Follow Slashdot stories on Twitter

 



Forgot your password?
typodupeerror

Comment Re:Except they don't (Score 1) 47

Just because it's a judge doesn't mean they understand basic math.

No, but they do understand the law, which considers an attempt to monopolize a crime even if unsuccessful.

They also understand that in the context of U.S. antitrust law, Apple's ~58% market share, at roughly twice the size of their next largest competitor (Samsung), is absolutely large enough to make Apple a successful monopolist, and they also understand that Apple is a twice-convicted monopolist — once involving Epic, and once involving the iBooks store — which makes their ongoing behavior worthy of extra scrutiny.

Comment Re:So... (Score 1) 9

Freelancers with coding skills comprising at least 25% of their work now earn 11% more for identical jobs compared to November 2022 when ChatGPT launched.

So what you're saying is that they're basically keeping up with GDP and using AI has no real benefit.

No, no. They're saying that the people who aren't using AI are getting less work because there are fewer jobs, and the people who are using AI are barely keeping up with inflation compared with the pre-AI world.

Comment Re:This is why (Score 5, Informative) 58

The argument against SMS is way overblown. For it to work an attacker would not only have to gain access to your account details but also spoof your phone on the phone network. Possible? Yes, likely? Unless a nation state is after you - no.

Actually, it's a pretty common strategy for breaking into the accounts of celebrities. It usually involves convincing someone who works for one of the phone companies that you've gotten a new phone, i.e. they already have enough personal info from you to impersonate you to the phone company. And then after that, all your accounts fall like a house of cards.

Comment Re:Only 20% for human doctors (Score 1) 66

I only skimmed the article, but am I the only person who thinks that, if we had a situation or field of diagnosis where doctors were only getting it right 20% of the time, we would throw some research/education/analysis at it? Because 20% correct (or 80% incorrect) seems kinda concerning and I would think would lead to a lot of brouhaha or lawsuits? Maybe it's just me.

I'm assuming this is based on edge cases, e.g. medical images where cancer was just barely starting to appear, situations where lupus is mistaken for rheumatoid arthritis, etc., in which case the human rate of correct diagnosis could indeed be very low, precisely because they were chosen from cases where humans had made mistakes before.

If that is the case, then the question becomes whether the model is over-trained on these edge cases and would generate false positives, would miss obvious diagnoses, etc.

Comment Re:Amazon (Score 2) 62

You're arguing with someone who's given you millions of dollars over decades about a single 2-dollar missing component on a massive order they made? You're insane. They're just going to go elsewhere. It's not even worth the time on the phone call to argue it.

You're assuming companies don't understand that. What you're missing is that the companies that do this tend to be the companies that have their customers over a barrel. You have a choice in where to buy random stuff online. You don't have much choice in airlines. Only a few companies go to both of the airports that you need to fly between. They can screw you as much as they want, and unless you're prepared to lawyer up, you're gonna accept whatever they give you and like it, or you're not gonna fly, because they're all approximately equally horrible.

Ultimately, the reason for bad customer service is that the customer has no power. Short of a class action, you're not going to change their behavior, and they usually write their contracts to make class actions hard. And governments are thoroughly in the pockets of these big corporations, so they're not going to do anything about the problems, either. And there's no competition, because a few big companies have cornered the market, in part because of high cost of entering the market, which in turn, is often because of high regulatory burden. But those regulations are essential for preventing other problems, e.g. safety issues, so removing the regulations won't help, either.

The right fix is to separate the customer-facing organization from the safety-critical organization. Have a few companies that own fleets of airplanes, and a hundred companies that rent planes from those companies and fly them and sell tickets. With that organizational model, all of these problems go away, because the customer-facing orgs have a low barrier to entry, so you'll tend to end up with companies competing to provide the best service at the best price, with some focusing on higher-tier service, and come focusing on lower prices, but everybody knowing that if they screw up, you'll go with one of a hundred other companies. And you'll get a higher diversity of routes, and you'll have aggregators combining routes from multiple airlines, etc.

Unfortunately, we won't see this, because regulators aren't interested in breaking up oligopolies these days.

Comment Amusing conjunction (Score 1) 41

Kind of funny to see how AI's improve by re-writing themselves, following immediately a story from earlier today about humans being driven into psychosis by AI's.

This claims it uses empirical evidence to judge improvement but why would an AI not be as much a cheerleader for anything it does as it is for any human?

Comment Re:Uh huh (Score 5, Insightful) 172

So basically this is a new version of "Listening to Judas Priest will make you commit suicide", the Satanic Panic and all the other utterly moronic moral panics that make people afraid of unlikely things.

If Judas Priest listened to what you said and wrote custom songs about you individually, sure.

Comment Re:She's not an AI (Score 3, Interesting) 83

Her experience is not unique and long predates the use of LLM chatbots.

To some extent, yes, but two minutes with one of those damn chatbots has me ready to strangle someone, and I'm a really nice person who basically never gets rage-level angry at actual human beings even when they're being complete a**holes.

Here's the thing: If someone explicitly asks to speak to an agent, your bot gets exactly one question after that to direct them to the right place. Any more than one question, and it becomes psychological abuse. This is doubly true when you have rage-inducing pieces of s**t that that ask you incessantly whether you want to do A or B when what you actually want to do is C, and they refuse to do anything but repeat the same question, and then they hang up on you rather than connect you to an agent over and over again.

These automation systems are quite literally designed to make customers get so fed up that they give up before getting help that would potentially cost the company money. That means when customers do get lucky and manage to get a human being, they are almost invariably absolutely fed up to an extent that they would not have been had they spoken to an actual human being the first, second, or even fortieth time they asked the sociopathic piece of donkey-s**t bot to let them speak to an agent.

So from a hostile working environment and corporate liability perspective, the fact that assholes will always find ways to be assholes doesn't change the fact that these systems make the problem orders of magnitude worse by turning normally calm, rational people into raging lunatics in under three minutes. Before those customer-abusing systems were installed, call center workers only occasionally dealt with customers who were already at their wits end, whereas now, customers exhibiting that level of post-traumatic stress psychosis are the default output from the automated call trees. The difference between a few bad customers per day and a few good customers per day is not insignificant to the mental health of workers. That makes those automated systems the proximate cause of the extreme psychological abuse that these call center workers go through.

BTW, you worked in the industry way too late to have seen call centers back when the system worked. By 2004, phone tree abuse was already in full swing. You'd have to go back at least another decade to get a call center environment that wasn't designed primarily to minimize the extent to which customer get actual service, and maybe longer. Ultimately, the root cause are the a**holes who run the companies, and customers and employees both need to take out their rage on them, not on the call-center workers. So the other thing you can do is give out your CEO's personal phone number when you get a particularly angry caller, and let the problem sort itself out.

Comment Yoda's wisdom best again (Score 4, Insightful) 172

Just another example of why having watched Star Wars is such an important aspect of lifetime mental health...

When exploring deep philosophy with an AI and ending up down rabbit holes, Yoda's warning was always there to moderate you ahead of time...

Luke: "What's in there?"
Yoda: "Only what you take with you".

Comment Re:Should have been rejected on procedural grounds (Score 1) 58

If a law is unconstitutional, it does not matter how long it takes for that fact to be determined, it does not suddenly become constitutional by default.

Ostensibly, but realistically, if the law were unconstitutional, it stands to reason that someone would have sued over it almost three decades ago, so the burden of proof required to argue against its constitutionality is basically insurmountably huge after so many years of no one doing so.

The only plausible grounds for arguing unconstitutionality after so long would be if there were new information that did not exist at the time, e.g. if you were able to collect statistics from a generation of people living under a school segregation law showing that separate but equal doesn't work, or if some new group of people were born who were somehow affected by some law who had some protected characteristic that did not exist at the time of the law's creation, or... something else new that could not have been determined at the time.

Absent that sort of pivotal change in reality or information, decades of presumption of constitutionality represents a high enough bar that such a challenge would almost inevitably fail. There's really no reason to not apply laches after such a long period of time, in the absence of new evidence, a new class of aggrieved party that didn't exist at the time the law was passed, or some other substantive change in reality that would make the constitutionality suddenly worth calling into question, where substantive change in reality is something much larger and more significant than a particular party winning the White House and deciding to throw their weight around in an effort to smash everything they don't like.

More to the point, in a reasonable world, the courts should have applied punitive damages for wasting the courts' time with such a frivolous, laches-doomed, decades-of-unchallenged-presumption-of-constitutionality-doomed legal challenge. It should be genuinely *masively* expensive to pursue a case like this about something that is not exigent. The courts have enough on their hands dealing with all the constitutional challenges from the actions of the current government without having their time wasted litigating the constitutionality of laws passed before a decent percentage of eligible voters were even born, and the fact that even a "win" would have been nothing more than a procedural unconstitutionality that limits how government does something rather trivial, rather than a meaningful unconstitutionality that truly deprives anyone of rights or creates any sort of grave societal harm makes the case doubly appallingly. Everyone involved in bringing that suit should be ashamed of themselves.

This was a waste of the courts' time. And although in some ways, it is nice that the courts shot down their absurd argument once and for all, it was still a waste of the courts' time to bother doing so.

The worst part of it, though, was realizing that the opposition's position, if taken to its logical conclusion, would also make it illegal for the IRS to accept credit card payments, accept direct withdrawals from banks, and lots of other similar situations where a private business ends up collecting a tax on the government's behalf. It would nationally eliminate red light cameras. It would nationally eliminate private parking enforcement contractors. And so on. The sheer number of things that would have to be rolled back and rethought if the other side of this case had won is mind-boggling to such a degree that a modicum of common sense makes it clear that their argument had no merit.

Estoppel by laches would have been too good for this case.

Comment Re:XLibre? Isn't that the Nazi fork? (Score 1) 126

XOrg might have been stuck in that anything new they bring to X would have broken existing implementations but if they did not make any changes, X remained stagnant in features. A fork really is the best thing strategically. Any implementation not wanting any changes like OpenBSD can stay on Xorg whereas others can use XLibre that gains new features.

Comment Re:XLibre? Isn't that the Nazi fork? (Score 1) 126

I thought Wayland was supposed to solve all the problems so nobody would have to touch that icky, unmaintainable X11 code any more. I had it on good authority (the same people everyone is trusting to develop an alternative to X11) that nobody could reasonably keep X11 working.

I do not know the technical reasons why the X Window system has largely been stagnant. I would guess that updating the X Window system these days would be like updating COBOL. Sure it can be done but there are few experts that are around anymore to understand the nuances of it.

With that said, IME the people who use DEI as a bad word are not serious people, so I expect XLibre to go nowhere.

Politics aside, I would say it is really old code that someone needs to dissect and understand. Maybe people talented enough just did not know that it was in need of major work.

Comment Re:She's not an AI (Score 2) 83

So when Lindsey starts reading from her AmEx-approved script

She's not an AI, she's a script. I suppose if she's a well written script she might be a decision tree.

Maybe if enough of these folks get angry enough, they'll finally get together and file a class action lawsuit against their employers for creating a hostile work environment.

The fact of the matter is that her bad experience is 100% caused by their employer designing a phone tree to waste people's time maximally and make it as hard as possible for someone to reach a real human being. By the time anyone reaches a person at most of these dirtbag companies, they're so angry that they are ready to shoot someone. And ultimately, that's a decision that their employers are making that is clearly creating a hostile work environment. It isn't just horrible for the workers. It's also hell on earth for the people calling in.

Unfortunately, the people calling in rarely have the power to do anything about it. But if all the employees got together and asked for injunctive relief in the form of forcing their employer to employ front-line human customer service reps with a maximum of one single routing question to get you to the right team, they absolutely *do* have the power to use employment law to compel these companies to stop that nonsense, and the world would be a better place for it.

So to all the customer service reps out there in the world, please, for your own mental health and the mental health of the people you're trying to help as well, hire a lawyer and sue your employers. It's the only way to rid ourselves of the hell that automated phone systems have become, for the good of literally everyone, including, ironically, for the pencil pushers who incorrectly think that they're saving money.

Comment Swift is more advanced... (Score 1) 44

Other than the SwiftUI framework, approximately everything that's in Swift was in Objective-C 5â"10 years ago.

Not the concurrency framework (GCD is not the same), SwiftUI doesn't have things like Swift structs, only supports integers enums, Generics, no guard statement. Also finer grained access control.

Mind you they have improved Objective-C over the years by bringing in some Swift features as Swift improved! Like nullability annotations.

I still do like Objective-C as a language but even with Swifts advanced areas and quirks, I still think it's more straightforward than Objective-C for newer users. And I think finally with the new beta version of Swift they have a concurrency model that is strong but also friendly enough for people to work with.

Slashdot Top Deals

It is masked but always present. I don't know who built to it. It came before the first kernel.

Working...