Explore a 34-minute conference talk from the Linux App Summit 2025 where Sriram Ramkrishna, a 30-year veteran of the GNOME community, examines how Large Language Models (LLMs) can reduce barriers to application development, particularly for resource-constrained projects. Discover practical insights from Sri's experiments with code-generating LLMs, including where they prove valuable and where they fall short. Learn about necessary guardrails to implement when using LLMs for code generation, especially as these tools increasingly contribute to pull and merge requests that can create additional work for maintainers. Sri brings his extensive experience as a community contributor, Linux App Summit organizer, and Intel program manager to provide a balanced perspective on integrating AI tools into the open source development ecosystem.
Overview
Syllabus
Using LLMs to lower barriers to write apps | Sriram Ramkrishna @ LAS 2025
Taught by
Linux App Summit