David Dalrymple

Research Scientist


MSc in Media Technology, 2008

Massachusetts Institute of Technology

BS in Computer Science and Mathematics, 2005

University of Maryland Baltimore County

Davidad studied computer architecture and programming languages at MIT, then biophysics and neuroinformatics at Harvard. He led an international collaboration to employ new methods to interrogate the smallest nervous system in nature (that of the worm Caenorhabditis elegans), funded by personal grants from Google cofounder Larry Page and PayPal cofounder Peter Thiel, which resulted in a publication in Nature Methods. He has also worked in machine learning and software performance engineering at major tech companies and startups. Davidad is one of the co-inventors of Filecoin, but now focuses primarily on longer-term research; he is a frequent visitor to Oxford’s Future of Humanity Institute and the Machine Intelligence Research Institute on questions around the beneficial development of artificial superintelligence. Davidad currently sits on the Industrial Advisory Board of the University of Birmingham’s School of Computer Science, and a joint China-U.S.-Australia AI governance panel.

Areas of Expertise

Distributed Systems, Economics, Type Theory, Category Theory, AGI Ethics, Neuroinformatics

Latest work

2019.9.6 / Talks

Dioptics: A common generalization of open games and gradient-based learners

David Dalrymple (Protocol Labs Research) talks about Dioptics at SYCO 5.

2019.9.5 / Publications

Dioptics: A common generalization of open games and gradient-based learners

Compositional semantics have been shown for machine-learning algorithms [FST18] and open games [Hed18]; at SYCO 1, remarks were made noting the high degree of overlap in character and analogy between the constructions, and that there is known to be a monoidal embedding from the category of learners to the category of games, but it remained unclear exactly what kind of structure they both are.

2018.10.23 / Talks

IPLD research and future directions

David Dalrymple (Protocol Labs Research) talks about IPLD research at Lab Day 2018.

2017.7.27 / Publications

Proof of replication

We introduce Proof-of-Replication (PoRep), a new kind of Proof-of-Storage, that can be used to prove that some data D has been replicated to its own uniquely dedicated physical storage. Enforcing unique physical copies enables a verifier to check that a prover is not deduplicating multiple copies of D into the same storage space.