Beginning a job search, it’s clear that to stay in technology means to be in the orbit of AI and AI-generated code, and I need to come to terms with this and figure out how to navigate it. As such, I’m going to start writing more about AI. I process information via writing, so here we are.

What I will be writing about:

• The hard truths that the existence of AI-generated content forces us to face.
• How to achieve goals now that content/code can be produced quickly that appears to work.
• Where and how humans can add value to systems that include AI-generated content?

That said, there will be guardrails. I cannot abide AI-boosting or AI-psychosis. I also just cannot engage an absolutist AI-doomerism.

For example, here are topics I’m going to try to avoid:

• “Golly Gee, look at what I did with Claude!” - I find these kinds of posts insufferable.
• “Here is how I set up 5 agents, git worktrees, and 4 harnesses to bump my productivity!” AKA Developer Productivity Pr0n. I have never liked these posts before AI and I find them even more uninteresting now. Coloring the output of ls doesn’t matter. Tweaking your harness doesn’t, either. Sure, I have a preference on coding fonts, and configure software, but I’m under no illusion that such tweaks matter in any real way.
• “The latest models are really good now!” - I lived through the 90’s where a bunch of dudes would brag about their video cards and obsess about front-side bus speeds in their homebuilt PCs. None of that ended up mattering at all to anyone, and so I am not interested in whatever some model does now that it didn’t before or whatever.
• All AI-generated output is slop and low-quality. While I can understand this viewpoint, this is a non-starter for my goals: trying to navigate a context where AI-generated output cannot be avoided. Before AI “slop” has always been a spectrum, and even high quality tools had some slop in them. I don’t see AI content-generation as being much different.
• AI is unethical so any discussion of it is bad. I don’t find this a useful perspective for learning, either, and I will re-state something I will continue to re-state: we can’t live a life free of compromise.

For what it’s worth, these LLMs were certainly created unethically. My own work was definitely stolen as part of creating them. Most of these AI companies are run by either horrible people, or people suffering from what I believe is a mass psychogenic illness.

But.

These tools are in wide use by many people who don’t care about these ethical issues. They are in such wide use by so many people in the context I must navigate to earn money that I can’t just ignore them. To come in as a technology leader and say “AI should never be used” is to make myself unemployable.

Bluntly, I can’t afford that right now. My healthcare costs equal my mortgage, I’m not part of an existing social safety net and I cannot relocate to somewhere more favorable to not earning an income. I have no children to take care of my wife and I when we get truly old. I need to stay employed and employable, and I can’t just start at the bottom of new career at 53.

We don’t get to live a life free of compromise.

Given all that, this post is a waypoint for anyone reading to stop reading if they don’t want to hear it. I get it. I’m sick of it, too. But I need to process it all, and I do that by writing.