AI doesn’t create in a vacuum - so where does that leave artists?
I’ve been thinking a lot about this since my last post here. I’m one of the co-founders working on this topic, and the more we dig into it, the more one thing feels pretty undeniable: AI can only generate what it has learned from human input before. That’s the whole point of these models. Which also means that a lot of existing art has been used to build systems that now generate value somewhere else, often without consent, credit, or compensation.
Call it what you want, but to many artists that just feels unfair.
That’s exactly the problem we’re trying to work on. We’ve built a service that attempts to detect whether certain works or characteristics might have ended up in training datasets, basically trying to turn that vague feeling into something more concrete and measurable. It’s still early and we’re definitely not catching everything yet, since this kind of analysis takes serious compute, but we’re continuously indexing more data and improving the system. (Thanks to funding from Google Start-up, we can activate a bit more computing power - shit is expensive) Let me state this upfront: We are worlds away from needing the computing power of AI models. It's simply not comparable...
The goal isn’t just a tool, it’s to move towards a situation where creators actually have some level of control again over what gets used and what doesn’t. If you’re curious, whether in general or with your own work or style, you can try it on NiftyIP.com, we would genuinely be interested to see what comes up and learn from it/improve.
At this point I’m mostly still trying to understand how people here feel about it: do you actually want that kind of control back, or does it feel like that ship has already sailed?
Big Thanks.