The world “may not have time” to prepare for AI security risks, a leading researcher says

The world “may not have time” to prepare for the security risks posed by cutting-edge AI systems, according to a senior figure at the UK government’s scientific research agency.

David Dalrymple, program director and AI security expert at the Aria agency, told the Guardian that people should be concerned about the growing power of the technology.

“I think we should look at systems that can do all the functions that people do to get things done in the world, but better,” he said. “We will be inferior in all areas where we must be dominant to maintain control of our civilization, our society and our planet.”

Dalrymple said there is a gap in understanding between the public sector and AI companies about the power of upcoming breakthroughs in technology.

“I would point out that things are moving very quickly and we may not have time to move forward for security reasons,” he said. “And it’s not science fiction…