tauon's recent activity

  1. Comment on Cards Against Humanity pays you to give a shit in ~misc

    tauon
    Link Parent
    They’re not paying people to vote for a candidate, they’re paying people to vote at all. Which in my understanding makes this okay. Technically someone could go out and vote for Trump after having...

    They’re not paying people to vote for a candidate, they’re paying people to vote at all. Which in my understanding makes this okay. Technically someone could go out and vote for Trump after having received their money, but given the pre-selection process and publicity it involves, I doubt many would go through that.

    From the bottom of the website:

    Paid for by Cards Against Humanity PAC. Not authorized by any candidate or candidate’s committee.

    10 votes

  2. Comment on US DOJ indicates it’s considering Google breakup following monopoly ruling in ~tech

    tauon
    (edited )
    Link Parent
    Or we do go back to the concept of paying for software one uses, if it’s too big of a task, in both complexity and time/resources necessary, for open-source volunteers. (I really like your concept...

    Or we do go back to the concept of paying for software one uses, if it’s too big of a task, in both complexity and time/resources necessary, for open-source volunteers. (I really like your concept though – I just don’t think it’ll become a reality within this or the next few generations sadly)

    Kagi’s Orion browser (beware: currently only available on Apple devices) is free to use, with an optional “subscription” fee or one-time purchase model that’s really a donation to the development effort. Not sure if it is profitable/sustainable yet, though.

    Edit: typo

    6 votes

  3. Comment on OpenAI is a bad business in ~tech

    tauon
    Link
    After recent discussion on how AI companies may use your data for profit, here is a breakdown of OpenAI’s financials… And while I knew they weren’t profitable yet, I wasn’t aware it’s quite that...

    After recent discussion on how AI companies may use your data for profit, here is a breakdown of OpenAI’s financials… And while I knew they weren’t profitable yet, I wasn’t aware it’s quite that bad, just a guesstimate feeling of “sure, every aspect of LLMs is expensive“.

    It’s a great read, but quite long if you’re only “adjacently interested“ in the topic, so here are some excerpts. Each separated quotation block means I left out a […] for the sake of reading flow:

    OpenAI's monthly revenue hit $300 million in August, and the company expects to make $3.7 billion in revenue this year (the company will, as mentioned, lose $5 billion anyway), yet the company says that it expects to make $11.6 billion in 2025 and $100 billion by 2029, a statement so egregious that I am surprised it's not some kind of financial crime to say it out loud. For some context, Microsoft makes about $250 billion a year, Google about $300 billion a year, and Apple about $400 billion a year. To be abundantly clear, as it stands, OpenAI currently spends $2.35 to make $1.

    (double emphasis not mine, but accurate)

    Collectively, this means that OpenAI — the most popular company in the industry — can only convert about 3% of its users.
    While there's a chance that OpenAI could have a chunk of users that aren't particularly active, one cannot run a business based on selling stuff you hope that people won't use.
    OpenAI's primary revenue source is one of the most easily-commoditized things in the world — a Large Language Model in a web browser — and its competitor is Mark Zuckerberg, a petty king with a huge warchest that can never, ever be fired, even with significant investor pressure. Even if that wasn't the case, the premium product that OpenAI sells is far from endearing, still looking for a killer app a year-and-a-half into its existence, with its biggest competitor being the free version of ChatGPT.
    And so, [OpenAI] has two options [for the necessary growth]: Either it relies on partnerships and external sales channels, allowing it to potentially increase the gross number of customers, but at the expense of the money it makes, or it can build a proper sales and marketing team.
    Both options kinda suck. The latter option also promises to be expensive, costly, and has no guarantees of success.
    Let’s go back to Twilio — a company that makes it easy to send SMS messages and push notifications. Over the past quarter, it made around $1bn in revenue. That’s what OpenAI made from renting out its models/APIs over the past year. Twilio also made roughly $4bn over the past four quarters — which is more than OpenAI’s projected revenue for the entirety of 2024. OpenAI, I remind you, is the most hyped company in tech right now, and it’s aiming for a $150bn valuation. Twilio’s market cap is, at the time of writing, just under $10bn.

    And I cannot express enough how bad a sign it is that its cloud business is so thin. The largest player in the supposedly most important industry ever can only scrounge together $1 billion in annual revenue selling access to the most well-known model in the industry. This suggests a fundamental weakness in the revenue model behind GPT, as well as a fundamental weakness in the generative artificial intelligence market writ large. If OpenAI cannot make more than a billion dollars of revenue off of this, then it’s fair to assume that there is either a lack of interest from developers or a lack of interest from the consumers those developers are serving.

    Around the halfway mark, before some of the above had even been mentioned, the thing that spoke to me the most:

    A note on “free” products: Some of you may suggest that OpenAI having 350 million free users may be a good sign, likely comparing it to the early days of Facebook, or Google. It’s really important to note how different ChatGPT is to those products. While Facebook and Google had cloud infrastructure costs, they were dramatically lower than OpenAI’s, and both Facebook and Google had (and have) immediate ways to monetize free users.

    Both Meta and Google monetize free users through advertising that is informed by their actions on the platform, which involves the user continually feeding the company information about their preferences based on their browsing habits across their platforms. As a result, a “free” user is quite valuable to these companies, and becomes more so as they interact with the platform more.

    This isn’t really the case with OpenAI. Each free user of ChatGPT is, at best, a person that can be converted into a paying user. While OpenAI can use their inputs as potential training data, that’s infinitesimal value compared to operating costs. Unlike Facebook and Google, ChatGPT’s most frequent free users actually become less valuable over time, and become a burden on a system that already burns money.

    A business model that has costs scaling with increased freemium user counts is nothing new, but it seems like their freemium cost is higher than the expected revenue a paying user will bring in, and that's definitely not how Big Tech got big. Adding on users in other tech companies is almost marginal (cost of storage and some computation) – here, the net value of adding a new user is a loss. (There’s a neat back of the envelope calculation in the blog post comments comparing OpenAI to Dropbox’s free users.)

    This business is a fundamentally different one, even if the product/service being sold had already proven itself over time in a big market.
    Once the hype starts to cool down (not die off – just relax a little), I’m curious to see what will happen to all the funding that they so desperately need.

    30 votes

  5. Comment on Your chatbot transcripts may be a gold mine for AI companies (gifted link) in ~tech

    tauon
    (edited )
    Link Parent
    To be 100% clear, I don’t know whether they specifically have done that. Didn’t mean to imply it, either. Let’s stick to the facts we do know, and my bad for the implication. While I agree, AI...

    If this is true, then it is indeed scummy that they've taken to selling data gleaned from user inputs given by API customers, and they deserve scorn and possibly legal action.

    To be 100% clear, I don’t know whether they specifically have done that. Didn’t mean to imply it, either. Let’s stick to the facts we do know, and my bad for the implication.

    None of the rest of what you said is specific to AI at all, though. This is all stuff we've needed to grapple with regarding the Internet and Big Data as a whole for decades. The only reason to single out AI companies for these things is because AI is the Thing to Be Outraged About du Jour™.

    While I agree, AI companies are unique in this so far as they don’t inherently print cash, or turn profits at all, really. It’s not even clear if what their business model offers is truly needed at this point.
    I recently started reading an excellent blog post/article (which come to think of it, I’m probably going to post here once I finish it – a bit lengthy) that basically blasts OpenAI’s business as one that will not be profitable for the foreseeable future. (Edit: now posted here)

    My point is this: If the regular business avenues don’t suffice, they may be more tempted to do something “monetizable” with user data than they would be otherwise.

    Outside the limits of any warranties they've given promising privileging the inputs of paid API users, we should have no expectations at all regarding how they may choose to use the data we freely give them. The assumption of privileged communications is generally legally reserved for very few interactions, such as those between medical practitioners or legal representatives and clients. It exists as a principle in law and medicine because the withholding of information undercuts the purpose of those endeavors entirely, and it's extremely difficult to regulate even in fields like those that do have strict licensure.

    This is your opinion based on probably the legal framework you feel at home with as well as a bunch of other factors in lived experiences. I have a completely different world view in this regard (and to be clear, that’s fine! We don’t have to be the same person :-)) – I’m a stark proponent of “all personal data is privileged* unless cleared for use otherwise”

    *Privileged specifically meaning, at a minimum, knowing what data is out there of you, and potentially having the option to delete it. Not necessarily that it needs to be protected as strictly as e.g. medical info, I’m content with a “right to know” and “right to delete” (similar to what EU legislation actually ensures). But the key point is that I would classify my conversations as “my” data in some form or another.

    I just don't see why we should expect the vast majority of Internet interactions to be privileged like that. It's as though people walk into a store marked "Information Merchants," ask a few questions regarding the wares, strike up some chit chat about the crazy thing their cousin did the other week, and then get all shocked Pikachu when the information merchants turn around and sell off any info from their conversation they think might be worth something. More than that, it's like doing this same thing repeatedly for 30+ years every time they change the coat of paint on the storefront.

    If that were known upfront, sure. But people don’t enter the (really liking this analogy, BTW) Information Merchants’ brick-and-mortar grounds, they use a business model that sounds like Look Something Up or Connect with Your Friends For Free, but it happens online, so in the process they are able to share data they don’t even know they possess.

    We definitely need to be devising workable rules regarding inferential data collection, data retention, data security, and transparency about how data is collected. Information that we freely give, literally unprompted, with no warranties of privilege, that's something else entirely.

    Isn’t this the same pair of shoes? Can data security be guaranteed transparently if some LLM is trained on your innermost thoughts? Would you assume a friend or even acquaintance (as that’s how people might use these chat bots) gives “no warranties of privilege” in IRL or digital conversation, or would they assume a default mode of “hey maybe they won’t be sharing my thoughts with dozens/hundreds of employees and potentially millions of people”?

    Either way, it’ll be a tricky topic and regulators worldwide will, like you said, probably have a bit of a hard time figuring out good solutions that work well. To me, this unresolved problem isn’t one of “fear” per se like it’d be with social media eroding democracy or information sources being monopolized… but I’d definitely give it the “concern” label.

    1 vote

  6. Comment on I'm getting a new Macbook Pro. What's your favorite apps and tips? in ~tech

    tauon
    Link Parent
    Thanks, wasn’t aware! Not as big of an issue nowadays anymore, but I like to use the builtin webcam for video conferences. So that’s not always a “true” option. And let’s be real here: The chip...

    Thanks, wasn’t aware!

    Not as big of an issue nowadays anymore, but I like to use the builtin webcam for video conferences. So that’s not always a “true” option. And let’s be real here: The chip absolutely could’ve handled more monitors at once had they prepared it to support that from the start…, it’s just Apple artificially separating their product categories at play here.

    1 vote

  7. Comment on What programming/technical projects have you been working on? in ~comp

    tauon
    Link Parent
    Very cool! Been meaning to (potentially) make one in that “web CLI” style myself. P.S.:

    Very cool! Been meaning to (potentially) make one in that “web CLI” style myself.

    P.S.:

    a

    Hi, I'm [Your Name]!

  8. Comment on Your chatbot transcripts may be a gold mine for AI companies (gifted link) in ~tech

    tauon
    Link Parent
    Very good to know indeed! I’d heard a bit online of HIPAA, but until now was under the very assumption you just clarified, too. Although I am not a US citizen, and my medical records are protected...

    Very good to know indeed! I’d heard a bit online of HIPAA, but until now was under the very assumption you just clarified, too.

    Although I am not a US citizen, and my medical records are protected (rights to information/correction/erasure) by EU GDPR no matter where collected or stored nonetheless :-)

    2 votes

  9. Comment on Your chatbot transcripts may be a gold mine for AI companies (gifted link) in ~tech

    tauon
    Link Parent
    If it works, the only reason not to use it would be privacy concerns. I’d probably use it too given sufficient trust in the provider company… As usual, I think the individual user cannot be at blame.

    If it works, the only reason not to use it would be privacy concerns. I’d probably use it too given sufficient trust in the provider company… As usual, I think the individual user cannot be at blame.

    2 votes

  10. Comment on Your chatbot transcripts may be a gold mine for AI companies (gifted link) in ~tech

    tauon
    Link Parent
    API users, at least of OpenAI, have long been promised/guaranteed to not have their data used for training or improvements, as far as I know about this topic. Patient data, mental health concerns,...

    Of course, OpenAI and its peers promise to keep your conversations secure.

    They don't, though. In all of my interactions with ChatGPT, the bot has been pretty upfront that OpenAI will use user input to "improve my responses." Most of my conversations with it have been me poking at its capabilities and finding them lacking in ways that totally undercut its makers' claims about its utility, so they're welcome to my inputs if they want them.

    API users, at least of OpenAI, have long been promised/guaranteed to not have their data used for training or improvements, as far as I know about this topic.

    Besides, isn't this the sort of data we want them collecting, as opposed to scraping every last unsecured scrap of data on the Internet without regard to copyright? The tone of this article seems to want people to get angry or upset about this, but I'm failing to see anything shocking or outrageous about it.

    Patient data, mental health concerns, other most intimate thoughts and essentially all mail are under protection by law in many jurisdictions for a reason.

    This is no different – even though you don’t care about this or may think you don’t have anything to hide, do you really want these “AI” companies working together with potentially, just to name a few examples, your employer, insurance(s), or law enforcement?

    People here in this thread are already reporting of using GPT as an at least basic doctor replacement, and while I’m not judging that in either direction, I think it’s clear that this has pretty easy paths to huge problems being created.

    “We” as a society probably want them to collect neither copyrighted works nor user input data unchecked. Not just one of the two.

    Mind you, I'm not a booster of LLMs or AI generally, but I recognize that the real danger they represent right now primarily isn't in the systems themselves, or even in how they steal intellectual property to generate their models, but in how people respond to and use the data they generate. This kind of weak outrage bait only obscures these issues.

    Agreed, but not with your takeaway. Again, both might be an issue: People believing potential hallucinations as gospel, and new companies gaining way too much insight into people’s minds.

    What if the next OpenAI is coming from or bought by a foreign state (or state-controlled actor)? Compare the situation to e.g. Instagram and others against Tiktok.
    Do we want a potentially adversarial entity to have insight into politicians, scientists, citizens’ heads and even more so than with social media? Would we even want a non-foreign company or government entity to have that power? People already believe too much in terms of fake news from social media and LLMs as you mentioned, now imagine that data not only being directly evaluated, but possibly even manipulated live with no potential for external supervision by someone with a specific goal. The dangers here are obvious if these conversations aren’t kept private/secured by design, IMHO.

    6 votes

  11. Comment on I'm getting a new Macbook Pro. What's your favorite apps and tips? in ~tech

    tauon
    (edited )
    Link Parent
    In that case, go for it if you want to. It’s obviously an outstanding computer, especially for a discounted price! Was just concerned about potential cost-to-performance/your money going wasted… I...

    In that case, go for it if you want to. It’s obviously an outstanding computer, especially for a discounted price! Was just concerned about potential cost-to-performance/your money going wasted… I totally overlooked just how much you’d get it on sale for in my first read of the original post :)

  12. Comment on I'm getting a new Macbook Pro. What's your favorite apps and tips? in ~tech

    tauon
    (edited )
    Link Parent
    Great list. I was going to duplicate your mentions of Raycast, Rectangle, Shottr and VSCodium, as those are the tools I used to or still use the most and like the best. Also, definitely use...

    Great list.

    I was going to duplicate your mentions of Raycast, Rectangle, Shottr and VSCodium, as those are the tools I used to or still use the most and like the best.
    Also, definitely use Homebrew to install your apps like @tonyswu said.

    Furthermore:

    • Hyperkey sets caps lock to all four modifiers ⌘⌃⌥⇧ at once (and back to caps lock on very short presses should you ever need it), meaning you can have an additional layer of easy-to-use hotkeys (configurable via e.g. Raycast or Apptivate if that’s still around): hyperkey+B opens my browser, hyperkey+M opens mail, hyperkey+C opens your VSC/other IDE, etc. I have nearly all letters mapped to an app, and I gotta say I use this over command-tab pretty much exclusively nowadays.
    • kitty is another terminal emulator recommendation, it has an excellent protocol for graphics support within the terminal! (in addition to a clean look/design by default which I find pretty nice)
    • Hide my notch for making the menu bar look like the black screen border above it, which I find very nice. Although it appears the app costs something now, which wasn’t the case back when I found it IIRC.
    • Ice for an extensible menu bar, since I tend to have a lot of items in it (see below). This is functionally basically the same as Bartender, which it has replaced for me ever since that app has been sold to a shady company? due to screen recording capabilities/concerns
    • Stats shows, well, System stats like CPU/RAM/network usage or battery level and wattage drawn while charging or using it.
    • AlDente (+ turning off Apple’s less reliable variant of this mechanism) keeps my battery at 85% whenever plugged in (pick your number, could be anywhere between 80-90% ideally AFAIK), and charged higher should I know I need it ahead of time.

    Finally @PetitPrince OP, a meta-question: Apart from the bit about potentially running LLMs locally, how certain are you that you’ll need the full heft of a MacBook Pro?
    I got the M2 Macbook Air (after diligently waiting to upgrade to a new form factor and not getting the M1 Macbook Air to replace my early 2015 MBP…) in 2022, and the chip is insane. And that’s with the processors from back then, I hear they’re at what, M4 now?

    I consistently and constantly have too much opened and going on simultaneously, and it’s handled it perfectly. Hell, I have tried and can play modded Minecraft on a fan-less machine!
    I expect the computer to last me for at least another 5-7 years computationally (if working memory suffices for applications until then, and if the SSD doesn’t die of a freak failure, and the battery keeps up). The only thing you must upgrade is RAM. Minimum 16, better 24/32/whatever you can afford and is available GB, though. (This would also apply to the potential MacBook Pro.)

    • Edit: forgot to mention Dropover since I don’t typically use it anymore, but for anyone in search of a better drag & drop model it’ll be excellent!

    Other potential Air killers include (Apple did play this rather smart):

    • No HDMI port
    • No (hardware) multiple external monitor support (you can get around this with e.g. a DisplayLink dock, but it’s not perfect. On the other hand, a single widescreen monitor is 100% supported, so it depends on your desk setup…)
    • Worse screen (I think, been a while since I’ve looked into and compared the MacBooks)

    Hope this helps!

  13. Comment on Using YouTube to steal your files in ~tech

    tauon
    Link
    Very nice write-up indeed! Really, it’s an exercise in creativity exploring all those possible paths, in addition to the obvious technical know-how required. I always wonder if, to work in...

    Very nice write-up indeed!

    Really, it’s an exercise in creativity exploring all those possible paths, in addition to the obvious technical know-how required.
    I always wonder if, to work in security/vulnerability research like this, you need to be of a certain “brain type” that enables you to go through and come up with all these ideas in search of an insecure one; or if it is a trainable skill like any other.

    Likely a combination of both things.

    9 votes

  14. Comment on The attempt to reform Intel in ~tech

    tauon
    Link Parent
    What does x86-64 do better than ARM except enabling legacy binaries to run unchanged? Genuine question, as a pretty much complete novice/layman on the topic.

    What does x86-64 do better than ARM except enabling legacy binaries to run unchanged?

    Genuine question, as a pretty much complete novice/layman on the topic.

    9 votes

  15. Comment on Los Angeles police raid goes bad after gun allegedly sucked onto MRI machine in ~news

    tauon
    Link Parent
    Police union’s insurance or something, I don’t know. I know, I jest.

    Police union’s insurance or something, I don’t know. I know, I jest.

    3 votes

  16. Comment on Los Angeles police raid goes bad after gun allegedly sucked onto MRI machine in ~news

    tauon
    Link Parent
    What are the odds of it being paid for with taxpayer money?

    Hopefully the fine/settlement will be enough to repair the machine, at least.

    What are the odds of it being paid for with taxpayer money?

    6 votes

  17. Comment on How do I sync my dotfiles between PC and laptop? in ~tech

    tauon
    Link Parent
    In addition to the previous praise, I’d like to note that even if you’re already storing everything in ~/.config (meaning a single git repo “could” be enough theoretically), and there are no...

    In addition to the previous praise, I’d like to note that even if you’re already storing everything in ~/.config (meaning a single git repo “could” be enough theoretically), and there are no differences from machine to machine, chezmoi can insert secrets such as API keys from password managers’ CLI tools. This means you can actually make your dotfiles repo hosted and public, able to be cloned and set up without any access tokens, only the (hopefully remembered by heart) password manager’s credentials required.

    I haven’t actually needed it yet, but I like to imagine I could get to work on any modern unix-y machine and feel comfortably “at ~” relatively quickly, provided it has internet access and a shell I can work with.

    To me this is kind of reminiscent of the losing all your stuff in a house fire thought experiment – having a config/setup repo also seems like good practice for keeping a “backup” of sorts that’s both off-site and always current, and gets all the more relevant the more time you spend in front of a computer.

    2 votes

  18. Comment on The iPhones 16 in ~tech

    tauon
    Link Parent
    There was some… I believe Russian company which circa 2014 came out with a model that had a regular screen in the front and an e-ink backside, defaulting to something like an always-on clock...

    There was some… I believe Russian company which circa 2014 came out with a model that had a regular screen in the front and an e-ink backside, defaulting to something like an always-on clock display, but capable of mirroring the regular screen, especially text.

    That phone was cool.

    Edit: Yotaphone 2

    5 votes

  19. Comment on Share your personal dotfile treats and Unix tool recommendations in ~comp

    tauon
    Link Parent
    This feels like one of those cases where ShellCheck would go “expands when defined, not when used” when trying to set it up as an alias – i.e., on shell session startup, it tries to set the...

    Can't remember why I defined this as a function and not an alias.

    This feels like one of those cases where ShellCheck would go “expands when defined, not when used” when trying to set it up as an alias – i.e., on shell session startup, it tries to set the directory that it would later cd to (so something like your home directory) just once and statically keeps it that way, ignoring working directory later.

    But I could also be totally wrong here, haha.

  20. Comment on Tildes should recognize the Gaza Genocide, and moderate accordingly in ~tildes

    tauon
    Link Parent
    Just so that this isn’t your last interaction or experience on the topic, I’d like to express that, as another German, someone saying this was not representative (just as me writing this isn’t),...

    Meanwhile, I was more or less told that I was a fascist who doesn't belong in Germany for criticizing the country for prosecuting pro-Palestinian speech -- and I didn't even directly mention the genocide there, merely criticized it on civil rights grounds

    Just so that this isn’t your last interaction or experience on the topic, I’d like to express that, as another German, someone saying this was not representative (just as me writing this isn’t), and you’re definitely welcome here!

    25 votes