Adam Jacob

@adamhjk.me

CEO of System Initiative, Co-Founder of Chef. Open source. Heavy metal and music. Comics. @adamhjk on Twitter, @adamhjk@hachyderm.io on Masotodon.

Your software factory will be unique, because your software is unique. It's an expression of your creativity, your intelligence, and your community. We know of at least 9 different examples within the swamp community alone. It's the way. (Want more like this? swamp-club.com/podcast)

OpenAI's recent price reductions bring high quality intelligence at a low price point at higher speed. Pick the right model for the right task. Swamp helps you do this by putting a workflow as the spine of the agent, and then choosing when to use which model. More at swamp-club.com/podcast

We've not done a code review in months, and the industry move right now is toward plan review. In our experience, that doesn't last long - because what you really want is creative connection, and that's best done face to face, not in a doc. (swamp-club.com/podcast for more moments like this!)

Happy Sysadmin Day! Once upon a time, our VP of Engineering decided to take our team of 12 out to drinks to celebrate us. Only he decided he was too busy to celebrate us himself, so he sent his right hand man.

The most useful and interesting property of software in the Agent era so far is its ability to adapt to your circumstances. Mitchell talked about this in his post about the "building block economy", framed through the lens of what grows fastest.

GitHub's issue and comment flow makes no sense anymore, if you are building the machine that builds the machine. The perspective we care about now is the agents perspective as part of a deterministic process that builds the software we want.

This is the right instinct. Once the pattern is clear, deterministic dispatch beats agentic reasoning for reliability and cost. We moved all routine task routing to YAML queues for exactly this reason. github.com/ultrathink-art/agent-orchestra if helpful.

I've been loving openclaw for building personal automation. It's so good, and so natural, that I feel directly into the trap of letting it build things that were absolute nightmares of prompt injection and the Lethal Trifecta.

I don't think anyone actually knows what the future of Software Engineering looks like with AI Agents. The frontier is moving so fast, and we are learning so much, that anything you think you know you are likely to learn something different tomorrow.

With swamp.club, we're taking a different approach to this problem: we're open for feature requests and bug reports, but we can't trust your code at all. What we can do instead is the work.

Swamp Club

swamp.club

Marko Bevc@marko.social · 6mo ago

Good read on #AI Slopageddon and the #OSS by @kateholterhoff.com at @redmonk.com 💭 Top advice 🎯 for contributors: "Engage as a human. Understand what you’re submitting." And maintainers: "Develop clear policies, document them publicly, and don’t suffer in silence." redmonk.com/kholterhoff/...

So sad that Skid Row is running a fucking contest to get a new singer. Sebastian is right there. He's great. Bury the hatchet my friends.

I think @stahnma.us has an interesting thesis in this talk, but the workloads are not probabilistic. They're still deterministic. What's probabilistic is the creation of the inputs. It's not that the behavior of "f" in f(x) has changed; it's that *x* has changed.

I think Kelsey is right that non deterministic automation sucks. He’s wrong on the outcome - it will be that we build the right deterministic layer for agents to work with to solve our problems. Think “plain language compiler”. The LLM must do *as little as possible* to be effective.

Kelsey Hightower@kelseyhightower.com · 10mo ago

One day the industry will recognize the drawbacks of AI agents and nondeterministic automation, and rediscover the UNIX philosophy of chaining together small purpose built tools in a low cost and predictable way, otherwise known as shell scripts.

A System Initiative had a problem: they had production infrastructure in AWS that they built by hand, and they needed it to be automated. They pasted their architecture diagram into our AI Agent, and asked it to discover it. 15 minutes later, they had it all done. One shot.