this post was submitted on 01 Jan 2024
182 points (92.9% liked)

Games

32681 readers
991 users here now

Welcome to the largest gaming community on Lemmy! Discussion for all kinds of games. Video games, tabletop games, card games etc.

Weekly Threads:

What Are You Playing?

The Weekly Discussion Topic

Rules:

  1. Submissions have to be related to games

  2. No bigotry or harassment, be civil

  3. No excessive self-promotion

  4. Stay on-topic; no memes, funny videos, giveaways, reposts, or low-effort posts

  5. Mark Spoilers and NSFW

  6. No linking to piracy

More information about the community rules can be found here.

founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
[–] funkless_eck 1 points 10 months ago (1 children)

You took all that to say what I was originally saying 4 or 5 comments higher in the chain: for idle chitchat yes AI could probably do it, but it can't make quests on the level of CP77's crucifixion quest, all it can do is choose from preselected parameters - but we don't need to do that. We have radiant quests already (as mentioned, Preston Garvey...)

Also your character creator randomizer doesn't need AI, it just needs a restrictive algorithm (maybe not even) not to get too random - ie don't put lizard hands on a human body, limit how big you can make the nose, ensure skin tone is even...)

Same for your canned goods creator, it doesn't need AI it just needs a pick list and an RNG. You don't even need AI to make the picklist, you can just scrape a list of most popular canned goods from Wikipedia or some stats site, and a list of your in-game races/affiliations/species

When you say that in-game characters, geography and props aren't needed, you're wrong — going to New locations, meeting new people and seeing unique things is part of what makes a quest interesting. You must of heard of gamers complaining about reused assets and reskinned characters and guns as rewards- adding a system that puts more "kill 5 goblins to get the same sword you've got but slightly more blue" will only make the game worse not better.

So, no, AI can't "do quests" as most people understand them, it can only create busy work, which is considered one of the things dragging the gaming industry down right now.

[–] [email protected] 1 points 10 months ago (1 children)

Ok, I'm done trying to educate you. You, by your own admission, are lay on literally every topic that you are speaking about and I am literally an expert. I have put in my thousand hours on modeling, animation, and VFX, as well as hundreds of hours on quest design for TTRPGs. I have a master's degree in Animation and a bachelor's degree in Applied Mathematics. I am professionally a developer and am actively designing AIs. Why the hell are you arguing with me on what is and is not possible? You know nothing but what you have read in a few pop-sci articles and blogs along with some observations from playing games. I have all of that too, plus everything else. Your opinions are based on incorrect assumptions that invalidate them from the start. If you can provide me a single scholarly article, published in a journal in the last 2 years that supports a single one of your assertions, I will yield on whatever you are able to support. Hell, I will even take a passage from a textbook on AI design published in the last year. I have already provided support for some of my assertions. Please provide some of yours. Again, scholarly publication. Not blog. Not pop-sci articles. A recent whitepaper from a research group is also acceptable.

[–] funkless_eck 1 points 10 months ago

I dont have to understand the deep complexities of those things because we haven't even got past the basics, and where the key issue lies.

I dont need a degree in animation to confidently tell you that AI isn't in the place where it can generate animations at scale for a playable experience without significant hand-holding. Is that true or not? In your reply above you agree with me saying AI can only pick from a list of already-created animations, which makes it no different to an RNG.

There aren't any scholarly articles about "AI can't make video games for you" because scholarly articles are about proving things can happen. Can you provide scholarly articles on consumer-ready video games that have been created with majority AI parts? After all extraordinary claims require extraordinary proof.

If you can do this- why haven't you? ("The technology isn't there yet!" I know that's been my argument from the start)