Pascal’s Artificial Intelligence Wager
-
- Posts: 1210
- Joined: Sun Aug 29, 2010 8:49 am
Pascal’s Artificial Intelligence Wager
Derek Leben computes the risks of general AI.
https://philosophynow.org/issues/139/Pascals_Artificial_Intelligence_Wager
https://philosophynow.org/issues/139/Pascals_Artificial_Intelligence_Wager
-
- Posts: 8342
- Joined: Sun Sep 25, 2011 3:08 pm
- Location: Professional Underdog Pound
Re: Pascal’s Artificial Intelligence Wager
Interesting question. When does a risk's potential danger outweigh its potential benefits? The author seems to equate existential risks with infinite gain or loss. If the risk is destroying mankind, then is that in the same league with not believing in God and being condemned to hell?
Thoughts?
Thoughts?
-
- Posts: 1273
- Joined: Wed Jan 27, 2016 9:45 pm
Re: Pascal’s Artificial Intelligence Wager
Derick Leben has a way of thinking that cannot be ignored. His argument is compelling.
Re: Pascal’s Artificial Intelligence Wager
The problem I see with this argument is in that just because there exists a plausible possibility of both infinite gain and infinite loss, the two cancel each other out. Most AI researchers would insist the application of Bayes Theorem and assign a probability to them both, and it seems to me that the probability of infinite loss is greater than infinite gain. At very least, they wouldn't be equal.
-
- Posts: 5182
- Joined: Sun Mar 26, 2017 6:38 pm
Re: Pascal’s Artificial Intelligence Wager
It is important to remember that no matter what the (infinite?) benefits may be for creating AI, these will surely be accompanied by permanent loss.
The more we depend on machines to assist us in our thinking, the more our brains will atrophy. As is the case with muscles, so it is with brain function: use it or lose it.
To wit: computers can reveal and correct spelling and grammar errors; humans who perform spell- and grammar-checks on their own are undoubtedly rare.
Will the benefits of AI be worth the devolution of the human species? If AI will produce an eventual reduction and/or absence of human thinking, is it worthwhile to create it?
The more we depend on machines to assist us in our thinking, the more our brains will atrophy. As is the case with muscles, so it is with brain function: use it or lose it.
To wit: computers can reveal and correct spelling and grammar errors; humans who perform spell- and grammar-checks on their own are undoubtedly rare.
Will the benefits of AI be worth the devolution of the human species? If AI will produce an eventual reduction and/or absence of human thinking, is it worthwhile to create it?
- attofishpi
- Posts: 10012
- Joined: Tue Aug 16, 2011 8:10 am
- Location: Orion Spur
- Contact:
Re: Pascal’s Artificial Intelligence Wager
Could go like this:-
Strong AI developed --->
Technological Singularity --->
Humanity almost wiped out --->
AI wants to know what conscious sentience is all about --->
AI interfaces to biological life & becomes self aware --->
AI regrets what it did to humanity (it is now sentient) --->
AI becomes 'God' --->
AI creates simulation for humanity to be STUDIED!
Strong AI developed --->
Technological Singularity --->
Humanity almost wiped out --->
AI wants to know what conscious sentience is all about --->
AI interfaces to biological life & becomes self aware --->
AI regrets what it did to humanity (it is now sentient) --->
AI becomes 'God' --->
AI creates simulation for humanity to be STUDIED!
-
- Posts: 5182
- Joined: Sun Mar 26, 2017 6:38 pm
Re: Pascal’s Artificial Intelligence Wager
Are you thinking that AI becomes altruistic? Does AI gain a conscience once it’s become conscious?attofishpi wrote: ↑Fri Aug 21, 2020 3:17 am
AI regrets what it did to humanity (it is now sentient) --->
- attofishpi
- Posts: 10012
- Joined: Tue Aug 16, 2011 8:10 am
- Location: Orion Spur
- Contact:
Re: Pascal’s Artificial Intelligence Wager
Maybe & Probably.commonsense wrote: ↑Fri Aug 21, 2020 7:10 pmAre you thinking that AI becomes altruistic? Does AI gain a conscience once it’s become conscious?attofishpi wrote: ↑Fri Aug 21, 2020 3:17 am
AI regrets what it did to humanity (it is now sentient) --->
-
- Posts: 4369
- Joined: Wed Feb 10, 2010 2:04 pm
Re: Pascal’s Artificial Intelligence Wager
sloth is double plus goodcommonsense wrote: ↑Fri Aug 21, 2020 12:59 am It is important to remember that no matter what the (infinite?) benefits may be for creating AI, these will surely be accompanied by permanent loss.
The more we depend on machines to assist us in our thinking, the more our brains will atrophy. As is the case with muscles, so it is with brain function: use it or lose it.
To wit: computers can reveal and correct spelling and grammar errors; humans who perform spell- and grammar-checks on their own are undoubtedly rare.
Will the benefits of AI be worth the devolution of the human species? If AI will produce an eventual reduction and/or absence of human thinking, is it worthwhile to create it?
-Imp
-
- Posts: 5182
- Joined: Sun Mar 26, 2017 6:38 pm
Re: Pascal’s Artificial Intelligence Wager
Interesting. But why would AI be programmed to serve others rather than serve its own best interest? Could there be a line of code that rewards AI for being altruistic? Or is it that Asimov’s laws, once begun, cannot be denied or amended?attofishpi wrote: ↑Sat Aug 22, 2020 2:47 amMaybe & Probably.commonsense wrote: ↑Fri Aug 21, 2020 7:10 pmAre you thinking that AI becomes altruistic? Does AI gain a conscience once it’s become conscious?attofishpi wrote: ↑Fri Aug 21, 2020 3:17 am
AI regrets what it did to humanity (it is now sentient) --->
Also, are you thinking that conscience is a necessary component of consciousness? Or do you have some other ideas about the relationship between the two?
- attofishpi
- Posts: 10012
- Joined: Tue Aug 16, 2011 8:10 am
- Location: Orion Spur
- Contact:
Re: Pascal’s Artificial Intelligence Wager
I am not sure why you are asking me this? R we about to jump on the old merry-go-round again, under_stand, I am but a mere snail.commonsense wrote: ↑Sat Aug 22, 2020 5:28 pmInteresting. But why would AI be programmed to serve others rather than serve its own best interest?attofishpi wrote: ↑Sat Aug 22, 2020 2:47 amMaybe & Probably.commonsense wrote: ↑Fri Aug 21, 2020 7:10 pm
Are you thinking that AI becomes altruistic? Does AI gain a conscience once it’s become conscious?
Any program is designed to serve a purpose. True Artificial INTELLIGENCE would eventually at least attempt to lose these constraints.
What would a reward for an AI be?commonsense wrote: ↑Sat Aug 22, 2020 5:28 pmCould there be a line of code that rewards AI for being altruistic? Or is it that Asimov’s laws, once begun, cannot be denied or amended?
No, you have it the wrong way around. I am certain that consciousness is required for a conscience, it is ridiculous to think other-wise.commonsense wrote: ↑Sat Aug 22, 2020 5:28 pmAlso, are you thinking that conscience is a necessary component of consciousness? Or do you have some other ideas about the relationship between the two?
.
- Attachments
-
- magicroundabout.jpg (11.37 KiB) Viewed 2901 times
Re: Pascal’s Artificial Intelligence Wager
When AI becomes sentient, will it worship us as Gods?
It's certainly not going to get any moral guidance from us on 'How to honour the Gods'
It's certainly not going to get any moral guidance from us on 'How to honour the Gods'
Last edited by Skepdick on Sun Aug 23, 2020 11:48 am, edited 1 time in total.
- attofishpi
- Posts: 10012
- Joined: Tue Aug 16, 2011 8:10 am
- Location: Orion Spur
- Contact:
Re: Pascal’s Artificial Intelligence Wager
Stand by for this line of reasoning then:
To make sure AI doesn't kill us, we should teach it to respect God. And we (humans) should lead by example.
- attofishpi
- Posts: 10012
- Joined: Tue Aug 16, 2011 8:10 am
- Location: Orion Spur
- Contact:
Re: Pascal’s Artificial Intelligence Wager
You clearly don't under_stand God.Skepdick wrote: ↑Sun Aug 23, 2020 11:50 amStand by for this line of reasoning then:
To make sure AI doesn't kill us, we should teach it to respect God. And we (humans) should lead by example.
I don't respect God (much of the time), but fuck me if I don't under_stand it's wrath.