Google Cloud, research units collaborate to speed up AI development


Google said it has seen rapid progress in artificial intelligence (AI) as a result of close collaboration between its cloud computing unit and its AI research lab.

Google Cloud has struck deals to incorporate its latest AI models into popular consumer products, the company said as part of its virtual Gemini at Work event, sharing more than 75 customer stories including social app Snapchat and storage company Pods. Have been done. The company also unveiled a new version of its AI model Gemini.

Officials said the collaboration between Google Cloud and the company’s research arm, Google DeepMind, could help speed the product to market. That’s important as investors pressure parent company Alphabet Inc. to show how AI investments can turn into new business.

It’s a close-knit relationship between two business units that both have something to prove. Google Cloud, which has long lagged behind Amazon.com and Microsoft in the cloud computing market, is trying to use the enthusiasm for generative AI to expand its business. Google’s AI researchers invented most of the technologies that underpin the latest wave of AI, but OpenAI overtook Google with the launch of ChatGPT, following which the search giant has launched products from research labs to consumers and businesses. There has been pressure to make the path easier.

As progress in the field accelerates, Google’s AI researchers have had to “step up the pace,” said Ellie Collins, vice president of product at DeepMind.

“Researchers are incredibly motivated to see their work in real products and in the hands of users,” Collins said. “It’s a cultural shift not only within Google, but also in the rest of the research community, where many of the major research labs at this point are actually product companies.”

In recent years, Google has lost a series of researchers who were impatient to launch the product. But one of the most vocal members of that camp, Noam Shazier, has recently returned.

In Silicon Valley, companies have struggled to unify research and product teams, which often have different incentives, said Sharon Zhou, CEO of Lamini Inc., a startup that helps companies use AI systems. It uses what is known as a large language model.

“It is extremely difficult for production AI and AI research teams to collaborate well,” Zhou wrote in a message. “Whoever figures out how to get these teams well supported will be the market leader. They are under huge stress in many companies from Google to their competitors.

In an interview earlier this year, Google Cloud Chief Executive Officer Thomas Kurian said that strengthening the relationship with DeepMind has been a major focus, noting that the teams in London, Seattle and the Bay Area are working together , sometimes even sit together. At a session in June, Google Cloud employees presented customers’ experiences with generative AI to DeepMind researchers, and some of the features discussed by the teams were included in Tuesday’s model launch, a Google Cloud spokesperson said.

To increase focus on the product, DeepMind researchers are working not only with Google Cloud but also with corporate customers, Collins said. In a conversation with customer Snap Inc., Snapchat’s owner, the company shared how users were often using the app for homework help. DeepMind tailored its models accordingly, Collins said.

“Cloud customers are our customers,” Collins said.

© 2024 Bloomberg LP

(This story has not been edited by NDTV staff and is auto-generated from a syndicated feed.)



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *