For me it’s the dead. Most days I’d listen to one or two shows from today in history on relisten.net/today.

I’m sysadmin not a dev but I also feel that managing an agent is a fundamentally different feeling than performing admin work like I used to. I used to build bridges out of software. I would learn how they work down to the nuts and bolts, and use them to make robust and occasionally clever solutions for our needs. Over time i was getting better at bash and powershell and regex and automating little tasks.

Gaining knowledge about kubernetes and building images and helm charts was some of the most fun I’ve had professionally. I found enjoyment and value in learning those things and enjoy knowing them for their own sake, much in the same way I enjoy being able to recite mechanics and knowledge about vanilla wow from memory. The knowledge is its own pursuit and obtaining it was fun and fulfilling.

Using AI is nothing like that. It’s not “fun” to me in any way. I don’t learn bash and powershell and templating. I don’t get to enjoy simple wins. AI does those all those for me.

I’m thinking of becoming an electrician. I can’t imagine babysitting an ai for another 30 years.

▲

davnicwil 4 hours ago | parent | next [-]

I completely understand the active/passive difference but I have a question for you that I think especially applies to sysadmin: isn't the knowledge you have and your continual learning just as important if what you're essentially doing now is using an LLM to help you find problems, propose solutions, and test the solutions?

Maybe it's even more important since you've gone from crafting your own solutions to reviewing the viablity of solutions from the LLM which occasionally are going to introduce new ideas, do things in a similar but slightly different way to how you would, or even be subtly wrong in ways that also exercise your ability to adapt and learn and nudge it in the right direction instead.

I don't really understand how the alternative could be true, in the sense of your knowledge not being relevant, as that would imply a couple of things that don't seem true as of now: that LLMs always produce optimal solutions in any given scenario, and (as a corollary) that sysadmin is entirely automated apart from some sort of thin beurocratic human approval/deployment layer.

In my experience so far I've found sysadmin type stuff to actually be what LLMs help me the least with, excluding text-in-text-out scripts, precisely because the setup often has to be so custom to the system, there's loads of side effects that are hard to communicate to the LLM so it oversimplifies and gets things wrong often.

So hopefully you can understand the implied challenge behind the question. I'm not saying your feeling of it being less fun is invalid. But I'm really curious about what you said specifically about the fun of applying knowledge and learning having been diminished. What do I have wrong?

▲

davkan 2 hours ago | parent [-]

So the biggest change is the day to day workflow change of just pumping out little powershell scripts for whatever I need and not building those scripting skills or performing the manual rote admin work. Rather than manually expand a disk in the hypervisor and then sshing in and expanding the partition and filesystem Claude spits out a one liner and I run it. While not challenging or particularly rewarding those trivial tasks are a soothing part of the constant flow of work that i didn’t dislike. Solid things in which i understood every aspect becomes a black box that does it for me and i have to reinterrogate or accept it, and in which it’ll never be worth the time to learn myself.

As for the deeper work, the most recent example I have is deploying an observability stack for our lab cluster. In the past i would have done way more upfront work on understanding the alternatives and the deployment. I would have known every line of config before pressing deploy including the ones I chose not to set. The why and how of everything. Now the most efficient way is just to tell Claude to stand up a monitoring stack in the manner that apps are already deployed on the cluster and iterate with Claude once it’s up and running. Why bother diving deep when you can be running a proof of concept in five minutes and then just see if it falls over.

Another part of it is that i really dislike learning through a prompt. I don’t like summaries, i like man pages. I don’t want to interrogate a result i want to provide a complete solution.

Part of it too is I’m fairly new to all this professionally. I’ve only been at it 5 years, though hobby much longer. Basically i can feel myself growing and getting better and now suddenly there’s this genius moron in my pocket that so much better than I ever dreamed I could be while simultaneously having huge shortfalls. It’s a just an instant paradigm shift and after the first week of glee it’s not been a positive feeling.

It’s not that my existing knowledge isn’t relevant at all, there’s still a huge base of networking and systems knowledge that’s necessary. It’s routinely surprised me in my career when i speak with a greybeard developer who is a an absolute wizard to me but dns/tcp etc are just a black box to them. Or when I talk to a friend new into the selfhosting hobby and i try to explain something to them that think is simple and i realize there’s 10 years of accumulated knowledge underneath that simple concept in order to actually grok it.

I don’t know if that answered your question, just an unstructured ramble. I think the heart of it is that googling and research and things being concrete and complete understandings of workings and interactions and variables and secrets and configs etc are giving way to a black box pumping out systems and it mostly working out of the box and often that is simply good enough for now and it’s on to the next thing.

	▲	davnicwil a minute ago \| parent [-]
		appreciate the response. I understand better now, and agree with a lot of what you've said about kinda the mechanics of scripting stuff 'by hand' being just quite fun and satisfying but maybe no longer the expected way to do things, and so in some sense the sweet spot of that being something you can be paid to do is largely over. Interesting times for our profession.

▲

deaux 3 hours ago | parent | prev [-]

I've learnt more, and more quickly, with LLMs than before them.

Not syntax, no. Bash and Powershell are syntex. But tradeoffs, concepts, understanding new domains I didn't know about.

	▲	davkan 2 hours ago \| parent [-]
		It’s more than just syntax for me though, it was learning how to code in general, through bash. I was slowly building those skills and starting to apply those lessons when looking at product code. I could just devote that time to study instead but I’ve always learned by doing in my career. Now LLMSs do the doing.