On August 5, 2026, a single announcement from Alphabet, Google's parent company, left the global tech community holding its breath. Demis Hassabis, founder of Google DeepMind and Nobel laureate in Chemistry, relinquished his day-to-day management duties to become Chairman of DeepMind and Chief Scientist of Alphabet. Jeff Dean, Chief Scientist who had served the company for 27 years, officially departed to co-found a new venture with three core researchers. On the day the news broke, Alphabet's stock price plummeted by over 5% intraday, closing with a narrower decline of around 4%, erasing over $100 billion in market value in a single day.

This was no ordinary personnel adjustment. It came amid a three-month delay of Google's flagship model, Gemini 3.5 Pro, after top talent had been poached by Anthropic and OpenAI, and at a time when the company's free cash flow turned negative for the first time since going public. Within 24 hours, a tech giant with a market value exceeding $4 trillion redefined the power structure of its AI business. Behind this lies a profound transformation in corporate organizational structures in the AI era—shifting from scientist-driven to commercialization-driven, and from decentralized academic labs to centralized product-focused organizations.

01. Two 'Pillars' Shaken Within a Day

The impact of this adjustment stems first from the stature of those departing.

Hassabis is the soul of DeepMind. When Google acquired the London-based AI lab for approximately £400 million (about $650 million) in 2014, Hassabis brought not just a top-tier team but also a belief—using artificial intelligence to solve humanity's most daunting scientific challenges. In 2024, he won the Nobel Prize in Chemistry for leading the development of AlphaFold2, propelling DeepMind to unprecedented heights of prestige. Dean, Google's 30th employee, built the distributed computing infrastructure underpinning Google Search and co-founded Google Brain. Together, they represented the 'brain' and the 'skeleton' of Google's AI.

Yet their simultaneous 'exit'—one stepping back from daily management, the other leaving entirely—was no coincidence. Sources close to Hassabis revealed that his idea to step down as CEO had been brewing for some time. Since the 2023 merger of DeepMind and Google Brain, Hassabis had been bogged down by team management, product delivery, and business alignment, which drastically diverted his focus from cutting-edge research and AGI strategy. Despite working 100-hour weeks for 50 weeks a year, the scientist, who had dedicated his life to pursuing artificial general intelligence (AGI), ultimately found no fulfillment in the role of a tech executive.

Dean's departure was even more resolute. He took with him Sanjay Ghemawat, Oriol Vinyals, and Quoc Le—collectively boasting over 80 years of tenure at Google. Their new venture, Discovery Loop, focuses on 'recursive self-improvement,' enabling AI to propose hypotheses, run experiments, and refine models with minimal human intervention. While Alphabet retains investment and cloud service partnerships, Google has lost the daily R&D capabilities of four core scientists.

02. Forced Restructuring Under Triple Pressure: Product Delays, Talent Drain, and Financial Strain

No organizational adjustment occurs in isolation. Google's 'major surgery' comes amid a confluence of triple pressures.

The most immediate trigger was product-related. Since its high-profile promise of a 'next-month launch' at the May I/O conference, Gemini 3.5 Pro has been delayed by three months, with internal tests showing it still lags behind competitors in core capabilities like programming. According to Semafor, Google's AI models trail the cutting edge in programming by about six months. Programming, however, is the primary driver of AI commercialization—Anthropic's Claude, for instance, convinced over 1,000 enterprises to spend over $1 million annually precisely because of its coding capabilities.

The talent exodus is equally alarming. Over the past eight years, core authors of over 20 landmark Google papers have departed. In 2025 alone, at least 11 AI executives left. This summer, Gemini co-lead Noam Shazeer joined OpenAI, while John Jumper, who shared the Nobel Prize with Hassabis, moved to Anthropic. All eight authors of the seminal Transformer paper have now left Google.

Financially, the situation is no less dire. Google just reported a strong quarter with $119.8 billion in revenue and an 82% surge in cloud business, yet its free cash flow turned negative for the first time since going public. The company raised its full-year capital expenditure guidance to $195-205 billion. With money burning and talent fleeing, Google's AI narrative has reached a point where it must be rewritten.

03. Power Shifts from London to Silicon Valley: A Dual Transformation of Geography and Culture

The most intriguing aspect of this adjustment lies in the geographic relocation of power.

According to the Financial Times, Google is transferring control of its AI business from London back to Silicon Valley. Koray Kavukcuoglu, who succeeded Hassabis in overseeing DeepMind's daily operations, relocated from London to Google's Mountain View headquarters last year and now leads Gemini's development from a desk beside co-founder Sergey Brin. Under the new structure, Kavukcuoglu reports directly to Alphabet CEO Sundar Pichai, with full responsibility for team management, model iteration, product delivery, and commercialization.

This marks a further erosion of DeepMind's autonomy since its 2014 acquisition. Multiple current and former DeepMind employees told the Financial Times and other outlets that they fear DeepMind's long-standing internal autonomy within Google will diminish further without Hassabis in daily management. A former DeepMind executive who now works for a competitor revealed that several employees have already reached out about leaving.

Meanwhile, Google co-founder Brin is reemerging as one of the most influential voices in AI strategy. Since ChatGPT's release, he has increasingly involved himself in Gemini's development, even convening a town hall meeting for hundreds of DeepMind employees in April to demand accelerated R&D iterations. Though Brin holds no formal executive title, he has become one of the loudest voices shaping AI strategy at the $4.4 trillion company.

04. The 'Coming-of-Age' of Organizational Structure in the AI Era

Google's adjustment reflects a fundamental challenge facing tech companies in the AI era: How should an organization rooted in scientific research evolve when forced to compete in commercialization?

Hassabis and Dean represented Google AI's 'scientist era'—long-term research, a lax culture, and a pursuit of breakthrough innovation. Kavukcuoglu, by contrast, embodies a 'product era' focused on efficiency, execution, and commercial proximity. This adjustment is essentially a division: separating 'visionaries' from 'executioners.' Hassabis retreats to top-level strategy, focusing on AGI's long-term research and scientific breakthroughs, while Kavukcuoglu oversees daily operations, model iteration, and commercialization.

The capital markets have already cast their initial vote—the stock price dropped about 4%, erasing over $100 billion in market value. Yet this may be a price Google must pay. When an organization faces product delays, talent drain, cultural clashes, and financial strain simultaneously, any CEO must pull the restructuring lever.

From a broader perspective, Google's predicament is not unique. OpenAI shifted from nonprofit to capped-profit, while Anthropic transitioned from a pure research institute to enterprise services. Every AI company is searching for its ideal organizational form. Google's adjustment simply pushes this question to its extreme—when a giant once proud of its 'Don't Be Evil' mantra and long-term research is forced to sprint in the AI arms race, its organizational structure must undergo a painful 'coming-of-age.'

In his open letter, Hassabis wrote that humanity stands at a critical juncture for AGI breakthroughs, and his life's pursuit is nearing reality. Yet he acknowledged the need to step back from daily management to focus on what truly matters. This may be the best reminder for all organizations in the AI era: Structures can adjust, power can shift, but what truly drives progress are those willing to sacrifice short-term comfort for long-term goals.

- End -