When I wanted to write my own game engine in 2015, people just directed me to Unity. Some people even said that unless I want to work for some company that actually develops the engines, I don’t need to learn API or anything like that. It took a while to people to point me in the direction of SDL for input and displaying a framebuffer (back then I used CPU rendering, as almost no one cared about pixel alignment, hence why my engine is called PixelPerfectEngine).

Police fans say the same thing about the police.

Nowadays:

“I’m about to start hunting mice, any tips?”

<output pasted from humanGPT>

“Ask HumanGPT to do it.”

“Bro, you need to start agentic hunting by yesterday, if you don’t want to be part of the permanent underclass. Sign up to our webinar on prompt hunting at <link>, the first few classes are free.”

Even worse is, many such “rational” use of AI often involves more AI usage than just posting slop.

They exist, but I’d like to pretend they don’t.

That ADHD meds are just fancy meth, so we have to rigorously select who gets them.

Some of us have to rawdog it, due to the meth myth.

When I wrote my numerical parser, people told me to use regexes to detect what kind of numerical value is being parsed. I just did it in one go. Only regex-like checking is for ISO datetime values.

/ˈlɪnuks/

https://en.wikipedia.org/wiki/Entryism

But not on FreeBSD!

Now I want to see a Bringus Studios video titled “Gaming on FreeBSD, because Linux now runs Fortnite”.

Now I have to move to FreeBSD…

And tells you it is your girlfriend and to shoot up your school.

The “open-weight models” argument just a fallacy IMHO, frontier models would require a pretty strong AI PC, not something that GTX1050 you put away as a backup could handle. The “just use smaller models” argument just gives leeway to just not using AI and using more classical methods of coding - use of dependencies, use of code generators, use of semantic search engines (can someone suggest me one?), etc. LLMs don’t do anything particularly new, they’re just mimic how humans talk.

Slop back then was using Electron at all cost, since it allowed them to hire cheaper web devs instead. Now it’s vibe-coded Electron apps.

> the end of software engineering

> looks inside

> overhyped slop generator

The goal of AI providers is to make humans unable to maintain code, so you have to rely on their expensive subscriptions and tokens.

It makes gatekeeper types in the Linux ecosystem happy, because it preinstalls you neovim, hyperland, and a bunch of bloatware, so they can get to their tinkering much faster.

Once I watched a Mr. Beast video on Twitter, my reaction to it was that it’s slop, way before the term was associated with genAI.

TBH sometimes I’m also guilty of this, if I forget about them. In one library on mine, I’ll have to put all error codes under a single enumerator, so I don’t need to make new ones every time I add functionality.