General Gaming News and Discussion

Malakriss

Malakriss

Golden Baronet of the Realm
13,101
12,471
AI has to not only do the work, it has to accomplish it at a much more efficient scale to be worth the time and energy consumption for that work to be practical. Otherwise we're still back to the principle of slave labor to build the pyramids being the go-to, except now it's slave wages for India and China if the incomplete AI we have now isn't good enough.

Industry wants the magic bullet, to do the myth of cracking encryption but in reverse. Instead of P = NP they want to start with non-polynomial hardware and have it produce exponential results.
 
Last edited:
  • 2Like
Reactions: 1 users
Kharzette

Kharzette

Watcher of Overs
5,978
5,978
It is good for textures, but I couldn't get it to do much else. I need to give it another try. I know people have ways to make it generate meshes, but modeling isn't that hard for me.

I'd really love to have AI generate animation data. That is probably the most difficult task at least for me. I can do janky temporary placeholder anims, but it is a true artform that few can do.
 
Zindan

Zindan

Bronze Baronet of the Realm
7,763
5,604
Cybsled said:
The best use of AI in game development is going to be in helping "fill in the blanks" that a human dev team experiences while making the game: Smoothing out animations, fleshing out flavor content, creating more meaningful "fluff" content.

Like for example, one complaint about Starfield was all the copy-paste POI on planets. What if you could use AI to actually generate more varied content using parameters you input, having a human give some final touches, and then adding that? You can still have your hand crafted POI, but then you have have more varied random POI stuff over "cave00, cave01, cave02" and that is it
Click to expand...
Like sophisticated procedurally created content?
 
Neranja

Neranja

<Bronze Donator>
3,007
4,865
TBT-TheBigToe said:
Radiant quests with the illusion of depth because they can be more varied and complicated than go to A, do B, collect C, wash, rinse, repeat?
Click to expand...
That sounds good in theory, but has a big problem: casualification.

People nowadays don't read complicated quest texts anymore (probably never have), and everything has to be a marker or colored area on the map, minimap, or the radar. WIth a to do list you can check off. You can look at WoW how they have simplified and dumbed down their quest design and presentation for years, because players had problems to find Mankrik's wife south of Crossroads in Vanilla.

It could work if you really make a distinct product aimed at a more hardcore audience. This will however shrink your target audience, so you have to balance your development costs around that. In prior years I would have said "yeah, it's impossible", but then Expedition 33 showed that a motivated team can build a proper game with less than $10 million that runs circles around AAA slop.

So I have upgraded my opionion to: maybe (with proper planning)
 
  • 1Like
Reactions: 1 user
Gavinmad

Gavinmad

Mr. Poopybutthole
45,326
55,114
Neranja said:
players had problems to find Mankrik's wife south of Crossroads in Vanilla.
Click to expand...

star wars GIF
 
You must log in or register to reply here.