We could use public libraries to do some kind of “are you a human” confirmation. I.e. you get a digitally signed token proving that you are a human, that you can use on the internet to show that you are human. Only the library has to be trusted by the public …
yes, but at least you can’t easily create 10000 bots at once. you could only create 1 bot at a time, and that would already do a lot.
I’m not so much worried about 1 person using AI to write their comments; I’m way more worried about large political parties making 10000 bots to support their agenda (propaganda) or companies creating a lot of bots to create hype about their products (advertisement).
walk up to your library in person and let them give you a chip card that contains an (anonymous) private key, or let them verify your fediverse account somehow, both methods would be without you showing them any kind of ID card.
They would at least need to have your ID on record, so they know you aren’t just getting a new private key at a different library every day. That said, I’d be happy with it if that record only showed that my ID had been used to acquire such a key, without any link between the ID and the key
So is this going to only work with fediverse accounts or are you gonna try and teach people about public/private keys?
And what happens when that database leaks and someone starts using your key to post illegal things? Or should we use password rules and have your library generate you a new key for every platform?
I’m not sure about the details yet. I think, libraries give you 1 chip card that contains a public/private key pair that you can use to “proof” yourself on the internet. Ideally, it would have widespread support and many platforms (both fediverse and commercial platforms) would support it.
I think it may or may not be tied to your person, i.e. you might get that chip card without any kind of ID. Then it wouldn’t matter so much if it leaks. But idk, just floating ideas here.
We could use public libraries to do some kind of “are you a human” confirmation. I.e. you get a digitally signed token proving that you are a human, that you can use on the internet to show that you are human. Only the library has to be trusted by the public …
And then you take that token and give it to your AI…?
Well, no, that wouldn’t work you see, because that would be lying.
yes, but at least you can’t easily create 10000 bots at once. you could only create 1 bot at a time, and that would already do a lot.
I’m not so much worried about 1 person using AI to write their comments; I’m way more worried about large political parties making 10000 bots to support their agenda (propaganda) or companies creating a lot of bots to create hype about their products (advertisement).
How would you verify that you are a human? The only ways I can really think of are enormously invasive
walk up to your library in person and let them give you a chip card that contains an (anonymous) private key, or let them verify your fediverse account somehow, both methods would be without you showing them any kind of ID card.
They would at least need to have your ID on record, so they know you aren’t just getting a new private key at a different library every day. That said, I’d be happy with it if that record only showed that my ID had been used to acquire such a key, without any link between the ID and the key
So is this going to only work with fediverse accounts or are you gonna try and teach people about public/private keys?
And what happens when that database leaks and someone starts using your key to post illegal things? Or should we use password rules and have your library generate you a new key for every platform?
I’m not sure about the details yet. I think, libraries give you 1 chip card that contains a public/private key pair that you can use to “proof” yourself on the internet. Ideally, it would have widespread support and many platforms (both fediverse and commercial platforms) would support it.
I think it may or may not be tied to your person, i.e. you might get that chip card without any kind of ID. Then it wouldn’t matter so much if it leaks. But idk, just floating ideas here.
Could just do MFA checks for all posts, comments, and upvotes. Imagine?
It would cut out the slop about 99%
Dissent detected, officers en route
Removed by mod
MFA helps boost security of an account. It would do nothing to prove a user is human or not.