AI explosion deserves regulation to curb threat, experts say

AUSTIN, July 12 (Reuters) – Rapid advances in artificial intelligence could exacerbate social problems and even pose an existential threat to human life, increasing the need for global regulation. , said AI experts at the Reuters MOMENTUM conference this week.

The explosion of generative AI, which can create text, photos and videos in response to free-form prompts, has made some jobs obsolete, upended the economy, There are growing concerns that it could even overwhelm humans. .

“We are flying down the highway in an AI car,” says Ian, co-founder and CEO of Protect AI, which helps companies secure AI and machine learning systems. Swanson said at a Reuters MOMENTUM panel on Tuesday.

“So what do we need to do? We need safety inspections, we need basic maintenance done properly, we need regulations.”

Seth Doblin, Director of the Responsible AI Institute, says regulators are turning to social media platforms to understand how the uncontrolled growth of new industries can lead to negative consequences such as the emergence of information echo chambers. Said it was enough.

“If we widen the digital divide, it will lead to chaos in society,” Doblin said. “Regulators need to think about it.”

Regulations are already in place in several countries to address AI issues.

For example, the European Union’s proposed AI law would categorize AI applications into different risk levels, prohibit uses deemed “unacceptable,” and subject “high-risk” applications to rigorous evaluation. is.

U.S. lawmakers introduced two separate bills focused on AI last month. One calls for transparency from the U.S. government when using AI to interact with people, and the other calls for the agency to determine whether the U.S. remains competitive with modern technology. It is to be installed.

One of the new threats lawmakers and technology leaders must watch out for is the potential for AI to make nuclear weapons even more powerful, said Anthony Aguirre, founder and executive director of the Future Life Institute, at a conference. said in an interview.

Developing even more powerful AI also risks eliminating jobs until it becomes impossible for humans to simply learn new skills and enter other industries.

“Ultimately, we’ll end up in a world where our skills are irrelevant,” he says.

The Future of Life Institute, a nonprofit dedicated to reducing catastrophic risks from advanced artificial intelligence, released an open letter in March calling for a six-month moratorium on training AI systems stronger than OpenAI’s GPT-4. was announced and became a hot topic. The report warned that AI labs are “trapped in an uncontrollable race” to develop “powerful digital minds that no one, even their creators, can understand, predict, or reliably control.” .

“It seems like the most obvious thing in the world not to introduce AI into nuclear command and control,” he said. “That doesn’t mean we don’t do it, because we do a lot of unwise things.”

Reporting by Sheila Dang in Austin, Editing by Deepa Babington

Our standards: Thomson Reuters Trust Principles.