Ai
Hello,
I’ve developed an AI-driven tool that reads the pixels on your screen. This tool is based on a 1:1 build. You need a custom build according to your current deck. It works amazingly with your deck, but as someone else mentioned, it's not perfect because there are too many cards to create consistently good outcomes. You can create combos, but it doesn't record the other player's hand. I haven’t inputted every single card in the game / can’t because I don’t own them
All.
Initially, it worked pretty well, but in ranked play, people use all sorts of advanced combos. If I were to release and sell this, I'd need a second developer to help refine the outcomes. I underestimated the time it would take. I can definitely produce a well-functioning AI that considers all the cards, but I'd need another developer or even two more. Of course, I'd have to pay them.
That's why I’m seeking feedback from others. I don’t want to invest all this effort only for the AI to turn out poorly.
Only way this would be worth it is if every card is accounted for.
But. In all honesty. In all honestly.
Not worth it. 1:1 builds would be expensive for the customer and time consuming for me. You wouldn’t be able to change decks. Not only that. I would charge a fair price for my time. So probably gonna ditch this because there isn’t enough people commenting and I don’t have people or help with identifying cards within the game.
I was gonna use the API call function but it’s unstable and again. Who wants a ****** AI that isn’t trained enough.
|