6 ms·

This may have been just a stupid delusion by Dalio but some day one of these sociopaths is gonna get their hands on the wrong AI model and end us all. The nece

• erulabs 3y ago
  I think you're seeing it wrong. When CEOs no longer need scientists and engineers, when they can speak their will into existence directly, they will immediately destroy themselves. This rating system, had the CEO gotten it's say, would not have made Bridgewater thrive. That stupid ideas are often impossible to accomplish is a huge barrier to stupid leadership screwing themselves sideways. Anyways, thats the optimistic take :P
  • hinkley 3y ago
    Sandbagging can just as easily be used for good or for evil. "How goes our plans to <do something morally reprehensible>?" Oh, we are still having trouble with the requirements. It's all blocked up in committee. Not much progress to report.
  • RGamma 3y ago
    That's assuming the effects of that stay contained within the respective organization which is not something I would bet on. Chances are a CEO type will be the first to get their hands on AGI or something close enough.
  • fzeroracer 3y ago
    Unfortunately I think this is sort of the optimistic take. They will destroy themselves, but also take down a lot of other systems with them, just look at Musk. These things have a very long tail end for all of the accumulated problems to turn from a ripple to a tide.
• immibis 3y ago
  This is already happening, and has nothing to do with LLMs. Look at the other thread about banks closing customer accounts for no reason.