Demis Hassabis, CEO of DeepMind (Google's AI research arm), isn't overly concerned about an "AI apocalypse" — he's more concerned about the risk of the technology falling into the wrong hands and the lack of safeguards to control complex, autonomous AI models.
“Both of those risks are important and challenging,” he shared in an interview with CNN at the SXSW event recently held in the UK.
Last week, the CEO of Anthropic, a prominent AI lab, issued a dire warning about the future of the job market. He predicted that AI could eliminate half of all white-collar jobs. But his biggest concern lies in the potential misuse of “artificial general intelligence” (AGI), a type of AI that is theoretically capable of achieving human-level intelligence.
“A bad guy could repurpose those technologies for malicious purposes,” he said. “So the big question is… how do we restrict access to these powerful systems for bad guys… but still allow good guys to do a lot of great things with it?”
In May, the FBI said hackers had used AI to create fake voice messages of government officials. A report commissioned by the US State Department last year also found that AI could pose “catastrophic” national security risks, according to CNN . AI has also facilitated the creation of deepfake porn.
Hassabis is not the first to raise such concerns. While AI has been hailed as one of the biggest technological advances since the internet, it also gives scammers and malicious actors more tools than ever before. Notably, it is growing rapidly without much regulation.
Hassabis believes there needs to be an international agreement on the basic principles of how AI should be used and how to ensure the technology is only used “for good use cases”.
“Obviously, given the geopolitical situation as it is, that seems difficult,” he said. “But, I hope that as things improve, and as AI becomes more sophisticated, I think the world will become more and more aware that it needs to happen.”
DeepMind's CEO also believes we're headed toward a future where people use AI "assistants" to perform tasks on their behalf, a vision Google is working toward by integrating more AI into its search function and developing AI-powered smart glasses.
"We sometimes call it an omnipotent AI assistant, which will go with you everywhere, help you in your daily life, do simple administrative tasks for you, but also enrich your life by suggesting great things, from books and movies to even friends to meet," shared the CEO of Google DeepMind.
However, AI is still susceptible to limitations like bias and “hallucinations,” which have caused some notable problems for companies using the technology. For example, the Chicago Sun-Times and Philadelphia Inquirer published an AI-generated summer reading list that included books that didn’t exist.
Hassabis doesn’t believe AI will make jobs obsolete. Like others in the AI field, he believes the technology can create new types of jobs and increase productivity. But he also acknowledges that society will likely have to adapt and figure out how to “distribute all the additional productivity that AI will create in the economy.”
He compares AI to the development of other technological changes, such as the Internet.
“Normally new, even better jobs emerge to replace some of the jobs that are displaced. We’ll see if that happens this time,” he said.
(According to CNN)
Source: https://vietnamnet.vn/ong-chu-veo-3-canh-bao-nguy-co-lon-hon-ca-viec-ai-gay-that-nghiep-hang-loat-2409366.html
Comment (0)