DeepSeek’s AI model is ‘the best work’ out of China but the hype is ‘exaggerated,’ Google DeepMind CEO says
Google DeepMind co-founder and Eminent Govt Officer Demis Hassabis provides a convention all over the Cellular Global Congress (MWC), the telecom business’s largest annual amassing, in Barcelona on February 26, 2024.
Pau Barrena | Afp | Getty Photographs
PARIS — DeepSeek’s AI fashion “is probably the best work” out of China, Demis Hassabis, the CEO of Google DeepMind mentioned on Sunday, however added that the corporate didn’t display any brandnew medical advances.
Utmost week, China’s DeepSeek discharged a analysis paper that rattled world markets nearest claiming its AI fashion used to be skilled at a fragment of the price of chief AI gamers and on less-advanced Nvidia chips.
DeepSeek’s announcement sparked an competitive book sell-off and sparked really extensive debate over whether or not massive tech companies are spending excess on AI infrastructure.
Hassabis praised DeepSeek’s fashion as “an impressive piece of work.”
“I think its probably the best work I’ve seen come out of China,” Hassabis mentioned at a Google-hosted tournament in Paris forward of the AI Motion Height this is being hosted by way of town.
The DeepMind CEO mentioned the AI fashion displays that DeepSeek can do “extremely good engineering” and that it “changes things on a geopolitical scale.”
Then again, from a era viewpoint, Hassabis mentioned it used to be now not a fat alternate.
“Despite the hype, there’s no actual new scientific advance … it’s using known techniques [in AI],” he mentioned, including that the hype round DeepSeek has been “exaggerated a little bit.”
The DeepMind CEO mentioned that the corporate’s Gemini 2.0 Flash fashions, which Google this month discharged to everybody, are extra environment friendly than DeepSeek’s fashion.
DeepSeek’s claims round its low value and the chips it makes use of were puzzled by way of professionals, who suppose the price of construction for the Chinese language company’s fashions is upper.
AGI 5 years away
The AI international has been debating for years when the arriving of synthetic common wisdom, or AGI, will occur. AGI widely refers to AI this is smarter than people.
Hassabis mentioned that the AI business is “on the path towards AGI,” which he describes as “a system that exhibits all the cognitive capabilities humans have.”
“I think we’re close now, you know, maybe we are only, you know, perhaps 5 years or something away from a system like that which would be pretty extraordinary,” Hassabis mentioned.
“And I think society needs to get ready for that and what implications that will have. And, you know, make sure that we derive the benefits from that and the whole society benefits from that, but also we mitigate some of the risks, too.”
Hassabis’ feedback replicate the ones of others within the business who’ve steered that AGI might be nearer to fact.
OpenAI CEO Sam Altman this time mentioned that he’s “confident we know how to build AGI as we have traditionally understood it.”
Nonetheless, many within the business have additionally flagged more than one dangers related to AGI. One of the vital largest issues is that people will lose keep watch over of the methods they created, a view shared by way of eminent AI scientists Max Tegmark and Yoshua Bengio, who just lately shared their issues with GWN over this mode of AI.
Correction: This text has been up to date to replicate that DeepMind CEO Demis Hassabis mentioned the corporate’s Gemini 2.0 Flash fashions are extra environment friendly than DeepSeek’s fashion.