by TimJRobinson on 8/12/2024, 8:49:36 AM
There's a few new proof of human systems out there. For each of these you have a private key and the service attests that "the holder of this key is a human".
Then when you sign up to a service you can create a ZK proof that this service has said you're a human and that you haven't created another account before. This allows for anonymous registration while also proving you're not a bot.
The two biggest services for this are:
- Worldcoin - scan your eye in an orb and it attests you're human. This is integrated with Reddit and Discord already.
- Gitcoin Passport - connect it to your social media accounts, proof of humanity services etc and it gives you a humanity point score.
by rolph on 8/11/2024, 3:42:27 PM
bots are more persistent than humans.
make a captcha sufficiently complex, and the bots will keep trying, while humans will stop trying and go away.
redirect the flow of humans to a "probablynotabotindex.html" page when they abdicate the captcha, and that good for now.
so if you want to seem human, learn how to be impatient, and quit a challange.
by mikewarot on 8/12/2024, 10:14:20 AM
The future is everyone having their own server and their own stuff, their own little BBS/Forum/etc... basically, the 1980s all over again.
In order to get there, we have to solve computer security, so everyone can run their own stuff without it getting owned in a heartbeat.
Without that step, we're all doomed to things like walled gardens and "servers" on Discord.
It's fortunate that computer security was solved in the 1970s. It's unfortunate that nobody listened for decades as Moore's law made hardware insanely better.
People running their own stuff, moderating it, and establishing reputation is the only way to fly.
I stumbled upon this site recently: https://subtlereply.com/ (they scan reddit for places to have an llm write a fake review of a product in the comments) and it's just so clear that everywhere is going to be full of bots using llms trying to sell/scam everywhere on the web. How can we maintain some level of knowing we are talking to fellow humans?
All I can think of right now is a much more involved captcha.