Eigen RadarAI
Analysis

Where AI capacity is concentrating

Google's lab reorganization, office concentration in King's Cross, and Anthropic's India appointment show how AI capacity gathers inside particular places and organizations.

Artificial Intelligence··Midday
In a bright architectural cutaway, research benches, flexible work floors, and a meeting area converge around a shared glass atrium and structural core.

From DeepMind toward Google's center

Google is concentrating Gemini work in the Bay Area while dissolving more of DeepMind's separate laboratory structure, according to The Decoder's report citing The Guardian. Demis Hassabis is moving to a chairman role and plans to leave DeepMind in the coming months, while chief scientist Jeff Dean has already departed. Koray Kavukcuoglu now runs day-to-day operations from the United States without a chief-executive title. The report quotes a former DeepMind employee describing the lab as an ordinary Google division and says cofounder Sergey Brin is expected to take a larger role. The change shows that capacity is measured through more than chips or models. Management authority, the institution that houses research teams, and the city where development happens all shape the system that turns ideas into products. Here, the geographical shift is visible in Gemini development moving to the Bay Area, while the organizational shift lies in DeepMind becoming more closely integrated into Google.[1]

The value of proximity in King's Cross

London's King's Cross offers a city-scale view of the same relationship. TechCrunch reports that AI startups have moved toward the district for proximity to talent and investors since Google DeepMind relocated there in 2016. The vacancy rate for conventional offices has fallen to 0.9 percent, while prime rents on the largest leases have risen 18 percent in three years. London has about 3,600 AI startups and, according to Dealroom figures cited in the report, they received about 12.1 billion dollars of the 14.8 billion dollars raised in the city since late July. Those figures describe London as a whole, while TechCrunch links the office squeeze in the district to companies' desire to be close to one another, capital, and a shared labor pool. King's Cross therefore operates as more than an address: scarce office space, fast rail links, and established laboratories form a cluster around which firms gather. Capacity here takes the form of a belief that physical proximity can ease recruitment and business connections.[2]

Enterprise reach in India

Anthropic's new India appointment shows capacity being built around customer and industry knowledge as well. CIOL reports that the company has appointed Anand Raisinghani as Head of Industries for India. Raisinghani previously served first as chief executive and later as president and chief business officer at Pine Labs-owned Setu, after senior roles at SAP India covering banking, financial services, insurance, telecom, and media. The appointment follows Anthropic's selection of Irina Ghose as country managing director and the opening of a Bengaluru office, its second in Asia. The company has also announced plans to provide in-country inference for Claude in India through Amazon Bedrock. Read together, the three reports describe AI capacity as an accumulation of decision-making authority, workers, offices, customers, and computing arrangements in particular locations. Google is creating such a focus through management and development in the Bay Area, King's Cross through physical clustering, and Anthropic through local enterprise roles and a service plan in India.[3], [1], [2]

References

  1. News sourceThe DecoderGemini work moves to the Bay Area as DeepMind loses its own chief executive↩1↩2
  2. News sourceTechCrunchOffice space in King's Cross has become a bargaining chip for London's AI startups↩1↩2
  3. News sourceCIOLAnthropic appoints an industries lead for India as it courts enterprise customers↩