Generative AI is Sucking the Joy Out of Work

It’s a bit of a negative title, and I don’t really like writing about negative subjects, but I really had to write this down.

Why I love programming

The thing that got me excited and invested into programming was the problem-solving aspect of it.

Figuring out how to get the result I expect and the behaviour I want is what ignited my curiosity. Putting together the pieces to a puzzle and solving the task at hand.

And to be frank, programming, in any form or language, was just a means to an end.

It always felt great solving another problem, structuring the data, making it all flow, and making readable and understandable code for others to work with.

I still get this feeling when I am working on subjects and problems I’ve not solved before. Though in “actual” work, this is rarely the case anymore.

Why I started disliking programming

On the other side of the coin there is “just work”. Tasks that need to be done, problems you’ve solved a million times before, but just in a different context.

Having worked as a developer for more than twenty years of my life, I have seen a lot of different projects. All with their own challenges to overcome.

One thing that started to bug me was the repetition of it all. And I got bored really quick.

Always looking to spice up my work, I tried to incorporate something new to learn. Which works great short term. But eventually you’ll run out of that too.

Every problem starts to look the same. Everything starts to look like “just work”. You’re solving the same thing over and over again.

We’re just producing

You’re building a lot of things as developers. And a lot of times I question why. What is the benefit? Does the end-user really want this? What is the goal? What is the expected outcome?

This attitude has saved me from building new features, sure, but more importantly I saved the client time and saved them more work to maintain.

I don’t think it’s healthy to make things that need not exist.

Being a developer often times feels like we’re just implementing features for the sake of it.

Struggling with motivation

This is a hard one for me. And I really think we should talk about this more.

I lack the motivation to “just work” when I am not engaged with or invested in the project I am working on.

Tasks that I’ve done a hundred times over will become harder and harder to do. Not because of the complexity, but because they’ve become monotonous.

It takes effort to complete the task, and I know I am more than capable, but it just takes concentration and focus to complete which I do not have without motivation.

I know I am the perfect person to implement a solution with the experience I have, but I’d rather have someone else do it.

Well then AI would be perfect, right?!

I thought so too. Wouldn’t it be perfect to offload the boring work to AI? Let the robots figure out the stuff I’ve done a hundred times over?

In the beginning I was quite impressed with what AI was making based on my input. I used snippets of the output as inspiration to tie everything together.

All seemed great and then it hit me again. I got bored.

Sure it took away some parts of the work I was doing I didn’t like, but it also took away the parts I did like.

It’s the same problem I already had, but now I reached the boredom point quicker. So I stopped using AI.

Cleaning up after and baby-sitting robots

Now that generative AI is being used everywhere, it’s hard not to get into contact with it.

It’s truly great to see people without coding experience being able to generate complete working applications by just typing prompts back and forth with a chatbot.

The nasty side effect is that that work now ends up on the plate of the people that actually understand what was generated. They get to review it and clean it up, or even fix some bugs that exist in the code.

We’re left looking at a huge pile of code which no-one can explain or take responsibility of. This will become your responsibility.

The people generating things feel great about being able to produce stuff. The people left to fix that stuff feel left devalued and frustrated. I have yet to meet someone who actually liked cleaning up after AI.

Remember when we were all complaining about out-sourced, cheaply made, crap code? It’s the same all over, but much quicker.

All the things I love about programming are being taken away and replaced by a prediction model.

What is left is just the part of programming I started to hate.

We’re just producing faster

All I’ve seen generative AI do is ramp up the rate at which we produce things. We’re just building stuff for the sake of it, because now we can.

We as humans have a tendency to do this. Make something more efficient? We’ll just use it more. Make something cheaper? Now we can buy more. Make something go faster? Then we’ll go further.

Generative AI is a perfect example of this. And the companies behind the chatbots are the ones benefitting the most of this.

(Not to mention that the code that inspired the generative AI models is based of free to use, human crafted, open-source work)

(Oh, and also the resources LLMs require in order to be used)

Un-humanification

It doesn’t really matter where you look online. More and more things are (sadly) being replaced by AI slop.

People I used to enjoy reading articles of, start sharing things which are very clearly written by AI.

More and more artwork is being replaced by AI generated overly shining stuff.

There are even music groups dominating music stream services.

And we seem to all consume it like it’s the greatest thing ever.

We are taking away the human connection to the things we do. And it’s happening at a really high rate.

When I notice something is written or made by AI I immediately loose interrest. You might agree with what the AI has written, but I want to hear YOU say it. I want to see YOU come up with art. I want to hear YOU make the music and all its imperfections.

Depending on it

If the latest big outages on “the cloud” should teach us something is that we need to prevent being dependent on these types of technological advancements.

The Cloudflare outage earlier this week showed a weakness in our way of work. A lot of websites and web-services run through this service and were unavailable.

This meant a lot of chatbots and vibe coding services were down too.

The same can happen to providers of LLMs and chatbots. Do you want to be dependent on just one party to solve your issues? What will happen if they hike up the pricing once they have reeled in enough users?

It might not be a problem only related to AI, but vendor lock-in and outsourcing the knowledge of your applications might bite you in the ass.

Old man yelling at cloud

I understand that this may feel like an old man complaining about the current state of the world. And maybe it is partly that. I probably had the same kind of people yelling at me for what I was doing when I started programming.

But I can’t shake the feeling there is something that needs to change here.

I want AI to help me program more and solve the parts I don’t want to do. Not the other way around.

Saw a comic post not too long ago that said: “I want AI to clean my toilet, so I can do art. Not do art, so I can clean my toilet.”
I do not remember where I saw it, but it really stuck with me.

Understanding my position in all this

I believe that there will always be a place for my skills and knowledge in this world. And I believe that I will not fully be replaced by AI.

Sure, some of the work I am hired to do at the moment will be done through generative AI, but I feel like most of that is not the work I wanted to do in the first place.

I’ll focus on the human connection to the projects I work on. I want to get a deeper understanding of the things I do. AI cannot help me achieve that.

Development has changed a lot, but fundamentally programming is still the same. I think the distinction lies here.

My goal is to keep investing time and effort in the things that I think matter. Whether it is for myself, or people I work for.

I still love helping people understand about how I approach development and how they can improve. Improving and honing my own programming skills is one of the things I like doing most.