Sam Altman, the CEO of ChatGPT’s OpenAI, told US lawmakers on Tuesday that regulating artificial intelligence was essential after its poem-writing chatbot stunned the world.
Lawmakers highlighted their deepest fears about AI developments, with a leading senator opening the Capitol Hill hearing with a computer-generated voice — remarkably similar to his own — reading a text written by the bot.
“If you listened at home you might have thought that voice was mine and the words were mine, but in fact that voice wasn’t mine,” said Senator Richard Blumenthal.
Artificial intelligence technologies “are more than just research experiments. They are no longer science fiction fantasies, they are real and present,” said Blumenthal, a Democrat.
The latest figure to erupt from Silicon Valley, Altman’s testimony before a U.S. Senate judiciary subcommittee, was far from the tantalizing whims Facebook or TikTok bosses got when they visited Washington.
“If this technology goes wrong, it could go pretty wrong,” Altman said.
Tipped as an opportunity to educate lawmakers, Altman used the session to urge Congress to impose new regulations on big tech, despite deep political divisions that have blocked legislation aimed at regulating the internet for years.
But governments around the world are under pressure to act quickly after the release of ChatGPT, a bot that can produce human-like content in an instant, went viral, stunned and shocked users alike.
Altman has since become the global face of AI, pushing both his company’s technology outward, including to Microsoft and a host of other companies, warning that the work could have nefarious effects on society.
“OpenAI is based on the belief that artificial intelligence has the potential to improve almost every aspect of our lives, but also that it carries serious risks,” Altman told the hearing.
He insisted that over time, generative AI, developed by OpenAI, will “address some of humanity’s greatest challenges, such as climate change and curing cancer.”
However, given concerns about disinformation, job security and other dangers, “we think regulatory intervention by governments will be critical to mitigate the risks of increasingly powerful models,” he said.
Altman suggested the US government might consider a combination of licensing and testing requirements before releasing powerful AI models, with the power to revoke licenses if rules are broken.
He also recommended labeling and more global coordination in regulating the technology, as well as the creation of a dedicated U.S. agency for artificial intelligence.
“I think the US should take the lead here and do things first, but to be effective we need something global,” he added.
Blumenthal underlined that Europe is already well advanced on its AI law, which will be voted on in the European Parliament next month.
A sprawling legal text, the EU measure could lead to a ban on biometric surveillance, emotion recognition and certain police AI systems.
Crucially for OpenAI, US lawmakers stressed that it is also trying to place generative AI systems such as ChatGPT and DALL-E in a category that requires special transparency measures, such as notifications to users that the output is machine-created.
OpenAI’s DALL-E last year sparked an online rush to create lookalike Van Goghs and has made it possible to generate illustrations and images with a simple request.
Lawmakers also heard warnings that the technology was still in its infancy.
“More spirits are coming for more bottles,” says New York University Professor Emeritus Gary Marcus, another panelist.
“We don’t have machines that can really improve themselves. We don’t have machines that have self-awareness, and we may never want to go there,” he said.
Christina Montgomery, chief privacy and trust officer at IBM, urged lawmakers not to be too broad when setting rules for AI.
“A chatbot that can share restaurant recommendations or compose an email has different effects on society than a system that supports credit, housing or employment decisions,” she said.
(This story has not been edited by DailyExpertNews staff and is auto-generated from a syndicated feed.)