Joel Lehman
Scientist In Residence at Second Nature AI
San Francisco, California, United States
Summary
👤
Senior🎓
Top SchoolJoel Lehman is a machine learning researcher and scientist-in-residence with 17 years of experience bridging academia and industry, including roles as a former team co-lead at OpenAI, founding member of Uber AI Labs, and employee #1 at Geometric Intelligence. He holds a PhD focused on evolutionary computation and has researched open-ended, creative algorithms, reinforcement learning, and AI safety across institutions from UT Austin to the IT University of Copenhagen. Joel is a co-author of the popular science book "Why Greatness Cannot be Planned," and his hands-on contributions include implementing ANN frameworks and visualization tools in the OpenCog AGI project, reflecting a sustained interest in creativity and integrative AGI architectures. Based in San Francisco, he combines deep theoretical insight with practical engineering and a curiosity for the psychology–ML interface that often informs his work on creativity and safety.
16 years of coding experience
10 years of employment as a software developer
Bachelor's Degree Computer Science, Bachelor's Degree Computer Science at The Ohio State UniversityDoctor of Philosophy (PhD) Computer Science, Doctor of Philosophy (PhD) Computer Science at University of Central Florida
Github Skills (7)
algorithm10
machine-learning10
machine-learning-algorithms10
cprogramming-language9
unsupervised-learning9
c-language9
python8
Programming languages (6)
Github contributions (5)
opencog/opencog
Jun 2009 - Aug 2009
A framework for integrated Artificial Intelligence & Artificial General Intelligence (AGI)
Role in this project:
ML Engineer
ML EngineerContributions:13 commits in 2 months
Contributions summary:Joel implemented and refined an artificial neural network (ANN) framework within the Opencog AGI project. Their work involved creating classes for simple ANNs, including node and connection definitions, and developing methods for input loading and signal propagation. The user also developed code for an ANN scoring system and integrated the ANN functionality into the existing moses framework of the project. Further contributions include the addition of memory nodes, and the creation of tools for visualising neural network structures.
unsupervised-learning-algorithmsroboticsnatural-language-understandingartificial-general-intelligenceintegrated
uber-research/atari-model-zoo
Nov 2018 - May 2020
A binary release of trained deep reinforcement learning models trained in the Atari machine learning benchmark, and a software release that enables easy visualization and analysis of models, and comparison across training algorithms.
Contributions:21 commits, 3 PRs, 13 pushes in 1 year 6 months
pytorchatarideep-learningreinforcement-learningcomparison
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.