Shulou(Shulou.com)11/24 Report--

Beijing, May 5 (Xinhua) A leaked Google internal document went viral in tech circles on Thursday. Google is losing its AI edge, but the winner will not be OpenAI, the ChatGPT developer, but AI created by the open source community, a senior Google software engineer said in the document.

The engineer believes that Google's competition with OpenAI distracts the company from the rapid development of open source technology. In the open source community, many independent researchers have made rapid and unexpected progress using AI technology.

"We worry too much about OpenAI. Who will cross the next milestone? what will be the next step?" "but the disturbing truth is that we don't have the ability to win this arms race, and neither does the OpenAI," he wrote. "while we are still arguing, the third faction is quietly defeating us. Of course, I'm talking about the open source community."

A Google spokesman confirmed that the memo was written by a senior software engineer at the company, but added that the article was not the company's official strategy document. The engineer has yet to comment.

As progress in generative AI accelerates, employees of Google and other tech giants are having a heated discussion internally and externally about the technology that is reshaping their industries. Google is under particular pressure because the popularity of ChatGPT has raised fears that Google may be losing its edge in AI. Google has long been a leader in this area.

The open source community threat, however, the Google senior engineer asserts in the document that Google's real threat comes from the open source community, where engineers are rapidly developing models of product quality comparable to those of large technology companies and at a lower cost. He says these models are faster, more customizable and more useful than Google's own.

"We don't have a secret recipe," the employee wrote. "our best hope is to learn and work with people other than Google." He is worried that customers are unwilling to pay for models that provide such high-quality technology for free.

In February this year, a large language model created by Meta was leaked to the public, boosting the progress of generative AI in open source forums. The model, called LLaMA, is smaller than the model that Google and OpenAI have been promoting, so it is easier to use. Researchers currently have to apply to Meta to access LLaMA.

The Google engineer believes it would be better for Google to shift its focus to smaller, more flexible models. "huge models slow us down," the engineer wrote. "in the long run, the best models are those that can iterate quickly."

A Google spokesman did not comment on the content of the post. In a recent earnings call, Google CEO Sandal Pichai (Sundar Pichai) said that "our investments and breakthroughs in AI over the past decade have put us in a strong position", noting that the company has made progress in developing models and working with developers and other partners. Mr Pichai has called for regulation of AI in the past, warning that the technology could be "very harmful" if not deployed in a thoughtful manner.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.