AI, please destroy us
Published: 24 Feb 26 01:06 UTC
Last Updated: 24 Feb 26 01:38 UTC
Knock knock, AI is Here
If you've been living on this doomed rock called Earth, you'll no doubt have heard of "AI" (as an aside, currently it's not really artifical intelligence, it's a fancy text prediction machine).
It's claim to fame includes:
It's looking decidedly unideal for humanity.
"We mourn our craft"
I've recently been reading a few articles about software engineers "mourning the end of their craft". E.g.
The general premise of these articles is that the authors (relatively experienced software engineers) who take great pride in the "craftsmanship" of their work believe that AI is removing the "craftsmanship" aspect of software engineering. Some choice quotes:
What if AI stops getting better and what if people stop caring?
What if most people truly just don't care about tech problems, about privacy, about Liquid Glass, about Microsoft's upsells, about constantly dealing with apps and features which just don't work?
I'm terrified that our craft will die, and nobody will even care to mourn it.
And my favourite:
If you would like to grieve, I invite you to grieve with me. We are the last of our kind, and those who follow us won’t understand our sorrow. Our craft, as we have practiced it, will end up like some blacksmith’s tool in an archeological dig, a curio for future generations. It cannot be helped, it is the nature of all things to pass to dust, and yet still we can mourn. Now is the time to mourn the passing of our craft.
I definitely sympathise with some of this. I truly do think that LLMs are transformational for the software engineering role. It does feel a little like the end of an era.
Even if AI doesn't evolve beyond LLMs, the amount of code that can be spat out in a very short amount of time really has changed the way that software engineers work.
We are expected to be more productive, and business stakeholders question whether we are even worth keeping around.
Big doomer energy
I find the general sentiment worthy of praise. These people really do care about their craft and the quality of their output. In a world of short-termism and profiteering, that's rare and really worth celebrating.
I am also surprised that so many of these blog posts and sentiments are being expressed now.
Here's my pithy take:
Most people have never given a shit. AI is just making it more obvious.
Yes, I am a doomer. But it is based on some real-life experience (normal disclaimer of sample size of 1, anecdotal evidence, etc).
What exactly is being "lost" here?
I am trying to figure out what is being mourned here. From my reading of it, it's bikeshedding.
When I was a data analyst and worked with software engineers to implement analytics tracking, I observed this "dedication to craft" many, many times.
As an analyst who did some coding, I generally understood the changes that needed to be made. I knew they could be done. It should not be hard. Insert some TypeScript code here and there.
But it was not being done. JIRA tickets just sat there In Progress. "The engineers were working
on it". I was told that the software engineers were deciding the best approach to take.
I got fed-up with the constant delays and attended stand-up (a red-flag, in my opinion). From the (poorly run) standups, I could see they were debating details about "structure". Planning an approach that works is fine. Going back and forth on it is also fine. But eventually the actual work needs to be done.
I did not care about these implementation details. I wanted the output of the code. I wanted the business value from it.
With the benefit of time and more personal experience in software engineering, I am even less sympathetic to those discussions now. They were, quite frankly, a waste of time and money. They blocked the delivery of value.
The joy of coding
As AI writes more and more code, humans are probably going to write less code. There is definitely something to be said about the joy of writing your own code. I think it's why many of us get into software engineering in the first place - it's fun.
I think we are definitely losing that, at least in corporate contexts. Coding for fun and pleasure still exists, but I think it will be in mostly personal pursuits. It is a shame, but ultimately the nature of our jobs are changing.
Humans in the machine
In my opinion, within the context of our jobs, we are being paid to solve business problems. Code should be written to meet that brief, and so long as the code can be easily read, is performant enough and can be extended relatively easily, then that should be sufficient.
I don't think that's going to be lost with AI. We now review code more than we write it. We ensure that the problem is solved in a good enough way - one that generates value for end users and gets out of their way.
You can still be thorough and dedicated in this process. There's still the software engineering craft there - we still look for performance optimisation, extensibility, logic, testing, security and all the other important parts. AI isn't perfect and there is definitely still a necessity for a human in the loop.
At least in production code, AI is not being asked to decide architecture or determine security posture. That's still up to humans to decide. It's a critical part of software engineering and I don't think AI is trustworthy enough in that respect.
But the fact that AI spits out code so quickly has removed some of the excuses for bikeshedding. Business stakeholders can code up things themselves. Even if it's ugly, partially complete and insecure, the value is mostly there. The timelines have shrunk. There is lower tolerance for hold-ups.
If that's the "craft" that software engineers are mourning, then I'm celebrating a little.
Back