AI itself has non-malevolent uses, but this doesn't. It's designed for negative uses.

ROT13 decryption does not have the same issues, and is not a good parallel here. Again, we're talking about a specific tool with specific use cases; a general algorithm being able to be used for negative ends is not the same as a tool being designed to be used for negative ends.

I also haven't mentioned banning anything; that's all on you. It really feels like you've jumped in with pre-prepared talking points rather than engaging with my comments.

To restate: "AI will benefit humanity" is a tough sell when so many people who are in to AI deliberately make tools that support the negative uses. This is an example of a project that superficially presents as high-minded, but is designed for exploitation.

what I've been trying to convey to you here with ROT13 analogy is that we're better off knowing that something is broken. ROT13 is a piece of shit, and so is SynthID. whether the author's intentions were noble or malicious is utterly irrelevant.

> we're better off knowing that something is broken

This is not a new insight, and it doesn't add anything to the discussion. It's effectively a platitude.

This project is absolutely not motivated by a wish to inform, and it's disingenuous to prevaricate around that; there are countless ways to inform that don't endorse.

there is no discussion. your opinion on the matter is irrelevant, same as mine. genai will be misused in every way imaginable, and there's nothing anyone can do about this. no regulations, no laws, nothing will stop it now.

for fuck's sake, Photoshop existed for how many years now? you would be a fool to trust any random photo to be genuine at any point in time. yes, yes, yes, back then only the people who had spent a few days learning it could do it, and now everyone can. that makes no difference whatsoever.

None of that is relevant to the conversation; it is genuinely like you are pasting from a random discussion a decade ago.

We're not talking about Photoshop, because it's not a parallel here. We're not talking about trust in authenticity. These are points you have plucked from the ether. Please reread my earlier comments.

it is very much relevant to the conversation because your assertion about this tool being harmful and unethical is rooted in your misguided belief that generative fakes that can fool a grandma on facebook are somehow worse than man-made fakes that can fool anyone.