I see fires of AI discussion is burning on HIVE again. As many of you probably know, I am a big proponent of AI both on chain and outside of it. But this doesn't mean I support every use of it.
With that out of way, let's talk what AI is in it's current iteration. Simply speaking AI is a tool and like any tool it can be used for nefarious means. We see this nefarious means on HIVE as posts written by AI posted by authors to deceive curators to get a cut of the reward pool. This should, of course, receive the same treatment as when we see plagiarized content, because in both cases intention is to deceive curators and defraud the chain.
But with every tool there are nuances, let's talk about AI that generates images like the one at the top of this post. Should we mention every time there are an AI generated image in a post that it was AI generated? I don't think so. If the main content isn't the generated image itself than not mentioning an image was generated by an AI is fine. At least it is fine in my book.
But what if the image is the main content of the post? Well, first of all, they should clarify that the image is generated by AI then we can move onto if we should reward it or not. Some AI generated images are result of a simple one sentence generation. We can judge these images as slops, like the one at the top of this post. But then there are AI image generations that have highly complex workflows that require quite a bit of work for to be as close to what the artist's imagination and intentions are. I think these kind of image generations are fine to be rewarded and you can adjust the reward if they include their workflow, talk about finer details of their workflow etc. There is a potential there for AI artists to flourish on HIVE rather than be ostracized.
What I see around all this AI discussion is that HIVE's current way of rewarding users (or distributing/decentralizing stake) is wholly unprepared. We are fine for now, some of us can tell when an image is AI or if a post is written by AI but AI is progressing when we are sitting on our laurels. Current reward dynamic of AI is wholly unprepared for when we can't tell if a post was written by AI or not.
Proof-of-Brain is dead, maybe it is time that we should look into other proofs.
