Meta's “Superintelligence” AI team appears to be made up primarily of Openai, cutting out their work for them

AI News


Meta's plans, comparable to Openai, seem to rely heavily on Openai itself.

Social media companies are bringing together “super intelligence” teams to boost AI chops, reach AGI and continue to dominate advertising. To provide that team this month Meta The New York Times reports that it attempted to poach more than 45 AI researchers on Openai alone, offering a starting package of as much as $100 million. Already, the social media giant has hired at least four Openai researchers, including Trapit Bansal, Lucas Beyer, Alexander Kolesnikov and Xiaohua Zhai.

“At least, none of our best people have decided to take it up so far,” Openai CEO Sam Altman said earlier this month.

Verge reports that at all internal hand meetings held in Meta this week, its employees asked questions about these reported $100 million offers Altman described on the podcast.

“Sam is just dishonest here,” Meta CTO Andrew Bosworth said at the meeting when asked about Altman's comments. “He suggests that we are doing this for each and every one…Look, guys, the market is hot. It's not that hot.”

According to NYT, Meta's leadership has raised “non-investment” from Meta's llama and instead embraced the AI ​​model from competitors, including Openai, but has yet to make a final decision.

Of course, Meta uses that deep pocket to gain talent other than Openi. This includes $14 billion investment in Scale AI and its founder Alexandr Wang, as well as hiring people from companies such as Google, Sesame and Safe Superintelligence.

Meta is also talking about AI Voice Startup Playai and potential companies and talent acquisition.

Low engagement llama?

Meta publicly boasts that it has 1 billion Meta AI users. This is a magic number that you're waiting to get hit before you monetize a new product. The company is spreading AI across its products, and released only the standalone “Meta AI” app a few months ago. However, new details coming out of meta-executives raise some red flags about how many people actually use it.

According to a report from the Verge of All Hands Meeting in the Meta this week, that may be less than they're saying publicly:

“The meta-executive who mentioned Openai at the internal meeting was not Bosworth. CPO Chriscox He also admitted that Meta AI has 1 billion users per month, but that engagement is “not as deep as the way people use ChatGPT.” The standalone meta AI app only has 450,000 users every day, he told employees, “Many people use it to manage Ray-Ban Metagrass.”

Cox also reportedly said at the meeting that he was not chasing AI-powered productivity tools. Googleof humanity, and Openai, focuses on “entertainment, connections with friends, how people live their lives, how we uniquely work.”



Source link

Leave a Reply

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