• Log in | Sign up@lemmy.world
    link
    fedilink
    English
    arrow-up
    7
    ·
    edit-2
    3 hours ago

    How else are they going to begin to recap their billions and billions of debt? Someone has to pay for all those data centres, all that hardware, all that power, etc etc etc. It will be the companies that have come to rely on AI.

    Sure, for now, AI is a lot cheaper than an intern, but it doesn’t become an expert like a human does. And Amazon used to be cheaper than other retailers right up until they had achieved vast market share.

    This cannot be the last 10x price multiplier they pull. Not even close. Firstly they’re way, way, way off from recouping their costs, and secondly, they’re still way, way off market value for an incompetent human intern who isn’t learning much.

    Uber didn’t enter the market to open up taxis to new drivers and bring down prices, that was marketing. They entered the market to take a cut out of ever taxi fare in the world, and drive up prices at peak times to many times the agreed fares, especially in regulated areas.

    Similarly, AI didn’t enter the coding market to drive down prices and enable greater access for folk to generate code. They entered the coding market to receive the wages of programmers and drive up prices in in-demand fields. They are not unaware of how much companies pay devs. Why else would they have spent all those billions in advance? Where is the payback coming from?

  • Wispy2891@lemmy.world
    link
    fedilink
    English
    arrow-up
    11
    ·
    5 hours ago

    So, if you subscribe to the $10/ month plan you get $10 worth of tokens to use within a month, and if you subscribe to the $49/month plan you get $49 worth of tokens to use within a month.

    At this point why subscribe? Just pay directly the tokens so if you use less you pay less.

    To me it doesn’t make sense to subscribe at $49 and maybe use $30 of tokens.

    People were subscribing at the $49 plan because they were using $1000 in tokens worth

    • FauxLiving@lemmy.world
      link
      fedilink
      English
      arrow-up
      8
      ·
      edit-2
      4 hours ago

      People were subscribing at the $49 plan because they were using $1000 in tokens worth

      Companies were just eating that $951/mo loss so they could gain market share by burning their nearly endless pile of capital.

      Now were on the cusp of the bubble bursting so the dominant players are now trying to extract revenue via rapid enshittification.

      It’s not price fixing, that would be illegal, it just so happens that they’re all changing their business model at the same time. If they were doing illegal things there would be an investigation.

      In completely unrelated news, didn’t you see the nice ballroom plans?

  • alpha1beta@piefed.social
    link
    fedilink
    English
    arrow-up
    4
    ·
    4 hours ago

    We’re not solidly into the enshitification stage of the AI bubble. But unlike most things, it’s never created value and only drains revenue.

  • melfie@lemmy.zip
    link
    fedilink
    English
    arrow-up
    21
    ·
    8 hours ago

    Just as open weight models are getting good. Qwen 3.6 27B just dropped with claimed performance approaching Opus 4.6, but it can run on a Mac with a M-series SoC. I tested it out today on a M4 Pro with Ollama and Cline and was impressed with its reasoning, but it was slow. Going to try with llama.cpp tomorrow and mess around tweaking it for speed.

    https://ai.rs/ai-developer/qwen-3-6-27b-local-coding-model

    AI coding agents are useful, but it’s time for the cloud-based models to chill out so we can get cheap RAM again to run our shit locally.

    • Pumpkin Escobar@lemmy.world
      link
      fedilink
      English
      arrow-up
      16
      ·
      8 hours ago

      That was intentional. They were trying to word the announcement to not make it sound like you’re now getting 1/5-1/9 as much AI for the same price.

      • trem@lemmy.blahaj.zone
        link
        fedilink
        English
        arrow-up
        7
        ·
        7 hours ago

        Yeah, I imagine that they did try. But it’s not just the intentionally misleading announcement post, they also have 5(?) different subscription tiers, which get different changes from this. And one of the subscription tiers is actually called “Pro+”, so that does not mean “Pro and more expensive tiers” like I wondered. And they have this ridiculous intermediate currency to make things even more confusing.

        Their offering itself is overly complex and confusing…

        • ID10T@programming.dev
          link
          fedilink
          English
          arrow-up
          1
          ·
          edit-2
          3 hours ago

          The whole “AI Credit” thing did strike me as odd to even introduce. They mention a few times that 1 AI Credit = $0.01, but then do the whole pricing table in $/1M tokens.

          The only place I saw them even use “AI Credits” as a unit was to say the $10/month plan includes 1000 AI Credits. Why even introduce a whole new unit if you only use it to say your $10 monthly plan includes $10 of usage?

          I suspect the answer is so that they can later muddy the waters by changing the number of included credits to be less than you’re paying monthly without directly saying “you’re now paying $20/month for $10 of usage”

          I guess it is coming from the same people who came up with the world’s most inscrutable billing scheme for compute…

    • qaz@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      4 hours ago

      It will suggest code completions while programming, can ask answer questions about code, and can edit and run code if asked.

    • lando55@lemmy.zip
      link
      fedilink
      English
      arrow-up
      16
      ·
      11 hours ago

      When the junior devs get too expensive they can outsource all of their software development to Bangladesh

      • Prox@lemmy.world
        link
        fedilink
        English
        arrow-up
        12
        ·
        8 hours ago

        When offshore-sourced code gets too shitty they can hire some senior engineers to rebuild it in a way that’s compatible with the rest of their ecosystem.

        • pdxfed@lemmy.world
          link
          fedilink
          English
          arrow-up
          6
          ·
          4 hours ago

          Which brings the model back to AI. All businesses have accomplished was basically add an extra step in actually forcing themselves to pay for what they actually need every so often. Instead of Sr>jr>offshore>sr. they now just added in AI to the loop.

          Not that charlatans care, as long as they get theirs in the interim.

  • UnspecificGravity@piefed.social
    link
    fedilink
    English
    arrow-up
    63
    ·
    14 hours ago

    Gonna be hilarious when the people who haven’t been paying attention realize that they just replaced workers with shit that doesn’t work AND actually costs more.

    • disorderly@lemmy.world
      link
      fedilink
      English
      arrow-up
      32
      ·
      13 hours ago

      Yep, I’ve been telling anyone who’ll listen that if you really want to drop juniors and give tools to seniors, then you have to pay the monthly cost (whatever it will be) and you have to be ready to foot the big bill in 5 years when your seniors (with no candidate replacements) say they’ll take a 50% raise or walk.

  • Eager Eagle@lemmy.world
    link
    fedilink
    English
    arrow-up
    91
    ·
    edit-2
    15 hours ago

    Users on annual Pro or Pro+ plans will remain on their existing plan with premium request-based pricing until their plan expires, however, model multipliers will increase on June 1 (see table).

    holy shit, 9x the previous cost. which was already not great. I was on the fence about cancelling it, but thanks for making up my mind, MS

    • Pumpkin Escobar@lemmy.world
      link
      fedilink
      English
      arrow-up
      11
      ·
      11 hours ago

      I love that their email made it sound like it was a wash, like “we’re just changing our billing model, you’ll get credits now, samesies!” but then this pricing chart buried 3-levels down from the announcement lays out just how much less you’re going to get for the same price.

      I wonder about all the startups who were bragging about their $10k/month AI coding bills being the best money they ever spent. When this new pricing kicks in and pushes it to $40k/month right around the time all the vibe-coded shit blows up their codebase, I wonder if they’ll still be so happy with their choices.

      I interviewed for a place a while back and started asking about quality and velocity and how they balance it with AI developer tools, he said something like “One of our developers closed 400 PRs last month” and I instantly knew it was definitely not the place for me.

      • a_non_monotonic_function@lemmy.world
        link
        fedilink
        English
        arrow-up
        4
        ·
        6 hours ago

        I wonder about all the startups who were bragging about their $10k/month AI coding bills being the best money they ever spent.

        WTF did they think was going to happen?

      • trem@lemmy.blahaj.zone
        link
        fedilink
        English
        arrow-up
        4
        ·
        9 hours ago

        So, did they use AI tools to type “LGTM” 400 times or nah?

        But yeah, I also find that frustrating. Management just looks at terrible metrics like PRs closed or lines of code produced.
        It’s not even novel that you can produce terrible code very quickly. Decades ago, our industry learned that it isn’t worth it, because you suffer for it later. Now the game is altered slightly and management demands that we throw all these learnings out the window.

    • Rhaedas@fedia.io
      link
      fedilink
      arrow-up
      27
      ·
      15 hours ago

      That’s been their business model for a while now. “Here’s something you also didn’t ask for”

    • [object Object]@lemmy.ca
      link
      fedilink
      English
      arrow-up
      4
      ·
      edit-2
      11 hours ago

      They should really describe this as you’re on the same plan, but your plan gives you 80-88% less use.

    • unitedwithme@lemmy.today
      link
      fedilink
      English
      arrow-up
      1
      ·
      edit-2
      15 hours ago

      You’re looking at Claude, I don’t See Copilot

      Edit: ignore me, I finally reviewed the article and is through GH, not actual MS 365 page.

      • terabyterex@lemmy.world
        link
        fedilink
        English
        arrow-up
        11
        ·
        edit-2
        7 hours ago

        copilot isnt a model. its a front end that lets you pick a model. that table shows usage costs.

    • XLE@piefed.social
      link
      fedilink
      English
      arrow-up
      44
      ·
      15 hours ago

      The good news is that none of the companies pushing these products have created the dependency yet, and they are running out of venture capital almost too fast to have the option.

      • henfredemars@infosec.pub
        link
        fedilink
        English
        arrow-up
        18
        ·
        15 hours ago

        I really hope you’re right. My employer is using it as a crutch. I don’t think they can stop using AI because they just don’t have enough skilled employees to deliver on their commitments. They would pay nearly any price, and I’m sure they’re not alone.

          • boonhet@sopuli.xyz
            link
            fedilink
            English
            arrow-up
            9
            ·
            edit-2
            13 hours ago

            That costs actual money though

            For the cost of one employee you can give 5 employees AI and tell them to work 10x faster while they have to wrestle the stupid AI.

            • XLE@piefed.social
              link
              fedilink
              English
              arrow-up
              7
              ·
              12 hours ago

              You aren’t wrong, but even if LLMs didn’t exist, employers would invent a scapegoat to make the same demands of their employees.

  • Fedditor385@lemmy.world
    link
    fedilink
    English
    arrow-up
    27
    ·
    13 hours ago

    This is the right way to go. This will incentivize many companies to rethink their strategy, and slow down or scale down AI adoption. After that and the revenue drops for many AI companies, they will back off purchasing all possible RAM and storage in existence which will drive down pricing. And when the prices get to normal again, we will simply buy more RAM for our local machines and run free models.

    This news kinda makes me happy. Shit’s starting to fall apart. Finally.