When I look at incredible code completion tools like TabNine (using deep learning), I'm not surprised that JetBrains is focusing more on ML techniques in their IDEs!
https://www.tabnine.com/blog/deep/
https://www.infoworld.com/article/3518429/jetbrains-taps-machine-learning-for-full-line-code-completion.html
miniblog.
Related Posts
Really cute approach to reporting type errors: when there's a type error, show an example of a runtime error that the type check has prevented!
Data-Driven Techniques for Type Error Diagnosis https://escholarship.org/uc/item/59s4h4pv
On how far programming language research has come, and the maturity of tools and techniques that make more ambitious projects viable:
Generated Code Generates Overconfident Coders: https://www.deeplearning.ai/the-batch/issue-180/
A study of programmers found that using a LLM for completion produced buggier code but users were more confident in it.
I wonder if this generalises to other completion techniques?
