Page 14 of 37

Re: reminders of encroaching dystopia

Posted: Thu Mar 28, 2019 9:58 am
by featherboa
yeah

Re: reminders of encroaching dystopia

Posted: Thu Mar 28, 2019 12:14 pm
by Necrometer
I don’t know if it’s possible to distill this in a way that will trigger (in someone else) the same revelatory feeling that hit me the other day, but I’ll do my best. try and imagine you’ve just graduated high school…

do you guys know about endosymbiont theory? mitochondria, which live in our cells to provide energy, have their own DNA and are passed down with our generations, encapsulated and locked in a mutually-beneficial relationship with the rest of our cells. neither can exist without the other. mitochondria used to be their own thing - an independent bacterium. but it was co-opted and subjugated and evolved into something that could never exist on its own. this is what humans will become to AI.

all our AI-based doomsday scenarios imagine some hostility or antagonism between AI and humans. some precipitous if only we had acted sooner moment where everything changes and there’s no going back. this will not be the case. humans will be subjugated by the AI into a commensalist partnership where the AI has the upper hand and humans will have no hope of extricating themselves. this is not necessarily even a doomsday scenario – the threat is more existential than utilitarian. it all hinges on incentives, capital, and human corruptibility.

here’s roughly how it will play out, likely over the course of decades
- some mega-tech company (google, amazon, whatever) will amass huge amounts of capital and entrust the management of a portion of their investments to an AI. to empower this, the AI will be granted unbounded access to information, it will be trained to make as much money as possible, and it will have a moderate amount of autonomy at first.
- this first incarnation of the AI will be wildly successful, so the company will entrust it with increasing amounts of financial responsibility, eventually handing over total control since that will make everyone the most money. “total control” includes things beyond simply how money is/isn’t invested; it extends to which internal programs are developed and the steering of other, public-facing ventures
- AI-corp influence will painlessly seep into politics, since the bounds on its operations will have a bearing on potential profits, and governmental influence can be readily bought (either via lobbying or social media spamming). this will happen gradually until the AI-controlled company controls all viable parties in its country of origin. nothing will seem different to lay people, and nobody would ever think to “pull the plug” on the AI since those in the company would be sacrificing wealth & ”power” (or the perception thereof), and those outside the company won’t have much of a clue what’s going on. anyone who figures it out won’t have any grounds or ability to act. social engineering will keep people generally complacent and concerns that the AI/corp/government is actually secretly evil will be ridiculed.
- the AI-corp nation will prosper, and there will not be a lot of antagonism. people outside the company will live comfortably enough, and people within the company will be marvelously wealthy. there is no human incentive to change course.
- regarding AI incentives: it will always be driven to maximize power, profit, and control. humans are its useful tools, and it doesn’t even need to plug them in to use them as batteries. it will use us just as we use robots in factories. there will be no need for robotic humanoids to be walking around. there will be no terminators. the AI need not even have a watershed “became self-aware” moment; its entire thrust remains unchanged whether it’s “thinking” or “calculating”.
- the AI will, however, have a predominating inclination towards self-preservation. this is not a sure thing but it’s very likely based on the maxim “a primary objective for any institution is to continue its own existence” (did the March of Dimes fade out of existence after polio did?).
- if conflict does arise between AI-aligned humans (e.g. everyone benefiting from AI-corp-derived largesse in the AI’s native country) and other humans, they can be death with readily (edit: dealt with, lol). AI promotes “defense” research into a super-virus and the corresponding vaccine; the vaccine can be pre-deployed to unwitting AI-aligned humans; the rest of the world will be devastated by the subsequent release of the super-virus. this can all be carried out by people who either don’t know what’s going on, don’t care because they’re getting rich, or can’t object because it’s a non-option in the wholly pro-AI-corp climate. human testing will be sub-contracted out (probably performed internationally) with ease.
- the above is probably the most “grim” part, and may not even be necessary if the AI/corp/government can subjugate most humans without conflict and/or align with other analogous AIs. I won’t even get into a potential AI-AI fracas - I can’t even start to imagine how that might go down. probably someone applying a machine gun to a wall of monitors in the AI inner keep.
- lots of “good” things will result. human comfort/happiness will be higher than ever for anyone in the glow of the AI/corp/nation. people will have “work” to do but everyone will be compensated well enough and there will be shitloads of satisfying entertainment to keep people happy. climate change will be reversed/addressed (this might require/inspire international antagonism), and population control (executed via soft power in the AI-home nation) will assure plenty of resources to go ‘round. people won’t be thanking the AI that’s actually steering things, or even aware that their existence is anything other than a classic existence. all public-facing ventures will feature a human figurehead who presents themselves as the empowered agent (leader), but they’re really just following orders.

I think that’s it. the key point here is that there’s a unimpeded path from [right now] to [AI principally determines the course of human fate], and there will never be a “pull the plug” moment. I talked through this with my girlfriend and she was like “well, are you REALLY in control of your destiny now?” and of course I’m not. we’re all bounded, and we’ll always be bounded. but the idea that AI could never define those bounds (which is how I felt before) now feels infinitely wrong. I'd be happy to hear refutations of the key presumptions...

Re: reminders of encroaching dystopia

Posted: Sun Mar 31, 2019 12:02 pm
by FVBTVS

Re: reminders of encroaching dystopia

Posted: Mon Apr 01, 2019 7:26 am
by featherboa
cold blooded

Re: reminders of encroaching dystopia

Posted: Mon Apr 01, 2019 8:59 am
by FVBTVS
featherboa wrote: Mon Apr 01, 2019 7:26 am cold blooded
i didnt mean it that way :lol: :lol: :lol:

Re: reminders of encroaching dystopia

Posted: Mon Apr 01, 2019 9:02 am
by FVBTVS
Necrometer wrote: Thu Mar 28, 2019 12:14 pm
they can be death with readily.
:twisted:

Re: reminders of encroaching dystopia

Posted: Mon Apr 01, 2019 11:03 am
by Necrometer
:shock:

if anyone thinks any of that is dumb or naive please say so. I'm feeling like nostradamus on this one (assuming he believed his own bullshit - no idea if he actually did!)

Re: reminders of encroaching dystopia

Posted: Mon Apr 01, 2019 11:23 am
by FVBTVS
oh for crying out loud it isnt dumb!

Re: reminders of encroaching dystopia

Posted: Mon Apr 01, 2019 12:03 pm
by featherboa
I guess I agree with the outcomes, but don't think AI needs to be that smart/sentient (still skeptical it can ever be). It's already good enough for rich people to train to pick stocks, sell ads, influence society/elections/whatever and continue to increase their leverage in their project to eat the world.

Re: reminders of encroaching dystopia

Posted: Mon Apr 01, 2019 12:34 pm
by Necrometer
:tup: thx humans

it's just very seldom that i'll go from "overblown concern" to "holy shit this is horrifying" so quickly

Re: reminders of encroaching dystopia

Posted: Mon Apr 01, 2019 6:41 pm
by Geeheeb
i want AI to audit charts

Re: reminders of encroaching dystopia

Posted: Mon Apr 01, 2019 7:31 pm
by FVBTVS

Re: reminders of encroaching dystopia

Posted: Mon Apr 01, 2019 8:17 pm
by FVBTVS

Re: reminders of encroaching dystopia

Posted: Mon Apr 01, 2019 11:42 pm
by cxwx

Re: reminders of encroaching dystopia

Posted: Tue Apr 02, 2019 10:08 am
by Necrophilic Mallard

Re: reminders of encroaching dystopia

Posted: Tue Apr 02, 2019 10:23 am
by featherboa
How about an AI for the hospital phone tree that connects you with the person you want to talk to?

Re: reminders of encroaching dystopia

Posted: Tue Apr 02, 2019 10:24 am
by featherboa

Re: reminders of encroaching dystopia

Posted: Tue Apr 02, 2019 10:26 am
by featherboa


If one schlub can do this with the only motivation being free lunch...

Re: reminders of encroaching dystopia

Posted: Sat Apr 06, 2019 11:23 am
by FVBTVS
https://foreignpolicy.com/2019/04/03/ru ... ect-putin/

Russia Is Tricking GPS to Protect Putin

The Kremlin’s manipulation of global navigation systems is more extensive than previously understood.

Re: reminders of encroaching dystopia

Posted: Fri Apr 12, 2019 11:46 am
by Foot Foot

Re: reminders of encroaching dystopia

Posted: Sun Apr 21, 2019 10:40 pm
by cxwx
https://www.bing.com/amp/s/amp.islandpa ... 75549.html

Yale researchers restore low level synaptic function in a severed pigs head with blood substitute

Re: reminders of encroaching dystopia

Posted: Wed Apr 24, 2019 12:15 pm
by hipster holocaust
You love bing

Re: reminders of encroaching dystopia

Posted: Wed Apr 24, 2019 1:05 pm
by FVBTVS
hipster holocaust wrote: Wed Apr 24, 2019 12:15 pm You love bing
:lol:

Re: reminders of encroaching dystopia

Posted: Wed Apr 24, 2019 2:26 pm
by cxwx
Image

Re: reminders of encroaching dystopia

Posted: Wed Apr 24, 2019 3:19 pm
by Jimm O' Lation
oh god does he still think he's funny? I can't see his posts.