irotsoma

@irotsoma@piefed.blahaj.zone

• ⁨Comment⁩ on ⁨Flock is canceled—and its competitors are profiting⁩ ⁨⁨5⁩ ⁨days⁩ ago⁩:

  Yeah, need to find a way to go after the government to stop using the tech rather than going after a single company. Eventually, Flock will just go bankrupt, the pieces will get bought up and either sold to a competitor or the VC that buys them will change the name and start a new company that’s basically the same and the executives from Flock will join that or go to competitors. There’s not much use attacking a single company these days. But it is good that attention is being brought to the technology in use and how the government is using it.

• ⁨Comment⁩ on ⁨Adobe to offer free access to AI tools in Saudi Arabia in $4 billion deal⁩ ⁨⁨1⁩ ⁨week⁩ ago⁩:

  I mean all of their products are subscription based and this is only a 1 year subscription. They’re just giving away a free trial to open up a new market is all. Not really big news. If they were making it free for the life of the product, that might be a bigger deal.

  But yeah, also F* SA.

• ⁨Comment⁩ on ⁨Employee revolt and failing agents forced Meta to scrap its AI layoff plan⁩ ⁨⁨1⁩ ⁨week⁩ ago⁩:

  Wish my company held off. So few with any experience left and “AI” not only isn’t going to replace an experienced engineer, only junior ones really, but the things I work on are proprietary and so “AI” has no way to understand them much less interact with them or deal with the proprietary devops, upgrades, migrations, etc. There’s basically 4 people left doing the jobs that used to belong to 20. Sure “AI” is a useful timesaver oncenin a while, but it can’t replace thinking, actually intelligent people, much less all of their experience that’s not documented anywhere for “AI” to even learn it.

• ⁨Comment⁩ on ⁨As Ubuntu embraces AI, Debian discusses banning all AI-generated code⁩ ⁨⁨1⁩ ⁨week⁩ ago⁩:

  Yeah and when teachers or employers use it to reject things outright, it’s really dumb. But if it’s presented as a probability to the user rather than a binary like most LLM results are presented to users, then it is useful. If something has a high probability of being LLM generated then it can cast suspicion and investigation can happen. Unlike how things like facial recognition are used to prove guilt without any investigation when the probability of a match is low, but no better matches were found and the police aren’t told it’s low probability because then the product would look useless in solving crimes for them which is what they want in a product, not an investigation tool. That’s not what they were sold.

  Point being. It’s a tool. It’s a metric. And used as such it can be helpful if all of a developer’s contributions get flagged as high probability, then looking more closely at the code can be warranted. If only LLMs themselves were configured to tell the user how probable that their answer is correct, it wouldn’t even be an issue needing investigation.

• ⁨Comment⁩ on ⁨As Ubuntu embraces AI, Debian discusses banning all AI-generated code⁩ ⁨⁨2⁩ ⁨weeks⁩ ago⁩:

  There are “signatures” that can be detected in code just like universities detect it in student work, online stores detect it in reviews they compensate people for, and some employers detect it in resumes and cover letters.

  That said it doesn’t detect it all, and it does rely on the developer to be honest. But it’s also a good indicator when reviewers post questions to the developer and they can’t answer them because they don’t understand the code and things like that. Also, if the developer did not follow coding standards, which LLMs can sometimes have trouble with. And there’s always been malicious code submissions in open source software, so there are other ways.

  Anyway, it’s more of a policy discussion at this point, and why LLM generated code is bad. I use them for manipulating data, figuring out syntax, and getting examples, but there are always problems with anything they generate, so I never would allow one to write any production bound code. I see use in the technology, but it’s not creative and it doesn’t understand what it’s writing so it can’t write code with consequences in mind. So it can write Jr level code, but not solid code, at least not without being properly trained on only well written code, and that proper training would be too expensive for an LLM company to actually do it.

• ⁨Comment⁩ on ⁨Amazon shuts down Mechanical Turk: Why Jeff Bezos once called it ‘artificial artificial intelligence’⁩ ⁨⁨2⁩ ⁨weeks⁩ ago⁩:

  Yeah, gotta keep your employees poor so they can’t afford to leave no matter how badly they’re treated.

• ⁨Comment⁩ on ⁨Amazon shuts down Mechanical Turk: Why Jeff Bezos once called it ‘artificial artificial intelligence’⁩ ⁨⁨2⁩ ⁨weeks⁩ ago⁩:

  Too bad LLM “AI” can’t pick up the slack which is the whole reason it was still around and so active.

• ⁨Comment⁩ on ⁨U.S. Citizen Who Deleted Phone’s Data Says His Prosecution Puts Privacy at Risk⁩ ⁨⁨2⁩ ⁨weeks⁩ ago⁩:

  Oh, yeah, but at least the courts would intervene back then. Now many of the courts have been corrupted and the ones that haven’t are bound by precedent which the supreme court no longer obeys and courts above them that are corrupted. So these specific interpretations are becoming more and more common and not just against poor people who can’t afford lawyers.

• ⁨Comment⁩ on ⁨Ads and tracking infiltrated TVs. Now they're coming for monitors.⁩ ⁨⁨2⁩ ⁨weeks⁩ ago⁩:

  I got a smart TV because the dumb ones were more expensive. But I’d never do that with a monitor. A monitor is not meant to be interacted with directly. Just display what the computer says to display. The computer is all that should be running software besides basic firmware for adjusting overall settings like brightness and contrast which computer OSes don’t have enough information to do most of the time. But that’s it.

• ⁨Comment⁩ on ⁨U.S. Citizen Who Deleted Phone’s Data Says His Prosecution Puts Privacy at Risk⁩ ⁨⁨2⁩ ⁨weeks⁩ ago⁩:

  The current administration has stated that even for US Citizens, border entry suspends all constitutional rights. They’ve also said they believe this applies to allow warrantless searches within 100 miles of a border, which includes most major US cities, thus allowing ICE to search and kidnap basically anyone, not just immigrants. And they have stated that searching an electronic device is not a “search” as mentioned in the 4th Amendment and so you can’t refuse to unlock the device or “destroy evidence”. And that once you give them that access, that you have also “freely given” them access to your private social media, bank, etc., records and so those no longer require a warrant to access. And it’s “gun to the head” tactics because they also believe they can detain you indefinitely in the 100 mile zone without a lawyer or any other usual rights, because you aren’t being arrested, just an extended interview or whatever they call it. All in the name of terrorism prevention, of course.

  Since they have corrupted the Supreme Court, this has been mostly upheld in various court cases so far. But any legal scholar will tell you it’s nonsense. But that doesn’t matter to authoritarians. They don’t work within the law.

• ⁨Comment⁩ on ⁨The FTC is cracking down on companies that charge you a "personalized price"⁩ ⁨⁨2⁩ ⁨weeks⁩ ago⁩:

  Disclosure isn’t “cracking down”. I think it’s becoming common knowledge anyway or the FTC would never even go that far, especially the current administration.

• ⁨Comment⁩ on ⁨Take-Two files subpoenas asking Microsoft and Discord for records in hunt for GTA 6 leaker⁩ ⁨⁨2⁩ ⁨weeks⁩ ago⁩:

  Instead of spending money on improving the product or mitigating the impact technologically.

• ⁨Comment⁩ on ⁨One In Five Drivers Switches Off The Safety Tech Meant To Save Their Lives⁩ ⁨⁨2⁩ ⁨weeks⁩ ago⁩:

  For me at least, and I love cool tech, so it has to be broken or too costly (in terms of money or privacy) for me not to use it:

  Adaptive cruise control: I live in a medium sized city and there’s rarely a case where I don’t have to be extremely quick to react to people who are aggressive and dangerous, so I need to change speed way too often for it to be useful. But even driving outside the city it’s rare that there are any significant stretches of road where there aren’t people camping in the left lane and causing lots of big groups of cars piling up together. And many highways don’t have enough long stretches of straight road. So it’s rare to be able to drive long enough stretches where I’m of constantly turning it off and on.

  Lane keeping: mine only works over 45 and so only on highways outside the city mostly or for short stretches inside the city. I do turn it on most of the time. But it has one really major flaw. If you’re driving in the right lane and come to an exit and there’s even the slightest bend in the road, it’s about a 50/50 chance that it’s not going to try to pull me off the exit and scream at me when I stop it from doing that. Also, the safety feature that determines the person has their hands on the wheel gets tripped if you hold the wheel in one place for too long. In the places I most would find it useful like long stretches of straight highway, it’s constantly complaining and shutting itself off thinking I’m not holding the wheel just because I’m going straight the whole time.

  Automatic braking: mine doesn’t actually brake, just screams at you and flashed a huge warning. Problem is it triggers when someone is turning and will be out of my way significantly before I get there because I don’t tailgate, so there’s no need to slow. Or it will suddenly trigger when there’s an odd narrow shadow across the road from something above the road like a pedestrian bridge, and it mistakes that for a car or something. Haven’t figured that one out. It’s not too common, but the fact that the car has both camera and radar my wild guess is on hotter days, the areas of asfalt in the shade are giving off less heat and possibly the air above has slightly different density than the areas in the sun. Wild guess though, but if the car did have auto braking, it would be braking randomly for no reason and causing accidents or near accidents too often.

  Voice assistants: screw LLMs

  Augmented reality: I don’t have that, but I might understand that being a distraction. I’m ADHD so I’m not sure how that would affect my distraction especially commuting when I’m too tired or not enough stimulants in my system.

  Massaging seats seem possibly distracting or too relaxing which then causes distraction. I don’t have it, but font blame buying a car with some feature and not using it since a lot of cars have packages of features and it might not be an option to not have that but still have some other useful feature like heated seats. I do love my heated seats in winter so if it was both or neither I’d choose both and ignore the massage feature.

  Remote start: I’d be afraid of opening it up to theft vulnerabilities and I don’t like wasting gas and generating pollution for some minor convenience.

  Backup cameras: I use mine in the daylight, but there’s no night vision so the quality is really poor at night if there’s not street lighting or if there’s too much glare from a bright light, basically useless in a lot of conditions.

  Driving Mode: I think this refers to the modes that change the sensitivity of brakes, acceleration, power steering, struts, etc. My partner’s car has 3 modes and I can definitely see switching being distracting. It would be like getting used to driving three different cars. Just choose one and use that. Might be useful if it’s an off road mode that increases the shock absorption and gives more torque while off road and gives more “feel of the road” while on well paves roads.

• ⁨Comment⁩ on ⁨Scientist says RAM pricing has risen to normalized 2007 levels, AI shortage undid 20 years of progress in a matter of months — memory prices had been falling exponentially for decades⁩ ⁨⁨4⁩ ⁨weeks⁩ ago⁩:

  And it’s all fake since neither the demand nor the supply exists, it’s all future promises to buy/sell what get manufactured. As soon as the data centers collapse and cancel their orders, the prices will plummet if no new fad takes it’s place. One of many ripples this bubble pop is going to trigger since so much is based on lies and promises not supply and demand.

• ⁨Comment⁩ on ⁨Is Myspace coming back? Owners confirm relaunch plans⁩ ⁨⁨5⁩ ⁨weeks⁩ ago⁩:

  It still existed, just not used by many people for a long time. Probably mostly just bots talking to each other. 🤣

• ⁨Comment⁩ on ⁨Copyright Is Not Enough: Why writers need new strategies to stand up to AI.⁩ ⁨⁨1⁩ ⁨month⁩ ago⁩:

  Copyright hasn’t benefitted creators in centuries, it’s for publishers who then gatekeep the money from the vast majority of creators.

  But even then it was never really meant to be a “welfare system” for creators, but a way to incentivize creators to keep creating by allowing them a short term boost in income while they create the next thing. Create one popular thing and never have to create again as well as your kids and grandkids never having to contribute to society was never meant to be the point.

• ⁨Comment⁩ on ⁨ICE shared Medicaid data it wasn't supposed to have with Palantir⁩ ⁨⁨1⁩ ⁨month⁩ ago⁩:

  Oh I totally agree that ICE should not be allowed to get the information no matter what. Just saying there are several ways to get medicaid without having an SSN and sometimes even being in the country “illegally” and this is why they want the info. They don’t care about those special situations like an emergency or cancer or being an abuse victim or other exceptions, they want any means to find anyone they can deport even if it’s extremely cruel to do so. So yeah, for anyone they deported from those lists it was especially cruel to do so and nothing can justify giving them the info in the first place, but especially knowing the extreme situations the people that would be on those lists would have endured, it was beyond unconscionable.

• ⁨Comment⁩ on ⁨ICE shared Medicaid data it wasn't supposed to have with Palantir⁩ ⁨⁨1⁩ ⁨month⁩ ago⁩:

  Depends on the state. Some check immigration status and some don’t. And situations like for children they don’t require as much information since there’s not a work requirement and they don’t have driver’s licenses or other identification requirements that might also require legal status to get. It’s better for a community to have as many people as possible, especially underprivileged children, having healthcare so that diseases don’t spread as quickly in a community.

• ⁨Comment⁩ on ⁨Suspecting AI cheating, Ivy League prof ordered an in-person final; scores fell 50%⁩ ⁨⁨2⁩ ⁨months⁩ ago⁩:

  It’s mostly just a piece of paper required for most white collar jobs to even pass on your resume to a real person. That’s the only reason I got a degree. I already had tons of experience, but couldn’t get a decent full-time job because the job postings all had a degree as a requirement and the automated systems take that as a hard requirement because it’s objective and thus easy to filter on.

• ⁨Comment⁩ on ⁨GTA 6 is a 'singleplayer experience,' GTA Online may be sold separately⁩ ⁨⁨2⁩ ⁨months⁩ ago⁩:

  I wish there were more games that were fully complete single player offline games with online multiplayer being an afterthought. I miss a good story and experience that isn’t grind for years or pay to continue.

• ⁨Comment⁩ on ⁨Hackers duped Meta AI support chatbot to steal celebrity Instagram accounts⁩ ⁨⁨3⁩ ⁨months⁩ ago⁩:

  They think LLMs are more intelligent than people because the LLMs tell them they are. So of course most companies who bought into “AI” have laid off staff, especially the most experienced and thus expensive ones, and have handed responsibilities to untrained contractors armed with “AI”. So yeah, most companies who did that are going to be falling apart over the next few years, including the one I work for. I was only safe, so far, because I’m grossly underpaid already.

• ⁨Comment⁩ on ⁨⁩ ⁨⁨3⁩ ⁨months⁩ ago⁩:

  Damn I hate the chatbots though.

  Just yesterday I was on a site to get some routine lab test results. There is no link for results on the hamburger menu or the top of the home page. There are a few links for things further down, one of which is “Test Results”. But those all open the chat bot and ask it instead of just going to the page. However, it just offers some FAQs that I could find just as easily on the FAQ page. But it doesn’t even link to any pages no matter what options you select whole chatting. I finally gave up and clicked the “Dashboard” link in the hamburger menu which I had originally assumed meant the homepage hoping I just missed something.

  Turns out the “dashboard” is a separate page, and I still don’t know if that’s where results go because it just asked me to verify my identity and despite completing the identity verification successfully (twice) it still keeps saying I’m not verified.

  So yeah, I completed my visit to the site while having accessed the chat bot for help and not interacting with a human, so it “solved” the issue in the sense that I no longer had any desire to get any additional information from them and just waited for my doctor to send the results. I was also never presented with an option to contact a human, nor would I consider it worth waiting on hold to call a call center to find the core info that the site should he helping people get.

  So, poor web design and broken services both made me use the chatbot unwillingly and made me leave the site without human interaction, so I bet they’d consider it a win for them.

• ⁨Comment⁩ on ⁨Gen Z Sabotaging AI at Work So It Won't Take Their Job⁩ ⁨⁨4⁩ ⁨months⁩ ago⁩:

  Don’t need to sabotage a “worker” who has been trained using 4Chan and Reddit. And refusing to use the tech is often because it does the work wrong and the human has to redo it anyway.

  I do use it for inline coding suggestions because it’s required, but almost never accept a line of code as-is, because there’s usually some mistake, subtle or otherwise. It does help me not have to google syntax sometimes. But the non-"AI” code suggestions used to do that just fine in the past, so it’s not much of an improvement. And I’d never let it write more than one or two lines at a time because that would mean debugging code I didn’t write which is much more difficult that writing your own code for most experienced coders.

• ⁨Comment⁩ on ⁨Solar is winning the energy race - The world’s cheapest power source is scaling at warp speed, pushing coal, gas and nuclear aside⁩ ⁨⁨5⁩ ⁨months⁩ ago⁩:

  Yeah, but most of the data centers recently brought online to feed the LLM/"AI” bubble have triggered a bunch of retired coal plants to be restarted as well as old “dirty” nuclear plant that generate fissile-material for the new nuclear weapons Trump ordered built and other nuclear waste that we already dont have anywhere to store longterm. Part of the excuse being that the demand of these centers is too volatile for green energy. Plus Musk and Trump killing off the programs to build a network of car charging stations mean electric car production for the US market has been drastically cut despite gains in other countries. And cutting the incentives for heat pumps and replacing natural gas furnaces and water heaters has reduced the boom that heat pumps were having as well as are having elsewhere.

  And the general public believes that natural gas in homes and gasoline in cars is cheaper than electric although that is not true, it’s only that

  Anyway, more “dirty” energy sources are in use than a few years ago, do any gains in clean energy have been outpaced significantly by increases in use of dirty energy in the US, though that isn’t the case in many other countries like China and many EU countries without such large tax subsidies for the general public to consume fossil fuels more cheaply out of pocket.

• ⁨Comment⁩ on ⁨Solar is winning the energy race - The world’s cheapest power source is scaling at warp speed, pushing coal, gas and nuclear aside⁩ ⁨⁨5⁩ ⁨months⁩ ago⁩:

  Too bad the US is owned by fossil fuel companies and weapons manufacturers who make money on wars fought over said fossil fuels. If we could at least eliminate the subsidies from taxes, then people might actually see how much more it costs. But as long as taxes pay for fossil fuel production, it will always appear cheaper to those not paying attention.

• ⁨Comment⁩ on ⁨Number of AI chatbots ignoring human instructions increasing, study says⁩ ⁨⁨5⁩ ⁨months⁩ ago⁩:

  More like training it wrong. It is just a mimicking engine, not intelligent. If it’s trained on data that includes bad information, it will periodically include that bad information.

  Also, wrong settings. Increasing the threshold of confidence in something before it presents it to the user would at least partly increase the accuracy, but also increase how often it would say it doesn’t know how to do something. And for corporate executives, admitting complete ignorance is unfathomable, so of course they don’t want their products admitting it.

• ⁨Comment⁩ on ⁨Windows 11 is finally getting a movable taskbar⁩ ⁨⁨5⁩ ⁨months⁩ ago⁩:

  Thank goodness. I often have to remote into remote servers to do work, it’s so confusing having stacks of task bars on the bottom. It’s been driving me insane since my company upgraded Windows.

• ⁨Comment⁩ on ⁨Nintendo sues to prevent Trump from dodging full tariff refunds⁩ ⁨⁨5⁩ ⁨months⁩ ago⁩:

  Yep, Americans pay more for products and then pay interest and court costs to companies. Those tariffs were such a genius idea, especially implementing them on a whim without knowing how tariffs work or how to legally implement them, right?

• ⁨Comment⁩ on ⁨Workers report watching Ray-Ban Meta-shot footage of people using the bathroom⁩ ⁨⁨5⁩ ⁨months⁩ ago⁩:

  Yeah, devices like this really should be designed to do most work locally or on connected phones and only send data to the cloud when necessary. But then they wouldn’t be able to farm the data for tracking, advertising, and “AI” training, which is where they make their money on these kinds of devices.

• ⁨Comment⁩ on ⁨GrapheneOS Collaboration With Motorola Mobility⁩ ⁨⁨6⁩ ⁨months⁩ ago⁩:

  That’s awesome considering Pixel devices may not be viable for much longer as Google has been making more and more code proprietary. Hope it works out!