Sure, it is not a big problem for SSH. It is a problem when you connect to a web site, especially as certificate lifetimes get shorter: you need the whole certificate chain from a root (that your browser trusts) to the web server, which means at least two public keys and signatures and often more.

The NIST-approved post-quantum options and PK/sig sizes (in bytes, for "security level 1", which is the lowest) are Crystals Dilithium 2 (1312 / 2420), Falcon-512 (897 / 666 but computationally expensive) or SPHINCS+-SHA2-128s (32 / 7856 for the smaller but more computationally expensive signatures; same for SPHINCS+-SHAKE-128s). This compares to 32/64 or 64/48 bytes for 256-bit ECC algorithms and 256/256 bytes for 2048-bit RSA. If you are fetching a few kilobytes of text or CSS, this additional overhead is huge.

Yup. I'm waiting for any quantum computer to actually break a non-trivial public key, even of a laughably small order (like RSA130, which was factored by classical computers 30 years ago). Lots of people get famous for papers based on theoretical quantum gates that nobody knows how to realize.

Elliptic curve crypto is vulnerable to the same kind of theoretical quantum attacks as integer-factorization cryptography. You currently need to use algorithms with unfortunate trade-off (large public keys or large signatures/key agreements) to get resistance to quantum attacks.

Assuming quantum computers ever factor numbers larger than 21 without cheating or falling back to deterministic algorithms, at least.

There's no such thing as technologically unable to comply.

If a nation state law enforcement insists, they will make you comply, and you and I will never hear about it.

A simple OS update with "If phone MAC == XXXXXXXXXX then send copy to FBI", targeted specifically at one phone, deployed only to that one phone, would go entirely unnoticed by the world.

And Official Secrets Act / equivalent, combined with a government-NDA and jail time for talking about it's very existence is literally routine. Has been since the days of black boxes in ISPs and them tapping Google's inter-datacentre links.

If someone like the FBI, NSA, MI5, GCHQ, etc. wants you to do something... you have literally zero choice in the matter. And talking about it will get you immediately jailed. And it really doesn't matter how big you are.

You think that Whatsapp end-to-end encryption is just going to make GCHQ etc. go "Oh well, nothing we can do?" No. If they need it, there'll be a guy knocking on your head office with a bunch of people, he'll only tell you why he's there in a closed meeting, you will comply, even if that means throwing everyone out of the datacentre and doing it yourself, and if anyone hears what he asked you to do, you will go to jail.

Been the same for decades. They just don't use it for ordinary crimes and petty stuff, mostly because of the resources they have to deploy to ensure that it stays quiet.

So we're giving up on CAPTCHA then, are we?

And the only alternative we can come up with is literal ID verification?

Something tells me that we skipped a step - i.e. a better CAPTCHA.

It's basically plugging the output of ChatGPT into a sudo terminal on your machine with write-access to all your data.

It's quite literally the dumbest thing I've ever heard of.

But then, even Slashdot are running obnoxious "generate apps with AI" ads in massive bars on my screen, and I paid to disable advertising and have ad-blockers.

What could go wrong?

LLMs have no business model.

That's why OpenAI is trillions in the hole, with no profitable tier of product in sight.

It's a cute toy that costs far too much to generate and maintain, and relies on basically stealing the data of the entire Internet to keep itself up-to-date and vaguely relevant, and the lawsuits on that have barely started yet.

And if an LLM was actually "AI"... it wouldn't need customers, as such. It could be left to wander off onto the Internet, given a credit card number and it would: set up its own company supplying goods that it obtains from others, answer customer queries, set up a fivr account and respond to every job on there, sell its own unique products, design its own 3D models for sale or production, trade on the stock market, bet on sports, or whatever... it would literally... just earn money for its owners. Directly. No need for a user to ask it to do so and then to give the result back to the user. Just... do the things that would directly earn it money.

Give it $100,000, a credit card, an Internet connection and... leave it to its own devices. It's "intelligent", right? And it has capabilities and capacities far in excess of any human, so we're told? So it could literally just start up a fake company, fill in the paperwork, register for tax, import goods, have a courier handle them, put them into a warehouse, set up a website, sell the product to the public, have a courier collect them from the warehouse, sell the product direct. Nobody would ever have to know that it wasn't human, and it could join the dots and just do what humans the world over do to make money directly.

If AI was any good... then IT would be the next billionaire.

https://www.sandlawnd.com/dui-...

(I don't understand the odd wording at the start of this quoted paragraph because it sounds like it's being set up for a contradiction when it's not)

"While the United States may seem like we have high numbers for DUI accidents every year, we actually are the third worst country when it comes to drunk drivingâ"which obviously isnâ(TM)t great. In 2015, South Africa was ranked number one as the worst country when it comes to drunk driving. With 58% of their fatal accidents involving alcohol in some way, they sit high above the second and third seats. The second seat goes to Canada, at 34% and the third to the United States at 31%. Countries on the lower end of the spectrum include Germany (9%), Russia (9%), India (5%), and China (4%)."

FYI, I've been on Slashdot nearly 25 years (maybe more? I can't remember).

You can read my FULL post history. Hell, I was a "paying subscriber". I used to run Slackware as a desktop for 10 years. Then I went to XP, then 7, then 10, and now... I'm back to Linux. Precisely because of... Windows 11, Microsoft's shovelling of shite into my OS, and increased frustration with it.

I manage Windows professionally, thousands of clients, and... at home... I'm now entirely Linux. Would you like an inventory (but I really gave one in the post above)? And I'm Linux BECAUSE of modern Windows. I literally went DOS, Linux, Windows, Linux, Windows, Linux over ~30 years. I did an awful lot of work on a single-floppy Linux router distro called Freesco. I ported code to the GP2X, a Linux-based handheld that sold mostly in Korea.

And now... I've gone back to Linux, the December just gone, because I've had enough of fighting with Windows 11 professionally and Copilot, Edge, taskbars and all the other nonsense necessary to make the OS do what I want to do. So I bought a Framework laptop. You can see posts from me going back a year or so talking about getting one. You can see YEARS of posts about the Steam Deck and, before it, my desire to have the Steam Machine/Box concept work better than it did going back to the original time of release of SteamOS back then.

I got my laptop for Christmas. I installed Linux on it, day one. I have then been praising it on here, and other tech sites I frequent, on my same username.

Tell me... how is that "made up"?

"You know, all of the AI-generated content increasingly looks similar, and they're all beautiful..."

Yeah, you and I are not on the same page.

Why does America have 150,000 drivers who can't be trusted to not drink and drive?

And why are they allowed a licence?

Chinese vehicles, both EV and ICE, are selling like crazy in every market where they can legally be sold. I've spent some time in Latin America recently and have ridden in several of the various models, and the reality is that they are all quite nice. The Uber drivers driving them invariably think that they got excellent value for their money.

In the United States we don't have access to these inexpensive brands. We can either buy expensive ICE vehicles, or even more expensive EVs where you pay a premium to not burn fossil fuels. In that situation it makes sense to want a vehicle that competes favorably with an ICE vehicle. After all, you can get a perfectly good ICE or hybrid vehicle for less than it would cost to buy a less capable EV.

The equation shifts dramatically when the Chinese vehicle you are looking at (whether it is ICE or EV) is 1/3 to 1/2 the price of a comparable vehicle. If I could get a Chinese EV for $13K I, personally, would be willing to put up with some of its shortcomings. As an example, I like the idea of the American made and designed Slate truck. However, it isn't available until next year at the earliest, and it is likely to cost $30K, very close to what a base model Ford Maverik, Nissan Frontera, or even a Toyota Tacoma currently cost. At that price it doesn't really make sense to purchase the far less capable electric vehicle.

However, if the Slate only cost $15K then it becomes far more interesting. That's the sort of price difference that Chinese brands are currently offering. I could learn to live with a range of 150 miles (that's supposedly the Slate's range, Chinese vehicles typically offer more than that), if it costs half as much as the competition. China is making vehicles that are more than competitive with what we currently have access to in the United States, and the prices are very low. The only thing keeping China from making huge inroads in the U.S. auto market is politics.

Sure there are some people that will never buy a Chinese vehicle, and there are other people that will never buy an EV. That's fine. I remember when the same arguments were made against Japanese (and later Korean) vehicles. If the politicians really thought that no one would be interested in these cars then they wouldn't need to protect us from them with tariffs.

The housing market is definitely another place where things have become ridiculously expensive. Fixing that issue is more difficult. Everyone is in favor of low cost housing, until they are building it in their neighborhood.

On the bright side, there is a ready source of inexpensive vehicles already for sale. The only problem is that, in the U.S. at least, our politicians won't let us buy them.

In the case of both cars and houses the solution is to remove existing barriers to supply. Right now it is impossible to build inexpensive housing in many parts of the country, and so we end up with expensive housing instead. It is likewise impossible to buy the inexpensive vehicles that I believe that consumers actually want.

My current daily driver is a 1996 Honda Civic (the base model with a 5 speed manual transmission, no AC, and manual windows). I say this to say that I really like the idea of the Slate. What I want is a basic electric vehicle without frills, and without extra technology that does nothing but break and drive up the price. The problem with the Slate is that it is not yet available, nor is it likely to be available in any numbers for a couple of years. What's more, there are already more capable Chinese vehicles selling in large quantities throughout the world that are available at a lower price. These vehicles come from companies that have already set up manufacturing and distribution channels, and they are selling vehicles in some of the most challenging markets in the world.

I've done a bit of traveling in Latin America in recent years and the reality is that there are several Chinese brands that are already powerhouses when it comes to actually selling, delivering, and maintaining vehicles. They make very competitive vehicles, and, at least in Latin America you can get these vehicles serviced and repaired ridiculously inexpensively. Uber drivers were quick to point out that their BYD (and other brand) Chinese vehicles weren't Toyotas, but they have invariably stressed that they would buy them again.

If it wasn't for the U.S. tariffs the Slate wouldn't even be a contender, and it isn't likely to be a contender when it is finally available. The only real advantage that it has is that it is comparatively affordable when compared to the other ridiculously overpriced EVs that you can currently purchase in the United States.

It is also worth noting that the projected base price of the Slate keeps going up. The first time I heard about it they were saying that it would cost around $12K with tax incentives. That would have put it under $20K without incentives. These days they say that it should cost less than $30K, but that puts it within spitting distance of the base model Ford Maverik, Nissan Frontier, or Toyota Tacoma, which, quite frankly, are far more capable vehicles, from companies with actual track records.

The reason that Chinese EVs are interesting is that they are essentially 1/3 to 1/2 the price of existing ICE truck models with compelling features and decent build quality. In the parts of the world where politics aren't getting in the way these Chinese vehicles are absolutely dominating. That's what I want.