Sergey Brin reveals Gemini was once on Google's internal coding 'no list'

Brin's admission about Gemini

Google co-founder Sergey Brin has revealed that the company's own artificial intelligence model, Gemini, was once on an internal 'no list' of tools that developers were asked not to use. The disclosure came as a surprise given that Gemini is one of Google's flagship AI products, positioned as a direct competitor to OpenAI's ChatGPT and other large language models.

The information emerged from Brin's comments during what is believed to be an internal company discussion. The 'no list' refers to a set of coding tools and platforms that Google's engineering teams were discouraged from relying on, often to encourage the use of internally developed alternatives or to avoid security and compatibility issues.

Why Gemini was blacklisted

According to Brin, Gemini's inclusion on the list was not due to any fundamental flaw but rather a reflection of the fast-moving nature of AI development. At the time, the model was still in early stages and may not have met the rigorous standards required for production-level coding tasks.

The list is a common practice in large tech firms where teams maintain a curated list of approved libraries and tools to ensure consistency, security, and performance. Having a product on the 'no list' means developers are officially advised against using it, even if it is available within the company.

Brin's comments suggest that the situation has since changed, and Gemini is now considered reliable enough for internal use. The company has been aggressively pushing Gemini across its ecosystem, including integrating it into Google Workspace, Android, and cloud services.

Broader context and implications

The revelation highlights the internal challenges Google faces in adopting its own AI tools. Even as the company markets Gemini as a cutting-edge solution, its own engineering teams were initially cautious about relying on it for core development work.

This is not unique to Google. Many technology companies maintain similar internal lists to manage the proliferation of tools and frameworks. However, the fact that a flagship product like Gemini was on such a list underscores the gap between external marketing and internal readiness.

Brin, who returned to Google in a hands-on role in recent years to work on AI, has been vocal about the need for the company to move faster. His admission may be seen as an effort to signal transparency and a willingness to acknowledge past missteps.

What happens next

Google is expected to continue refining Gemini and expanding its use across the company. The 'no list' incident is now in the past, but it serves as a reminder that even the most advanced AI models require rigorous testing before they earn the trust of their own developers. Industry watchers will be keen to see if any other internal restrictions are lifted as Gemini matures.