Google's Co-Founder Returns to the Front Lines of an AI Coding War
Sergey Brin, Google's co-founder, is no longer on the sidelines. He has taken direct charge of a specialized team within DeepMind, tasked with a single objective: closing a performance gap in...
Sergey Brin, Google's co-founder, is no longer on the sidelines. He has taken direct charge of a specialized team within DeepMind, tasked with a single objective: closing a performance gap in AI-assisted coding that has seen rivals like Anthropic's Claude pull ahead. Internal assessments reportedly show Claude outperforming Google's Gemini models, particularly in complex, multi-step programming tasks.
Brin's involvement is operational, not ceremonial. According to reports from The Information and Android Authority, he issued a directive mandating that every engineer working on Gemini must use Google's internal AI coding tools, a practice known as 'dogfooding.' There are no exceptions. The team includes senior figures like research engineer Sebastian Borgeaud and DeepMind CTO Koray Kavukcuoglu, all focused on developing advanced models trained on Google's proprietary code—tools considered too valuable to release publicly.
The push reflects a broader strategic reality. During a recent earnings call, Google CFO Anat Ashkenazi noted that roughly half of the company's new code is now AI-generated. However, competitors are moving faster; Anthropic claims near-total automation of its coding, and Spotify's lead developers are said to have abandoned manual coding months ago. To accelerate adoption internally, Google is tracking usage of tools like its 'Jetski' assistant through token-spending leaderboards and enforcing training sessions.
At its core, this isn't just about developer efficiency. The belief is that superior AI coding tools create a self-reinforcing cycle: better models build better systems, which in turn train even more capable models. Falling behind in this area could mean losing ground in the wider contest for AI supremacy. While Brin's team works quietly, the pressure is evident. The outcome will likely be measured first in internal adoption metrics and, eventually, in the capabilities of Google's public-facing AI models.
Source: Webpronews
Ready to Modernize Your Business?
Get your AI automation roadmap in minutes, not months.
Analyze Your Workflows →