ChatGPT maker OpenAI is losing even more executives — right as Reuters reports that it's looking to restructure its core business.
  • halcyoncmdr@lemmy.worldEnglish
    3622·
    2 个月前

    You know guys, I’m starting to think what we heard about Altman when he was removed a while ago might actually have been real.

    /s

  • N0body@lemmy.dbzer0.comEnglish
    3235·
    2 个月前

    There’s an alternate timeline where the non-profit side of the company won, Altman the Conman was booted and exposed, and OpenAI kept developing machine learning in a way that actually benefits actual use cases.

    Cancer screenings approved by a doctor could be accurate enough to save so many lives and so much suffering through early detection.

    Instead, Altman turned a promising technology into a meme stock with a product released too early to ever fix properly.

    • patatahooligan@lemmy.worldEnglish
      683·
      2 个月前

      No, there isn’t really any such alternate timeline. Good honest causes are not profitable enough to survive against the startup scams. Even if the non-profit side won internally, OpenAI would just be left behind, funding would go to its competitors, and OpenAI would shut down. Unless you mean a radically different alternate timeline where our economic system is fundamentally different.

      • rsuri@lemmy.worldEnglish
        581·
        2 个月前

        I mean wikipedia managed to do it. It just requires honest people to retain control long enough. I think it was allowed to happen in wikipedia’s case because the wealthiest/greediest people hadn’t caught on to the potential yet.

        There’s probably an alternate timeline where wikipedia is a social network with paid verification by corporate interests who write articles about their own companies and state-funded accounts spreading conspiracy theories.

      • Petter1@lemm.eeEnglish
        12·
        2 个月前

        There are infinite timelines, so, it has to exist some(wehere/when/[insert w word for additional dimension]).

      • mustbe3to20signs@feddit.orgEnglish
        345·
        2 个月前

        AI models can outmatch most oncologists and radiologists in recognition of early tumor stages in MRI and CT scans.
        Further developing this strength could lead to earlier diagnosis with less-invasive methods saving not only countless live and prolonging the remaining quality life time for the individual but also save a shit ton of money.

        • T156@lemmy.worldEnglish
          373·
          2 个月前

          That is a different kind of machine learning model, though.

          You can’t just plug in your pathology images into their multimodal generative models, and expect it to pop out something usable.

          And those image recognition models aren’t something OpenAI is currently working on, iirc.

          • mustbe3to20signs@feddit.orgEnglish
            213·
            2 个月前

            I’m fully aware that those are different machine learning models but instead of focussing on LLMs with only limited use for mankind, advancing on Image Recognition models would have been much better.

            • Grandwolf319@sh.itjust.worksEnglish
              6·
              2 个月前

              I agree but I also like to point out that the AI craze started with LLMs and those MLs have been around before OpenAI.

              So if openAI never released chat GPT, it wouldn’t have become synonymous with crypto in terms of false promises.

          • Grandwolf319@sh.itjust.worksEnglish
            3·
            2 个月前

            Not only that, image analysis and statistical guesses have always been around and do not need ML to work. It’s just one more tool in the toolbox.

          • Petter1@lemm.eeEnglish
            2·
            2 个月前

            Fun thing is, most of the things AI can, they never planned it to be able to do it. All they tried to achieve was auto completion tool.

          • TFO Winder@lemmy.mlEnglish
            64·
            2 个月前

            Don’t know about image recognition but they released DALL-E , which is image generating and in painting model.

        • msage@programming.devEnglish
          14·
          2 个月前

          Wasn’t it proven that AI was having amazing results, because it noticed the cancer screens had doctors signature at the bottom? Or did they make another run with signatures hidden?

          • mustbe3to20signs@feddit.orgEnglish
            10·
            2 个月前

            There were more than one system proven to “cheat” through biased training materials. One model used to tell duck and chicken apart because it was trained with pictures of ducks in the water and chicken on a sandy ground, if I remember correctly.
            Since multiple medical image recognition systems are in development, I can’t imagine they’re all this faulty trained with unsuitable materials.

            • msage@programming.devEnglish
              6·
              2 个月前

              They are not ‘faulty’, they have been fed wrong training data.

              This is the most important aspect of any AI - it’s only as good as the training dataset is. If you don’t know the dataset, you know nothing about the AI.

              That’s why every claim of ‘super efficient AI’ need to be investigated deeper. But that goes against line-goes-up principle. So don’t expect that to happen a lot.

    • Petter1@lemm.eeEnglish
      2·
      2 个月前

      Or we get to a time where we send a reprogrammed terminator back in time to kill altman 🤓

  • kippinitreal@lemmy.worldEnglish
    1642·
    2 个月前

    Putting my tin foil hat on… Sam Altman knows the AI train might be slowing down soon.

    The OpenAI brand is the most valuable part of the company right now, since the models from Google, Anthropic, etc. can beat or match what ChatGPT is, but they aren’t taking off coz they aren’t as cool as OpenAI.

    The business models to train & run models is not sustainable. If there is any money to be made it is NOW, while the speculation is highest. The nonprofit is just getting in the way.

    This could be wishful thinking coz fuck corporate AI, but no one can deny AI is in a speculative bubble.

    • somethingsnappy@lemmy.worldEnglish
      100·
      2 个月前

      Take the hat off. This was the goal. Whoops, gotta cash in and leave! I’m sure it’s super great, but I’m gone.

        • frunch@lemmy.worldEnglish
          25·
          2 个月前

          It honestly just never occurred to me that such a transformation was allowed/possible. A nonprofit seems to imply something charitable, though obviously that’s not the true meaning of it. Still, it would almost seem like the company benefits from the goodwill that comes with being a nonprofit but then gets to transform that goodwill into real gains when they drop the act and cease being a nonprofit.

          I don’t really understand most of this shit though, so I’m probably missing some key component that makes it make a lot more sense.

          • sunzu2@thebrainbin.org
            81·
            2 个月前

            A nonprofit seems to imply something charitable, though obviously that’s not the true meaning of it

            Life time of propaganda got people confused lol

            Nonprofit merely means that their core income generating activities are not subject next to the income tax regimes.

            While some non profits are charities, many are just shelters for rich people’s bullshit behaviors like foundations, lobby groups, propaganda orgs, political campaigns etc

    • Kalysta@lemm.eeEnglish
      41·
      2 个月前

      If you can’t make money without stealing copywritten works from authors without proper compensation, you should be shut down as a company

    • trollblox_@programming.devEnglish
      21·
      2 个月前

      ai is such a dead end. it can’t operate without a constant inflow of human creations, and people are trying to replace human creations with AI. it’s fundamentally unsustainable. I am counting the days until the ai bubble pops and everyone can move on. although AI generated images, video, and audio will still probably be abused for the foreseeable future. (propaganda, porn, etc)

      • kippinitreal@lemmy.worldEnglish
        1·
        2 个月前

        That is a good point, but I think I’d like to make the distinction of saying LLM’s or “generic model” is a garbage concept, which require power & water rivaling a small country to produce incorrect results.

        Neural networks in general that can (cheaply) learn on their own for a specific task could be huge! But there’s no big money in that, since its not a consolidated general purpose product tech bros can flog to average consumers.

  • JustARaccoon@lemmy.worldEnglish
    134·
    2 个月前

    I’m confused, how can a company that’s gained numerous advantages from being non-profit just switch to a for-profit model? Weren’t a lot of the advantages (like access to data and scraping) given with the stipulation that it’s for a non-profit? This sounds like it should be illegal to my brain

    • berno@lemmy.worldEnglish
      54·
      2 个月前

      Careful you’re making too much sense here and overlapping with Elmo’s view on the subject

    • ipkpjersi@lemmy.mlEnglish
      43·
      2 个月前

      I’m confused, how can a company that’s gained numerous advantages from being non-profit just switch to a for-profit model

      Money

      • TachyonTele@lemm.eeEnglish
        44·
        2 个月前

        Money doesn’t have any advantages in other countries? When did that happen?

            • affiliate@lemmy.worldEnglish
              21·
              2 个月前

              the person that you’re replying to said something that’s true about the USA. they didn’t say anything about other countries.

              for another example, i can say “if you’re in the USA, then the current year is 2024” and that statement will be true. it is also true in every other country (for the moment), but that’s besides the point.

              • TachyonTele@lemm.eeEnglish
                23·
                2 个月前

                And I replied that it’s also true in other countries, it’s not a problem only the US has. It’s not besides the point. It’s acting as if only the US has the problem.

                • floofloof@lemmy.caOPEnglish
                  31·
                  2 个月前

                  And I specifically mentioned the USA because that’s the country where OpenAI operates and where the events in the article take place, so if someone asks why it’s so easy for OpenAI to go from being a nonprofit to a for-profit company (this was the issue I was responding to, not some general question about whether money has influence around the world), it’s the laws of the USA that are relevant, not the laws of other countries.

    • FatCrab@lemmy.oneEnglish
      11·
      2 个月前

      Their non-profit status had nothing to do with the legality of their training data acquisition methods. Some of it was still legal and some of it was still illegal (torrenting a bunch of books off a piracy site).

    • gencha@lemm.eeEnglish
      2·
      2 个月前

      These people claimed their product can pass the bar exam (it was a lie). Tells you how they feel about the legal system

  • barnaclebutt@lemmy.worldEnglish
    1374·
    2 个月前

    I’m sure they were dead weight. I trust open AI completely and all tech gurus named Sam. Btw, what happened to that Crypto guy? He seemed so nice.

    • dan@upvote.auEnglish
      109·
      2 个月前

      It’s amusing. Meta’s AI team is more open than "Open"AI ever was - they publish so many research papers for free, and the latest versions of Llama are very capable models that you can run on your own hardware (if it’s powerful enough) for free as long as you don’t use it in an app with more than 700 million monthly users.

      • a9cx34udP4ZZ0@lemmy.worldEnglish
        2·
        2 个月前

        That’s because Facebook is selling your data and access to advertise to you. The better AI gets across the board, the more money they make. AI isn’t the product, you are.

        OpenAI makes money off selling AI to others. AI is the product, not you.

        The fact facebook release more code, in this instance, isn’t a good thing. It’s a reminder how fucked we all are because they make so much off our personal data they can afford to give away literally BILLIONS of dollars in IP.

        • dan@upvote.auEnglish
          4·
          2 个月前

          Facebook doesn’t sell your data, nor does Google. That’s a common misconception. They sell your attention. Advertisers can show ads to people based on some targeting criteria, but they never see any user data.

            • wischi@programming.devEnglish
              3·
              2 个月前

              Selling your data would be stupid, because they make money with the fact that they have data about you nobody else has. Selling it would completely break their business model.

              • Knock_Knock_Lemmy_In@lemmy.worldEnglish
                1·
                2 个月前

                Depends why they are selling it, to whom, and under what restrictions.

                Yes, they don’t make the majority of their money from selling actual data, but that doesn’t mean they don’t do it.

  • Kyrgizion@lemmy.worldEnglish
    1073·
    2 个月前

    Canceled my sub as a means of protest. I used it for research and testing purposes and 20$ wasn’t that big of a deal. But I will not knowingly support this asshole if whatever his company produces isn’t going to benefit anyone other than him and his cronies. Voting with our wallets may be the very last vestige of freedom we have left, since money equals speech.

    I hope he gets raped by an irate Roomba with a broomstick.

    • ipkpjersi@lemmy.mlEnglish
      4·
      2 个月前

      But I will not knowingly support this asshole if whatever his company produces isn’t going to benefit anyone other than him and his cronies.

      I mean it was already not open-source, right?

    • eatthecake@lemmy.worldEnglish
      4·
      2 个月前

      Good. If people would actually stop buying all the crap assholes are selling we might make some progress.

    • Dkarma@lemmy.worldEnglish
      1533·
      2 个月前

      I love how ppl who don’t have a clue what AI is or how it works say dumb shit like this all the time.

      • zbyte64@awful.systemsEnglish
        172·
        2 个月前

        I also love making sweeping generalizations about a stranger’s knowledge on this forum. The smaller the data sample the better!

      • Ragnarok314159@sopuli.xyzEnglish
        152·
        2 个月前

        There is no AI. It’s all shitty LLM’s. But keep sucking that techbro cheesy balls. They will never invite you to the table.

        • WindyRebel@lemmy.worldEnglish
          32·
          2 个月前

          Honest question, but aren’t LLM’s a form of AI and thus…Maybe not AI as people expect, but still AI?

          • whats_all_this_then@lemmy.worldEnglish
            5·
            2 个月前

            The issue is that “AI” has become a marketing buzz word instead of anything meaningful. When someone says “AI” these days, what they’re actually referring to is “machine learning”. Like in LLMs for example: what’s actually happening (at a very basic level, and please correct me if I’m wrong, people) is that given one or more words/tokens, it tries to calculate the most probable next word/token based on its model (trained on ridiculously large numbers of bodies of text written by humans). It does this well enough and at a large enough scale that the output is cohesive, comprehensive, and useful.

            While the results are undeniably impressive, this is not intelligence in the traditional sense; there is no reasoning or comprehension, and definitely no consciousness, or awareness here. To grossly oversimplify, LLMs are really really good word calculators and can be very useful. But leave it to tech bros to make them sound like the second coming and shove them where they don’t belong just to get more VC money.

            • slackassassin@sh.itjust.worksEnglish
              2·
              2 个月前

              Sure, but people seem to buy into that very buzz wordyness and ignore the usefulness of the technology as a whole because “ai bad.”

              • whats_all_this_then@lemmy.worldEnglish
                1·
                2 个月前

                True. Even I’ve been guilty of that at times. It’s just hard right now to see the positives through the countless downsides and the fact that the biggest application we’re moving towards seems to be taking value from talented people and putting it back into the pockets of companies that were already hoarding wealth and treating their workers like shit.

                So usually when people say “AI is the next big thing”, I say “Eh, idk how useful an automated idiot would be” because it’s easier than getting into the weeds of the topic with someone who’s probably not interested haha.

                Edit: Exhibit A

                • slackassassin@sh.itjust.worksEnglish
                  1·
                  2 个月前

                  There’s some sampling bias at play because you don’t hear about the less flashy examples. I use machine learning for particle physics, but there’s no marketing nor outrage about it.

  • NeoNachtwaechter@lemmy.worldEnglish
    981·
    2 个月前

    Altman downplayed the major shakeup.

    "Leadership changes are a natural part of companies

    Is he just trying to tell us he is next?

    /s

    • xavier666@lemm.eeEnglish
      16·
      2 个月前

      Sam: “Most of our execs have left. So I guess I’ll take the major decisions instead. And since I’m so humble, I’ll only be taking 80% of their salary. Yeah, no need to thank me”

    • Avg@lemm.eeEnglish
      9·
      2 个月前

      The ceo at my company said that 3 years ago, we are going through execs like I go through amlodipine.

    • Wanderer@lemm.eeEnglish
      31·
      2 个月前

      They always are and they know it.

      Doesn’t matter at that level it’s all part of the game.

  • Chaotic Entropy@feddit.ukEnglish
    96·
    2 个月前

    The restructuring could turn the already for-profit company into a more traditional startup and give CEO Sam Altman even more control — including likely equity worth billions of dollars.

    I can see why he would want that, yes. We’re supposed to ooo and ahh at a technical visionary, who is always ultimately a money guy executive who wants more money and more executive power.

    • toynbee@lemmy.worldEnglish
      24·
      2 个月前

      I saw an interesting video about this. It’s outdated (from ten months ago, apparently) but added some context that I, at least, was missing - and that also largely aligns with what you said. Also, though it’s not super evident in this video, I think the presenter is fairly funny.

      https://youtu.be/L6mmzBDfRS4

      • Melatonin@lemmy.dbzer0.comEnglish
        9·
        2 个月前

        That was a worthwhile watch, thank you for making my life better.

        I await the coming AI apocalypse with hope that I am not awake, aware, or sensate when they do whatever it is they’ll do to use or get rid of me.

        • sunbeam60@lemmy.oneEnglish
          4·
          2 个月前

          You will be kept alive at subsistence level to buy the stuff you’ve been told to buy, don’t worry.

        • toynbee@lemmy.worldEnglish
          1·
          2 个月前

          My pleasure! Glad it helped. Also, I like your username.

          I’m still not sure how much to fear AI, as I’m not knowledgeable on the subject (never even intentionally interacted with one yet) and have seen conflicting reports on how worryingly capable it is. Today I did see this video, which isn’t explicitly about AI but did offer an interesting perspective that could be compared to the paradigm: https://youtu.be/fVN_5xsMDdg

          (Warning, the video was interesting, but I got invested about halfway through when I started comparing it to AI, then was disappointed in the ending)

    • assassin_aragorn@lemmy.worldEnglish
      772·
      2 个月前

      paid for entirely by venture capital seed funding.

      And stealing from other people’s works. Don’t forget that part

        • raspberriesareyummy@lemmy.worldEnglish
          405·
          2 个月前

          When individual copyright violations are considered “theft” by the law (and the RIAA and the MPAA), violating copyrights of billions of private people to generate profit, is absolutely stealing. While the former arguably is arguably often a measure of self defense against extortion by copyright holding for-profit enterprises.

        • exanime@lemmy.worldEnglish
          141·
          2 个月前

          Right, it’s only stolen when regular people use copyright material without permission

          But when OpenAI downloads a car, it’s all cool baby

    • sunbeam60@lemmy.oneEnglish
      910·
      2 个月前

      Barely usable results?! Whatever you may think of the pricing (which is obviously below cost), there are an enormous amount of fields where language models provide insane amount of business value. Whether that translates into a better life for the everyday person is currently unknown.

    • flo@infosec.pubEnglish
      810·
      2 个月前

      barely usable results

      Using chatgpt and copilot has been a huge productivity boost for me, so your comment surprised me. Perhaps its usefulness varies across fields. May I ask what kind of tasks you have tried chatgpt for, where it’s been unhelpful?

      • wholookshere@lemmy.blahaj.zoneEnglish
        107·
        2 个月前

        Literally anything that requires knowing facts to inform writing. This is something LLMs are incapable of doing right now.

        Just look up how many R’s are in strawberry and see how chat gpt gets it wrong.

        • boonhet@lemm.eeEnglish
          1·
          2 个月前

          Okay what the hell is wrong with it

          It took me three times to convince it that there’s 3 r’s in strawberry…

          • wholookshere@lemmy.blahaj.zoneEnglish
            1·
            2 个月前

            Because that’s not how LLMs work.

            When you form a sentence you start with an intent.

            LLMs start with the meaning you gave it, and tries to express something similar to you.

            Notice how intent, and meaning aren’t the same. Fact checking has nothing to do with what a word means. So how can it understand what is true?

            All it did was take the meaning of looking for a number and strawberries and ran it’s best guess from that.

        • interdimensionalmeme@lemmy.mlEnglish
          112·
          2 个月前

          Yes it says aim for the brain stem but like most things it says, I already knew that. Finally quietness from the hearing the same thing over and over and over and over

                • interdimensionalmeme@lemmy.mlEnglish
                  12·
                  2 个月前

                  I suggest you touch grass if you think remembering some social media server web address that the phone remember.

                  But also if you want to discriminate based on what server a user used to sign up, then it’s already too late for you

          • Flying Squid@lemmy.worldEnglish
            41·
            2 个月前

            but like most things it says, I already knew that

            So how long have you been putting glue on your pizza?

            • interdimensionalmeme@lemmy.mlEnglish
              11·
              2 个月前

              That’s Google and it’s also called being able to tell reality apart from fiction, which is becoming clear most anti ai zealots have never been capable of.

              • Flying Squid@lemmy.worldEnglish
                1·
                2 个月前

                You seem to have forgotten your previous post:

                Yes it says aim for the brain stem but like most things it says, I already knew that.

                So either you already knew to put glue on pizza or you knew that the AI isn’t trustworthy in the first place. You can’t have it both ways.

      • Clent@lemmy.worldEnglish
        181·
        2 个月前

        That’s not the incentive you think it is.

        Make sure you go deep. Need to get the whole thing to real show you’re serious.

  • pjwestin@lemmy.worldEnglish
    75·
    2 个月前

    I really don’t understand why they’re simultaneously arguing that they need access to copyrighted works in order to train their AI while also dropping their non-profit status. If they were at least ostensibly a non-profit, they could pretend that their work was for the betterment of humanity or whatever, but now they’re basically saying, “exempt us from this law so we can maximize our earnings.” …and, honestly, our corrupt legislators wouldn’t have a problem with that were it not for the fact that bigger corporations with more lobbying power will fight against it.