Demis Hassabis, co-founder and CEO of Google Deepmind, will speak on February 26, 2024 at Mobile World Congress, the largest annual rally in the telecom industry in Barcelona, Spain.
Pau Barrena | AFP | Getty Images
GoogleAccording to audio from the All-Hand Conference held in Paris on Wednesday, Artificial Intelligence Chief Demis Hassavis is not worried about China’s deep shake, and Google has excellent AI technology, according to the audio of the All-Hand Conference held in Paris on Wednesday. He spoke to.
At the meeting, Alphabet CEO Sundar Pichai read aloud questions about Deepseek, a Chinese emerging lab that recently launched the US market when the app shot to the top of Apple’s App Store. Deepseek released a research paper in January claiming that AI models were trained at the cost of some of the other major models.
A question that was submitted by employees asked how as Google trains its future models, “what lessons and meanings” can Google gather from Deepseek’s success.
Hassabis, CEO of Google Deepmind, was asked to provide the answer.
“After looking into the details, some of Deepshek’s claims are “exaggerated,” Hassavis said.
Hassabis said the reporting costs for Deepseek’s AI training are likely to be “a small percentage” of the total cost of developing the system. He said Deepsee probably uses far more hardware than it allows, and it relies on Western AI models.
“In fact, we have a more efficient and performant model than DeepSeek,” says Hassabis. “So we’re very calm and confident in our strategy and we have all the elements that will help us maintain our leadership this year.”
However, he seemed impressed by Deepseek’s achievements.
“It’s definitely the best team I’ve seen coming out of China, so it’s something that’s something that’s something that’s something that’s taken seriously,” Hassavis said.
He pointed out that Deepseek has “security” and “geopolitical” meanings. Several US agencies have banned staff from using Deepseek, citing security concerns.
Google declined to comment. Deepseek did not respond to requests for comment.
Google Deepmis Hassabis, Left and Google CEO Sundar Pichai have announced May 14, 2024, Tech Titan’s annual focus on how artificial intelligence is woven into search, email, virtual meetings and more Hold an I/O developer meeting.
Glen Chapman | AFP | Getty Images
Google executives have also been questioned by many employees about the company’s recent decision to change the “AI Principles” to eliminate the pledge to use AI for weapons or surveillance.
Pichai ended with “Why did you remove this section?” and read aloud the version of the question that hindered the AI.
Pichai turned the question to Kent Walker, president of Google’s Global Affairs. He said he began efforts in 2024 with Hassabis, James Manyika and others.
Google established the AI principles in 2018 after refusing to renew a government contract called Project Maven, which helps to analyze and interpret drone videos using AI.
“Some of the strict bans that were in V1 of AI principles don’t work for more subtle conversations than we have now,” Walker said, referring to the 2018 rules.
“So many things have changed over the last seven years,” says technology, “used in very subtle scenarios.”
Watch: China Tech Rally by Deepseek
