“The project claims that the World ID will prove they are not robots”

Well no, obviously.

Individual level —

It doesn’t at all show that you didn’t simply hand over your authorization to a robot, to do whatever on your behalf. So at the individual level, no, it’s beyond useless at definitively proving this negative.

I.e. this CANNOT prove someone INNOCENT of using a robot.

Once they force everyone on OpenAI to authenticate through Worldcoin, though, then it could in help to prove the positive dual — i.e. proving it WAS you who used an AI to help you write that scathing article about some politician.

I.e. it CAN prove you GUILTY, but only if you’re not a sophisticated criminal.

= Can only hurt you, never help you, at an individual level.

Aggregate level —

Ok, so what about at the aggregate level, e.g. for preventing cheating for voting, product reviews, and the like?

Well, here it wouldn’t be totally useless in principle, as it creates a financial burden for a single entity to pretend to be multiple people, and rig a vote.

Only problem? That financial burden is already tiny, if not totally collapsed.

Black markets already showing that financial burden to have an upper-bound of maybe $4 per identity max, much cheaper if you rent, and the lower-bound on that is effectively zero, if any of several very easy hacks happen.

How for about proving OpenAI innocent or guilty? —

Nope. Can’t do that at all. The cost for them to rig this system against everyone else is, obviously, potentially $0. Hugely net-profitable even, in many cases.

So, what’s Worldcoin good for?

Helping the central powers,
while hurting you.
That’s it.