Britain's most tattooed man says UK's age check told him to "remove his face"

But the law doesn't force them to have a fallback.

Anyone that the AI doesn't like is therefore excluded from those services.

I've personally been unable to pass AI 'liveness' detection (which was a high-stress situation when it related to something my new employer asked me to do after I already resigned from my previous role) despite repeated attempts and all I have is alopecia areata affecting my eyelashes / eyebrows (a relatively common condition).

These models are discriminatory for a lot of people, I'd say, and shouldn't be allowed.

And this is the issue. 100%.

Edge cases are fine. Not having simple, easy ways to resolve it should be illegal for any form of ID verification.

One example? Some of these apps require a Google Play account to install. Or an Apple account. This is unacceptable.

Why would anyone be gatewayed to their, for example, government services by having a valid Google account? Or their bank?! Google bans people. Cuts them off.

Doesn't matter why, or how often, all that matters is that it can happen.

ID verification should not rely upon firebase, Google's push notifications, a Google account, and so on. And yes, there are ways to avoid these things for ID verification.

I don't have a Google account. I do have an Android device.

But the law doesn't stop them from having a fallback either.

The problem lies squarely with the companies contracting these AI services not adding a "facial recognition doesn't work? click here to try something else" button.

The problem lies squarely in the fact that some fundamental thing is getting contracted out to some company.

> If that's all they offer, it's on the companies to implement a fallback for edge cases like these.

These news articles and the adjacent online discussion are textbook warfare psyops 'nudging'.

Doesn't matter if you are real/bot, being payed or not. The discourse is now changing the goalposts to focus on the details of OSA implementation, not OSA itself. Mission acomplished.

It's on governments to stop pushing legislation that slow boil us into autocracy. It's on us to not be ok with that.

Everything else is noise.

I think these models are fine for people who they do work on, but it's idiotic to assume facial recognition works for everyone. I should be able to use a website if my webcam is broken.

The practical problems are all caused by AI companies lying through their teeth and making bold claims and their customers being dumb enough to believe them.

The actual problem that needs solving is the fact that you need to validate your age without a form of solid proof being available in the first place. In cases where everyone has digital ID already there are technical solutions to solve that problem, and until those are available for free, it's idiotic to require the use of such technology in the first place. The UK doesn't have common, accessible digital ID yet they expect digital identification of some kind to just happen.

> "doesn't work on a man that looks more like an injured smurf than a human"

What the actual fuck is up with this statement? Human faces are complex, there are so many different cultures and all kinds of complex body modifications (including facial tattoos) have a rich history going back thousands of years. Some shitty cobbled together AI not being able to recognize that is the real problem. Not the facial tattoos. And the injured smurf comment was completely uncalled for.

Sorry, but the pure-black eyes, bright blue skin with red patterns gave me a visceral response. In his own words, people are afraid to sit next to him on the bus, so I'm hardly the only person who had such a reaction to his appearance. This is more than just "facial tattoos" or body mods, this is quite literally an extreme case.

I believe anyone should be able to alter their appearance in any way they want, but if you choose to ink your skin to be more blue than skin-tone (linked articles have pictures without the Instagram filter that make the blue-ness much more clear, like https://i2-prod.mirror.co.uk/article34897342.ece/ALTERNATES/...), you should not be surprised when people and people-made programs think you're wearing a costume. This isn't just an AI program, according to the articles about him, he's had the same issues with real people at airports.

The core of the problem still remains the same; AI is not capable of reliably recognising people (kids used video game characters to bypass age verification), let alone determine their age reliably, and I believe most of the OSA should be abolished.

at least in germany, once it becomes evident that an account is needed in order to use services that are vital for the participation in society, banning effectively becomes illegal.

though hopefully they will also enforce that such services must not rely on foreign companies that will share personal data outside of the EU.