Can AI be a TOOL for ARTISTS? My workflow + Pros & Cons

We will see HOW I use AI as a TOOL to create exactly what I want. And review AI to see it's PROS and CONS.
Mainly from my point of view, as a beginner artist. Since this workflow has been developed to cover for my shortcomings. Though I'll share tips and ways to use AI that can be helpful to EVERY skill level.
Feel free to discus what you think in the comments!


Join our Discord:   / discord  

------------ Links used in the VIDEO ---------

Main checkpoint used: https://civitai.com/models/997769/nai...

Lora Used: https://civitai.com/models/654564/sun...

AfterDetailer Extension: https://github.com/Bing-su/adetailer.git

Download the Ultimate SD Upscale: https://github.com/Coyote-A/ultimate-...


Download main CN Models: https://huggingface.co/lllyasviel/Con...
Download CNXL models: https://civitai.com/models/136070/con...
Download Controlnet for Illustrious models: https://civitai.com/models/929685/noo...

ControlNet Extension: https://github.com/Mikubill/sd-webui-...


------------ Useful links ---------

Similar workflow with less drawing:    • BREAK Posing Limitations with Stable Diffu...  

Older but useful info: https://github.com/lllyasviel/ControlNet


------------ Social Media ---------

-Instagram:   / not4talent_ai  
-Twitter:   / not4talent  


------------- Explanation for AI beginners --------------


I'm using Stable Diffusion, a free, opensource, uncensored gen-AI for images. More specifically the Automatic1111 UI, and Forge UI whenever I need to use FLUX (You can use just Forge UI, since it's basically an upgraded version of A1111).
Stable Diffusion is the base AI, and it has 3+ variations that people use.
-Sd 1.5 (the oldest version, low prompt comprehension and generates at low resolutions, but has the largest amount of extensions and trained LORAs) .
-SDXL (the second oldest version. The base is decent but people have re-trained the checkpoints in a way that vastly improve the quality of it. Mainly for anime generations. Though these new models have some specifics to them. Like the need of their own controlnet models, or LORAS. Since they are not compatible with the base SDXL models). The models I use in the video are mainly these. One is an Illustrious based checkpoint. And another (the one needed to use the LORA) is a PonyXL based checkpoint.
-Flux (The newest version for Stable Diffusion. Great prompt comprehension and resolution. Probably the best so far when it comes to just text to image. Though there is a current lack of LORAs and controlnet models for it. Which makes me not want to use it all that much except for specific cases).

I explain Controlnet and Image2Image more in these 2 videos:

Image2Image Explained:    • STOP making BORING AI art  

Controlnet Explained:    • Get FULL creative control over Stable Diff...  

If you want to use StableDiffusion locally like I do, you'll need a decent PC.
How to install forge and some explanations by ‪@pixaroma‬    • How to Install Forge UI & FLUX Models: The...  



Make sure to subscribe if you want to learn about AI and Art. Grow with the community as we surf the AI wave :3

0:00 Intro
0:54 Sketching the base Idea
1:22 Controlling AI explained
2:35 What will we need?
3:38 Important parameters explained
6:52 Generating
7:46 Combining 1
8:08 Character's Face
9:20 Drawing daggers + cleanup
9:40 AI-refining and combining 2
10:32 A "problem" with this workflow
13:27 Combining and re-drawnig 3
13:40 Adding detail with AI
14:28 CleanUP and postprocessing
15:00 Final Result + Review
15:15 PROS
16:15 CONS
18:40 Best way to use AI (for artists)
20:22 WDYT? Thanks for watching




#aiairt #digitalart #automatic1111#stablediffusion #ai #free #tutorial #prompting #sd #digitalart #control #controlnet #digitalart #artist