Nearly one-third of corporate executives surveyed did not understand how their operating costs relate to AI deployment.
    • mcv@lemmy.zipEnglish
      3 months

      They’re not selected for intelligence, but for sociopathy.

  • SocialMediaRefugee@lemmy.worldEnglish
    3 months
    1. Build up reliance on AI, which looks really cheap
    2. You can now replace employees with AI so fire away!
    3. You are now completely dependent on AI and a handful of employees
    4. AI company sees they have you and start jacking up rates. If you could afford paying for people before then you have the $ to pay high rates.
    5. Company now wonders why costs are back to where they were before and the AI isn’t working out as expected.
    • monotremata@lemmy.caEnglish
      3 months

      It’s particularly funny because I’m pretty sure AI companies are still selling the service below cost to try to retain market share (and drive small competitors out of business). They just aren’t taking quite as big a loss on every token with the increased prices.

      • Kairos@lemmy.todayEnglish
        3 months

        So, they’re earning money on token generation but not overall (including training)?

        • DeadDigger@lemmy.zipEnglish
          3 months

          Openai had 2025 6billion in revenue and 20 billion costs on compute. So just to run the models to get 6billion they need to pay 20billion r&d and marketing etc get on top of that

          • Pacattack57@lemmy.worldEnglish
            3 months

            I’m sure there’s a term for it but this is like when a company keeps securing funding from investors so they keep growing to try to outpace costs with the illusion that you’re profitable when in reality you’re not. Just like WeWork.