I want to share my bad experience—maybe it will be useful (Daz or stable diffusion)

[render9000]

I want to share my bad experience—maybe it will be useful.

So, the main heroine. She has to be as well-developed as possible, right?
I redesigned her hundreds of times. Every single detail had to be perfect.

Realizing that my 3060 is not meant for 3D at all, I dived deep into AI.
And then, I redesigned the main heroine many more times.



Don't think it's easy—just press a button and it's done. AI is just another tool, like DAZ. It only seems simple at first. Once you dig deeper, you’ll completely drown in it.

In total, 6 months to create renders for one game day in DAZ. And another 6-8 to completely rework all the content two more times in stable diffusion. The first time on XL models and then again on pony models.

Result: complete burnout and abandonment of the project.

Don't devalue the work of AI artists. Please. Believe me, it's not as easy as you think. Yes, with experience everything will be easier and faster, but it is also a lot of work.

 

[NeonSelf]

So, the main heroine. She has to be as well-developed as possible, right?
I redesigned her hundreds of times. Every single detail had to be perfect.

That is common mistake for new developer - you dont have to make everything perfect, you dont have to make a huge project to have success. One released mediocre game is better than 10 years of polishing and never releasing anything.

Just keep the same level of detail and polish on everthing and people wont notice any problems. Because consistency is more important than a single perfect model.

There is always space for improvement, but you dont have to make everything perfect. There are successful games with standard models from outdated editors, games without sounds and aminations, with worst possible UI or insane grind. But still they've got their fanbase and keep updating and improving one step at a time.

 

[Icarus Media]

One released mediocre game is better than 10 years of polishing and never releasing anything.

This ^

 

[dontcarewhateverno]

I want to share my bad experience—maybe it will be useful.

So, the main heroine. She has to be as well-developed as possible, right?
I redesigned her hundreds of times. Every single detail had to be perfect.

Realizing that my 3060 is not meant for 3D at all, I dived deep into AI.
And then, I redesigned the main heroine many more times.

View attachment 4659070

Don't think it's easy—just press a button and it's done. AI is just another tool, like DAZ. It only seems simple at first. Once you dig deeper, you’ll completely drown in it.

In total, 6 months to create renders for one game day in DAZ. And another 6-8 to completely rework all the content two more times in stable diffusion. The first time on XL models and then again on pony models.

Result: complete burnout and abandonment of the project.

Don't devalue the work of AI artists. Please. Believe me, it's not as easy as you think. Yes, with experience everything will be easier and faster, but it is also a lot of work.

I think this is true for people who care about their work, their consistency and quality. Which you obviously do. For some though, it's really just pumping out quick, cheap slop with 6 fingers where the model changes every single pic.

 

[tanstaafl]

One released mediocre game is better than 10 years of polishing and never releasing anything.

Ah, the good ol' work life balance vs. profit debate. Can't argue this sentence for most devs or even small dev shops, but I'll just ask a single question. What if FromSoft thought this way?

 

[MarshmallowCasserole]

What if FromSoft thought this way?

Actually FromSoft pumps out games at a very respectable pace, and some of them to fairly mixed reception. So they are in fact thinking somewhat along these lines. Now, Valve... But Valve have the biggest cash cow in the gaming space.

 

[Dark Lord Nova]

One released mediocre game is better than 10 years of polishing and never releasing anything.

Ah, the good ol' work life balance vs. profit debate. Can't argue this sentence for most devs or even small dev shops, but I'll just ask a single question. What if FromSoft thought this way?

And yeah look at Take-Two Interactive and Rockstar. They've been building GTA 6 for over a decade—not to just drop a game, but to craft something truly special. After watching the second trailer. The visuals are insane, the world feels alive, and the attention to detail is next-level. You can tell they're not rushing this—they're taking their time to get it right.

They're clearly not in it for just a quick cash grab. They want GTA 6 to be a massive success, and honestly, I respect that. I wouldn't be surprised if all the money from milking GTA 5 Online is helping fund this beast. And hey, if that's what it takes, fine by me.

If waiting means we get a game that actually lives up to the hype instead of some half-baked mess, I'll wait. Every time. I'd take one masterpiece over five forgettable releases. Hopefully, we really do see it on May 26, 2026—even if it's taking them one extra year polishing it.

 

[NeonSelf]

And yeah look at Take-Two Interactive and Rockstar. They've been building GTA 6 for over a decade—not to just drop a game, but to craft something truly special. After watching the second trailer. The visuals are insane, the world feels alive, and the attention to detail is next-level. You can tell they're not rushing this—they're taking their time to get it right.

They already have experience of releasing the first 5 GTA's before GTA 6. Have you seen GTA 1? That game was needed mostly to test the concept. Making something like GTA 6 as your fist or second game is a huge risk, that even experienced developers have failed to achieve. Well known example is Duke Nukem Forever, that took 12 years to make.

My advice was for an indie developer, not for a giant corporation, that has stable income from previous games to support the development of a new game for decades.

I'm making games for more than 20 years now. Even with such experience I would not risk making a comlete game myself (again). It takes not only programming skills, but also art, testing, advertising, dealing with feedback etc. While being dependant on its financial success and running on personal savings or borrowed money.

 

[rk-47]

"AI Artists" - ahh love a good oxymoron

 

[suprisedcrankyface]

What if FromSoft thought this way?

Dark souls was released with a half finished tail end, so it kinda proves the point. Good enough is good enough, if time is a luxury you don't have.