korrents
Chris Wellons

What Chris Wellons thinks about coding agents

@chris-wellons · 15 positions · On ppll ↗

Software engineer who writes about low-level and systems programming at nullprogram.com; author of the Elfeed feed reader, the Endlessh SSH tarpit and Enchive.

Chris Wellons did not write this page.

We collected these quotes from things they published elsewhere, and every quote links to where it was said. They have no account here and have not endorsed this site. Quotes are word for word; the short line under each one is our own restatement, not their wording. Their own site. Is this you? Claim it or ask us to remove it. Or tell us what is wrong here.

Filter coding agents · Newest first
Topic
All 15LLMs 5code generation 4coding agents 4design 4developer tools 3AI and jobs 2compilers 1crypto 1cybersecurity 1
4 more topics
innovation 1open source 1software quality 1vibe coding 1
Sort
NewestOldest

Only Chris Wellons on coding agents. See everyone on coding agents

4 dated positions, 2026, in their own words. Our reading of what Chris Wellons has said — not written or endorsed by them.

4 positions so far — this page is not yet offered to search engines.

  1. Our readingFrontier AI coding agents can now fix complex compiler bugs like LTO issues in about fifteen minutes.

    compilers

    Their wordsStrap a frontier AI into a good coding harness, show it what's broken, give it all the sources (source.tar), and ~15 minutes later I have a patch.

    ↗What's been going on in w64devkit the past yearnullprogram.com 3rd of 3 in this piece

  2. 5 months earlier

  3. Our readingAI coding assistance now lets a programmer complete in a day work that used to take weeks.

    Their wordsThough now that I can accomplish ~3 weeks of old-way work in a new-way day, this sort of project becomes that much easier to start and finish.

    ↗I have officially retired from Emacsnullprogram.com

  4. 3 weeks earlier

  5. Our readingFrontier AI models still cannot handle certain deep technical implementation details without human guidance

    Their wordsThis took some special care, in particular avoiding any C++ standard library I/O functionality. Current frontier AI cannot handle this detail on their own.

    ↗dcmake: a new CMake debugger UInullprogram.com 2nd of 2 in this piece

  6. 9 days earlier

  7. Our readingFrontier AI now outperforms most human programmers at most programming tasks, with very low defect rates.

    Their wordsAI is now highly skilled at programming, and better than me at almost every programming task, with inhumanly-low defect rates.

    ↗2026 has been the most pivotal year in my career… and it's only Marchnullprogram.com 2nd of 3 in this piece