Tech News: Did Samsung Fudge TV Benchmark Tests?

companies regarding user location and health data. As it stands, Warrent argues, data brokers can buy and sell that kind of information, which obviously compromises people's privacy. And in light of the expected Supreme Court decision to overturn Row versus Wade, there's also a concern women who live in states that outlaw abortion might be persecuted should they seek an abortion

in another part of the country. In fact, some states might consider passing laws that would make it illegal for citizens to leave the state in order to get an abortion, which is a complicated topic. It would undoubtedly face multiple tests in court, but in the meantime, Senator Warren is concerned that people will be struggling while the political system

actually hashes everything out. So part of the impetus for this legislation is the desire to protect those who need to travel in order to seek out a medical procedure that would otherwise be illegal for them. But generally speaking, there's been a need for better protections for data for

a very long time for US citizens. We've seen other parts of the world, notably the European Union past broad legislation aimed at protecting citizen data, but here in the US that really hasn't been a thing, at least not at the federal level. Warren's proposal would give the FTC the power to enforce rules regarding location and health data, and it would also give citizens a path to sue

data brokers that are found to violate that law. Of course, a lot has to happen before a proposal becomes law, so there's no guarantee that this will actually get to a point where you know, we can rely on it. We'll have to see. While we're on the subject of data brokers and user privacy, let's talk about a change

that Firefox has made recently. The browser has been employing numerous protective features for several years now, back in Tighten it launched the Enhanced Tracking Protection or e t P feature, and initially that was an option you could activate, but Firefox later would turn it on by default. ETP blocks

web cookies from known trackers. These are entities that attempt to track user brow your behavior from site to site in an effort to build out dossiers on those users, typically so that the tracker can then sell that data to data brokers, and they in turn typically sell the information advertisers. Of course, there are a lot of other things that can happen with your data, most of them are not good. But the most recent feature that Firefox has turned on by default is one that protects against

cross site tracking. That means this feature prevents cookies from tracking where you came from and where you're going to across different sites, and it's called total cookie protection. The concept is that each website cookie will be siloed from

everything else. In fact, bleeping computer dot com has a great way of putting this that each websites cookies will be kept in that website's own cookie jar and they won't mix with the other websites, so you'll still be able to take advantage of what cookies can offer on an individual site. So for example, uh, it could keep your browsing history on a specific site, like a shopping site, so that way you can easily navigate to stuff that you had shown interest in, or it might keep you

logged into a site between sessions. Those are useful things. So cookies are not all bad, but it will prevent trackers from being able to build a browsing profile or

fingerprint on you. So this has led to some people declaring Firefox as being the most secure browser when it comes to user data, that there are the most protections in place, many of them on by default that will prevent your data from falling into the wrong hands, especially without your knowledge, because a lot of that stuff happens without us necessarily being aware of it. I don't often talk about benchmark tests on this show, but I've got

a couple of stories that relate to them today. So a benchmark test is one in which the test administrator puts whatever technology it happens to be through a series of tasks to measure it's actual output. Because sometimes actual output and what was, you know, theoretically possible on paper are two very different things. So Our first benchmark story is about Apple's mto processor. Apple had been relying on Intel up until a couple of years ago to provide

the processors for Apple's computers. But now the second generation of Apple's own processor, the M two UH, is out, and the tests say that the new one is twelve percent faster than the M one processor in the same sort of machine if it's used in a single core processor mode, but when used in multi core functions, the M two processor is twenty percent faster. That's actually even

better than what Apple was saying. And it's w w d C presentation about its upcoming computers it had more conservatively placed the M two as being eighteen percent faster than the M one, So this is really good news for folks rely on Apple machines. Now we just have to hope that Worth's law doesn't muck things up. And in case you're unfamiliar, Worth's law is actually a kind of a cheeky observation that says software is getting slower at a rate that outpaces how computers are getting faster.

Essentially that programmers, when they're given the leverage of better hardware, end up making even more bloated software. The other benchmark test story I wanted to cover is about Samsung and it's television's Vincent Teo of h D TV Test discovered that the s nine five b q d O led TV that he was testing in gosh, I just I love model names, don't you. It's so easy anyway, that this particular television he was testing was acting a little peculiar.

You see, Typically, the way that test administrators check on televisions is that they create a test window that only takes up a certain percentage age of the full screen of the television. You know, they're focusing on a sector of the TV screen in order to do their tests, and the typical test uses ten percent of a television screen.

These tests tend to measure everything from luminosity as in how bright the images can get, to contrast which is the range of colors between true black and true white, and resolution, among other things. Teo documented that the Samsung TV he tested appeared to be boosting performance within the test area, which meant that the television was inflating its score. It was cheating, in other words, was using an algorithm to give unrealistic results that could not be replicated on

the television as a whole. This actually reminds me of Diesel gait, which was the Volkswagen scandal from a few

years ago, if case you don't remember that. In that scandal, people discovered that several diesel powered Volkswagen models would switch into kind of a low emissions mode when the car's computer system detected that the car was being tested for emissions, But then once it was disconnected from the test system, the car would then switch into a performance mode, presumably one that would not have passed the emissions test because it was going to create more emissions, it was going

to output more power. Well, Samsung sounds like it was doing something similar, and when tested, the television would send more power to that little test region to boost stuff like luminosity, But it wouldn't be able to do that for the television screen as a whole, at least not

sustainably without risking damaging the TV itself. Teo and others found that by using a non standard test window, one that didn't take up ten percent of the screen, say just nine of the screen, that the television would perform and its normal pacity, and that this the so called cheating algorithm, would be foiled. So Samsung came under fire for trying to fudge its television benchmarks, at least allegedly trying to do that. The company's response has not exactly

been an admission of wrongdoing. Rather, when responding to a query posted by flat Panels h D, the company repped from Samsung wrote quote, to provide a more dynamic viewing experience for the consumers, Samsung will provide a software update that ensures consistent brightness of HDR contents across a wider range of windows size beyond the industry standard end quote. Now what they're saying is, oh, we didn't account for people using other sizes besides ten percent window because that standard.

So we're just gonna update this so that when you do your test window, no matter what the sizes, you're gonna get the right results. It kind of sounds like they're updating the cheating algorithm, as flat Panels a D called it, in order to fool any kind of test. But you know, maybe maybe that's just a misinterpretation on

my part. You know, we'll have to wait and see if there's any follow up to this, and you know, if if it turns out that the whole TV screen actually is capable of the same things that the benchmark tests were saying. There's a moot point, but flat panels, HD and others have suggested that according to their tests, that just wasn't what they were seeing. So we'll have to see if this story develops. We have some more stories for today. Before we get to those, let's take

a quick break. We're back. It's time to get boring. Yeah. The Boring Company, which if you weren't familiar, is an Elon musk Endeavor, has received the go ahead to day more tunnels beneath the city of Las Vegas, Nevada. So the Boring Company previously built a tunnel system under the Las Vegas Convention Center with the goal of making it easier to get across the lvc C during busy conferences

like C s now way back in the day. In the earliest days of the Boring Company, we largely associated the idea of the Boring Company with the idea of the hyper loop transportation system. But these days that vision has been whittled back to essentially underground roads, at least in the case of the Las Vegas Convention Center, because they're rather than having passengers and cars moved through tunnels on trains or even on sleds within tunnels where much of the air has been pumped out in order to

reduce air resistance and increase the rate of travel. Instead, passengers hop into human driven Tesla vehicles. So really it's ride hailing, but underground and with only Tesla vehicles driven by humans. So this isn't even that first compromised vision where we went from the hyperloop train to a moving sled, so that you would drive your vehicle onto a sled which would then take you and your vehicle to wherever

it was you wanted to go. Instead, this is just getting into a vehicle and writing to a stop, at least at the convention center. I don't know that that's how it's going to be for Las Vegas as a whole, but the City of Las Vegas has approved the boring company's plan to dig a system that will spend nearly thirty miles and will include more than fifty stops on it.

And maybe that system will incorporate the sled devices we heard about in the past, or maybe it'll just be an underground road system that adds more traffic to Las Vegas. Is just traffic that's happening underneath the city. Many critics have argued that an alternative like a train system would have been a much better option. But I suppose we're just gonna have to wait to find out like it's Once it happens, we'll be able to say more clearly if it was a success or not. Of course, by

then it might be too late. But who knows. Maybe the tunnels will be really effective. I'm a little skeptical, but I hope that I'm proven wrong. The Washington Post recently published an article about how the US National Highway Traffic Safety Administration, or n h t s A, released a report showing that within the last year or so, Tesla vehicles in autopilot mode have been involved in two

hundred seventy three crashes. That's a lot more than had been previously reported, and according to the nh t s A, Tesla vehicles made up almost seventy percent of all crashes that happened while a vehicle was operating in an advanced driver assistance mode, which is what we would call Tesla's autopilot system. Now, I want to walk through a few

things before we really draw any conclusions. Y'all know, I am not a big fan of Tesla the company, but I think it's only fair that we really consider all the different factors before we just say, oh, that means Tesla vehicles are inherently unsafe. I don't think that's a fair conclusion, because one is that you know, of all

accidents in driver assistance mode being Tesla vehicles. That is huge, you know, saying that nearly three quarters of all vehicles that are involved in crashes, all vehicles that were in advanced driver assistance mode while being involved in crashes are Tesla's. That does paint a bad picture. But does that actually mean that Tesla system is inherently more dangerous than any other manufacturers system. Not necessarily, because to draw that conclusion

we would need a lot more data. For example, you know, Tesla is a really popular car model that has those sorts of features. Not every car has advanced driver assistant features, right, So you would first need to know what percentage of the overall market of vehicles with these features does Tesla occupy, right, Because if Tesla has of the cars with these features on the road, but only accounts for seventy of accidents involved with these features, enabled. That's different than saying, oh, well,

Tesla makes up of the market. Right. You have to know all these things in order to be able to draw some conclusions. Tesla maintains that its systems reduced the likelihood of crashes, and it could be that Tesla's driver assistance features have actually prevented way more accidents than otherwise would have happened. So like, there's no way for us to know that, right. We can't look back and say

how many accidents didn't happen because of autopilot. It could be that if autopilot had not been a factor, lots more people would have been in a lot more accidents. It's impossible for us to say. So. Again, we can't draw firm conclusions because we lack the data to be able to do it. One other thing I did find interesting. According to the Washington Post, regulators discus heard that Tesla's autopilot system would automatically shut off a second before impact

when the system detected an imminent crash. So the implication there is that Tesla may have built in kind of a plausible deniability switch that technically, when a Tesla vehicle gets in a crash, autopilot wasn't in charge because autopilot shut off a second before the truck the crash actually happened. However, the nh T s A accounted for this by demanding data for any accident that happened within thirty seconds of an advanced driver assistant system being active. So you couldn't

ignore those cases. Right, if the autopilot shut off one second before an accident, that would still qualify for the nh T s A. UH and a thirty second gap is long enough where you could reasonably say, well, the auto pilot system or whatever the driver assist system was couldn't have been responsible because there was a long enough gap there between when it was on and when it was switched off. So it had to be you know, human responsibility, not not machine responsibility. So that was how

the nh T s A kind of got around that issue. Now, this information about these accidents have prompted some lawmakers to say they need to look more closely at this sort of technology, and that does make sense. I don't think anyone is going so far as to say these features are a bad idea, but rather that the implementation and deployment can put people at risk, particularly if people rely too heavily on systems that are meant to assist, but

not replace, human attention and action. And now for a couple of stories about censorship the web, misinformation abuse, that kind of thing. And first up is Turkey, which may soon pass a law that will make it illegal to read fake news and misinformation on the Internet, primarily targeting social media companies. And here's where things get really tricky. So on the one hand, yes, misinformation is a very

real and global problem. We frequently talk about it on this show, and there are countries all around the world grappling with ways to hold platforms accountable for how they respond to reports of misinformation. I myself am in favor of measures that can detect and respond to misinformation. That being said, we also have to remember that you always have to ask the question who determines what is and

isn't misinformation? Because in Turkey that falls to a government that has for years cracked down on various communications platforms in an authoritarian approach to control the narrative. Now, this is the sort of scenario that free speech absolutists often cite, that any frictions on speech are going to be dictated by an authority, and that authority might not have your

best interests at heart. So therefore restrictions on free speech are bad and certainly, should this proposed legislation become law and Turkey, I expect the government will use it as a justification to censor any voices that oppose that government

in any meaningful way. However, we also have seen the incredible harm that misinformation campaigns can cause, and shrugging your shoulders and saying well, people just need to make up their own minds, as Zuckerberg once said, that absolutely ignores the truckloads of evidence we have gathered about how things

like echo chambers can help radicalize people. Anyway, long story short, I am concerned about this law, although I also think the Turkish government has effectively been censoring stuff online for years, so in a way you could just say that this is, you know, just a formality, and I, you know, I also struggle with this. This is a complicated issue where I definitely want to see misinformation get pushed back, but at the same time you have to worry about who's

doing the pushing. Meanwhile, over in Japan, there's a new punishment for people who are found guilty of insulting someone online. And first, by insulting, I mean insulting with the intent to harm someone. So presumably the way Ben Bolan and I insult each other would not qualify, because actually we like each other, we're friendly with one another, but we do insult each other on occasion. Secondly, I need to

say this isn't a new law. Instead, it's more of an expansion of an existing law that's meant to deter people from insulting each other and for those insults to ultimately cause harm. So this expansion means that now if you were found guilty of insulting someone with the intent to harm that person, you can face up to a year are in jail. Yikes. Now, there's already a plan in place to revisit this law in three years, to review it to see what kind of impact it has

on free speech versus civility. So it's possible that in the future Japan will remove the jail time sentence. But yeah, you gotta be careful who you call baca, y'all. All right, we've got a couple more news stories to go before we get to that. Let's take another quick break. We're back so some so nos customers got a little more than what they bargained for, literally after they ordered products from the company. So in a few cases, so nos

Is ordering system accidentally sent some customers extra products. So, in one case reported on by the Verge, a customer who ordered like six different things from so Nos ended up receiving six of each of those items. Also, Sonas charged people for the extra stuff it's sent. So this isn't a case of so no Nos makes an error in your favor or anything like that. That's not like Sonas made uh an accidental mistake and as a result

charged you for one stereo system but sent you six. No, it's more like it charged you for all six of those. The company said that the system update it recently had on its internal systems accidentally caused quote some orders being processed multiple times end quote. Sonas has said it will refund customers their money, which I mean, of course sons

is going to refund customers their money. They have to do that because otherwise they would be breaking a federal law that says you cannot charge customers for products that they didn't order. Now, according to the Verge, for that poor soul who ended up with more than thirty boxes of stuff. It gets more complicated because so Nos, according to the customer who was not named, Sons, has said it will not issue a refund until it receives all of the products that it accidentally sent out, which come

on so NOS. I mean, I understand wanting to get the product back because that is a real cost to the company. But this was so nos Is mistake. It was not the customer's mistake, and it sounds to me like the burden of getting everything set right is falling more on the victim of this rather than the company that made the dumb mistake to begin with. So this is me officially wagging my finger at you, Sons. I

don't know if you can hear it. It's good audio. Finally, you might have seen some news reports saying that Chinese scientists thought they had detected radio signals that may have come from aliens using a telescope RAO telescope called fast f A s T, and that initial report, which originally published on state sponsored media in China, has subsequently been removed from said media source, so it's no longer up there. Secondly, before we start flipping out that we've picked up I

don't know, TV signals from Alpha Centauri or something. We need to employ some critical thinking. Radio antenna often pick up odd signals. Sometimes those radio signals are extraterrestrial in nature, but by that I mean the radio signals originated somewhere other than here on Earth. And there is a lot of stuff out in space that generates radio waves. In fact, that's why we have radio telescopes, is to you know, measure those kind of things and learn more about our galaxy.

It's a natural phenomenon. In other words, it's not being generated by an intelligent species out there. And several scientists have already said that what this Chinese group most likely detected was radio frequency interference or r f I, and that was actually something the scientists themselves said was a possibility. So our i f I is just that's just radio frequency interference that comes from Earth. It's something that all

radio observatories have to take into account. That's why a lot of radio observatories, most of them in fact, are located in pretty remote areas where they're far enough out from sources of radio frequencies that though that doesn't end up, you know, interference isn't as big a concern. But that's

certainly something that does happen. And if we employ Acam's razor here, which is where we say, if you've got more than one potential explanation for something, it's best to go with the simplest explanation or the one that requires the fewest unproven assumptions. You know, if if you feel a cool breeze in your house, you could think, well, I might have a window open. You know, there could be a draft here, or maybe it's a ghost. Well, drafts,

you know, that's something we know that exists. It really does exist. We we are aware of it. Ghosts are unproven to exist, So it makes way more sense to assume that it is a draft, like maybe there's a crack in the wall, or maybe there's a window or door open or something. Then it makes sense to say, oh, it's old Aunt Bertha and she has unfinished business here

on this plane. So if you detect strange radio signals and your conclusion is that it could either be aliens or it's radio interference, then r if I is the simplest solution. We know rf I is a thing, we encounter it all the time. It's far more likely that we're picking up radio signals from here on Earth then picking up some sort of radio signal of alien origin.

For us to conclude that the signal actually is alien, we would need some sort of incredibly convincing proof, proof that secured our confidence that it could not have originated as our fi. You've probably heard the saying extraordinary claims require extraordinary evidence. This is one of those cases. So while I would love to say we picked up evidence that there are alien intelligences out there that communicate via radio,

it's too too early to have said that. It is uh an incredibly unlikely that that is what the Chinese radio telescope picked up. So sad to burst that bubble because I would. I would also love to find out we're not alone out here. I don't think we are alone,

by the way. I just also happen to know. Space is so large that the odd of us picking up radio signals from another intelligent species are just incredibly low, because if that intelligent species is on the other side of the galaxy, it will take hundreds of years for any signal to make its way over to a point where we can detect it. So the odds are very much low for that to happen. All right, that's it for the tech News for Thursday, June two thousand twenty two.

If you have anything you would like to send me, Maybe it's a suggestion for a future episode, maybe it's commentary about aliens. Who knows. Then you can do that in one of a couple of ways. One way is to use the I Heart Radio app. It's free to download. You can just go straight to the tech Stuff page on the Heart radio app and there's a little microphone icon there and if you if you use that, you can leave a voice message up to thirty seconds in length, and if you like, you can even let me know

if I can use that audio in an episode. I prefer opt in to opt out, so just let me know. Otherwise I will likely just use the audio to you know, launch an episode or answer a question, but I won't actually use the audio within an episode. The other way, of course, is just to send me a message on Twitter. The handle for the show is tech Stuff H s W and I'll talk to you again really soon. Tech

Stuff is an I Heart Radio production. For more podcasts from I Heart Radio, visit the I Heart Radio app, Apple Podcasts, or wherever you listen to your favorite shows,