the concern about concentration of AI. Our stance is that the only path forward is to heavily democratise AI and to put these tools in the hands of people, even if it comes with some downsides, even if it means that society has got to wrestle with some big challenges," he said at a select briefing on the sidelines of AI Summit.
OpenAI, he said, invented the strategy called iterative deployment that allows people to become familiar with AI innovations, use it even when it's imperfect, even when it has flaws.
"And it doesn't mean we aren't responsible how we do it, doesn't mean that we don't start conservatively, but it does mean that we empower people to do things with the technology that we ourselves might not like...It means that we try to encourage a robust ecosystem being built around the world," he said.
It means that we we take the trade of empowering people and accepting that society is going to have to wrestle with a new thing, rather than try to hold on all the power for ourselves and say, we can guarantee this or that outcome," he said.
New Delhi | Sam Altman, Chief Executive Officer of OpenAI, on Thursday defended use of copyright news articles and opinion pieces in training AI models like ChatGPT, saying it followed the US principle of fair use.
OpenAI, the company behind ChatGPT, aims to balance content use with collaboration with creators, he said.
"Our stance is that when we're displaying a quote from a news article or an opinion piece or something like that, we want to do that in a way where we can explore new business models with creators for training," Altman said.
He added that OpenAI relies on the US principle of fair use, noting that "models can learn like people can learn, but models also can't play tricks that people can't play."
Altman framed this approach as a pathway for AI to responsibly learn from existing content while respecting intellectual property rights.
His remarks come amid ongoing legal challenges from publishers, who allege that AI models have used their work without permission.
OpenAI has defended its practices as consistent with fair use, while also exploring licensing and partnership opportunities to ensure creators benefit from AI training.