← The AI Power Map

About This Project

What this is

If you are named in the map

If something about your profile is inaccurate, missing context, or uses information you believe should not be public, click the Contribute button on the map and submit your request. We will review it and correct within 1 week.

Methodology

Who was included: People with visible influence in the modern AI industry (founders, researchers, investors, executives, operators, safety voices, infrastructure builders) identified through public sources and algorithmic clustering, then manually reviewed.

What "public tie" means: A relationship documented in public sources: co-authorship, co-founding, employment, board membership, investment, marriage/family (when professionally relevant and publicly known), mentorship, and repeated professional collaboration.

Sources: Public company filings, Crunchbase, LinkedIn (public profiles), published interviews, news coverage, academic papers, conference talks, and publicly available bios.

Limitations: This is a snapshot, not a live feed. Relationships change. Some ties are stronger than others. The map inevitably over-represents people with large public footprints and under-represents people who are influential but private.

For media

Two findings that may be useful for coverage:

Finding 1: Google/DeepMind alumni appear as founders or key leaders across an unusual share of major AI companies. The Transformer paper alone produced 8 authors who went on to 6+ different companies.
Finding 2: Several of the largest AI funding rounds went to teams with zero revenue but deep prior institutional trust. The map shows which specific trust chains (labs, co-authors, prior cofounders, repeat investors) predicted those rounds.

Screenshots available on request. Project URL: the full map.

Contact: @yumi_willems on X · LinkedIn

About the author

Yumi Kimura. Founder of BehaviorGraph, building the organizational-behavior layer for enterprise AI agents. Columbia University.

Explore the Map Read the Book