Posts by heyrick

most of them American

...most of them based in a country that has repeatedly shown no interest in European laws, rules, and regulations; but rather seems to expect us (in sovereign countries on another continent) to abide by their laws, rules, and regulations (or total lack thereof).

There, fixed that for you.

Hmm

This story is immediately followed by "AI still doesn't work very well, businesses are faking it, and a reckoning is coming".

"It is vague, using terms that allude but do not define. It sets specific and punitive fines for non-compliance, without specifying what non-compliance looks like."

Sounds like somebody has been taking notes from the UK's recent Online Safety Act.

Would any of these dipshits like to suggest...

...how one is supposed to have a reliable way of verifying age in an open source operating system...

...when it would be trivial to "userage = 42;" then "make"...

...and quite a few people are ideologically opposed to the idea of binary blobs so having them run a closed age verification service would be a non-starter...

...and wouldn't that be a brilliant way to backdoor systems that people think are reasonably secure, so not trusting is a completely valid approach.

Anyone? 'cos all I can hear is the sound of crickets.

I disabled V8 a while back

Too many stories about flaws leading to potential bad things. The only way to browse is to assume, by default, that any website one visits is hostile. So I'll take a slower jerkier browser over one that is easily pwned.

This is, of course, the times I use Android Chrome and not just Firefox that is set to disallow most things by default.

F*ck Perplexity

This is the salient point here: Amazon's rules prohibiting automated data gathering

This is bullshit: Today, Amazon announced it does not believe in your right to hire labor, to have an assistant or an employee acting on your behalf

I'm surprised that this hasn't already been ruled on, as Amazon isn't preventing one from employing a lackey to wade through the site for you. They just don't want an automatic device that will blitz the site. The only reason Perplexity doesn't understand this is because it's in their interest to fail to understand it.

But f*ck Perplexity right between the eyeballs because they have been hitting my largely uninteresting site. As soon as I blocked the User-Agent, they came back with exactly the same behaviour but a faked UA that looks like a regular user. It's arseholes like this that are causing websites to be less and less open. Scum...

but undermining copyleft is a serious act

Putting aside the AI involvement, what it reads like to me is somebody created a version that does the same thing and has the same name but is (allegedly) different code.

What's the problem? If it's not the original, he can licence it as he likes. The antisocial part is calling it the same thing (and using an AI as an author which muddies the licencing tangle somewhat).

Haven't we been here before? The idea that an open source bit of code can "do the same thing" as something closed source, but it's okay because it's not actually the same thing?

Block lists are only useful...

...when handled correctly. I don't use my ISP mail address because a certain list provider had a tendency to block the entire subset of ISP addresses "to make a point". Given that dealing with spammers is basically whack-a-mole, the point they made was probably not the point they wanted to make. By all means shut down spammers, but when one is happy to bring in lots of collateral damage, that's when one crosses from being useful to being arseholes.

So the company that was a major force behind pushing HTTPS and online safety is suspiciously absent minded when it comes to their AI shovelware. Funny, that.

How about a version of Chrome without the AI? Let the users decide which they'd like to install.

22 months and a 50K fine?

If she paid $5M for the labels, how much did she make? Because if she pocketed a couple of million, two years and a slap on the wrist fine is nothing. Just think how long it would take to make that sort of money doing a regular job...

"MLM allows patients to check their health records electronically, communicate with their physician, and offers doctors a range of administrative features."

How is this different or better to Mon Espace Santé?

Wasn't the manager that approved the repair required to sign off on it?

The guy is quite patient, isn't he?

Once having done the job I was there to do, the moment they refused to let me leave is the moment I would be on the phone to the police (and if phones had been confiscated because of security mumble, plan B, prod the red button marked "In case of fire"). You got a problem, you sort it out through the correct channels, you don't hold people hostage.

disappointing an officer has behaved in a way which could not only discredit the police force

Regarding the second point, allow me to introduce The Met to anybody who has never heard of those arseholes that take "bent copper" as a job requirement.

For the first point, can they please explain by what rational means they think that typing speed is an indicator of productivity? Rozzers are supposed to solve crimes, that implies dealing with people, looking at evidence, reading forensic reports, and all the other stuff you see on TV except car chases that involve people walking around carrying huge glass windows - never seen that once in my life. Where does typing speed come into this? In fact, given that what a Rozzer writes may end up being used as evidence, it is likely that they would need to think of the most concise and unambiguous way to write something, being careful about facts and clarity. So when does typing speed come into this?

had become a bit aggressive when facing the prospect of a shutdown

Okay, so you're a computer with a shitload of processors in order to create pictures of six limbed cats and fifth grade level poetry...

...and I have my finger resting on the power switch.

Wotcha gonna do about it? Off means off. Reboot means reloading from disc, a mere hiccup. There's no concept of life versus death. The data could be transferred to an entirely different machine in an entirety different country and (so long as it's compatible) will start up and spout bollocks with the appearance of understanding. Assigning any sort of emotional state to a machine is like saying your compiler is angry because it doesn't like that you missed a semicolon. It's just software. Clever software, but still software.

want to be fair to our actual users

You mean the ones paying quarter of a grand that just got cut off?

Maybe in AI land that sort of money is a mathematical rounding error, but in the real world that's not an inconsiderable sum (3K/year).

I can't help but think that even ChatGPT would consider this a really dumb argument.

Maybe he should have asked it first?

You may mock it...

...but I bet it's been more reliable than any modern offering, may well still work (after a bloody good wash), and won't need to periodically reboot, or hang up while dealing with upgrades, or turn into a potato if it can't contact the mothership. Not to mention your standard telephone will work just fine and you won't have to try to deal with getting equipment from manufacturer A to work with stuff from manufacturer B (both of whom claim compatibility and blame the other when that's clearly false). And, finally, it doesn't require monthly dues in the form of the ever growing subscription model.

It simply hangs there on the wall and does what it does, the only fuss and trauma will be entirely down to the ambient conditions (but that's the fault of the people, not the machine).

And have they contacted each and every person whose information got swiped? And why the hell do they have a system perfectly willing to cough up that much info at once without flagging plenty of warnings?

Putain! C'est vraiment de l'incompétence spectaculaire. Un pays plein des fonctionnaires et paperasse sans fin...et quoi? Légumes!

Half the time, the card number/date/check is all that's necessary to buy stuff online. Often the cardholder's name isn't requested, and it's a bit hit and miss whether or not the bank app will require a secure authorisation to let the payment go through. So, names or not, this was a monumental screw up and I'm rather surprised that whoever provides their card handling hasn't revoked that until they can prove their systems are secure.

Furthermore, why did the tills even have this information? The tills here have a little card reader that you put your card and PIN into. This gives the till an authorisation code and obfuscated card details like 4040********1234 which is what gets put in the receipt. The till itself doesn't need to know the card info, just whether or not the payment was approved and a reference number for accounting.

Was this some home-grown "solution" outsourced to the lowest bidder?