Slop.
I came across a post showcasing a fancy new AI tool by a company called Supabase. I like Supabase.
In the showcase, AI takes a prompt and spits out an app database based on the prompt. You know, because databases are grunt work idea people don't need to do. What is the amazing idea though? What is the problem for the awesome power of AI?
“a project management app”
What the fuck is wrong with people these days? Is that seriously the best people can do with years of “AI”? Like, god damn. It is literally a todo list in disguise. Never mind the thousand and one project management app (todo lists) templates on the internet. What a rubbish example.
Look, I am cynical for sure. We know this to be true. We also all know there are a bunch of kids (adults?) with VC money who are “disrupting” industries they know nothing about, like project management apps, by asking AI to spit out the aforementioned templates. They are blown away. It is code from the internet that works. Who knew! It is almost like art from the internet you didn't make would work too if not for copyright laws.
I am just so disappointed that with a tool like LLMs and gen AI the best examples people have are “an xyz thing which already exists” I cannot tell if it is because AI is just so bad it cannot do anything more interesting or if these idea people with VC money really do have so little creative juice.
I know, I know. What are my amazing ideas right?
Ok, how about this for a prompt, “a better jira” or maybe “a better linear?” or better yet “shortcut but without all the shortcuts?” Ya? That one is pretty good.
You see, in those 3 examples I have already shown a better understanding of the respective industry space as I know the names of some existing products. I also had an idea that puts a spin on something one of the products is known for. Is removing shortcuts a good idea? idk, but it is damn more original than disguised todo list app #9001 or spotify clone #102.
I am being cheeky here, but everything in the dev space these days is all about shoveling out as much slop as possible. People brag about pushing work out so fast they forget to update the favicon for their web app. They say you are a bad “builder” if you decided that having a unique favicon a priority before shipping.
Perhaps it is my game dev side hobby background which prevents me from being a true “builder” but the idea of having Vercel's fucking favicon on even a “good enough tm” something I build is just, unthinkable. I think this is, in part, because it takes all of 5 minutes? to grab a free Material Design icon (or something) and slap it in your public folder. Those 5 minutes could have been used bragging on the socials I know but I cannot help but to think that the tiny little bit of “extra” effort makes your slop look just a teeny tiny bit more…betterer? Polished? Like you give a shit?
What really sells my point is that an LLM could add a favicon in like 5s or less. It will even pick an icon for you based on your riveting prompt, if you only asked it to.
None of this has anything to do with poor'ol Supabase. They're just trying to sell some shovels.
Here is a little succinct nugget, AI used by uninspired people will always result in uninspired output. The entire “builder” space today has less creativity in it than it had back when crypto was in vogue… ok maybe that joke is a line too far. You get the point.
Be inspired. Go the extra mile. You don't realize it, but people notice the effort, we really do, even those of us that say we don't.
I could leave this as a simple text blog rant but I simply must show off some silly text effects I've been twerkin with.
Before you ask, yes Balatro was NOT the only inspiration but certainly the one that got me wanting to scratch the wiggly text itch. And yes again, you bet your ass I used the fuck out of LLMs, gen AI, IA, homomorphic gaussian splats, the whole shabang.
First, we have FancyText.
Next, a more recent electric? effect. I had an idea in my head that took way to long to get working and ended up looking like shit. Layered shadows moving towards the middle point of the text content. Ugly result. It did give me another idea which was the foundation for this effect. Never know where the css dragon will take you I'll tell you what.
Oh, and we have a combo component, the SnazzyCard. Give em a tap.
Remember kids, if you ever want to add a little spice to life, just multiply by a -1, it makes all the difference.
I'd share the source code for these but let's be real, after about a month, just ask your local gypity to spit out an "inverted snazzy card". I have complete confidence you will get an identical result as you see here.
Until next time.