Boaz Barak

Boaz Barak is a Professor of Computer Science at Harvard University's John A. Paulson School of Engineering and Applied Sciences. He is also a member of the technical staff at OpenAI, focusing on AI safety and alignment. Barak's career includes tenure at Princeton University and research roles at Microsoft Research New England. He is a co-director of undergraduate studies (DUS) in Harvard's Computer Science department, alongside Stephen Chong, and is involved in non-profits like AddisCoder and Jam Coders for coding education in Ethiopia and Jamaica.

• Education: Ph.D. from the Weizmann Institute of Science, postdoctoral fellowship at the Institute for Advanced Study in Princeton.

Barak's research interests span theoretical computer science, machine learning foundations, cryptography, and quantum computing. His recent work emphasizes AI safety, deep learning mechanisms, and computational complexity. He co-organized the Harvard Machine Learning Foundations seminar and serves on advisory boards for Quanta Magazine and the Simons Institute.

The articles highlight Barak's interdisciplinary work. Recent publications focus on AI safety (e.g., watermarking, alignment), deep learning theory (double descent, SGD dynamics), and computational complexity (planted clique, unique games). These reflect his dual emphasis on theoretical rigor and applied machine learning.

• Scientific Awards:
  • ACM Dissertation Award
  • Packard and Sloan Fellowships
  • Foreign Policy's 100 Leading Global Thinkers (2014)
  • Simons Investigator
  • Fellow of the ACM
  • FOCS Test of Time Award

Barak advises current students like Gustaf Ahdritz and Natalie Abreu, with former advisees including Preetum Nakkiran and David Xiao. He is supported by grants from the Simons Investigator Fellowship, Kempner Institute, NSF, DARPA, and DOE, among others.