Alex Fogelson

Alex Fogelson

I'm a former software engineer turned researcher, working on how we measure progress in AI, algorithmic efficiency, and their downstream implications on math and science. I'm currently at MIT CSAIL (FutureTech group).

I believe in changing the norms around charitable giving by discussing it publicly: I've signed the Giving What We Can pledge to give 10% of my income to effective charities, and encourage others to do the same!

Education

  1. 2021 – 2022 M.S. in Data Science Carnegie Mellon University MS-DAS Program
  2. 2018 – 2021 B.S. in Mathematics
    Double Major in Computer Science
    Minor in Chinese Studies     Carnegie Mellon University Completed in 3.5 years, concurrent with Master's program.
  3. Jun – Jul 2019 Study Abroad – Chinese Language and Culture Shanghai International Studies University Two-month Chinese immersion program.

Career

  1. Apr 2024 – Present MIT CSAIL (FutureTech Lab) Research Assistant
  2. Jun – Jul 2025 Wolfram Institute Summer Research Program
  3. Feb 2023 – Mar 2024 MFS Investment Management Software Engineer
  4. Aug 2022 – Jan 2023 Meta Software Engineer
  5. Jan – May 2022 PNC Financial Services Group Master's Capstone Project
  6. Jun – Aug 2021 Qualcomm Software Engineer, Intern
  7. 2020 – 2021 Carnegie Mellon University Teaching Assistant
    21-128 / 15-151, 21-241
  8. Jun – Aug 2020 Audition Technology Software Engineer, Intern
  9. 2019 – 2020 Expii, Inc. Content Developer, Intern

Publications

  1. Paper2026 On the Origins of Algorithmic Progress in AI NeurIPS 2026 — Main Track ICML TAIGR 2026 — Best Paper Award Hans Gundlach, Alex Fogelson, Jayson Lynch, Ana Trisovic, Jonathan Rosenfeld, Anmol Sandhu, Neil Thompson
  2. Paper2026 Science Is Falling Behind the Frontier: Foundation Model Adoption Across Half a Million Papers NeurIPS AI4MetaScience 2026 Ana Trišović*, Alex Fogelson*, Janakan Sivaloganathan, Neil Thompson
  3. PaperApr 2026 Two AI Metrics Diverged: Will it Make All the Difference? ICML TAIGR 2026 Alex Fogelson*, Zachary A. Brown*, Hans Gundlach, Jayson Lynch, Neil Thompson
  4. PaperDec 2025 Four Facts About U.S.–China AI Competition in Science SSRN — Pre-print Ana Trišović*, Alex Fogelson*, Neil Thompson

Writing

  1. EssayOct 2026 How Much of AI’s Mathematics Ability Is Selection Bias? Companion: Live Unsolved Math Problems Tracker
  2. EssayMay 2026 How Much Would It Have Cost Claude Shannon to Train Claude Opus?
  3. EssayMar 2026 Can Academic Science Keep Up with the AI Frontier?
  4. EssayJan 2026 On the Origins of Algorithmic Progress
  5. EssayJul 2025 The Computational Limits of Language Models with Wolfram Research

Talks

  1. Talk2026 AI in the Economy: A Technical Perspective AI Safety Foundation — AI Safety Week, Toronto (upcoming, Nov 10)
  2. Talk2026 The Rapid Growth of Foundation Model Usage in Science Wharton School of Business — AI and the Future of Work Conference
  3. Talk2025 The Computational Limits of Language Models Wolfram Research — Wolfram Technology Conference
  4. Talk2024 The AI Risk Repository Elisava School of Design and Engineering