51–60 of 88 posts
Earlier quoted context omitted.
Wizards of the coast owns the patent on collectable cards, so they should be excited, since there's no way to break into the market without them suing you. They can fire all of their artists and save on costs, but still produce good working cards with AI.
That patent expired several years ago.
https://en.wikipedia.org/wiki/Collectible_card_game#Wizards_...
As a former MtG player I think the difficulty lies not in creating cards, but in ensuring there's a level of balance between existing cards and new sets.
That is where AI-based play automation to test the cards should come in place.
See for instance amateur Go player armed with loophole consistently beat top Go playing software which otherwise easily crushes world champions.[1]
[1] https://www.ft.com/content/175e5314-a7f7-4741-a786-273219f43...
[0] https://hardwork.party/#/entropy/ … recursively-generated MTG-style cards generated by GPT-3 (not 3.5t) and VQGAN+CLIP. The recursion brings out the crazy!
I ended up scrubbing for every non-dictionary word in the MTG corpus - find a ‘nono’ word? Regenerate.
Also did cosine vector similarity on the body, title, type, and flavor text fields against existing MTG corpus to throw out anything too similar.
Also my partner on the project ZNO did all-original card BG, back, and icons.
In the end the thing was an art piece so it didn’t suffer from deviating widely from MTG, although I’d say it probably would have been a bigger commercial success as a straight imitation.
Why not allow the players to 'create' new cards on the fly and the LLM balance it automatically? So that one would feel like an 'real' wizard. Curiousy I was thinking on how to actually implement something like that tonight.
Bad Lightning Bolt is so hilariously perfect that I’m surprised it isn’t already an un- set card.
Earlier quoted context omitted.
ChatGPT's June update added support for "function calling", which in practice is structured data I/O marketed very poorly: https://openai.com/blog/function-calling-and-other-api-updat... Here's an example of using structured data for better output control, lightly leveraging my Python package to reduce LoC: https://github.com/minimaxir/simpleaichat/blob/main/examples...
How has your experience with function calls been? We tried doing code generation and making ChatGPT generate diffs and it seems to perform worse than the March edition.
This post was originally made in 2022. In early 2023, I spent a lot of effort training a model with a new approach that made the generation quality extremely good... Then the ChatGPT API came out and made all my effort obsolete. In one hour, I was able to create a Notebook ( https://colab.research.google.com/github/minimaxir/chatgpt_a... ) that was able to create mechanically valid and relatively balanced cards given…
Earlier quoted context omitted.
> only to realize All of it? Some people would immediately recognize the typical StableDiffusion output but not a Midjourney.
Midjourney is a limited set of models and (to me at least) much easier to spot than Stable Diffusion output, with it's giant array of custom models, LoRA's, textual embeddings and tricks like HiRes fix and upscale script.
If anyone put SD down for a good period of time, they should really give it another go. Use the full suite of options and make sure you are using a model fitted for your purpose.
As a former MtG player I think the difficulty lies not in creating cards, but in ensuring there's a level of balance between existing cards and new sets.
That is where AI-based play automation to test the cards should come in place.
One of the shocks of the games industry is in many respects it doesn’t learn from the wider world but in others it is quietly so far ahead that you can throw certain ML papers at former game devs to be met by confused faces as to why what is claimed is considered surprising.