I really avoid writing publicly about AI.

Why?

For a few reasons - I'm as concerned as everyone else about rapid tech expansion harming the environment, about applying AI to everything and further enshitifying the world, about IP infringement that really only hurts individual artists and authors, about what our brains do if we rely on it. Hell I'm still sitting with my concerns about what the INTERNET is doing for us. - Anything you say is out of date in like a month. People are also constantly trying to predict the future and we are all probably wrong and right in unexpected ways - In the States at least, we really like to divide across binaries and argue at each other... and AI optimists and skeptics are pitted against each other and of course I'm both in a nuanced and individual way.

So why now?

I spent more than a year thinking about it, learning about the environment, re-learning about data centers since my last cloud provider job, and listening to other software engineers, musicians, artists, teachers, doctors, my mom, other people's moms, basically anyone who would talk to me or at me about it.

And at this point I feel it's important I speak to other people like me: crafts people... software engineers. People who can work hands on with this stuff.

Now that tech has so fully embraced AI, the tools have gotten so good, I'm afraid it's far past time to engage with it. In fact if you're still getting bad results, you probably just have more learning to do. Learning to use AI coding tools is about as complex as learning an editor for the first time or maybe like a second programming language.

If you want the more hands-on version of that learning curve, I wrote Learning Agentic Coding On a Budget.*

The engineers I'm seeing go through the most pain are folks who've been on the backend of the adoption curve. Who are sharing their first impression but haven't learned more deeply about tradeoffs yet.

Staying neutral, wishing things were different, expressing your frustration but continuing on the same path isn't a neutral choice. And at this point, it seems to have consequences for software engineering careers at the moment.

Why are you telling me this?!?!

A couple years ago I read this book "Collaborating with the Enemy, How to work with people you don't agree with or like or trust." It was authored by Adam Kahane who has mediated between people who had literally made assassination attempts on each others' lives and between people who've oppressed and people who have been oppressed. Here's what I took away that applies here.

You can: - Embrace the conflict and collaborate. This is going to mean moving forward. Engaging with it. Learning about other perspectives. Being willing to be wrong some of the time. Being a part of it. - Escape. Walk away and do your own thing. I'm not sure what that means for you. If it's inside tech it might mean finding something other people value that doesn't ever require touching AI. It'll be HARD but it might be really meaningful and a net positive. Or you could learn some hard lessons. Probably both! - Fight it tooth and nail and be totally at odds with it. Asking other people to do something different while working for them isn't really fighting. You have to actually oppose it. I personally know people who refuse to touch AI, are angry with me for working with it, try to boycott every company they learn is using it, try to find ways to MAKE it stop. I have friends who are angry at me or at least uncomfortable about my job. Totally opposing AI, to me, is the hardest of the paths in this situation and not one I would choose personally. It sounds hard and lonely. If you really believe in it, I wouldn't be surprised to find there's value in it. But it's not for me and it's not for the faint of heart.

My choice is to collaborate. I chose to write software. I was early in working in the music industry when I decided to. I had options. And writing software was no more a neutral and non-political choice 15 or 20 years ago than it is now. I chose to do it so I could be involved in making things that made people's lives SLIGHTLY better and to make things that could connect people. Sometimes my jobs where I work for other people take me far away from that mission and when they do I tend to change gears and do something else.

Frankly I'm having fun collaborating. It's hard. It requires learning not just how to use AI, but actually what are the environmental impacts, how can I make decisions that can move the needle on problems, how do we maintain our humanity and our connection to each other in such a connected and yet totally disconnected world at the same time. It requires being a part of debates, disagreeing with people, holding space for difficult conversations and feelings...

But I'm having fun with it because collaboration is connecting. I end up with more compassion and love to give other people. More patience. I hit days where I'm tired or I want to quit... but I actually think me showing up and trying to do right by others can matter. You can think I'm silly. I think meaningful things are often pretty silly.

What hasn't changed is...

AI is changing a lot of technology and surprising us with capabilities. Randomness does that. We are used to building deterministic and predictable systems and it's made us think that by wrapping enough process, metrics, and planning around software we can control every outcome.

We can't. We couldn't really before. And a lot about working with software has changed.

But people haven't changed. What it takes to understand big complex systems, systems that include people, hasn't changed. And what it takes to show each other patience and love in a world full of hate and conflict and competition hasn't changed.

Disclaimer

A disclaimer before I get comments: this post was hand written and not AI generated. I'm just autistic.

Footnote