• intensely_human@lemm.ee
    link
    fedilink
    arrow-up
    13
    ·
    9 months ago

    So what if you had the capability to be good at that heart pump software, and by not doing it you’re letting in someone who’s not as good as you at it.

    Not trying to trap you or anything, but at a certain point somebody’s gotta do that high stakes stuff.

    Maybe you know you’re not good at that kind of precision stuff, and I respect that. But maybe the difference between you and the guy who did step up to do that work isn’t that he has more skill, but merely less humility.

    Maybe you, the one scared shitless of the ramifications of a mistake, are the one for that job.

    • Scrubbles@poptalk.scrubbles.tech
      link
      fedilink
      English
      arrow-up
      17
      ·
      9 months ago

      Maybe. I’ve just shipped too many bugs to prod to trust myself. But you’re right for every senior engineer who is paranoid about that there’s 100 devs who will ship it and then take vacation for a week.

      • intensely_human@lemm.ee
        link
        fedilink
        arrow-up
        10
        ·
        9 months ago

        Perhaps the reason you ship bugs to prod, is because you know it doesn’t matter. Maybe your bug rate isn’t a pure function of the coding task complexity, but is context-dependent on the goal being coded toward.

        I don’t even know why I’m pushing you on this to be honest. Trust your gut man

    • morry040@kbin.social
      link
      fedilink
      arrow-up
      15
      ·
      9 months ago

      There’s quite a difference between rapid prototyping on software/hardware versus the human body.
      Musk’s approach to developing engineering advances has worked well in the software, aerospace, and vehicular industries. Development on inorganic things is much more predictable, we can isolate variables, and it is easier to understand cause & effect. If you screw up some software on an inorganic system, your program might crash, your rocket might explode, or your car won’t start. These risks can be anticipated and costed fairly well, therefore rapid prototyping has an acceptable risk/reward ratio in that environment.

      The human body, on the other hand, is an extremely complex system that we still don’t fully understand. Each person is a unique variation on the model and that changes over time depending on upbringing, diet, exercise, and life experiences. Applying the same engineering approaches from inorganic industries has a much higher risk once you cross into the medical realm. If you have errors in a medical situation, you risk sickening, injuring, or even killing a person. The risk/reward ratio is skewed towards ensuring that human life is protected at all costs.

      Using SpaceX as an example, the first three launches failed spectacularly and a fourth failure would have ended the business but fortunately the fourth test was a success. If you’re suggesting that we apply the same risk-taking to Neuralink, are you suggesting that it’s acceptable for the first three patients to die, as long as the fourth is a success?