News
Google flagship model Gemini 3
2 min read
Source: Telegram AI频道
Google's flagship model Gemini 3.5 Pro was delayed, and brain drain and cooperation with the military triggered an internal morale crisis. According to the news website Axios, six current and former Google DeepMind employees revealed that low employee morale was one of the reasons why Google AI Lab delayed the release of its flagship model Gemini 3.5 Pro. Bloomberg previously reported that the model is months behind schedule, and the market response to a series of small models recently released by Google has been mixed. Meta AI head Wang Tao even ridiculed "Gemini? Never heard of it" on X. Top talents are leaving one after another, and military agreements have become a centrifugal force. Google is experiencing a serious brain drain. Gemini co-lead Noam Shazer has switched to OpenAI, and Nobel Prize winner John Jungper has switched to Anthropic. Several employees resigned over the military cooperation agreement Google reached with the U.S. Department of Defense in April, which was frequently mentioned in employee exit interviews. Alex Turner, a former research scientist who resigned due to a military contract, said that CEO Hassabis’s internal appearances were far less frequent than those of rival CEOs Altman and Amodei. Some employees bluntly said, "We are lagging behind." The Gemini series has failed to rank among the top ten in the OpenRouter large language model rankings. AI investment has caused free cash flow to turn negative, and Google denies morale problems. Google's total expenditure this year is expected to reach 190 billion US dollars, and free cash flow has turned negative. The market has intensified concerns about the return of AI capital expenditures. Google denied that employee morale issues caused its models to lag behind, saying that the AI talent turnover rate in the first half of this year was lower than the same period last year, and more than 90% of those who were hired accepted the position. A Google spokesperson expressed satisfaction with the Flash family model and future roadmap. However, the report pointed out that Google can afford to be briefly surpassed by competitors, but it cannot afford the loss of talent and investors losing patience. via AI News (author: AI Base)