Live data from Hacker News

Worldcoin: A solution in search of its problem

newsletter.mollywhite.net

101–110 of 176 posts

Re: Worldcoin: A solution in search of its problem

#101
post #71
post #9

> While Worldcoin may have developed an algorithm that reliably distinguishes unique irises among its test pool, it’s not clear that it will work with a pool orders of magnitude larger. This is a really key problem. The main application of biometrics today is in cell phones, and for those we operate on the assumption that we're trying to distinguish between the authorized user and a quite small pool of non-authorized…

This is essentially the Birthday paradox. You need a quadratically better false match rate to deduplicate than to authenticate. It’s also why Worldcoin went with Irises (highest entropy among biometrics), custom hardware, custom optics and an in-house trained algorithm.

The Iriscode (if it's the same as Daugman's iriscode) has tested so far to have 249 bits of entropy, or over 10^74 combinations [1]

So even with the birthday paradox you'd need 10^37 people before having a good chance of a collision, which is rather more than we are likely to have in the next few centuries.

Of course, it's possible that there are some subpopulations who don't have this amount of entropy in their irises, most obviously the small number of people who have a birth defect such that they are born without eyes.

[1]https://www.cl.cam.ac.uk/~jgd1000/binomdata.html

Re: Worldcoin: A solution in search of its problem

#102
post #71
post #9

> While Worldcoin may have developed an algorithm that reliably distinguishes unique irises among its test pool, it’s not clear that it will work with a pool orders of magnitude larger. This is a really key problem. The main application of biometrics today is in cell phones, and for those we operate on the assumption that we're trying to distinguish between the authorized user and a quite small pool of non-authorized…

This is essentially the Birthday paradox. You need a quadratically better false match rate to deduplicate than to authenticate. It’s also why Worldcoin went with Irises (highest entropy among biometrics), custom hardware, custom optics and an in-house trained algorithm.

This is a really good comment. But of course the problems in authentication and de-duplication are also different in that you care about adversarial false-positives much more once authentication is the goal. As I understand things, Worldcoin claims the retina scans won’t be used to control funds (or other services.) I am skeptical that many of those users will retain their non-biometric wallet credentials long-term, which will leave you with a database of biometric credentials that will have to be used for authentication if you want to use those credentials for anything important in the future.

Re: Worldcoin: A solution in search of its problem

#103
post #40

Earlier quoted context omitted.

Not ‘All’. Just like 90% of startups fail, it is the same with 90% of crypto startups, and AI startups will fail. The 10% remaining will continue to exist. Just stop this absolutist nonsense.

The absolutism is a bit extreme, but your comparison isn't really fair either. Outside of crypto, we don't regularly see startups get loads of hype and money only to explode so spectacularly that the founders end up in jail.

> Outside of crypto, we don't regularly see startups get loads of hype and money only to explode so spectacularly that the founders end up in jail.

It is already known that the majority of unprofitable startups take tons of VC money and have regularly played the Silicon Valley playbook of 'faking it' until they are caught out in the open [0] [1] [2]. It has gotten so common to the extent where their favourite bank (SVB) went under with all these unprofitable startups crying over payroll when generating little to no revenue with inflated valuations.

We were supposed to learn from that VC pyramid scheme that has caused SVB to collapse which was so seismic that all those unprofitable startups would all have gone bust had capitalism just run its course without government intervention.

There really is no defence for continuing the constant dependence on raising VC money in unprofitable startups for years anymore after over-leveraging and injecting more cash at unjustified valuations in these startups. For this scam to be revealed so late shows how long many startups were able to get away from 'faking it'.

[0] https://www.theguardian.com/media/2023/feb/23/ozy-media-foun...

[1] https://abcnews.go.com/US/startup-founders-alleged-175-milli...

[2] https://www.cnbc.com/2022/10/14/nikola-nkla-founder-trevor-m...

Re: Worldcoin: A solution in search of its problem

#104
post #96
post #84

Earlier quoted context omitted.

I’m not saying it’s a scam, just that it seems their system isn’t really Sybil resistant due to the issue of account selling. And if that’s the case they either didn’t think this would be an issue, or they expected it but don’t really care, and the claim of Sybil resistance is just marketing. Is this wrong in some way you’d like to explain?

I can't explain anything about this project because I haven't invested the effort to understand it. Maybe I'll spend the 20 minutes to read the 5000 words in Molly White's article, and then another couple hours to read the whitepaper, then another who knows how many hours researching the claims and counterclaims to make my own judgment. But probably I'll never do any of those things, and I'll still have high confiden…

Once again, I never claimed this was a scam. I'm simply pointing out that the system has an inherent flaw regarding account selling, which makes it not sybil resistant, as claimed by the founders. I usually give projects the benefit of the doubt, but I've never heard Sam or the other founder talk about this major issue with their system design.

Re: Worldcoin: A solution in search of its problem

#105
post #98

Earlier quoted context omitted.

Can you name a concrete counterexample of a crypto venture that provides actual value to society?

Yes. Moneygram, Circle and Stellar launched and aid program (Stellar Aid) that allows near instant, very low-fee and cross-border P2P payments worldwide to those in need of it. [0] [0] https://thefintechtimes.com/stellar-aid-assist-creates-new-r...

Wait, but Moneygram already does that by itself. What (except of more complexity) does all the crypto stuff add?

Re: Worldcoin: A solution in search of its problem

#106
post #78
post #67

Earlier quoted context omitted.

I can’t imagine how naive you have to be to not see account selling as a massive problem. Especially when your plan is to first onboard users in poor countries. Why would these people do anything other than signup and immediately sell their accounts, for what may amount to a month or two of normal pay, just for waiting in line and getting scanned. I find it crazy that they claimed their system is Sybil resistant when…

> sweet sweet VC money Sure, yeah, maybe that's all it is, a scam that's totally transparent and obvious to every random internet commenter but totally non-obvious to the simple and gullible marks on Sand Hill Road. Or maybe, just possibly, there might be more to it. Perhaps it takes a little more effort to understand than just piling on to the reflexive groupthink cynicism which passes for conventional wisdom around…

> but totally non-obvious to the simple and gullible marks on Sand Hill Road

You mean the ones who also invested in Theranos, WeWork, and FTX? They don’t always make the wisest investment decisions. They’re not complete idiots, but they’re sucsceptible to same biases and misjudgements as the rest of us. And I’m sure they‘re aware of some or all of WC’s flaws, and are just investing b/c Sam Altman, or because they invested back in the 2018-2021 crypto bubble. There’s also some time pressure, they’ve gotta put that money somewhere within a year (standard VC LP contract), or give it back.

> Perhaps it takes a little more effort to understand than just piling on to the reflexive groupthink cynicism which passes for conventional wisdom around here.

Molly already put extensive effort into just that in the OP, as did several other folks she references and links to. If you have any critiques of her critiques and why she may be wrong, love to hear it. But without that, it’s not her or us that’s being reflexive here.

Re: Worldcoin: A solution in search of its problem

#107

Earlier quoted context omitted.

I don't know how it works outside of Italy but here we can make instantaneous bank transfers for free or a really small (and capped fee) to anyone just knowing their bank account without leaking any secret information. How does it work in the US?

You go to your bank's website and input the other person's bank routing information to make a transfer. But with the same information someone can also make a withdrawal , which is problematic.

Whoa... that's totally broken...

Re: Worldcoin: A solution in search of its problem

#108

Earlier quoted context omitted.

And? I was talking to the op about their story... don't need a link to a worldcoin announcement, thanks.

This is an article about Worlcoin, Sam Altman's pet project, and OP is sharing an anecdote about Sam Altman. Not sure why you find this so confusing, but I'll be happy to walk you through it slowly.

It's not clear whether or not he talked to Sam Altman in his/her "pre-interview".

If I had a pre-interview for a job at Amazon, does that mean I have a Jeff Bezos story?

Re: Worldcoin: A solution in search of its problem

#109

Earlier quoted context omitted.

I am all for privacy and encrypted communications. Until a legitimate law enforcent body gets a legitimate warrant against a specified organisation or individual. Outsourcing some, or all of that to some VC backed companies, facilitating access for the various intelligence agencies, is what I have a problem with. And a crypto start-up setting out to build a global database of peoples retinas is exactly that.

> I am all for privacy and encrypted communications. Until a legitimate law enforcement.... So you're against encryption then. Simple as that. > Outsourcing some, or all of that to some VC backed companies, facilitating access for the various intelligence agencies, is what I have a problem with. I also have a problem with that. We should aim to build systems that are resistant to all attackers, even intelligence agen…

What mechanism prevents the government from “un-person-ing” someone if Worldcoin was used as the global identity system?
Post reply on HN