Content
@
0 reply
0 recast
0 reaction
kripcat.eth 🎩
@kripcat.eth
Calling @polymutex.eth This cast reminded me of your story from a couple days back. I’m no fan of Musk and this screenshot seems believable based on verifiable past behaviour from Musk, so there is a threat of confirmation bias here. It could be a fake account or a fake image entirely. Short of me trawling through Twitter myself to confirm something which is pretty immaterial to my life. How could the original caster use EAS or ZK Proofs to provide verification that this image is genuine when casting? If it’s not possible today, what are the roadblocks?
3 replies
0 recast
29 reactions
polymutex
@polymutex.eth
All the pieces to make this work exist and work in theory, but no one has hooked them up all together to make this a reality in practice. You need: 1: A way to retrieve tweet data from X 2: A way to post them onchain in a way that everyone can tell they came from X 3: A way to prove that they match a screenshot ⤵️
2 replies
0 recast
2 reactions
polymutex
@polymutex.eth
Steps 1 and 2 are handled by @tlsnotary or zkTLS or @reclaimprotocol. Any of them allows you to make an HTTPS connection to a website (in this case X) and post the contents of what you retrieved through this connection onchain, in such a way that all can see that the content was verifiably retrieved from X. ⤵️
1 reply
0 recast
1 reaction
polymutex
@polymutex.eth
One challenge in this screenshot's case is that you don't need just one tweet, you need 2, and need to prove their relationship to each other (one is a reply of the other). This is difficult because X has forbidden logged-out users from retrieving more than 1 tweet at a time, and shut down the API for doing so. ⤵️
1 reply
0 recast
1 reaction
polymutex
@polymutex.eth
That said, you can still prove the existence of the tweets individually, and probably to prove that the second one is a reply (because the "reply to" data may still be on single-tweet pages). So let's assume we get past this problem. Having an onchain proof is good, but how do we surface this to actual humans? ⤵️
1 reply
0 recast
0 reaction
polymutex
@polymutex.eth
This is where it gets a bit hairy. The image is a set of pixels and it's pretty blurry. It's not obvious to a computer that these pixels are text, and that the text is two tweets, and that these two tweets correspond to onchain proof data that these tweets exist. Multiple ways we can go about that problem. ⤵️
2 replies
0 recast
0 reaction
