ML on Code: Machine Learning Will Change Programming
summary
Machine learning on source code treats code as the input and reuses data‑mining and NLP techniques but with stricter correctness constraints. Workloads train on all of the repositories on GitHub and require heavy cleaning and unification, rely on fancier machine learning algorithms, and accept that models are probabilistic so predictions can be wrong. Practical systems must surface suggestions that compile or fail safely and include error handling when generated code is incorrect.
AI-written summary · about
related talks
45mProgrammer’s Apprentice Season 2: Future Directions in AI-assisted Coding
46mCode + AI: Will Robots Take Our Coding Jobs? ML Applied to Programming
26mHow Machine Learning Can Help
43mMachine Learning for Developer Productivity
01h 28mFuture of Software Programming with AI Bots at Hand
33mPractical Evaluation of LLMs for Code
22mAdapting to the Future of Work in Tech
29mHow the Future of Software Will be Different