AI/ML Archives - Rainer Gerhards

2026-04-102026-04-10

AI codegen is growing up. So are the price tags.

Over the last couple of days, I noticed something that feels small on the surface but is actually part of a much bigger shift.

OpenAI introduced a new higher Codex-focused tier. At the same time, I could clearly see that the practical 5-hour limit on the lower tier had become noticeably tighter. Even light documentation work burned through it faster than I was used to. I saw similar trends and implementations at other AI tools, e.g. the Google codegen ratelimit reductions.

I do not think this is wrong.

2026-03-172026-03-17

Adding Project Policy Checks to CI

If you want the computer science summary: it is at the end.

Many open source projects struggle with “AI Slop”, and some have begun to judge a PR if it is AI generated or not. And treat it differently or do not accept it if it is AI-generated. This in my opinion is generally not the right move. It is more important if a PR / patch has good quality or not. So checks should focus on this.

Cubic AI review interface showing custom policy agents evaluating PR quality, policy compliance, and AI probability in rsyslog — Policy-compliance rating prompt in Cubic Review Agent as of 2026-03-17.

2026-02-272026-02-27

AI slop. AI is hype. AI will replace us all.

Or maybe not.

If you are fighting low-quality AI code, or you think the whole thing is just marketing noise, here is a radical idea: fix the environment instead of blaming the tool. Doing AI right is not rocket science. It is mostly common sense. And, as so often, discipline.

I keep repeating this because it matters. There are three simple pillars that make AI code generation actually work. I can prove it, I do it every day.

2026-02-122026-02-12

AI Slop – or Human Laziness?

It’s strange – we have great new tools, but many folks are using them in such a sloppy way, that the tools get discredited. And this also tends to boil down that serious users get into trouble. You probably guess what I am talking about: AI tooling.

I am for sure nobody who jumps on the latest hype. So I resisted AI quite long for complex things. Until it was ready, which for me was around summer 2025. That was for coding. For some doc writing it was ready earlier (and I used it to cover my weak spots). Now, AI has evolved to also help with video and audio.

2026-02-072026-02-07

Using Local AI Review to Save CI Time, Money, and Nerves

You know, I like efficient processes. After all, that was one reason that I wrote rsyslog. Which, btw, nowadays is increasingly useful and cost-saving as an ETL/ingestion issue for its speed. So no surprise, I also like efficient workflows.

Terminal showing iterative local AI code review with cubic CLI, code fixes, rebuilds, and test runs before final commit.”

We strongly believe in CI. Especially with AI code generation, it is your ultimate safeguard. However, CI is costly, and AI review usually runs max once per CI run.

So I have paired that review with some local test execution and review. Nowadays of course AI assisted. I usually use CLI tools for their efficiency. As part of the post-build process, I make the AI run various checks. The last one is a full review. I often use cubic for that, because it provides very good results to me.

2026-01-222026-01-22

AI Code Generation in a 200k LOC C Codebase: What Actually Worked in rsyslog

If you want the CS summary: it is at the end.

I keep seeing the same take pop up: “AI is overhyped. Mostly money burning.” Sure. There is hype. There is also a whole lot of low-effort “vibe coding” that produces low-quality output at impressive speed.

Illustration of a robotic hand carefully adjusting gears in a large rsyslog C codebase machine, symbolizing AI-assisted maintenance of mature software.

But there is also something else: if you treat AI as a serious engineering tool and you are willing to do the unglamorous work, it can make measurable difference and boost productivity and quality.

So here is a concrete case study: agentic code work in rsyslog.

2025-11-172025-11-17

The Real Scope Behind the rsyslog Documentation Overhaul

For a concise Computer Science summary of this effort, see the section at the end of this article.

When I began the current documentation overhaul, the objective was never limited to cleaning up a few pages. From the beginning, the plan was to prepare rsyslog for the AI era. And the truth is simple: without modern AI tooling, this work would not have been feasible at this depth or speed.

Symbolic illustration showing documentation, an AI head, and a graph structure representing RAG.

2025-10-292025-10-29

Having a good Time with DO Gradient AI

I thought I provide a quick “happy camper” status update. Even though this was only a side-activity, I had an exceptionally good day working with Digital Ocean Gradient and open source models. I am working on support Agents for the rsyslog ecosystem.

rsyslog site with DO Gradient-based AI Agent. Right now, we do not auto-expand it, but this might change.

2025-08-212025-08-21

When Humans and AIs Overthink: a “complex” rsyslog crash that wasn’t

I chased a rare crash in highly-threaded code. It popped up now and then; earlier fixes didn’t stick. I suspected an advanced concurrency issue. I also asked Gemini, Copilot/Codex, and Claude for help. They agreed with me: surely something subtle—epoll, re-queueing, ownership flags…

Human and AI thought bubbles full of tangled lines; a small check mark off to the side. — My AI use on images as inferior, as you can see here. I hope you like that fact ;-)

We were all wrong—and, importantly, I was wrong in the same way the AIs were. Their analyses reinforced my initial hypothesis. The fact that the static analyzer reported nothing reinforced it even more—after all, that’s “proven non-AI tech.” In hindsight, if I had thought earlier about the limits of these tools (AI and non-AI), I might have changed direction sooner—but I was also primed by experience: in this part of the codebase, bugs are almost always complex.

2025-08-202025-08-20

Dogfooding the rsyslog Commit AI Assistant

I’ve been using AI to help with commit messages for a while now. Yesterday, in a discussion with co-workers, it became clear that this may not just be a convenience feature — it’s turning into a real time saver.

That was the background for creating the new rsyslog Commit AI Assistant. It directly addresses a problem we ourselves face in daily development. True to dogfooding, we now use it internally whenever we craft a commit message — myself included.

The “rsyslog commit assistant” in action. You can even see my typos ;-) (Screenshot: Rainer Gerhards, actual session)

Want to give it a try: use the rsyslog commit Assistant.