Its high level developer Google He wrote an internal memo for the company, warning that the tech giant was losing any advantage in its research artificial intelligence, and the primaries are being demanded by the “open source” community, as many independent researchers use artificial intelligence. Rapid and largely unexpected progress.
A software engineer, Luke Cernau, shared his concerns on Google’s internal platform in early April, and in recent weeks his text has been shared thousands of times among Google employees, according to a source familiar with the matter. He spoke to Bloomberg anonymously, as she was not authorized to discuss internal matters. for the company.
Thursday, The paper was published by consulting firm SemiAnalysis and has been making the rounds in Silicon Valley.
According to Cernau’s analysis, Google’s competition with startup OpenAI (with billions of dollars in funding from Microsoft) has distracted from the rapid developments taking place in open source technology. “We looked a lot behind us, to OpenAI. Who will achieve the next stage? What will be the next step? But the inconvenient truth is that we’re not in a position to win that race, and neither is OpenAI. For us We were fightingAnother person was “quietly eating our meal”. I’m talking, of course, about open source.
Sernau did not respond to Bloomberg’s request for comment.
As advances in artificial intelligence accelerate, employees at Google and other tech giants are vocal, in private and in public, about the technology reshaping their industry. Google and Microsoft have come under fire from experts, because of the sudden success of ChatGPT, they have taken greater risks in terms of ethical guidelines while partially ignoring the risks.
Sernow believes the real threat to Google comes from open source communities, where engineers are developing AI models at a speed that rivals the quality of big tech companies’ systems, and can be built more cheaply. These forms can be faster, more adaptable, and more useful, Sernau says, than Google Forms.
“We don’t have a secret recipe. The best thing for us is to learn and collaborate with what other people outside of Google are doing,” he comments, predicting Customers will not be willing to pay for a company’s AI models when high-quality systems are offered for free.
A Google spokesperson declined to comment on the post. In the company’s recent earnings call, Alphabet CEO Sundar Pichai painted a different picture: “Our investments and successes in AI over the past decade have positioned us well,” noting progress in model development and collaboration with developers and other partners. Pichai has called for regulation of AI, warning that the technology could become “very harmful” otherwise.
in february, A large language model created by Meta Leaked, which stimulated the development of artificial intelligence in open source forums. The LLaMA model is smaller than the models announced by Google and OpenAI, and therefore easier for developers to work with. Scholars pursuing the legal path must currently apply to Meta to gain access to LLaMA.
“Google would be better off turning its attention to smaller, more flexible models,” Sernow argued. “Giant models slow us down,” the software engineer judges. “In the long run, the best models are the ones that can be replicated quickly.”
Source: Bloomberg
⇒ News of the day
follow her kathimerini.gr on Google News And be the first to know all the news
Find out the latest news from Greece and the world on kathimerini.gr
More Stories
“Recycling – Changing the water heater”: the possibility of paying the financing to the institution once or partially
Libya: US General Meets Haftar Amid Tensions Between Governments
New tax exemption package and incentives for business and corporate mergers..