They ruined everything: agentic coding, AI slop and losing the joy of making things

I wish I never opened Pandora’s box of agentic coding. If I could reset the universe and remove all traces of LLM chat bots and coding agents as though they never existed, I would do it in a heartbeat.

I say this and it feels paradoxical because I have been extremely productive in terms of my output using these new tools. Both at work and in my spare time I have been blazing through a number of projects, delivering results at breakneck speed, but also delivering code I am happy with and happy to take ownership of, including the long-term maintenance and keeping the agents in check.

This morning I got up, had coffee and jumped around a few codebases. I refactored the web app code in my music site diig, adding an LLM to validate cross-referenced links in its database and improving the efficiency of a nightly job. I also managed some digital music files on my USB sticks for CDJs, with some supporting scripts. At the same time I was iterating with the Blender MCP and Claude to design a wooden speaker enclosure for my smart speaker project oi-speaker. I was doing all of that on the side while working on a little game project and now I am writing this blog.

So why do I feel so unsatisfied? The fact I am doing all of these things simultaneously illustrates how bored I am. I am able to do so much so quickly I need to do something else to keep me occupied. I have always been this way to some extent. Before LLM agents I was a pretty prolific coder at work and at home, able to really min-max my effort to output ratio. This made me great at startups and small studios, but now there is an even more prolific coder in town and anyone can use it.

I have a lot of ideas, and this has always driven me to start side projects. I used to start them and never finish them. Now I can realise these ideas even more quickly. The end result of something more complete and polished is somehow less enjoyable. Previously, the effort and time I put in would get me into a flow state, and the process was ultimately what made the whole thing worthwhile and fun.

But it is difficult to embark on the journey when you have the easy-mode, fast-travel button. It is very hard to sit down and justify starting something and doing all of the gnarly plumbing to even get started on some new feature in a codebase. So instead I jump to the interesting new bit, complete it quickly and then also take the easy-mode route on the way out to finish it off and clean it all up. But this means you can churn through lots of interesting work more quickly and you have to make a lot more decisions. The effort is spent in a different place and it is a lot more tiring, even though it feels like doing less work. The cadence of thinking / planning to slow “busy work” has changed.

There is currently an explosion of slopware coming out, which is accelerating rapidly. Everyone is one-shotting games, entire software suites, reverse engineering, cloning and such. The playing field has been levelled. Work that used to take effort, time and care no longer does. Even if you struggle along and go back to basics, doing things the old way, it isn’t impressive. You don’t stand out as much anymore; someone with more tokens probably will.

What’s worse, most of the slop merchants are just copycats with no skill. They can’t do anything themselves and haven’t put in the work or effort, but armed with a $200 plan of tokens, they can copy your product and release it first. They don’t care about the small details, they don’t have the same love and passion, they just want notoriety or to make money. I think a different kind of person benefits from this new paradigm at the expense of the true creators.

This happened to one of my projects. The “developer” contacted me originally for access to my app (I assumed it was so they could copy it, because they had already released their own). Their version was slightly different in the UI/UX so I thought, fine, and ignored it. Then I noticed they started copying incremental updates I had made. I sent them a terse email and even the reply was an LLM-generated slop response. These people don’t have ideas, skills, or a voice. They made a snide remark about me complaining about them copying and my code being open source. This bit is where these losers still miss the point: open source is not about stealing, it’s an invitation for collaboration. I noticed this sentiment in some recent discussions about reverse engineering where someone who paid to successfully reverse engineered a competitor’s products suddenly realised they were vulnerable to the same thing. Later in the discussion someone else said they went to reverse engineer something and realised someone else had done it already and put the code on GitHub so they said they “stole” the other person’s code. Stealing from the stealers, it’s all just plagiarism now. But we learned from the best, the AI companies stole the sum of human knowledge and now sell it back to us and eventually will replace us all with no remuneration.

Gamedev is my passion, I have always just wanted to make my own games. I have a few projects I am currently working on, but I now feel like sharing progress just risks the ideas being leeched. This destroys chances to build community, engagement and the fun of sharing the journey. It also just ruins any prospect of there even being a worthwhile path to releasing something, first risking being ripped off and then trying to stand out in a sea of slop in an already competitive landscape.

Between these headwinds I just feel empty and unenthusiastic about it all. It should be an exciting and incredible time to have this kind of technology at our disposal, but I’m just not really enjoying it. All the while underneath all of this is the constant barrage of “AI is going to take your job”. Somehow they managed to make the most amazing technology and make it totally miserable at the same time.

I suppose this starts to get into deeper philosophical territory. Ultimately, reaching the destination won’t make you happy. Having those half-finished projects I would daydream about coming back to was maybe a thread to hold onto. Now, when the ideas are realised, that hope evaporates and I have to think of a new idea. The journey and the struggle are the real parts that give fulfilment. Perhaps this links back to our distant ancestors, when every day was a struggle just to stay alive. We have lost the struggle, and we are being sold the lie that if we just put our faith in the AI companies, the magic imagination machine will make us happy.

To anyone out there still doing it slowly, taking their time and perfecting their craft: keep going, you have my respect. The streets won’t forget the real makers, doers and grinders.