CNN
–
Demis Hassabis, CEO of Google's AI Research Arm Deepmind and Nobel Prize winner, is not too worried about AI's “Jobpocalypse.”
Instead of worrying about AI replacing jobs, he worries that technology will fall into the wrong hands. And there is a lack of guardrails controlling sophisticated autonomous AI models.
“Both of these risks are important and challenging,” he said in an interview with CNN's Anna Stewart at the London SXSW Festival this week.
Last week, the CEO of the high-profile AI Lab Humanity had a harsh warning about the future of the work situation, claiming that AI could wipe out half of the entry-level white-collar jobs. But Hasavis said he is most concerned about the potential misuse of what AI developers call “artificial general information.”
“Bad actors can reuse those same technology for harmful purposes,” he said. “And one big thing… how do you limit access to these systems and limit strong systems to bad actors… but does a good actor allow them to do a lot, a lot of amazing things with it?”
Hackers used AI to generate voice messages pretending to be US government officials, the Federal Bureau of Investigation said in a public recommendation in May. A report commissioned by the US State Department last year found that AI could pose “devastating” national security risks, CNN reported. AI also facilitated the creation of Deepfake porn. However, the Take It Down Act, which President Donald Trump signed into law last month, aims to stop the spread of these deepfakes.
Hassavis was not the first to call such concerns. However, his comments further highlight both A promise with AI I'll alarm it It comes as technology It will improve handling of complex tasks such as writing code and generating video clips. AI has been announced as one of the biggest technological advances since the Internet, but it offers more tools than ever for scammers and other malicious actors. And it is moving forward rapidly without many regulations as the US and China compete to establish control on the ground.
Google removed the language from the AI Ethics policy website in February and pledged to not use AI for weapons and surveillance.
Hassavis believes there should be an international agreement. The basis for how AI should be used and how technology is used is used is only used “for the right use case.”
“Obviously, the current state of geopolitics seems difficult today,” he said. “But as you know, I think as things improve and AI becomes more refined, it becomes clearer that it needs to happen.”
The CEO of Deepmind believes people are heading towards a future where they use AI “agents” to perform tasks. Google is working to integrate more AI into search capabilities and develop smart glasses with AI.
“We call it a universal AI assistant who will go around with you everywhere, help you in your everyday life and enrich your life for you, but even from books and films, you will enrich your life by recommending your wonderful things, perhaps even friends,” he said.
The new AI model shows progress in areas such as video generation and coding, increasing the fear that technology can eliminate jobs.
“AI is beginning to get better than humans on almost every intellectual task. We as a society are working on that,” Humanity CEO Dario Amodei told CNN after telling Axios that AI can do entry-level work. April, Meta CEO Mark Zuckerberg said he expects AI to write half of the company's code by 2026.
However, a future focused on AI is closer to promise than reality. AI still embraces weaknesses such as bias and hallucination. A company using technology. For example, the Chicago Sun-Times and the Philadelphia Inquirer have published a list of AI-generated summer readings that contain books that do not exist last month.
Hassavis says that AI will change the workforce, but he doesn't think that AI will abolish work. Like others in the AI space, he believes this technology can lead to new types of jobs and increase productivity. However, he also acknowledged that society is likely to have to find some way of “distributing all the additional productivity that AI produces in the economy.”
He compared AI to the rise of other technological changes, such as the Internet.
“There will be a big change,” he said. “Usually, what happens is a better job arrives to replace some of the jobs being exchanged. We'll see if that happens this time.”
