Hey. I’m just one idiot. Who else are you talking about?

What improvements have there been in the previous 6 months? From what I’ve seen the AI is still spewing the same 3/10 slop it has since 2021, with maybe one or two improvements bringing it up from 2/10. I’ve heard several people say some newer/bigger models actually got worse at certain tasks, and clean training data is pretty much dried up to even train more models.

I just don’t see any world where scaling up the compute and power usage is going to suddenly improve the quality orders of magnitude. By design LLMs are programmed to output the most statistically likely response, but almost by definition is going to be the most average, bland response possible.

You’re not just one, you’re one of many. All saying the same shit.

Lots and lots of people have told me that.

That tracks.

It’s assuming the ai output isn’t very good. It assumes it can create a framework that necessarily still needs the actual writers, but now they don’t have to come up with 100% of the framework, but instead work on the actual content only. Storyboarding and frameworking is a hodgepodge of nonsense anyway with humans. The goal is to achieve non-linear scaling, not replace quality writers or have the final product Ai written.

This sounds like it takes away a huge amount of creative freedom from the writers if the AI is specifying the framework. It’d be like letting the AI write the plot, but then having real writers fill in details along the way, which sounds like a good way to have the story go nowhere interesting.

I’m not a writer, but if I was to apply this strategy to programming, which I am familiar with, it’d be like letting the AI decide what all the features are, and then I’d have to go and build them. Considering more than half my job is stuff other than actually writing code, this seems overly reductive, and underestimates how much human experience matters in deciding a framework and direction.

Well yeah, I’m not going to argue that a well made game that respects the player isn’t going to do well. But that doesn’t matter to the publishers and their shareholders when they can pump out AI slop garbage year after year and still have people that drink it up. Just look at the yearly shooters and sports games, they sell enough.

Besides, what happens when this sort of slop has been normalised? Look at the mobile market, no one bats an eye at the intensely predatory microtransactions, and you’ll even find people defending things like gacha games.

There was a time where people scoffed at the notion of paying $2~ for some shitty cosmetics, but now people don’t even blink at the idea. Hell, it’s downright cheap in some cases. The AAA industry just has to slop things up for long enough for people to stop caring, because they will stop caring and then continue to shell out for the dubious privilege of guzzling their mediocre, uninspiring garbage.

Not even necessarily a human being! I’d appreciate the fuck out of art if any species made it. But there must be more than uncaring, unfeeling, probabilistic interpretation of input data.

I think the reason so many AI bros are conservative is that conservatives have historically had really bad taste in art/media, so they see the drivel AI creates and think, “oh wow, it looks just like what the artists make,” not realizing that they don’t have the eye to see what it’s missing.

I remember reading a longer post on lemmy. The person was describing their slow realization that the political beliefs they were raised with were leading down a dark path. It was a process that took many years, and the story was full of little moments where cracks in his world view widened and the seed of doubt grew.

And someone who was bored/overwhelmed with having to read a post over three sentences long fed the story into AI to make a short summary. They then posted that summary as a “fixed your post, bro” moment. So basically all the humanity removed. Reminds me of that famous “If the Gettysburg Address were a PowerPoint” https://norvig.com/Gettysburg/

All your base are belong to us.

I should have been the one to fill your dark soul with liiiiight!

Sounds like an interesting read, got a link to said post?

I doubt that. A lot of the poor writing quality comes down to choice. All the most powerful models are inherently trained to be bland, seek harmony with the user, and generally come across as kind of slimy in a typically corporate sort of way. This bleeds into the writing style pretty heavily.

A model trained specifically for creative writing without such a focus would probably do better. We’ll see.

Even in programming there are common feature frameworks. Having a system enumerate them based on a unified design vision from a single source architect rather than 50 different design ideas duct taped together could help a lot. I’ve seen some horrendous systems where you can tell a bunch of totally separate visions were frankenstein’d together, and the same happens in games where you can tell different groups wrote different sections.

Tangentially related, the easiest way to come up with a unique and cool idea is to come up with a unique and dumb idea (which is way easier) and then work on it until it becomes cool. (Think how dumb some popular franchises concepts are if you take the raw idea out of context.)

You’re absolutely right, Lisa!

Aand you end up with… ta-da-m, same old things, just rebranded. Very creative (no)

Wake me up when that happens. Like literally, @mention me somewhere