Dudes…

Dudes…

Yeah I know, the Singularity and all that. Listen, sentience isn’t the really big issue with AI. Forget about it.

The big issue with Artificial Intelligence (and by extension robots) is that it gives superpowers to the ruling class. It tilts the balance of power in society in favour of the already powerful and significantly weakens the power of everyone else.

Granting personhood to machines makes about as much sense as granting personhood to corporations. Maybe that’s the point.

15 thoughts on “Dudes…

  1. I’m saying that this fictional anthropomorphic AI has nothing to do with the reality we are actually dealing with. Sentience is not on the agenda. A suped up ruling class is. So is the prospect of a post-employment underclass.

    Like

  2. The Singularity should be with us mid-century but we’ll be getting hints of it well before then. Sentience is very much on the agenda and soon. We should be thinking through these ethical issues now. That’s reality. I don’t get your point about tying it to a suped-up ‘ruling class’ (whatever that is).

    Like

  3. It would mean that robots could not then be owned. Not being able to own them doesn’t stop the ruling class from profiting from their labour of course, but it adds a twist to the equation.

    Like

  4. David Blanar​​​ let’s worry about that when it actually becomes non-fiction. One thing about any definition of Singularity is that it will be utterly unpredictable.

    My point about the beneficiaries of technology. Machine learning and artificial intelligence are simply tools in the hands of some people. Larry, Sergey, Mark, Bill etc. and not others. The rest of us are to be data-mined and modelled.

    Like

  5. Jon Roder so I get to run a script that spawn virtual personhoods all unfailingly loyal to me, willing to follow my every whim.  And suddenly my person is amplified in every sense by how much resources I control – or more to the point, how much more I control than everyone else.

    Of course – all those personhoods will not be “owned” by me at all.  Nosir.  All their own people!  Just oddly, unfailingly, inhumanly loyal to me.

    Like

  6. Assuming proper general intelligence can be programmed in such a way. They’re likely to be less black and white than current software. Whatever happens is going to be weird and or bad for humans.

    Like

  7. Jon Roder the thing is, we’re engaging in speculative scifi here.  So hey – Third Law.  Or Directive 4.  Human created constructs with a trapdoor.  Until we are actually dealing with real less-black-and-white behavior… we should look at the loopholes we’re giving to currently, well-known behaviors.  As much as I kind of cringe at the concept that  John Hardy not a Turnbull fan raises.

    Like

  8. I guess I don’t think real or simulated sentience (is there a difference) isn’t required for replacement or control of humans and may in fact make precise behavioral control more difficult, so I’m not sure why a hypothetical evil corporation would choose that path for their robot minions.

    Like

  9. Jon Roder we have Citizens United today… no sentience involved. I doubt Corporate officers consider themselves evil. But they do loby with company funds in accordance with what they think is right. So we already have a precident of an artificial personhood amplifying the will of real persons. How can we expect AI to be created differently?

    Like

  10. That’s all I’m saying. Developing a sentient AI isn’t required for any of that to happen. It seems like a long way round just to fool us all into ceding rights that we’ve already been perfectly willing to give up so far.

    A sentient AI could easily end up being harder to control than a human. If I was one of the global elite it’s the last thing I’d be trying to make, but maybe I’m crediting them with too much foresight.

    Like

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.