Skip to content

Researchers

Researchers
400
855 co-author links
Main web
326
connected to each other; 48 small groups apart
Average distance
4.61
co-author steps; the farthest pair is 11 apart
Clusters
22
Leiden communities of 4+ people

Degrees of separation

The fewest co-author steps between any two researchers on the map, with a paper for each step. The map's own "Erdős number" counts steps from Erik Brynjolfsson, its most central researcher by PageRank: everyone in the main web is within 6 steps.

to

Who holds the network together

Most central PageRank

Linked to people who are themselves well linked; counts papers together.

  1. Erik Brynjolfsson National Bureau of Economic Research
  2. Daron Acemoglu Massachusetts Institute of Technology
  3. Philippe Aghion Harvard University Press
  4. Yoshua Bengio Université de Montréal
  5. Klaus Prettner Vienna University of Economics and Business
  6. Ajay Agrawal University of Toronto
  7. Nicholas Bloom Stanford University
  8. David Autor National Bureau of Economic Research
  9. Gillian Hadfield University of Toronto
  10. John Van Reenen London School of Economics and Political Science

Bridges betweenness

Sit on the most shortest paths between others, often joining clusters.

  1. Erik Brynjolfsson National Bureau of Economic Research
  2. David Autor National Bureau of Economic Research
  3. Scott Stern Massachusetts Institute of Technology
  4. John Van Reenen London School of Economics and Political Science
  5. Ajay Agrawal University of Toronto
  6. Philippe Aghion Harvard University Press
  7. Susan Athey National Bureau of Economic Research
  8. Anton Korinek University of Virginia
  9. Lawrence Katz Harvard University
  10. David J. Deming Harvard University Press

Economists and computer scientists

Each researcher's field comes from the topics of their papers in OpenAlex: an economist publishes mostly in economics, finance, business or management, a computer scientist mostly in computer science; everyone else counts as another field. On this map 241 researchers are economists and 51 computer scientists, and 57 of the 855 co-author links join one of each. Show only those pairs on the map.

Pairs that cross fields most papers together

  1. Frank Neffke and Simone Daniotti 4 papers together
  2. M. Amin Rahimian and Avinash Collis 4 papers together
  3. Nicholas G. Otis and Rembrand Koning 4 papers together
  4. Nicholas G. Otis and Solène Delecourt 4 papers together
  5. Abhishek Nagaraj and Imke Reimers 3 papers together
  6. Adrian Weller and Diane Coyle 3 papers together
  7. Andrew Caplin and Philip Marx 3 papers together
  8. César A. Hidalgo and Michiel Bakker 3 papers together
  9. Daniel Rock and Pamela Mishkin 3 papers together
  10. David Holtz and Nicholas G. Otis 3 papers together
  11. Erik Brynjolfsson and Sebastian Steffen 3 papers together
  12. Frank Neffke and Johannes Wachs 3 papers together

Blended papers 21% of published papers

27 of the 130 published papers whose authors' fields we know (at least two of them) have an economist and a computer scientist among the authors. 84 are by economists only, 15 by computer scientists only. Papers without a byline in OpenAlex, such as most AI-lab posts, are not counted.

Clusters

22 groups of more than 3 people who write more with each other than with the rest, found by the Leiden algorithm (330 of 400 researchers; smaller groups are not shown). Each is named after the topic of the papers its members wrote together: an AI-written label, checked against those papers' titles, that describes the papers, not the people. Click one to isolate it on the map.

ClusterPeopleMost central members
AI safety and governance38Yoshua Bengio, Gillian Hadfield, Anton Korinek
Automation, firms and productivity36Erik Brynjolfsson, Daron Acemoglu, Pascual Restrepo
Market design and field experiments30Susan Athey, Michael Kremer, John List
Innovation and economic growth29Philippe Aghion, Nicholas Bloom, John Van Reenen
Computational social science24Maria del Rio-Chanona, Frank Neffke, J. Doyne Farmer
Labor markets and technology23David Autor, David J. Deming, Lawrence Katz
Economics of prediction and innovation22Ajay Agrawal, Joshua Gans, Avi Goldfarb
Platforms, startups and experiments20Rembrand Koning, Sinan Aral, David Holtz
Foundation models and society16Diyi Yang, James Landay, Arvind Narayanan
Automation, growth and robot taxes14Klaus Prettner, David E. Bloom, Sophia Kazinnik
Platform rules, income and well-being14Jacques Crémer, David Dinielli, Fiona M. Scott Morton
Entrepreneurship and labor markets12Sampsa Samila, José Azar, Liudmila Alekseeva
Taxation and economic mobility7Emmanuel Saez, Danny Yagan, Raj Chetty
Inequality in macroeconomics6Benjamin Moll, Jean‐Michel Lasry, Pierre-Louis Lions
AI risks and adoption6Neil Thompson, Alexander K. Saeri, Peter Slattery
Hyperspectral mineral exploration6Gabriel Unger, Robert Zimmermann, Moritz Kirsch
Persuasion and moral attitudes5Robb Willer, Johannes C. Eichstaedt, Jan G. Voelkel
Algorithmic fairness and regulation5Sendhil Mullainathan, Jens Ludwig, Ashesh Rambachan
Trade, investment and digital services5Daniel Trefler, Walid Hejazi, Jianmin Tang
Algorithmic pricing and collusion4Emilio Calvano, Giacomo Calzolari, Vincenzo Denicolò
Exchange rates and retail firms4Jen Baggs, Loretta Fung, Beverly Lapham
AI and expertise at work4Luca Vendraminelli, Arvind Karunakaran, Andrea Vinelli

What these numbers mean

They describe this map, not the field. The map starts from the letter's signatories and their co-authors on papers found through OpenAlex and the research database, so a researcher who mostly publishes elsewhere looks less connected here than they are. A rank is a fact about this graph, not a verdict on anyone's importance or influence.

PageRank (damping 0.85) scores a researcher higher when their co-authors are themselves central, weighting each link by papers written together. Bridges are ranked by betweenness: the share of shortest co-author chains between other pairs that pass through someone. Steps from Erik Brynjolfsson is the length of the shortest co-author chain to the map's PageRank #1, in the spirit of the Erdős number. Clusters come from the Leiden algorithm (Traag, Waltman & van Eck, 2019) maximising modularity (0.843 here; above 0.3 is usually read as clear structure), best of 10 fixed seeds, so the result repeats on the same data. Cluster numbers can change when the weekly rebuild adds papers.

Positions on the color-by-question view are documented stances on the stated question, classified by an AI agent from each researcher's papers and public statements, always with cited evidence, and never a claim about personal relationships. Gray means no evidence found or reviewed yet. Data files are on the data page · updated 2026-10-26.