John Langford

Computer scientist

American Capricorn

Quick Info

ProfessionComputer scientist
NationalityAmerican
Date of Birth02/01/1975
Age51 years
Zodiac SignCapricorn

Latest News about John Langford

06/11/2026

John Langford announced on November 6, 2026 that he had left Microsoft and that the AI Frontiers group he co-founded within Microsoft Research was over. The roughly 50-person group had been moved into Microsoft AI in September 2026. In his blog post he said his research could no longer reach practical use in that environment.

Do you like John Langford?

You can vote only once.

88% like · 12% dislike

Bio/Wiki

SummaryComputer scientist working in machine learning and learning theory. Known for work on contextual bandits, learning reductions, Cover Trees, and the Vowpal Wabbit project, and author of the hunch.net blog.
Fields
  • Computer science
  • Machine learning
  • Learning theory

Educational Qualification(s)

Bachelor'sDouble bachelor's degree in Physics and Computer Science, California Institute of Technology, 1997
DoctoratePhD in Computer Science, Carnegie Mellon University, 2002

Career

Workplaces
  • Microsoft Research New York (founding member)
  • Yahoo! Research
  • Toyota Technological Institute at Chicago
  • IBM Watson Research Center
Professional Service
  • Program co-chair, International Conference on Machine Learning (ICML) 2012
  • General chair, ICML 2016
  • President, ICML 2019 to 2021
Notable Work
  • Vowpal Wabbit (principal developer)
  • Contextual Bandits (term coined by him)
  • Learning reductions
  • Cover Trees for nearest neighbor search
  • Isomap embedding algorithm work
  • CAPTCHA challenges work

Social Media

Xhttps://x.com/johnclangford

Some Lesser Known Facts

1. Describes machine learning as shifting from an academic discipline to an industrial tool
2. Was a postdoctoral visitor associated with Michael Kearns's group
3. Maintains the hunch.net weblog

Also born on 2 January

Information on this page is collected from public sources and may not be 100% accurate. Report an error
Scroll to Top