Comment by nullbio
3 days ago
Feels more like a joint effort given the way the world is going with AI. Likely, everyone who has access to the internet will require a passport to authenticate and access all websites, including AI bots. That way, any illegitimate activity from AI bots can be traced back to individuals.
If there was to be a small silver lining out of all this, then a way to verify human comments would be nice.
But it's not worth the privacy impacts and chilling of free speech.
Yeah, I have no idea what a good solution is that solves what is coming when malicious actors start flooding the web with agents at hyperscale, because that is certainly going to come.
I'm really hoping smart people can figure out a good middleground before the government does something stupid. Because right now, their two cannons are either:
- Ban open-weight/open-source models and self-hosted LLMs
- Require internet passports to access any website and have everything recorded and monitored for signs of abuse that can be tied back to an individual and enforced in real life
What does a good middle-ground look like? I can't think of one. Neither of these outcomes is good and neither of them are enforceable. You ban something for law abiding citizens and only criminals will use it, and if you restrict everything behind internet passports then the criminals will just use the millions of stolen identities they already have access to.
Maybe biometrics that generate short-lived tokens that only the government can tie back to an individual? Like a "biometric session". Still though, you can kiss freedoms goodbye. It's no surprise Sam Altman started Orb so long ago, he obviously saw this one coming.
> You ban something for law abiding citizens and only criminals will use it
Combine it with harsh enough punishments and few criminals will be willing to risk it. That's how we deal with other crimes that are "hard to enforce" already.
And remember, the optimum amount of crime is not zero. Both because the cost of getting it to zero is too high to bear and because sometimes the morally correct thing to do is to break the rules-as-written.
It's also not the responsibilities of internet users to provide a way for websites to verify that comments are human. If they want to make that a part of their commenting process that's a choice they can make without requiring legislation.
Maybe zk proof of government documents.
I'm not sure that verifying a human identity for comments is even that useful as there are already plenty of people comfortably with putting their own name on slop.