Jun. 19th, 2023

walkitout: (Default)
I’ve been wrestling with an ongoing collaboration problem, which is one of the collaborators introducing an irrelevant or distracting assertion that there is a better choice than the one or ones currently being contemplated.

If this occurs during bluesky / brainstorming / long range planning, this is actually fine. It is not irrelevant or distracting; it is a plea for more options. The problem occurs when I’m attempting to implement a choice, and someone wants to return to that earlier phase, _usually_ because they have some kind of issue with the price.

Recent examples include: the price of renting a bike at the bike rental place I had made certain was in walking distance of the hotel we would be staying at, the cost of an airport parking reservation vs. car service. I wrote a really long post to generate a script to validate input while also staying on task, because there are very real problems that I run into when I’m doing something that I _thought_ was going to take a few minutes and be easier than telling my son, once again, Not Right Now, but that turned into a half hour research project that is poorly defined and coincides with trying to prepare and eat dinner.

But I don’t really like the script. It’s unwieldy. If I’m going to be dealing with the burden of figuring out how to do something AND paying for it, input from other parties on cost is just irritating. The rental bike issue was resolved without me having to do any rental bike price comparison; I limited my commentary to, “I made sure that was in walking distance, and a kayak rental place too, so that we only need one rental car. Make sure you don’t save money on the rental bike in a way that requires us to rent another car.”

In general, tho, I think the correct response when anyone in this family says, but there’s a cheaper choice, is, “There isn’t a cheaper choice that is as good or better” and carry on without giving it any further thought. I don’t need to research or assess their assertion. It’s just an assertion. Someone in my family made an assertion that I do not give a single fuck about (cost). Acknowledge, deny and move on.

They _might_ eventually come up with a cheaper, better choice, but I’ll deal with that when it happens. I’m uninterested in cheaper and worse and I’m the one paying for everything.

For completeness: cheaper and otherwise equivalent is totally uninteresting to me. If it takes more time to find it, I don’t want it. I saw Good Enough value in what I was about to implement. I am not an optimizer on anything; I’m a satisficer. And I’m _really_ not an optimizer on price.
walkitout: (Default)
https://www.bloomberg.com/news/articles/2023-06-15/the-eye-scanning-orb-that-lets-humans-prove-they-re-human

You can listen to it. You can watch the video. That link also includes a lightly edited transcript.

I generally feel like people who are doing crypto or crypto adjacent stuff show up on Odd Lots and 1-3 weeks later they are in a whole lot of trouble with somebody. This is a crypto adjacent thing, so I’m kind of expecting that to happen here, altho I have no idea precisely how that will manifest.

But I’m also interested in using iris texture recognition for identification purposes. I think it is fair to say that absolutely anything you use for identification purposes can fail in really unfortunate ways, either inadvertently or because humans are amazingly scammy.

I did a little digging specifically on cataracts, because cataracts are ludicrously common and as our global population ages, this is only going to happen more. And cataract _surgery_ would appear to be another way in which eye-related recognition could fail. The short answer here is: about 10% of cataract surgeries will result in an iris texture that is meaningfully different from pre-operative texture and will require re-enrollment in the identification system. That’s both an “inadvertent” fail and also an amazing opportunity for scamming.

Each eye has a different iris texture. A lot of people have two eyes. But not all people have two eyes. In fact, some people don’t have eyes at all, due to being born that way, or trauma. Even people with eyes, but who suffer eye trauma and require iris reconstructive surgery may need to be re-enrolled. Less common than cataracts, probably, but definitely something that will need to be taken into account.

It is absolutely banal to say that a problem with crypto is people forgetting/losing the thing they need in order to access their account. To the extent that you build a system that requires iris recognition to access, and a person dies, has eye trauma (including cataract surgery) and there is no alternative way to access it, then the “access” component of this system is gonna be pretty bad! And to the degree that there _is_ an alternative way to access it, then that’s where security problems will proliferate.

Recently, we’ve seen law enforcement use the immutability of the bitcoin ledger to connect people to crimes and prosecute them _a decade or more later_. Using iris recognition to connect people to accounts and systems in a trustless way would seem to introduce another opportunity for this kind of delayed FAAFO.

Finally, to the degree that accounts attempt to build a financial system that is outside of jurisdictions, users will find themselves deeply unhappy that the resources flowing through that system are not subject to the laws of jurisdictions. I mean, that’s obviously the goal, but people are stupid. They’ll be all happy at the prospect of evading taxes and then get mad when they get scammed. They might be happy at being able to move money across borders freely and then get mad when they have no way to claw back money that their soon to be ex-spouse moved beyond the reach of a divorce settlement or child support payment judgment.

In any event, Odd Lots did not get into any of that. The person they interviewed said some things like this:

“So you build systems that are not dependent on a small group of people, can be completely verified and can run over decades without being interrupted.”

Statements like that are _deeply_ hilarious (and infuriating!), but honestly, if you are talking about software and using a phrase like “completely verified” then you have just indicated that you know absolutely nothing and I’m still trying to figure out why I’m wasting my time listening to you.

Anyway. Worldcoin and the Orb are stupid, and I expect something to crater in 1-3 weeks, but iris texture recognition as a biometric is probable, and we should be thinking about how to handle cataracts and re-enrollment in general. With surgery, you can identify before surgery and re-enroll after surgery and _presumably_ build a nice, solid protocol for making sure that the surgery isn’t an opportunity for a person to become a different person or whatever, from a digital identification perspective. But with trauma, that can happen to anyone at any time. There’s gonna be a huge potential for loss of access unless there’s a mechanism for access that is Other Than Iris, and as soon as you have an alternate mechanism, all the security attacks will glom onto that, too.

September 2025

S M T W T F S
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 2627
282930    

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Sep. 27th, 2025 11:49 pm
Powered by Dreamwidth Studios