I expect a lot of comments on this article to be varieties of "this is terrible"... but it's really not, and I happen to have significant knowledge here. There is a big caveat, though, which I'll explain below.

First, the basic thing that makes strong, reliable age verification possible in the EU is national ID cards. In every EU country, as far as I know, you can get a national ID card basically from birth. A few issue at birth by default, but even those that don't allow parents to apply for cards for their kids at basically any age, and it's not uncommon.

I get the widespread American resistance to a national ID card, but I really think it's misplaced. There are risks, yes, but on balance the benefits are far larger.

Second, when the EU says you can verify your age without revealing your identity, they seriously mean it. I worked on the ISO 18013-5 mobile driving license standard, and its protocol is the basis for the age verification scheme (18013-5 also supports privacy-preserving age verification). The protocol enables cryptographically-secure privacy-preserving age verification, providing, essentially, a single cryptographically-verifiable bit answering the question "Is this person over age X", for specific legally-important ages. A great deal of effort goes into ensuring that the keys used to sign the bit cannot be linked to the identity of the person. One important element of that is the signing keys are single-use, so if your prove your age to two different web sites, they can't compare notes and notice that your proof of age used the same signing key, thereby proving that whoever you are, you visited both.

Note that under the 18013-5 design, if the verifier (e.g. the web site receiving proof of age) could collaborate with the issuer (the government), they could deanonoymize the holder (the person proving their age). Work is ongoing to devise protocols using group signatures or other cryptographic constructs that make verifier/issuer collusion fruitless. It's been a couple of years since I worked in this space, so I don't know if those new approaches have gone into production, but if they haven't, they will.

The big caveat I mentioned at the top is that there is no way for these systems to verify that the person who is providing age verification is the legitimate holder of the national ID upon which it's based. That is, a kid can steal their dad's ID and use it. Because the age verification is truly, strongly anonymous, there is no way for anyone to detect or prevent this... yet.

The "yet" is because people are working on incorporating privacy-preserving biometric authentication into the scheme. This is a little tricky because to provide privacy it's critical that the biometric acquisition and matching happen entirely in the user's device (or in the chip in the national ID card). But it can be done. Making it sufficiently secure, sufficiently reliable and sufficiently cheap is a significant engineering challenge, but it's being worked on. In another decade or so, the caveat may be removed.

If all of this seems silly to you... well, the age verification for porn may be, but the privacy-preserving selective proof technologies are general-purpose, and able to answer any age verification question any many other useful questions in a strongly privacy-preserving way. In any case where you need to prove something about yourself (age, city of residence, driving privileges, etc.) right now you need to provide the complete contents of your ID, which reveals far more about you than is necessary. The combination of cryptography, secure hardware and clever protocols used in this age verification can fix that, generally, enabling us to identify, authenticate or prove things about ourselves with only the minimal information absolutely necessary. It's a good thing.

And, honestly, it's a good idea to keep very young children away from porn.

I can think of four reasons ticket sales have dropped.

1. Much of what Hollywood has been putting out is crap.
2. Politicians increase taxes constantly, resulting in less disposable income for potential movie-goers.
3. Ticket costs increasing - Not much can be done there, theaters do it to compensate for the loss in sales.
4. Streaming is a thing. Why go to a theater to sit in a sticky, uncomfortable seat when one can wait a little while for the movie to hit $STREAMING_SERVICE and watch it from the comfort of one's own home?

Absolutely not. The charge/discharge round trip losses will be a few percent, maybe 10% if the batteries are in bad shape. The price difference between peak and off-peak is often 5-10X. Commercial users also get hit with demand surcharges based on the peak draw during the month and those can really make a huge difference. Using batteries to smooth out those peaks can be a bigger savings even than avoiding draw during peak times.

Even for residential use, the savings can be significant. I have batteries and I'm on a time-of-use plan that charges me 5X as much during peak hours (6-10PM) as the rest of the day. I make sure the system is set up so that I never draw any power during peak.

Incredibly unlikely. If the claimed violations are legitimate, and webXray reported them to the state plus the attempt to lean on them, Google would get slammed, hard, both legally and in the press. No way in hell Google would risk that.

What? You're not going to advocate for it??? I thought you were invoking some kind of principle or something.

You're down with Russia killing far more Ukrainians, whom they claim are their own people?

You're down with what China's doing to the Uyghurs, whom they claim are their own people?

And while we're on the topic, how many Iranians should we be willing to kill to save them from their leaders? Nuclear extermination would surely do it... do you advocate that?

But maybe it won't take that much. Regime change in Afghanistan only cost 2000 American lives, 175,000 Afghan lives, and 2,000,000,000,000 dollars, but we sure got rid of those sorry... What? They're back in power?

Only the simplest minds think intervention automagically yields the intended result. In fact the current sorry situation in Iran is a direct result of us trying to "fix" things more to our liking in the middle of last century.

No I think the proposers of this think the only effect is that people won't be able to print guns on a 3D printer. They are wrong (both about it stopping the printing of guns, and that this being the only effect), but I really don't think they are planning some different objective.

So... Should we attack every country that slaughters its own people?

No this is classic conspiratorial thinking. The government SUUUUUUUUCKS at controlling information. But they are dicks and are very good at acting like it.

No one can make session tracking with form variables or URL arguments as reliable as it is with cookies.

I have a lot of success with obscure, mostly-undocumented systems. Which models are you using? There's an enormous difference in capability level between the top-tier models and the next step down. Also a pretty big cost difference.

It can be, sure, but it's less reliable and more painful to work with.

If the law is about sale or sharing, not collection, then Google doesn't have to change anything, because Google doesn't sell or share data. That would be wasteful; Google's ad business is all about monetizing the data at Google, not giving someone else a chance to monetize it.

PS I am not a gun nut. If guns were illegal then printing parts of them using a 3D printer would also be illegal and you could get in a lot of trouble if you do it. But after the fact, don't try to prevent it first.

Is cooking equipment supposed to not work if it figures out it is being used to make drugs?

This is obviously impossible unless almost any moving parts are prevented, or printing with some materials is prevented. That would make it impossible to 3D print lots of stuff that is not illegal.

The effect on open software and hardware is also disastrous.

As others pointed out, currently you can make better gun parts on some non-computerized equipment, so this also does not seem to prevent anything.