Jonathan Shafer

Postdoctoral Associate, MIT; Researcher, RESI; Incoming Assistant Professor, Weizmann Institute of Science

  • Jonathan Shafer is a Postdoctoral Associate at MIT and a researcher at RESI. Starting fall 2027, he will be an Assistant Professor at the Weizmann Institute of Science. He earned a PhD from UC Berkeley, advised by Shafi Goldwasser.

    A theoretical computer scientist by training, Jonathan does theoretical work on learning theory and its connections to cryptography, as well as practical research on applications of AI for mathematics.

  • We will work on two related projects. Candidates may work on either or both:
    1. Theoretical: Developing theoretical definitions and foundations for AI alignment.
    2. Applied: Developing AI harnesses and tools to help accelerate theoretical work on AI alignment.

    See resi.org for more about our approach to AI safety.

  • For theoretical work, candidates should have a strong background in theoretical computer science or mathematics (ideally, a track record of publications at conferences similar to STOC, FOCS, ITCS, COLT, CRYPTO, SODA, etc., or mathematical journals).

    For applied work, candidates should have strong working experience in developing and deploying AI agents for mathematical proof generation and/or Lean formalization, or comparable experience building agentic AI tooling.