|
eXXon posted:Since this is apparently troubling you deeply, I apologize for my split-second decision to derisively refer to the head of an organization as a capo - an Italian mafia slang term that also literally means head - in order to exaggerate for effect, mistakenly leaving the impression that I consider their activities to be genuine organized crime rather than just mildly sociopathic. I hope we've all learned a valuable lesson about how choosing to "exaggerate for effect" tends to make an absolute loving mess of a discussion, as people try to puzzle out why you're saying things that are both false and entirely unreasonable just for the sake of dumb shock value.
|
#
¿
Dec 12, 2020 04:29
|
|
|
|
| # ¿ May 21, 2024 07:15 |
|
|
Cicero posted:I don't think that was a serious accusation, they were just being insulting Well, don't throw out bullshit accusations for the sake of being insulting. D&D doesn't need people making up inflammatory fiction about each other.
|
|
#
¿
Dec 15, 2020 03:39
|
|
|
The content of the email doesn't really matter. The core issue is that GoDaddy's employees are in an economic situation such that a plausible-sounding $600 bonus goes right into "CLICK THIS RIGHT NOW" territory. Like, it's a legit subject for a phishing test, because an actual scammer can dangle money just fine. But the solution for people biting on offers of money isn't retraining, it's paying them more in the first place.
|
|
#
¿
Dec 28, 2020 16:32
|
|
|
BiggerBoat posted:This is a great point. I don't think telehealth is going to replace in-person doctoring. If anything, the goal is for it to become part of the pipeline to in-person doctoring, dragging people into the system to rack up further bills. The theory is that since telehealth is easier and quicker than regular doctor's appointments, people might be more willing to schedule them for things they might bother setting up a real doctor's appointment for, allowing the medical system to collect some money for things that otherwise would have gone unseen. It becomes a funnel to draw people into the system. But if it turns out your issue isn't obviously minor enough to be be solved with over-the-counter medicine or a basic catch-all prescription, then all the telehealth doctor can really do is tell you to come see a real doctor - often an urgent care or ER. So in the case of a real issue, the medical system at large gets paid for the telehealth, and then gets paid for the actual in-person visit the telehealth doc tells you to make.
|
|
#
¿
Oct 16, 2021 16:28
|
|
|
Let's look at the two possibilities for the Mastodon lawsuit:
Somehow, 1 seems a little more likely than 2. After all, it's not like Trump is having some kind of crusade against copyleft or anything. He just wants his base to think his team built the site themselves. Fighting a high-profile lawsuit all the way to the Supreme Court, only for a judge to rule that he used Mastodon's code but doesn't have to follow the license terms, seems like it would only draw more attention to the very part he doesn't want attention on.
|
|
#
¿
Oct 24, 2021 19:32
|
|
|
It's gonna be real interesting to see the specifics of that deal, because Tesla's total production output in 2020 was barely 500k vehicles. So the Hertz deal is probably somewhere between one-third and one-fifth of Tesla's entire yearly Model 3 production. How long is it even going to take them to deliver that?
|
|
#
¿
Oct 27, 2021 06:10
|
|
|
Mister Facetious posted:So they hired the AI ethics people Google fired right? Nope. They hired Accenture's senior Responsible AI researcher and Facebook's technical head of privacy. And on top of that, this team is explicitly a advisory research group with no authority to force change. All they can do is make suggestions, which the engineers are free to ignore with no consequences. quote:What the META team doesn't have is serious enforcement power. They say they don't want it at the moment — "You can't really drive change through fear of enforcement, but for long-term investment in change you do much better by growing education," according to Williams — but at the end of the day, META is a knowledge-creating team, not a police force. While they can research and propose changes, they cannot necessarily force other teams to fall into line. Their work is democratic, not authoritarian. i don't exactly have high hopes for this approach
|
|
#
¿
Oct 29, 2021 01:35
|
|
|
How are u posted:
The future of the videogame industry is endless microtransactions, with collectibles, cosmetics, and gachas constantly tempting players to keep coming back to the game every day to be bombarded with opportunities to feed money into the system. Shouldn't be hard to see where crypto bullshit fits into that.
|
|
#
¿
Nov 5, 2021 22:43
|
|
|
Heck Yes! Loam! posted:As someone that got yelled at once when passing through Oregon for trying to pump my own gas, yeah it's real. Originally, all gas stations were full-service with the gas being pumped for you. The development of reliable gas pumps and various safety improvements made self-service viable, and various economic pressures made it competitive: by firing the attendants and making drivers pump their own gas, self-service gas stations could lower prices and undercut their full-service competitors. When oil prices rose in the 70s, that was the deathblow for full-service, as gas stations all over the country fired their attendants in order to try to counteract the big rise in the sticker price of gas. The exception is in NJ and Oregon, which banned self-service gas in 1949 and 1951 respectively. Officially, the point of the laws was safety, and the explanation was convincing enough for the NJ Supreme Court to uphold the law. Many have suggested that it was actually protectionism, with Big Full-Service lobbying the legislature to pass laws protecting the inefficient full-service stations that were unable to compete with the cheaper self-service stations. Personally, I'd say both are credible explanations - pumping your own gas wasn't as easy or as safe in 1949 as it is now, but it's usually fair to assume that lobbyists had a hand in something. Either way, those laws surviving all the way to 2021 likely has a lot to do with economic issues, though politicians in both states are quick to say that residents seem to genuinely prefer full-service as well.
|
|
#
¿
Nov 14, 2021 18:11
|
|
|
It's worth noting that Uber is subject to a very different financial context from standard taxi companies: they can operate without regard for profit. Q3 2021 was Uber's first profitable quarter ever, after accumulating more than $22 billion in losses over the course of more than a decade. With investors keeping them afloat, Uber and Lyft are free to experiment without worrying too much about costs or revenue impact. That's how they've managed to implement innovative systems and unusual side projects while still maintaining price parity with standard taxis or even undercutting them. The taxi industry certainly has been stuck in its ways, but having to operate at a profit has also restricted its ability to change things compared to Uber. I'm bringing this up, because Uber is commonly cited as an success for innovation and deregulation, but to me it looks more like a sign that an organization has more freedom to focus on improving service and creating a more efficient system when it's subsidized and doesn't have to worry about being run for profit. Which sounds like a sign that we should be putting more money into publicly-run transit systems!
|
|
#
¿
Nov 17, 2021 22:00
|
|
|
PT6A posted:Uber's success is innovation, not deregulation. They want to convince you that it's both at the same time, but it's all about the app, and there was never any real obstacle to taxi companies getting together and saying "hey, let's do this" other than the fact they thought they didn't need to. The key to Uber's success isn't the app, it's pretending that drivers are independent contractors and making them provide their own vehicles. The massive potential workforce this enables is crucial to their quality of service.
|
|
#
¿
Nov 18, 2021 08:04
|
|
|
Owlofcreamcheese posted:No, I'm pretty sure the success is the app. No actual consumer cares one bit on the employment details of the guy driving.. Everyone cared a huge amount they made taxis a thing where you click a map then it charges your credit card seamlessly instead of taxis being a weird complicated set of antagonistic interactions from start to finish The employment structure is how Uber eliminated the antagonistic interactions from the taxi system. They're able to fire workers practically at a whim, allowing them to maintain a tyrannical system in which drivers are desperate to maintain customer satisfaction lest they be fired at Uber's whim. Even fast food companies, notorious for high turnover and quick firings due to a large potential workforce, can't just fire anyone whose customer satisfaction rating falls below 4.6 out of 5. Uber's independent contractor model allows them to easily maintain significantly more active workers than they need in an area without incurring extra costs, which not only helps to maintain quick service but also makes individual workers extremely disposable. The secret to ensuring absolutely no bullshit whatsoever from the drivers isn't the app - it's the employment environment Uber created in which they have overwhelming power against the driver.
|
|
#
¿
Nov 18, 2021 15:14
|
|
|
BigRed0427 posted:The gently caress? Why? Why is that loving possible? They weren't completely locked out. The key still worked. But since they could unlock the car via the Tesla app, many owners stopped bothering to carry their physical keys around, since they already had something in their pocket that could unlock the car. The app doesn't directly communicate with the car, though. It sends the request via the internet to Tesla's servers, and then Tesla's servers authenticate it and forward it along to the car. If Tesla's servers go down, then everyone who didn't bring their backup physical key with them is in trouble. It's just another example of how Tesla designs stuff that tends to fail poorly, and then markets it in a way that leads users to overly rely on the best-case scenario and neglect backup options.
|
|
#
¿
Nov 20, 2021 23:48
|
|
|
Owlofcreamcheese posted:Is this post based on confusing this with the Facebook thing, or is someone actually trying to take a black excellence in technology conferences software down a peg? Roughly thirty seconds after Facebook announced their Metaverse product, "metaverse" has become the media's favorite phrase for every single even slightly virtual experience, even if it's 00s-tier "walk an avatar around" stuff. There's professional thinkers around every corner acting like they've never seen a website before in their lives as they furiously bang out tens of thousands of words on the amazing potential of metaverses everywhere.
|
|
#
¿
Nov 27, 2021 07:14
|
|
|
Sagacity posted:Perhaps they did some market research and it seems like an investment that'll pay off to them? In the B2B technology market, there's a lot of weird poo poo that gets sold because it appeals to the top execs who sign the checks and make the purchasing decisions, even though the line employees who'll be forced to use the stuff will absolutely hate it. This Metaverse stuff sounds like the perfect example. The C-levels will love the idea of screwing around in VR during their meetings, at least for a few weeks...and that'll be enough to get them to buy several hundred VR headsets for the whole company along with a year-long subscription to something, regardless of whether it's any use to the people who do actual work. "Is it useful?" and "will top executives buy it?" often have different answers.
|
|
#
¿
Nov 27, 2021 17:14
|
|
|
How are u posted:Play to... earn?? Imagine if there were capped numbers of good items in MMORPGs, and once that cap is hit, the only way to get your hands on those items is to buy them from other people. Also, imagine that instead of using an in-game currency, the game only uses outside currencies like Bitcoin and Etherium or a bespoke cryptocurrency, so it's encouraged to plow real money into buying the currencies necessary to engage in these inter-player sales. Now imagine crypto nerds and MMO poopsockers being able to tell their parents that they're spending eighteen hours a day grinding out MMO items as an investment, and that all that videogame time will surely lead them to massive profits in the player economy. Of course, in the end it boils down to the same dilemma as every other crypto scam - if the whole system is designed around letting early adopters and dedicated devotees hoard up the vast majority of the stuff while forcing later adopters and casual dabblers to buy it from that first gen at a huge markup, well...how many people do you really expect to get in on the bottom rung of Pyramid Scheme: The Game? Play-to-earn is basically play-to-win, except that instead of paying the game maker to not have a miserable experience, you have to pay the older players.
|
|
#
¿
Nov 29, 2021 22:06
|
|
|
Sage Grimm posted:Which is already happening in the Philippines and Argentina through essentially gold-farming schemes. Except of needing just computer and internet access, you also need someone to stake the initial ~$1500 USD to get started. Yeah, but imagine if there was a cap on the total amount of gold that could exist in the game at once, or on the total amount that can be generated gamewide in a given time period. So if people gold farm hard enough then they could literally own all the gold in the game, making it impossible to get gold except by buying it from gold farmers, who could use their monopoly on gold to charge insane prices. That's what crypto bros want to do - take the exploitative dynamics that already exist in gaming-for-profit, and add even more artificial scarcity to further increase the power of the sellers by allowing them to outright monopolize rare items. Of course, in practice, it suffers from the same issue as every other dream of making stuff based around crypto: it's so focused on creating theoretical massive profit opportunities for the haves that there's no thought given to preserving the fun for the have-nots. In practice, most people will abandon the game as soon as the supply of useful stuff gets bottlenecked enough that have-nots are forced to buy their stuff from the haves. Which of course tanks the value of all that carefully-hoarded stuff, and likely causes the game itself to die out as well. This leads to the second, much less realistic, pillar of crypto-based pay-to-earn: storing the items on the blockchain so that they exist independently of the game, so that people can retain their item hoards even after the game dies out. Sometimes the idea is that even though the NFTs are no longer connected to a useful item, they will become collectibles because the game that issued them no longer exists, or something. I can see game companies getting into that, honestly - it's no problem of theirs if it turns out no one wants to buy those item NFTs from the rubes who paid out for them. Sometimes the idea is that games will create ingame items only accessible to people with NFTs from other games, allowing them to take their collection of hoarded items from game to game: getting bonuses from each one, and using those advantages to farm up even more items in the new games, building up an incredible dragon's hoard of items from a variety of games that allow them to be incredibly powerful in all NFT games forever or sell off that ultimate gaming power to other people for megabucks. This is pretty much just fantasy, and there's no incentive for game devs to cater to that fantasy unless they're scammers seeking to quickly lure in easy marks.
|
|
#
¿
Nov 30, 2021 19:15
|
|
|
Owlofcreamcheese posted:It's honestly actually kinda surprising there isn't some indie game API to make some shared dumb database. Like dead cells now has weapons from hollownight, shovelknight is in everything, the crypto part would be pointless but "get a sword in this game, it shows up as a bonus skin in these other games" feels like it'd be a standardized thing by now. For that sort of game. There's not really much reason for a dev to give players bonus poo poo for buying games made by different devs. And there's even less reason to restrict it only to the smaller subset of players who've obtained a particular item or cleared a particular hurdle in that other game. As a general rule of thumb, an indie dev wants to devote their resources to making stuff for the people who bought their game and played it, instead of rewarding people who bought some other game and played it. It's been done occasionally, but there isn't enough demand to build a whole framework around it. It usually involves a $$$ cross-promotional marketing deal, not just something devs do for the hell of it.
|
|
#
¿
Dec 1, 2021 18:03
|
|
|
sinky posted:Oh no who could have foreseen this Motronic posted:Well that was quick. CommieGIR posted:Literally started hours after the announcement. Twitter is effectively dead as a way to expose the Alt-Right. I know that article confirms everyone's preconceived notions about the new Twitter changes this week, but it's two months old, and therefore doesn't have anything at all to do with this week's Twitter changes.
|
|
#
¿
Dec 2, 2021 23:34
|
|
|
Mercury_Storm posted:Like Crain, I also greatly disagree with this. Right wing propaganda is many times easier and cheaper to spread, and people seem to take to it like heroin. I would love if facts and logic would have an equal shot by default but this has not been the case. It's not that right-wing propaganda is easier to spread, it's that emotional clickbait that promotes a black-and-white view of the world and makes people furious is easy to spread, especially if it plays to existing distrust of systems and convinces people they know the secrets The Man isn't telling you. Left-wing propaganda that plays to the same themes spreads reasonably well. The problem is that leftist propaganda playing to those themes is less effective. Conservative ideologies calling for a maintenance of the status quo with only slight alterations, or a return to an idealized version of the past, fundamentally have an easier time spreading via vague messages of hate and distrust.
|
|
#
¿
Dec 3, 2021 21:05
|
|
|
Court documents have given us a glimpse at how hosed-up and abusable Youtube's automated copyright management system is: https://twitter.com/torrentfreak/status/1466728172918128645 Two men formed a fake music company, and managed to convince a company that handles Youtube copyright claims that they were the real publishers and copyright holders of over 50k music tracks. That company then helped them use Content ID to claim all monetization payments from anyone who used any of that music on Youtube - including the actual creators, owners, and copyright holders of that music. They pulled in more than $20 million of monetization payments that should have gone to the actual artists, and despite plenty of complaints to Youtube, were able to keep doing it for four years before any consequences appeared. Interestingly, there's no word of them getting booted from Youtube, nor is there any indication of whether Youtube's going to do to compensate the victims. The reaction is instead coming in the form of a grand jury indictment for various federal crimes (conspiracy, wire fraud, money laundering, and identity theft).
|
|
#
¿
Dec 3, 2021 22:42
|
|
|
WAR CRIME GIGOLO posted:You don't have to worry about that anyway. Franchises are under strict requirements to call themselves McDonald's. As I., Don't have X item on the menu screen by Tuesday? Your not a McDonald's anymore. Don't have all the signage updated to x standard? You lose your McDonald's name. That's about in line with restaurant profit margins in general. Paying real people to cook real food and serve it to real customers just has inherently crappy profit margins.
|
|
#
¿
Dec 8, 2021 07:46
|
|
|
There Bias Two posted:Does a car even come with a terms of service agreement that can be retroactively modified? In the case of Toyota, it wasn't retroactive. It's been in the fine print when you buy affected models all along, complete with the details and specifics of the free trial period. The article (dunno why it wasn't linked in the first place!!!!!) is playing it up because it's the apparently the only example they could find of a non-Tesla company charging for something software-locked that wasn't tied to the mobile app or the infotainment system... ...which suggests they didn't look very hard, because there's plenty of other instances of car companies software-locking something and making people pay to unlock it. For example, Mercedes limiting the rear-wheel-drive unless you buy a $500+/yr subscription is much more relevant to the article's fears of locking basic car performance behind subscriptions.
|
|
#
¿
Dec 15, 2021 16:41
|
|
|
PT6A posted:The one aspect I've not heard much about is... what are we doing to boost our electrical generation capacity to account for most vehicles being electric? We already have requests to save energy and/or brownouts whenever it's too hot or too cold, what happens when every petrol vehicle becomes electric? That's a more pressing issue than "where" in my opinion. Typically, most EVs charge at off-peak times when energy use is lower, not peak times when the grid is being pushed to its limit. So it's not as big an issue as you'd think.
|
|
#
¿
Dec 31, 2021 03:17
|
|
|
Baronash posted:I think some of your underlying assumptions are wrong. Online communities in the 90s and 00s were shaped by moderation. By the early 2000s, you even had tv shows making jokes about draconian modding of discussion forums. There were some notable exceptions, such as 4chan, but the pseudo-libertarian approach came much later. The mid-to-late 2000s is when you really started to see websites begin to abrogate their responsibility to moderate content on their platforms, with the rise of sites like Reddit (still moderated, but just at the community level with few site-wide restrictions) and the explosion of social media. Social media may have pre-dated Facebook and Twitter, but the News Feed/Twitter Feed concept was a radical departure from what came before, and it has shaped the platforms that have arisen since. I think it's even deeper than that - it's a shift to a decentralized, anything-goes siloed approach. It's an overall migration from a "this is my site and my community and I as the owner am responsible for keeping it clean" to "we're just a ~platform~, everyone can go make their own communities and run them how they want, and our job is mostly to keep those communities separate". It's a philosophy built around "don't like it, don't read it". It starts with stuff like Reddit which focuses heavily on dividing the site into completely separate subcommunities with no central oversight, and then comes the rise of algorithmic platforms built specifically to try to give everyone their own personal echo chamber where they only see what they want to see. It's not a clean-cut split, since there were platforms built around encouraging subcommunities from the start, but it's become far more mainstream now. That way of thinking has even spread into old-web mainstays; SA is certainly a lot more about "every forum and thread is its own separate little subcommunity" (complete with hands-off admins leaving them to self-police) than it was a decade ago. The thing is, I don't think all this was originally intended to encourage radicalization. The basic concept of showing people stuff they want to see regardless of their interests, instead of trying to pursue a single content stream with wide appeal, makes sense - especially in the heyday of stuff like Cracked Dot Com. And I'm sure that using algorithms to silo objectionable content so that only people who want to see it can see it, instead of using moderators to remove it completely, comes with double benefits since it not only cuts human moderation costs but also retains those objectionable groups as a niche audience that's easy to appeal to since they're rejected from most other outlets. The various feedback effects that have come to light since then, such as the high engagement rate of controversy and extremism causing the algorithm to preferentially push it, were probably unexpected...but as long as the money kept flowing, the platforms weren't especially bothered by it.
|
|
#
¿
Dec 31, 2021 05:40
|
|
|
PT6A posted:What happens when every car in America is scheduled to charge from midnight to 6, rather than a handful? Does our grid still handle that without problems, over a heatwave or a cold snap where heating/AC is still running overnight? The Texas grid problems were because their power generation equipment wasn't winterized and a significant portion of it froze, and also because Texas intentionally isolated their power grid from the rest of the country so they wouldn't be subject to federal regulation. Power generation will need to be scaled up somewhat to handle electric cars, but the ability to schedule charging to off-peak times when power usage is at its minimum will help a LOT, especially as electric car usage slowly ramps up.
|
|
#
¿
Dec 31, 2021 21:08
|
|
|
Tuxedo Gin posted:It's funny because they paid money for what amounts to a third party turning their bank statement into an Instagram "memories of the year" post but on the blockchain! and then, of course since this is crypto, the third party was a scammer and did a pump and dump It was also supposed to give people tokens based on how high the numbers on their "memories of the year" post were that year. The more active you were on NFT marketplaces, the more tokens you'd get. Which is a pretty obvious warning sign that someone's targeting whales, but hey, no one said crypto folks weren't gullible.
|
|
#
¿
Jan 1, 2022 01:49
|
|
|
Arsenic Lupin posted:https://twitter.com/ThePracticalDev/status/1477452084693942284 Unironically posting a joke article from a teenager's blog here as something to complain about sure is an interesting negative anecdote about social media and human nature, but the teenager isn't the one who comes out of it looking the worst.
|
|
#
¿
Jan 2, 2022 06:47
|
|
|
BiggerBoat posted:What would the thread say about tech's overall impact on the quality of news and journalism? Better, worse or the same? On the subject of journalism specifically... I'd say the overall impact is negative, because of a factor even more fundamental: the ability to know exactly how many people were interested enough to look at any given article, exactly how many people read through the whole article, and exactly how many people decided to send the article to someone else. Of course, it's been possible to kind of estimate that by looking at paper sales or TV viewership at particular times. But the data online is far more granular, and breaks down to the specific article level. That really has an impact on the incentive structure of news, driving things toward clickbait, misleading headlines, and controversial opinion pieces. Combine that with the ability to publish instantly at any time of the day, without even waiting for the 7pm news, and reporters are incentivized to rush in a constant battle for scoops and rumors that might get people stirred up. On the subject of news as a whole, it's a mixed bag. It breaks our dependence on well-funded mainstream news organizations, which can be both a blessing (see: much better info about police shootings) and a curse (conspiracy alt-media can click-farm even better than mainstream news).
|
|
#
¿
Jan 5, 2022 03:11
|
|
|
Mr. Fall Down Terror posted:the tunnels are short enough now as designed that they can be ventilated by just moving cars through them, and picking up fresh air at the stations which are only a few minutes stroll apart at best. there is one underground station and both other ends of the tunnel are open air Relying solely on the piston effect has important implications in the case of a service-stopping emergency. Let's take a look at what conditions look like after a British subway crash: quote:Platform 9 was 21 metres (70 ft) underground, and fire and ambulance crews had to carry all the equipment they needed through the station and down to the scene of the accident. The depth at which they were operating, and the shielding effect of the soil and concrete, meant their radios could not get through to the surface. Messages and requests for further supplies were passed by runners, which led to mistakes: one doctor requested further supplies of the pain-killing gas Entonox, but by the time the request reached the surface, it had been garbled to "the doctor wants an empty box".[38][39] The fire brigade deployed a small team with "Figaro", an experimental radio system that worked in deep locations.[34] Working conditions for the emergency services became increasingly difficult throughout the day.[34] The crash had thrown soot and dirt into the air from the sand drag, and from between the two metal layers of the tube carriages. Everything was covered with a thick layer of the residue which was easily disturbed.[40] The lamps and cutting gear used by the fire brigade raised the temperature to over 49 °C (120 °F) and oxygen levels began to drop. In the deep lines at Moorgate, ventilation is produced by the piston effect, created by trains forcing air through the tube lines. With services stopped since the crash, no fresh air was reaching platforms 9 and 10.[41] A large electric fan was placed at the top of the escalators in an attempt to remedy the situation, but soot and dirt was disturbed and little draught was created; the machine was soon turned off.[42] And that was at a subway station with an exit to the surface right there, in a system with larger tunnels than the Boring Company builds, and the train never caught fire.
|
|
#
¿
Jan 7, 2022 20:17
|
|
|
https://twitter.com/BleepinComputer/status/1480182019854327808 Some popular open source JS libraries run by Marak Squires were updated with seemingly malicious code that would cause any project using them to output "LIBERTY LIBERTY LIBERTY" and the American flag over and over again in an infinite loop, blocking other functionality. The natural assumption was that his accounts were hacked (compromises of widely-used libraries have been a growing problem in recent years), and Github suspended his account while NPM rolled the libraries back to previous known-good versions. So there's been some debate about the ethics of the platforms doing this. Do they have the right to lock him out of his accounts and roll back changes that his accounts made? However, a little digging around brings up a Github issue post from November 2020 titled No more free work from Marak - Pay Me or Fork This, in which he complains that corporations are using his freely available open-source libraries and not paying him a six-figure salary. Given that he's been on social media yet hasn't commented at all on the library compromises, people naturally moved to the assumption that the damaging code commits were intentional, perhaps a method of striking back against what he sees as ungrateful users of his free software. So that of course sparks a lively debate on the economics of free open source software, the ethics of people using free software in for-profit applications, whether the platforms have the right to prevent him from intentionally screwing over his users on his way out, and so on. So that kinda raises some interesting questions, overall! Probably the one I'm most interested in is one that's increasingly becoming a problem in the library-driven tech industry: if someone's made a very popular tool that's automatically incorporated into a lot of software, what happens if they decide to take their ball and go home? Does it matter whether they simply yank the package as the left_pad author did a few years back or push a change that actively sabotages users? What about selling the package to someone else, who may very well be a malicious software writer themselves? Does the platform it was published on have the right to undelete the package or rollback the breaking change? What if it's open-source software that anyone can fork or duplicate, does that change the answer to the previous question? It's a complex ethical landscape, overall. There's balancing programmer rights vs user rights (or whether they have any at all) vs platform rights, there's the question of how open sourcing (which involves the programmer giving up many of their rights and controls) affects things, there's the question of whether the considerable real-world harm makes a hardline "the users should suck it up" stance troublesome...there's really a lot there! As for some side context that isn't really relevant to the philosophical questions here, Marak first started complaining about unpaid use of his libraries shortly after a reclusive programmer and crypto investor named Marak Squires showed up in the news for accidentally setting his apartment on fire while building bombs. Rumor has it that his insurance refused to pay for the damage (I suspect bombmaking accidents aren't covered), and Marak claimed to have lost access to his gold stockpile in the aftermath of the fire.
|
|
#
¿
Jan 14, 2022 21:53
|
|
|
BiggerBoat posted:So am I in the minority thinking that huge TV screens and the ubiquity of cell phones are directly responsible for this? The spike they're talking about is one that began in 2020. In 2017-2019, the number of traffic fatalities fell each year, only to suddenly shoot up in 2020. And although we don't have 2021 full-year numbers yet, early indications are that the rise has continued. Economic shifts and the resulting social effects have a huge impact on the traffic fatality rate. The three top single-year declines in traffic fatalities since WWII were all in major recession years: 1974, 1982, and 2008. And the NHTSA itself credits the 2008 decline to economic factors: the younger generations who were most likely to lose their jobs in the Great Recession also tended to be the riskiest drivers, so when they lost their jobs and stopped driving as much, there were a lot fewer major accidents. quote:Examination of the 2007 and 2008 FARS data shows that the fatality declines were driven by significant reductions in fatalities in crashes involving younger drivers 16 to 24 years old. Also, there were significant reductions in the number of children 15 and younger killed in traffic crashes. The reductions in fatalities in crashes involving younger drivers manifest significantly in corresponding reductions in multiple-vehicle crashes and those occurring during the weekend. Of course, the 2020 pandemic has caused significantly different dynamics from a normal recession, but it's still not hard to see the impacts on driving behavior. There's still not a ton of good data, but early statistics suggest that the 16-24 age group is by far the least likely to be working from home, which skews traffic toward that riskiest group as older workers drop the daily commute and set up their home offices. The general reduction in road traffic also tends to cause riskier driving behaviors, as people feel more safe and confident in the less-crowded roads and drive more aggressively as a result. And while I can't find any studies on this yet, the massive rise in delivery orders (delivery app demand has more than doubled compared to pre-pandemic years) and online shopping is almost certainly a contributor. "Driver/sales workers and truck drivers", the BLS category that includes delivery workers of every kind, had one of the highest fatality rates even before the pandemic. Now there's way more demand for people to shuttle food and packages to people's front doorsteps in a hurry. So to sum it all up, the social and economic dynamics of COVID have taken the safest groups of drivers off the road, while putting more and more of the riskiest groups of drivers on the road to provide and deliver services to those safe drivers sitting in their home offices.
|
|
#
¿
Jan 27, 2022 18:17
|
|
|
BiggerBoat posted:I could very well be mistaken but I was surprised that the ubiquitous use of cell phones while driving wasn't even mentioned or looked at, either way, in those studies. My "data" is not empirical in any way but my day to day anecdotal observations reveal a LOT of people drifting into bike lanes, crossing into mine and having to make very fast stops at intersections. I'd estimate the % of "distracted drivers" at (conservatively) somewhere around 20 - 25% and these people seem even easier to spot than drunk drivers are. The way they drive is very similar. Here's the thing: cellphones have been around for a fair few years at this point, so it's very unlikely that they were responsible for a sudden massive spike that started in 2020. There's plenty of people studying the use of cellphones and their impact on distracted driving, but that's a longer-term trend that didn't start the year COVID came around.
|
|
#
¿
Jan 27, 2022 21:44
|
|
|
CommieGIR posted:Speaking of NFTs: Generally if someone's a racist neo-Nazi, you can point to instances of them actually being racist. This sort of symbology-diving stuff can occasionally be useful to draw further insight into already-questionable behavior, but it's not very useful in making the case all on its own. When an item like this is in the top of the list, it feels like they don't really have a lot to go on. quote:4. The other co-founder goes by Gordon Goner, who says he picked the name because it sounds like "Joey Ramone".. being that the group frequently uses anagrams and that it doesn't at all sound like Joey Ramone, maybe this is an anagram. Sure enough its an anagram with only one solution, Drongo Negro, Drongo is common 4chan and Australian slang for stupid, its in the dictionary as such, second definition.. So, his name means "Stupid Negro." The probability of picking a name that is also a racist anagram is about 1 in 750 billion. Writers frequently use anagrams for character names and they are also often incorporated within videogames, some of which Yuga Labs have stated they play. At the very least, it's dubious enough that it feels like a distraction from the many better-supported reasons to hate NFTs. And there's plenty of obviously-racist NFTs out there, like Floydies.
|
|
#
¿
Jan 28, 2022 22:28
|
|
|
Doggles posted:Using suicide hotline data to help train customer service chatbots to be more sympathetic. Even worse than that - the for-profit company was founded by the suicide hotline company as a spinoff, for the express purpose of using the gathered data for for-profit usage. https://www.crisistextline.org/blog/2018/03/12/what-is-loris-ai/ quote:Over the past five years, we’ve trained 12,000+ people to demonstrate empathy and compassion in conversations with people in crisis. And they’re great at it! Our self-reported texter satisfaction rate is 85%. Crisis Counselors tell us they use these skills in other ways too–in their family life, work life, etc.
|
|
#
¿
Jan 29, 2022 21:32
|
|
|
BiggerBoat posted:Why does a dishwasher need to be connected to the internet? I've somehow managed without it my entire life. What if you load your dishwasher in the morning, and then forget to turn it on? What if you don't know what water temperature and wash intensity settings to use, and need an app to tell you? Do you want a dishwasher that's smart enough to turn itself off when it's done washing your dishes? By golly, Home Connect is the dishwasher for you! https://www.home-connect.com/global/smart-home-appliances/dishwashers quote:Did you forget to turn on the dishwasher in the morning? It's no problem thanks to Home Connect. If you have activated remote start, just send the command through the app to your smart Home Connect dishwasher while you are already sitting at the office. Your clean dishes will be waiting for you when you come home. Your dishwasher will shut itself off automatically after its job is done. ...in other words, it doesn't really do anything genuinely useful or interesting, but Wi-Fi chips are cheap and a lot of marketing buzzwords can be implemented in software.
|
|
#
¿
Jan 31, 2022 20:48
|
|
|
withak posted:Sorry to interrupt the valuable IT guy website service planning discussion, but The Algorithm has started recommending products for suicidal people: Looks like the algorithm hasn't identified a product that people are using to commit suicide. It's just noticing that a fair number of people buy certain things together, and suggesting that anyone who buys one of those things should also buy the other things. The algorithm has no idea these things are being used for suicide, and it's not treating them specially at all. Which is a growing problem in general: most of these algorithms don't have any smarts at all, they're highly generic and treat everything equally, so they're incapable of accounting for real-world situations. That ranges from petty inconveniences, like buying a TV and getting bombarded with TV ads and recommendations for weeks afterward, to serious issues like Facebook ads' repeated discrimination scandals because they didn't think to exclude race from the auto-generated list of categories advertisers could target with. Of course, the real problem in this case isn't the algorithm at all. It's that the human managers at Amazon are consciously choosing to keep the product listed and subject to the system's normal behaviors, even after being repeatedly warned of the issue.
|
|
#
¿
Feb 6, 2022 18:45
|
|
|
Owlofcreamcheese posted:I feel like a huge advantage of a machine ordering system is that it could inherently speak any language the software supports if they wanted to. Like that feels like a big deal. Being able to just roll up only speaking french or spanish or chinese and expecting to have a way to interact with a drive through. Ordering food in a language you don't speak is a charming magical experience like 8-10 times that grows extremely frustrating rapidly. Speaking to siri in your native language would be a thousand times easier for a lot of people that trying to pantomime "cheese burger" to a bored teen Is there any voice recognition software that not only supports multiple languages, but just automatically detects what language someone is speaking instead of needing to be explicitly set to a language? Given the current state of the art, I kinda doubt it. I don't have an iPhone, but I'm pretty sure you can't just roll up and start speaking French to Siri. You have to go into a menu and change its language setting first.
|
|
#
¿
Feb 17, 2022 13:58
|
|
|
BigRed0427 posted:Wait, what the hell did i mess? Never thought I'd see the day Elon Musk deleted a tweet. Sad!  And I just remembered that, per the terms of a settlement with the SEC, Musk is supposed to get all his tweets approved by a Tesla lawyer before making them. That's very obviously not happening, and the SEC has so far failed to successfully enforce it in court (though it appears that the SEC still hasn't given up hope on that), but it's still funny to think about the CEO of Tesla running his Hitler meme tweets past a company lawyer.
|
|
#
¿
Feb 17, 2022 20:34
|
|
|
|
| # ¿ May 21, 2024 07:15 |
|
|
Mister Facetious posted:you're not making songs with a singer that sounds like them, you're taking their actual voices as source samples, and using machine learning to copy the sound/style. Is that really any different from having a real human engaging in voice training to make their voice sound more like them?
|
|
#
¿
Feb 22, 2022 05:19
|
|