Unable to load image

rDrama consensus: Aislop in vidya

What is your take (opinion) on Aislop (AI generated art) in vidya (videogames)

!poll_voters

24
Jump in the discussion.

No email address required.

Procedural generation has been used in games for a long time, like for generating maps. It's often meh, but if it works well with the gameplay that make it better. It makes the most sense when you could never have the same quantity of hand-made content. Language-based AI or voice generation has similar applications on a large scale, like thinking of giving unique (-ish) dialogues to all pedestrians in a city.

Jump in the discussion.

No email address required.

>unique (-ish) dialogues to all pedestrians in a city

It goes way beyond that too, ai architecture to design an entire city and its buildings' interiors, on the fly quest and npc generation, etc, etc, etc.

Jump in the discussion.

No email address required.

The thing with AI art in vidya is that when it actually works it will be incredible, basically unending worlds and shiet. But for now it is utter shite for the most part. Like look at AI art, if you zoom in you find that it's nonsense which gives you an eye rolling sensation.

Jump in the discussion.

No email address required.

If you thought games were samey already, AI is a literal copy paste

Jump in the discussion.

No email address required.

Games are good when they are made by passionate people and AI will empower passionate solo developers to accomplish things 500 people going through the motions following blueprints designed by committee never could.

Let's say you wanted to make a deep not wide rpg that takes place in like one small city/town. You could use ai voice acting if you don't want the characters to be text only for example. You could write a quest description like "Write potential quests in a story line about two fueding families for my rpg" then take the best ones and discard the rest, prompt on them to determine what the quest might look like such as "Come up with some potential quest objectives/solutions to a quest about[Sam Boulder the young child of the boulder family being upset that his rival Billy Rock got a new Sword for his birthday] or something r-slurred like that. When it tells you one possible objective like convincing Billy's mother that Swords are dangerous and she shouldn't let her son play with one and instead sell it to you, you could use AI by telling it to write a conversation tree that revolves around this conflict. Where 1/5 outcomes ends with you getting to buy the Sword, one ends with her calling you a swordlet and a strag etc and then fine tune the outcome.

The human is there to prune and hold the reins, the idea isn't that you tell the AI "hey make a game for me" and then get mad when it's slop. You use the AI to empower you to accomplish your passion project, you supply the SOVL.

Jump in the discussion.

No email address required.

OUT!

Jump in the discussion.

No email address required.

I think drafting with AI is fine, but you really need to finish it with human touch and vision. Especially since I think overuse of AI might end up homogenizing lots of art and media since it relies so heavily on existing styles. You could be a genre-defining aesthetic, but AI won't invent your personal style for you.

Jump in the discussion.

No email address required.

Seems like it could be very useful for super small-time developers who need to generate art assets quickly and cheaply.

Jump in the discussion.

No email address required.

I got plans, that I cannot tell you about, because the haters will sabotage me.

again, I got plans. That I cannot tell you about because the haters will sabotage me.

Jump in the discussion.

No email address required.

:#marseygrover:

Jump in the discussion.

No email address required.

Link copied to clipboard
Action successful!
Error, please refresh the page and try again.