2 ms·
> > verify [that you are human] on the internet IMO it's crucial that we keep in mind the policy/engineering differences between: 1. Prove you're a human and
by Terr_ 2y ago
> > verify [that you are human] on the internet
IMO it's crucial that we keep in mind the policy/engineering differences between:
1. Prove you're a human and share your global permanent ID.
2. Prove you're probably human.
3. Prove that if you're a bot you're at least the high-quality kind that isn't part of an army and whose operators are likely to follow the rules.
In many cases #1 is overkill, difficult to engineer, has reams of nasty privacy implications, and requires a massive amount of bureaucracy for edge-cases.
Most websites are happy with #2 or even #3.
[Edit: Some prior musings in comments here https://news.ycombinator.com/item?id=41714245 https://news.ycombinator.com/item?id=41714245 , where someone also mentions Worldcoin.]
- CatWChainsaw 2y agoThe finance industry has strict KYC laws to prevent fraud and funding of other criminal pursuits. Now we have several converging trends: - we need to know if anyone ever says anything doubleplus ungood so we can punish them with a lifetime exile from society. what qualifies depends on the day and if mercury is in retrograde. - data is useful for making money - human verification on an AI-enshittified net becomes practically "required" to get any "legitimate" content, by which i mean the dreck we had pre-2022 as opposed to post-2022. - everything as a subscription service - also the ghost of web 3.0 More parties are incentivized to hoover more data. So although #1 is overkill and everything else you listed, it's the path more companies will be taking in a neverending arms race.