AI-driven Program Synthesis
AI-driven Program Synthesis


Program synthesis, the ability to generate code from very high-level specifications, has been a dream of computer science dating back to the early 80s. Over the last decade, though, advances in machine learning and formal methods have made program synthesis practical for a variety of domains. Intriguingly, as program synthesis technology has advanced, the community has discovered possibilities that go beyond the obvious applications in software development.

In this talk, I will give an overview of the state of the art in program synthesis, from some of the landmark results in the field to new developments that demonstrate the use of program synthesis as a tool for generating interpretable models from data.


Armando Solar-Lezama leads the Computer Aided Programming group at MIT. His research group has done pioneering work on the application of learning and automated reasoning to automate challenging aspects of programming, as well as the application of program synthesis techniques to problem domains beyond software development, including robotics, learning, and computer-aided design. He holds a PhD from the University of California at Berkeley, and BS degrees in Computer Science and Mathematics from Texas A&M University.