Joint Keynote Address: Accelerating Software Development: The LLM (R)evolution
summary
A concrete recipe emerges: evolve existing tools, exploit what they uniquely know, and ensure outputs are validated. Examples: the scaline profiler++ suggests AI optimizations that sometimes hit ~90× speedups; chat DBG uses an enriched dynamic stack plus language servers to raise fix rates from <25% with plain LLM queries to >60% on a single prompt and ~75–80% with dialog; Coverup auto-generates tests to raise often-38% coverage substantially above prior state-of-the-art; Flowco builds dataflow checks and postconditions for safer notebook-style analysis.
AI-written summary · about
chapters (6)
- 0:00The LLM revolution
- 5:00Scalene profiler
- 10:14ChatDBG debugger
- 20:11CWhy compiler
- 26:12CoverUp test generator
- 34:57Flowco notebooks
related talks
44mAccelerating Software Development: The LLM (R)evolution
01mSpeed Up Your Code With LLM's!
33mOpening Keynote: LLMs, between fantasy and reality - the poin...
43mFrom Lambda to LLMs: AI Coding and the Future of Programming
32mWhat can we expect of LLMs as software engineers?
01h 15m10x Development: LLMs For the working Programmer
43mAI Development Lifecycle: Learnings of What Changed with LLMs
57mKeynote: Using LLMs and generative AI to fix software vulnerabilities