In the latest episode of the "20VC" podcast, Google DeepMind CEO Demis Hassabis detailed the lab's critical transformation path over recent years. He noted that during the past two to three years, Google executed a significant internal restructuring to consolidate its dispersed computing resources and research teams into a unified entity, substantially boosting overall R&D efficiency and model scale capabilities. Hassabis stated that compute power has long been one of the core bottlenecks constraining AI research. By concentrating resources and aligning directions, Google enabled different teams to "push toward the same goal," eliminating redundant development of multiple model versions internally. This integration strategy allowed the company to focus on building larger-scale, more competitive model systems. In 2023, DeepMind officially merged with Google Brain to form Google DeepMind, which now serves as the primary developer behind several core AI products, including Gemini. Hassabis emphasized that this consolidation essentially involved "recombining existing resources while advancing with greater intensity, focus, and speed," bringing team operations closer to those of startups. He also highlighted that many foundational breakthroughs in the current AI industry stem largely from accumulated research within Google's ecosystem. Facing competition from leading labs such as OpenAI, Anthropic, and Microsoft, Google DeepMind continues to solidify its leadership position in AI by strengthening execution efficiency and organizational synergy. Industry observers note that adopting startup-like mechanisms to enhance speed and focus is becoming a shared strategic choice among major tech companies competing in the next wave of AI advancement.