20/03/2026
We used Higgsfield. Damn tool from hell, it’s actually great. It looks like this. (I’m blurring some images because it’s a shared account and there are projects from other people and my colleague.)
Well, it did what it could. How about just one page…? Better, but a lot of inconsistencies.
How about panel by panel? Much better; I can work with this. And I can upload reference images. ‘Done,’ I said. We uploaded 3D reference images or sketches and had it swap them for the scene I requested.
And in fact I did use a couple of these.
Ok, ok. I can admit the right one (with DLSS5 turned on by IA 
) looks better, point for the AI. Even more, it can make this style and another much more complex, even realistic, with the same effort. 2 points for the AI.
Even more it’s actually great at making Stylesheets for reference, and you use the same Stylesheet so the character stays consistent! 3 points for the AI. But…
Take this comparison:
1st image: is the ref I passed the AI
2nd image: is what it returns (although is the same pose, like it’s not the same idea.)
3rd image: I decided to do it myself
4rd image: Final result (It’s a matter of taste, but I think point for humanity)
And that is something happened all along the comic, specially when is more explicit, it has problems with the NSFW. It can’t handle many angles because it tells me they’re obscene. I understand there are other tools like ComfyUI where I could do NSFW, but it’s not as automated as NanoBanana2 in Higgsfield. So, I’d waste time trying to get it to do what I say.
So I basically had to redraw everything because in the parts where is explicit I had to do it manually. But in the previous where it has no problem I cannot use them either because I have to make an uniform style along all the comic. When they are dressed and when they are nude.
The final panel is unthinkable for the Nano Banano in Higgsfield.
At the end, I fed it with the first pages of the comic asking for a cover. without too much explanation, like ‘Get creative, AI,’ and it gave me something. Let’s be honest, I was just lazy here. I’d have to describe the scene better, like: “Rear view of Cammy protecting her face, Leona in the background in Orochi mode unleashing her power, things flying here and there, some people on the balcony. Manga style.”
But I wanted something like, surprise me AI, And they are weird but ok. So I made this one
Again a matter of taste, but I like it more mine. But…I’m not a pro using AI. I’ve seen people using it better, for comics too.
There is much more to tell, but for the moment my story with AI was kinda I ask for an scene, it would make them, and maybe that would be enough for many, but I don’t like compromising on an idea. I asked for something, and it gave me another thing. So, I had to redo again.
And througth this, I realized I truly love drawing. I felt a bit strange using it… all the time.
I saw a YouTube post a while ago where someone commented that when AI started, people thought it would kill off monotonous jobs like filling out forms and leave us free to create and do the fun stuff. But the opposite happened: it kept the fun of drawing for itself and left us filling out forms (typing prompts).
It can be fascinating trying to get the AI to do what you ask—I’m not being categorical about it—I know the tools exist, I’m a bit better now with it, it has some good points.
I don’t rule out keeping an open mind and using it in some way in the future. As long as it doesn’t make me invisible.
Because in the midst of this particular chaos, I did end up feeling like I am where I need to be… drawing.
THANK YOU GUYS, SPECIALLY IF SOMEONE REACHED OUT THIS POINT….NO…NOBODY… 
Well let’s continue living.
