Google DeepMind CEO Demis Hassabis recently revealed on a podcast that in order to stay ahead in the fierce AI competition, this tech giant is breaking down internal barriers. By integrating the company's computing resources and top talents, DeepMind has successfully transformed from a "follower" to a "leader" in the past two or three years, with its operational efficiency now comparable to that of startups.

Resource Integration Reshapes Competitiveness

Hassabis pointed out that computing power has become the biggest bottleneck in current AI research. By centralizing scattered resources, Google ensures core projects receive maximum support. This transformation allows the company to focus on high-efficiency development rather than being bogged down by multiple inefficient versions internally, enabling rapid iteration of large-scale models like Gemini.

As a major contributor to breakthroughs in modern artificial intelligence, the synergistic effects of DeepMind's merger with Google Brain are gradually becoming evident. Hassabis proudly stated that about 90% of fundamental breakthroughs in the industry originate from Google-affiliated labs, and the current organizational structure enables them to more flexibly respond to challenges from competitors such as OpenAI and Anthropic.

Management Philosophy of Speed and Standards

This trend of "giant startup" is not unique to Google. Executives from Amazon and Apple have also repeatedly expressed similar management philosophies. Amazon CEO Andy Jassy once emphasized that fast action and high standards are not contradictory, while Apple founder Steve Jobs once defined Apple as "the largest startup on Earth," aiming to maintain organizational agility.

For Google, this transformation is not just an emergency measure to deal with external competition, but an inevitable choice to reestablish its leading position in technological frontiers. As AI enters deeper waters, Hassabis' DeepMind is trying to maintain dominance in multiple cutting-edge fields through a culture of simplicity and speed.