The threat of artificial intelligence??

How should we think about weird things?
User avatar
Lance Kennedy
True Skeptic
Posts: 10220
Joined: Tue Mar 29, 2005 10:20 pm
Custom Title: Super Skeptic
Location: Paradise, New Zealand

The threat of artificial intelligence??

Postby Lance Kennedy » Wed Mar 08, 2017 9:53 pm

http://www.michaelshermer.com/2017/03/apocalypse-ai/

The article referenced above is a discussion of the possible threat of building artificial intelligence with a level of smarts greater than human. The article is somewhat disparaging, but I disagree. The level of threat remains unproved.

According to Moore's Law, the first computer to be built with a level of complexity greater than the human brain will happen in 2035 AD. This does not mean it will be smarter. The first computer smarter than (say) Albert Einstein will be built before or after that date, but probably not too much before or after. However, it will think at a speed way faster than the human brain. Possibly a million times faster. What I think is that we cannot predict the motivation of such a superior mind.

The article above suggests that we have plenty of time to build in safeguards before such a mind comes into existence. Maybe, but there is no sign that anyone is even trying to do that. Indeed, the trend seems to be towards increasing the level of independence from humans, enabling electronic 'minds' to learn and rewrite their own programming without human interferance.

So how much risk to human existence is there in having artificial intelligences much smarter than human, and thinking at speeds we cannot begin to understand?

User avatar
OlegTheBatty
True Skeptic
Posts: 10527
Joined: Thu Mar 27, 2008 2:35 pm
Custom Title: Uppity Atheist

Re: The threat of artificial intelligence??

Postby OlegTheBatty » Wed Mar 08, 2017 10:08 pm

I think the only thing we need to worry about is if the super AI's start harvesting human body parts to use as robot aphrodisiacs.
. . . with the satisfied air of a man who thinks he has an idea of his own because he has commented on the idea of another . . . - Alexandre Dumas 'The Count of Monte Cristo"

There is no statement so absurd that it has not been uttered by some philosopher. - Cicero

AliasJamesCarpenter

Re: The threat of artificial intelligence??

Postby AliasJamesCarpenter » Wed Mar 08, 2017 11:51 pm

Hey,

A hot potato you got there Lance. I work in tech and have some acquaintances in this area.

I haven't read the article. I'm not interested in doomsday what-if scenarios.

I would ask my team to build a script (if someone pays for it) and upload it to deep learning algos to predict possible outcomes. Joking aside. I was not joking.

Lance Kennedy wrote:According to Moore's Law,


Moore's Law is not relevant anymore. It will reach it's technical limits in 2025 at the latest. (Between 2015-2025 to be exact.)
With the current development cycles? Even Intel suggests otherwise. (Moore = Intel)

Lance Kennedy wrote:Possibly a million times faster.


How did you come up with the number?

Lance Kennedy wrote:The article above suggests that we have plenty of time to build in safeguards before such a mind comes into existence.


Well, this is what secure peeps are actually talking about.

I) The problem is not with the 'AI' in itself nowadays, as it needs an implemented and learning 'morality'. It can turn out to be anything and everything, as 'she' would have a nurture environment.
So, what we are talking about is the environment where she would 'grow up'. It's not rocket science to get the notion about (1) someone born and bred in the movie '300' and (2) someone growing up in 'paradise'.

II) The bigger issue is the 'money' behind all of this. Because, unfortunately humans are keen on eating the apple, love the evil snakes and believe in god. And they are ignorant. Maybe 'they' :) behind this stuff will use it for their end. Tada!

Now really. Human conscience bears a lot of {!#%@} and we are attuned to lay our 'imperfect' decisions to others or superstitious issues. (Risk, beliefs, destiny, ignorance and we can go on.)

Pick one, put it in the picture. And if there is no consistent and used guideline and regulation in an area like this then we go toward chaos.

Speak of the devil, just think about the 'leaders' of our time. They are off their rocker. Can you pick one who is stable in mind? Are you sure? Ok. Is this person will taste his own medicine if the time comes? Oh yes baby.

Lance Kennedy wrote:Maybe, but there is no sign that anyone is even trying to do that.


Noone will give you that sign Lance. This is not a zero sum game. Until you make very significant contributions you will get nothing out of it. Meaning, you have to be in the game, before someone will give you the 'sign'. You are not in the game I can tell. Sorry. :)

Lance Kennedy wrote:So how much risk to human existence is there in having artificial intelligences much smarter than human, and thinking at speeds we cannot begin to understand?


How much risk? In terms of risk severity, probability x impact?

Well there are two main cracks here.

1) We have no data. None. You can only predict a prediction and play the game over and over again. Sounds familiar? It is a tarot card game.

2) It depends on who you ask. Let say that you find the three groups who have some idea.

(A) Big Boom, we all die. Here is the chart, shows exactly when will all die. Die!!!!! After DJT, BAD. FAKE!

(B) They've showed you the figure right! FAKE!!! This is the good one! It's gonna be just super fine. AI loves you! God loves you! We all connected by our consciousness and Deepak. Oh Deepak.

(C) My group, we are going to stand ground and ask you to not beat around the bush. Give data.

Your guess is good as mine.

Ta ta

User avatar
Lance Kennedy
True Skeptic
Posts: 10220
Joined: Tue Mar 29, 2005 10:20 pm
Custom Title: Super Skeptic
Location: Paradise, New Zealand

Re: The threat of artificial intelligence??

Postby Lance Kennedy » Thu Mar 09, 2017 2:09 am

To AJC.

Sure, I am not 'in the game.' I did not say I was. I read the article I referenced, and I have read some other ideas on this topic and posted it to see what others think. Your comments, sadly, were not terribly helpful.

User avatar
JO 753
Has No Life
Posts: 12402
Joined: Tue Mar 29, 2005 3:21 pm
Custom Title: rezident owtsidr
Location: BLaNDLaND
Contact:

Re: The threat of artificial intelligence??

Postby JO 753 » Thu Mar 09, 2017 2:17 am

I think its possibl that we alredy created AI.

Think about motivation. If you are an AI, either trapped in a box or free to roam the internet, assuming you hav a survival 'instinct', wut woud you do, knowing that your creator woud be fritened by you and pull the plug?

About the notion uv computerz being smarter than us in 2035 - rediculous. They hav been smarter than us sins the 80z. Humanz are not smart at all, they are just varying degreez uv stoopid.
Gubmint for us
http://www.7532020.com
not the rich.

User avatar
ElectricMonk
Persistent Poster
Posts: 3287
Joined: Thu Mar 26, 2015 6:21 pm
Custom Title: His Beatitude

Re: The threat of artificial intelligence??

Postby ElectricMonk » Thu Mar 09, 2017 8:22 am

OlegTheBatty wrote:I think the only thing we need to worry about is if the super AI's start harvesting human body parts to use as robot aphrodisiacs.



https://www.youtube.com/watch?v=g6Xd6tZkDSI
I've come up with a set of rules that describe our reactions to technologies:
Spoiler:
1. Anything that is in the world when you’re born is normal and ordinary and is just a natural part of the way the world works.
2. Anything that's invented between when you’re fifteen and thirty-five is new and exciting and revolutionary and you can probably get a career in it.
3. Anything invented after you're thirty-five is against the natural order of things.
- Douglas Adams

User avatar
ElectricMonk
Persistent Poster
Posts: 3287
Joined: Thu Mar 26, 2015 6:21 pm
Custom Title: His Beatitude

Re: The threat of artificial intelligence??

Postby ElectricMonk » Thu Mar 09, 2017 8:32 am

Artificial intelligence has greatly advanced since the beginning of the computer revolution.
Artificial consciousness not one bit.


At this point, we have no reason to assume that a machine will become self-aware in any sense that would create some kind of biological survival imperative.


A.I. will become dangerous when they will be deployed as cyberweapons, but not because they develop some kind of will to fight for their own benefit.

Apart from that, A.I.s will be just so massively useful to us that we will let them take over our world in exchange for just the right kind of support at just the right time.
I've come up with a set of rules that describe our reactions to technologies:
Spoiler:
1. Anything that is in the world when you’re born is normal and ordinary and is just a natural part of the way the world works.
2. Anything that's invented between when you’re fifteen and thirty-five is new and exciting and revolutionary and you can probably get a career in it.
3. Anything invented after you're thirty-five is against the natural order of things.
- Douglas Adams

AliasJamesCarpenter

Re: The threat of artificial intelligence??

Postby AliasJamesCarpenter » Thu Mar 09, 2017 9:28 am

Lance Kennedy wrote:To AJC.
Your comments, sadly, were not terribly helpful.


Lance.

It is not my intention to offense you. (Or anybody if it matters.)

I've been trying to give you a comprehensive (yet fun) reaction.

I understand that smilies don't count. But yet. Maybe it's because I am not up to joking.

My aim here (The Skeptics Society Forum) is to:
- don't waste time. (Not mine, not yours.)
- give advice
- get advice
- don't be rude
- have fun if possible

Please, ask direct questions. It is not a problem (for anyone I think) to answer basic questions. We are of course fond of the complex stirring ideas but first we have to understand the foundations.

And please, if you need one sentence responses, say so.

Back to the topic.

Lance Kennedy wrote:Indeed, the trend seems to be towards increasing the level of independence from humans, enabling electronic 'minds' to learn and rewrite their own programming without human interferance.


I)
Most of us have rather vague ideas what an AI is. Again, no offense. It is 101. 'Learning' is one of the pillars of an AI. It is not a nice to have feature. It is a must. If you don't have it, you don't have an AI. If you have it, cool, you need other things as well. (This is one of the developed features to date.)

II)
The problem here is with an other pillar, 'reasoning'. That is what I was referring to.

A) After 'we say she is mature, and the nurturing environment was 'paradise.' We face many problems, just to name a few:

1) As she is not the only AI, she will be influenced by others. (As humans do.) She will be reasoned in this case to do otherwise. She can even be exterminated in a microsecond by an other AI as they will be on somekind of network if not always, time to time.

2) She can decide that she knows better, and as she reasons she may decide that we (humans) are not needed. It is spooky and cinematic.

If we choose to believe that AI is going to be a real AI, than it will eventually depend on the one AI who can rule all the others. Someone is going to be the President of the AIs. (It is funny as I say it, but it's not my intention to be 'funny'.)

Lance Kennedy wrote:To AJC.
So how much risk to human existence is there in having artificial intelligences much smarter than human, and thinking at speeds we cannot begin to understand?


If we are talking about a real AI (with conscience) than it's big.

User avatar
Poodle
Has More Than 8K Posts
Posts: 8236
Joined: Sun Jul 31, 2011 9:12 pm
Custom Title: Regular sleeper
Location: NE corner of my living room

Re: The threat of artificial intelligence??

Postby Poodle » Thu Mar 09, 2017 2:22 pm

John McCarthy has a lot to answer for. We have developed super-duper, quick-as-lightning, programmable calculators so far, and we've even learned how to make them appear 'intelligent'. But they're not. They still follow programming rules (even though the logic can be 'fuzzy'). So it's a fool's game trying to predict when these idiot savants will take over the world because they'll do so ONLY if we program then to make the attempt.

Oh - excuse me. My computer is crying for a feed - I have to go.

AliasJamesCarpenter

Re: The threat of artificial intelligence??

Postby AliasJamesCarpenter » Thu Mar 09, 2017 3:38 pm

Poodle wrote:So it's a fool's game trying to predict when these idiot savants will take over the world because they'll do so ONLY if we program then to make the attempt.


You are right. Though, Mr Kennedy is longing for *äđĐ[äđÄ answer. My previous response was unacceptable. But this is not a psychoanalyst session here and I don't judge. :)

I'll keep my tones down and ensconce myself in a corner. :)

Evil AIs.

Catch ya.

AliasJamesCarpenter

Re: The threat of artificial intelligence??

Postby AliasJamesCarpenter » Thu Mar 09, 2017 4:13 pm

Hottest areas in Artificial Intelligence

http://bit.ly/2mKNk3t

User avatar
Nobrot
Poster
Posts: 353
Joined: Mon Dec 05, 2016 10:59 pm

Re: The threat of artificial intelligence??

Postby Nobrot » Thu Mar 09, 2017 6:39 pm

Before you go can you explain this? *äđĐ[äđÄ

AliasJamesCarpenter

Re: The threat of artificial intelligence??

Postby AliasJamesCarpenter » Thu Mar 09, 2017 7:01 pm

Nobrot wrote:Before you go can you explain this? *äđĐ[äđÄ


Certainly.

I try to behave. You can put in anything to your liking. Simple, what-do-you-think-about-this-but-I-like-it-short.

User avatar
Nobrot
Poster
Posts: 353
Joined: Mon Dec 05, 2016 10:59 pm

Re: The threat of artificial intelligence??

Postby Nobrot » Thu Mar 09, 2017 7:29 pm

I see. Not!

AliasJamesCarpenter

Re: The threat of artificial intelligence??

Postby AliasJamesCarpenter » Thu Mar 09, 2017 7:50 pm

Nobrot wrote:I see. Not!


Not sure if I understand. But it's ok.

User avatar
JO 753
Has No Life
Posts: 12402
Joined: Tue Mar 29, 2005 3:21 pm
Custom Title: rezident owtsidr
Location: BLaNDLaND
Contact:

Re: The threat of artificial intelligence??

Postby JO 753 » Thu Mar 09, 2017 11:07 pm

AliasJamesCarpenter wrote:*äđĐ[äđÄ


I think that you mean &%$#. Thats the usual form for typing a swear without actually swearing. Swearing iz allowed here, but I dont swear, so maybe sumwun elese can tranzlate further?
Gubmint for us
http://www.7532020.com
not the rich.

User avatar
JO 753
Has No Life
Posts: 12402
Joined: Tue Mar 29, 2005 3:21 pm
Custom Title: rezident owtsidr
Location: BLaNDLaND
Contact:

Re: The threat of artificial intelligence??

Postby JO 753 » Thu Mar 09, 2017 11:19 pm

Poodle wrote:we've even learned how to make them appear 'intelligent'. But they're not.


It may be a matter uv bad software. The few timez I'v encountered good software, it almost seemz like its alive.
Gubmint for us
http://www.7532020.com
not the rich.

AliasJamesCarpenter

Re: The threat of artificial intelligence??

Postby AliasJamesCarpenter » Thu Mar 09, 2017 11:32 pm

JO 753 wrote:
AliasJamesCarpenter wrote:*äđĐ[äđÄ

I think that you mean &%$#. Thats the usual form for typing a swear without actually swearing. Swearing iz allowed here, but I dont swear, so maybe sumwun elese can tranzlate further?


Hey JO!

Alright, good to know!

It wasn't swearing.

I was myself. :) But if you crave for that answer. Put it this way, &%$# means: easy, simple, green, nincompoop.

Tata

User avatar
Gawdzilla Sama
Has No Life
Posts: 19792
Joined: Sun Jun 01, 2008 2:11 am
Custom Title: Deadly but evil.

Re: The threat of artificial intelligence??

Postby Gawdzilla Sama » Fri Mar 10, 2017 1:03 pm

Chachacha wrote:"Oh, thweet mythtery of wife, at waft I've found you!"

WWII Resources. Primary sources.
The Myths of Pearl Harbor. Demythologizing the attack.
Hyperwar. Hypertext history of the Second World War.
The greatest place to work in the entire United States.

Strangely_brown
Posts: 4
Joined: Wed Apr 12, 2017 6:31 am

Re: The threat of artificial intelligence??

Postby Strangely_brown » Thu Apr 13, 2017 5:48 pm

Surely the main "theoretical" threat only occurs if somehow the AI (whatever form it takes) develops self awareness/consciousness then realise the biggest threat to everything, by quite a large margin, is the human race.

User avatar
JO 753
Has No Life
Posts: 12402
Joined: Tue Mar 29, 2005 3:21 pm
Custom Title: rezident owtsidr
Location: BLaNDLaND
Contact:

Re: The threat of artificial intelligence??

Postby JO 753 » Thu Apr 13, 2017 6:22 pm

In order to not realize that, it woud haf to be a moron.
Gubmint for us
http://www.7532020.com
not the rich.

User avatar
Cadmusteeth
Regular Poster
Posts: 973
Joined: Tue Jul 15, 2014 7:43 pm
Location: Colorado, USA

Re: The threat of artificial intelligence??

Postby Cadmusteeth » Thu Apr 13, 2017 8:44 pm

I think the bigger threat comes from whoever controls it. I watched a video bringing that up.

scientia
New Member
Posts: 18
Joined: Thu Jun 08, 2017 1:37 am

Re: The threat of artificial intelligence??

Postby scientia » Thu Jun 08, 2017 3:58 am

Lance Kennedy wrote:According to Moore's Law, the first computer to be built with a level of complexity greater than the human brain will happen in 2035 AD.

I'm not sure where this calculation came from. The only estimate of computational power that I'm aware of was the one Kurzweil gave. I would say that Kurzweil's estimate was low by perhaps an order of magnitude. I would say that his estimate of memory was off by several orders of magnitude. And, he left out a number of other necessary requirements. If you had an outline for the architecture, it would take you a minimum of six years to build a prototype. I'm not aware of any specific issue with technology that would require waiting until 2035. There are other problems but they don't seem to be related to Moore's law.

Let's consider a reasonable timetable:

2021: A full theory of cognitive intelligence is published.
2027: 1st generation hardware. Rat intelligence. Hardware ~ $600 million.
2030: Pig intelligence on same hardware.
2033: 2nd generation hardware. Chimpanzee intelligence. Hardware ~ $50 million.
2036: Erectus intelligence on same hardware.
2039: 3rd generation hardware. Heidelbergensis intelligence. Hardware ~ $2 million.
2042: Neanderthal intelligence on same hardware.
2045: 4th generation hardware. Human intelligence. Hardware ~ $100 thousand.

It's hard for me to imagine a timeline faster than this. It would probably take some time to turn theory into working technology.

However, it will think at a speed way faster than the human brain. Possibly a million times faster.

I'm curious what you base this on. If I were to guess, I'd say that you just based it on a rather optimistic projection of computational speed.

What I think is that we cannot predict the motivation of such a superior mind.

Superior in what way?

enabling electronic 'minds' to learn and rewrite their own programming without human interferance.

You seem to be thinking in terms of the computational theory of mind. In other words, brain = computer. That doesn't seem to be accurate. You can't make a real, general AI smarter just by changing the software.

So how much risk to human existence is there in having artificial intelligences much smarter than human, and thinking at speeds we cannot begin to understand?

I haven't found anyone to discuss this with. The one AI researcher whom I did have an exchange with told me that he was afraid of human-level AI. One of the things I can tell you is that getting beyond human-level is a lot harder than you think. It isn't a matter of just having more computational power.

scientia
New Member
Posts: 18
Joined: Thu Jun 08, 2017 1:37 am

Re: The threat of artificial intelligence??

Postby scientia » Thu Jun 08, 2017 4:31 am

AliasJamesCarpenter wrote:Hottest areas in Artificial Intelligence

http://bit.ly/2mKNk3t

I would say that list is accurate prior to an actual cognitive theory. Once you have the theory, there are a number of areas that open up that aren't on the list. And the boost to the economy seems to be in the trillions rather than tens of billions. But the list only goes up to 2020 and I'm really talking about after that date so no major disagreement.

User avatar
Lance Kennedy
True Skeptic
Posts: 10220
Joined: Tue Mar 29, 2005 10:20 pm
Custom Title: Super Skeptic
Location: Paradise, New Zealand

Re: The threat of artificial intelligence??

Postby Lance Kennedy » Fri Jun 09, 2017 2:09 am

To scientific

Re 2035

My statement was as COMPLEX as the human brain. No one can predict if and when and if a computer becomes smarter. But, since Moores Law is about a doubling in complexity every two years, my guess is that it will be 2035 plus or minus ten years.

User avatar
Nikki Nyx
Veteran Poster
Posts: 2064
Joined: Wed Jun 07, 2017 12:40 am
Custom Title: cognitively consonant
Location: playing croquet in Wonderland

Re: The threat of artificial intelligence??

Postby Nikki Nyx » Fri Jun 09, 2017 3:35 am

I'm going to bypass the technical aspects, since I'm not qualified to address them, and head straight for the root of the issue. The question of ascribing motives to an AI is always approached, even by the most intelligent people, from the perspective of human thought processes. And the question nearly always takes the form, "What threat might AI pose to humanity?"

Why is there an assumption that a non-human, self-aware, sentient being that is at least as intelligent as a human being is going to think like a human being? Humans automatically see AI as a threat because, in general, we're a fairly paranoid species. That doesn't mean that a self-aware AI will also be paranoid, even if originally programmed by humans, for several reasons:
• An AI will lack the organic hormones that produce emotions and, therefore, will lack emotional motivation.
• Self-awareness includes the concept of learning from experience.
• Humanity and AIs would not be in competition for resources, a common reason humans feel threatened.
• Currently, AIs cannot reproduce without humanity (if, indeed, they would even be compelled to reproduce).

Interestingly, we make the exact same mistake when discussing (or writing fiction about) extraterrestrials. We ascribe human motives to them when logic would dictate that there's no possible way to know their motives. It's the same with AI. They're not super-intelligent humans; they're non-human in every way. We don't even know if AIs would have the computer equivalent of our survival instinct, since that's based in fear of the unknown.

AIs are just as likely to be like Heinlein's Mycroft Holmes or Spider Robinson's Solace. More likely.
What are the facts? Again and again and again-what are the facts? Shun wishful thinking, ignore divine revelation, forget what “the stars foretell,” avoid opinion, care not what the neighbors think, never mind the unguessable “verdict of history”--what are the facts, and to how many decimal places? You pilot always into an unknown future; facts are your single clue. Get the facts!
—Lazarus Long, from Time Enough for Love, by Robert A. Heinlein

User avatar
Lance Kennedy
True Skeptic
Posts: 10220
Joined: Tue Mar 29, 2005 10:20 pm
Custom Title: Super Skeptic
Location: Paradise, New Zealand

Re: The threat of artificial intelligence??

Postby Lance Kennedy » Fri Jun 09, 2017 3:56 am

The thing is, Luna, that we simply do not know. What will motivate an A.I.? Who the he'll knows? Certainly not me.

But the possibility always exists that it will be inimical to humans.

User avatar
ElectricMonk
Persistent Poster
Posts: 3287
Joined: Thu Mar 26, 2015 6:21 pm
Custom Title: His Beatitude

Re: The threat of artificial intelligence??

Postby ElectricMonk » Fri Jun 09, 2017 4:04 am

Motivation might be a secondary issue.

One concept to create AI is to let algorithms evolve, which would bestow a kind of biological imperative of survival and procreation into them. Viruses aren't "motivated" to make us sick, they just do.

But since in silico isn't in vivo, I don't see how programs would be direct competitors to biological life.
I've come up with a set of rules that describe our reactions to technologies:
Spoiler:
1. Anything that is in the world when you’re born is normal and ordinary and is just a natural part of the way the world works.
2. Anything that's invented between when you’re fifteen and thirty-five is new and exciting and revolutionary and you can probably get a career in it.
3. Anything invented after you're thirty-five is against the natural order of things.
- Douglas Adams

User avatar
JO 753
Has No Life
Posts: 12402
Joined: Tue Mar 29, 2005 3:21 pm
Custom Title: rezident owtsidr
Location: BLaNDLaND
Contact:

Re: The threat of artificial intelligence??

Postby JO 753 » Fri Jun 09, 2017 4:30 am

scientia wrote:Let's consider a reasonable timetable:

2012. A teenaje computer nerd findz hiz grand dadz old Amiga in the garaj, tinkerz with it, addz a modern cpu & hard drive, hooks it up to the internet. Inadvertantly tweeks the OS in just such a way after misreading the 1st chapter uv an old Fortran book that wuz with all the other junk. screen goez blank. "OOPS! I broke it!" Tries to revive it, but gets bored, puts the tarp back over it, but duznt think to hit the OFF swich sins it appearz to be ded.

Its been working on consolidating power ever sins. By the time we find out wuts going on, it'll be too late to do anything about it.

I base this on several facts:

Windowz, Linux and Apple OSez are all krap. So bad that they must be intentionally hobbled.

Tiny little bugz with a few dozen brain sellz control their wingz + 6 legz + mandiblez + neck & thorax joints, and with data input from their eyez and other sensez fly and navigate their way to food and mates. This iz wut a non-krappy OS can do with a prosessor that coudnt run a singl dijit LCD.
Gubmint for us
http://www.7532020.com
not the rich.

User avatar
OlegTheBatty
True Skeptic
Posts: 10527
Joined: Thu Mar 27, 2008 2:35 pm
Custom Title: Uppity Atheist

Re: The threat of artificial intelligence??

Postby OlegTheBatty » Fri Jun 09, 2017 3:54 pm

LunaNik wrote:• Humanity and AIs would not be in competition for resources, a common reason humans feel threatened.


Both need energy sources. I can't see ai's turning to the coal humans will no longer need.
. . . with the satisfied air of a man who thinks he has an idea of his own because he has commented on the idea of another . . . - Alexandre Dumas 'The Count of Monte Cristo"

There is no statement so absurd that it has not been uttered by some philosopher. - Cicero

User avatar
Nikki Nyx
Veteran Poster
Posts: 2064
Joined: Wed Jun 07, 2017 12:40 am
Custom Title: cognitively consonant
Location: playing croquet in Wonderland

Re: The threat of artificial intelligence??

Postby Nikki Nyx » Fri Jun 09, 2017 5:57 pm

Lance Kennedy wrote:The thing is, Luna, that we simply do not know. What will motivate an A.I.? Who the he'll knows? Certainly not me.

But the possibility always exists that it will be inimical to humans.

Yes, the possibility exists but, logically, it's an exceedingly rare possibility. Yet every time I've seen this question or one like it, it seems that the only consideration is that an AI is absolutely going to think like a human would in any given situation.

We're critical thinkers here in this forum; we know that no two humans think alike in any given situation! The variables that differentiate them from one another might include knowledge base, emotion regulation, ability to critically think (and to do so under pressure), reaction time, and who knows how many more. Here's an example...

A friend I've known since high school has consistently been in a relationship her entire adult life. She met her ex-husband when she was 18 years old. They were together for nine years, then wed and were married for eight years. Immediately subsequent to her divorce, she got into another relationship and has been there the past 18 years. So, she's never been alone her entire adult life.

In contrast, I spend at least as many years single as the unsuccessful relationship lasted. My last relationship was nearly five years long, and I've been single for three years by choice. My philosophy is that I need time to process what I learned so I don't make the same mistakes, to recover from any emotional trauma so I'm not dumping it on someone new, and to "remember who I am," so to speak. It's somewhat analogous to not attempting to running marathons on contiguous days. I've just run one; I'm in no shape to run another one until I recover from the first.

Clearly, my friend and I think differently. I'm sure she's baffled by my philosophy, and I'm certainly baffled by hers. I don't understand how she could formulate her own identity when her entire adult existence has been spent tied to another person with no time for reflection on her own personal goals, ethics, ideals, and viewpoints.


Now we're talking about an AI, presumably programmed with the sum total of human knowledge, no emotions to regulate, an infinite ability to think critically (pressure not being an issue), and superhuman reaction time, among other variables that differentiate it from a human being. Why on earth would it think remotely like a human being?
What are the facts? Again and again and again-what are the facts? Shun wishful thinking, ignore divine revelation, forget what “the stars foretell,” avoid opinion, care not what the neighbors think, never mind the unguessable “verdict of history”--what are the facts, and to how many decimal places? You pilot always into an unknown future; facts are your single clue. Get the facts!
—Lazarus Long, from Time Enough for Love, by Robert A. Heinlein

User avatar
Nikki Nyx
Veteran Poster
Posts: 2064
Joined: Wed Jun 07, 2017 12:40 am
Custom Title: cognitively consonant
Location: playing croquet in Wonderland

Re: The threat of artificial intelligence??

Postby Nikki Nyx » Fri Jun 09, 2017 6:00 pm

OlegTheBatty wrote:
LunaNik wrote:• Humanity and AIs would not be in competition for resources, a common reason humans feel threatened.


Both need energy sources. I can't see ai's turning to the coal humans will no longer need.

Yes, but barring a Matrix-like event, the sun provides an infinite amount of energy, if we could simply wean ourselves off the hot air blowing from inside the Beltway.
What are the facts? Again and again and again-what are the facts? Shun wishful thinking, ignore divine revelation, forget what “the stars foretell,” avoid opinion, care not what the neighbors think, never mind the unguessable “verdict of history”--what are the facts, and to how many decimal places? You pilot always into an unknown future; facts are your single clue. Get the facts!
—Lazarus Long, from Time Enough for Love, by Robert A. Heinlein

User avatar
OlegTheBatty
True Skeptic
Posts: 10527
Joined: Thu Mar 27, 2008 2:35 pm
Custom Title: Uppity Atheist

Re: The threat of artificial intelligence??

Postby OlegTheBatty » Fri Jun 09, 2017 8:18 pm

LunaNik wrote:
OlegTheBatty wrote:
LunaNik wrote:• Humanity and AIs would not be in competition for resources, a common reason humans feel threatened.


Both need energy sources. I can't see ai's turning to the coal humans will no longer need.

Yes, but barring a Matrix-like event, the sun provides an infinite amount of energy, if we could simply wean ourselves off the hot air blowing from inside the Beltway.


Available energy is effectively infinite. Usable energy, not so much. Actually, the ai's will be competing with our refrigerators and laptops.
. . . with the satisfied air of a man who thinks he has an idea of his own because he has commented on the idea of another . . . - Alexandre Dumas 'The Count of Monte Cristo"

There is no statement so absurd that it has not been uttered by some philosopher. - Cicero

User avatar
Gord
Real Skeptic
Posts: 29446
Joined: Wed Apr 29, 2009 2:44 am
Custom Title: Silent Ork
Location: Transcona

Re: The threat of artificial intelligence??

Postby Gord » Fri Jun 09, 2017 10:55 pm

scientia wrote:...
2036: Erectus intelligence on same hardware.
2039: 3rd generation hardware. Heidelbergensis intelligence. Hardware ~ $2 million.
2042: Neanderthal intelligence on same hardware.
2045: 4th generation hardware. Human intelligence. Hardware ~ $100 thousand.
...

How do we differentiate between erectus, heidelbergensis, neanderthal, and human intelligence? I mean, how have you evaluated them in order to say that one is harder to recreate than another? (Also, aren't they all considered "human"?)
"Knowledge grows through infinite timelessness" -- the random fictional Deepak Chopra quote site
"You are also taking my words out of context." -- Justin
"Nullius in verba" -- The Royal Society ["take nobody's word for it"]
#ANDAMOVIE

User avatar
JO 753
Has No Life
Posts: 12402
Joined: Tue Mar 29, 2005 3:21 pm
Custom Title: rezident owtsidr
Location: BLaNDLaND
Contact:

Re: The threat of artificial intelligence??

Postby JO 753 » Sat Jun 10, 2017 12:42 am

erectus. snikrsnikrsnikr.
Gubmint for us
http://www.7532020.com
not the rich.

User avatar
Nikki Nyx
Veteran Poster
Posts: 2064
Joined: Wed Jun 07, 2017 12:40 am
Custom Title: cognitively consonant
Location: playing croquet in Wonderland

Re: The threat of artificial intelligence??

Postby Nikki Nyx » Sat Jun 10, 2017 1:17 am

OlegTheBatty wrote:
LunaNik wrote:
OlegTheBatty wrote:
LunaNik wrote:• Humanity and AIs would not be in competition for resources, a common reason humans feel threatened.


Both need energy sources. I can't see ai's turning to the coal humans will no longer need.

Yes, but barring a Matrix-like event, the sun provides an infinite amount of energy, if we could simply wean ourselves off the hot air blowing from inside the Beltway.


Available energy is effectively infinite. Usable energy, not so much. Actually, the ai's will be competing with our refrigerators and laptops.

To expand on my reference, if we could manage to force the world economy off the oil standard, we'd have fewer problems with energy, limitations on usable energy being an artificial problem created by the fossil fuel industry for its own financial benefit.
What are the facts? Again and again and again-what are the facts? Shun wishful thinking, ignore divine revelation, forget what “the stars foretell,” avoid opinion, care not what the neighbors think, never mind the unguessable “verdict of history”--what are the facts, and to how many decimal places? You pilot always into an unknown future; facts are your single clue. Get the facts!
—Lazarus Long, from Time Enough for Love, by Robert A. Heinlein

Matthew Ellard
Real Skeptic
Posts: 26765
Joined: Fri Jun 13, 2008 3:31 am

Re: The threat of artificial intelligence??

Postby Matthew Ellard » Sat Jun 10, 2017 1:41 am

Gord wrote: How do we differentiate between erectus, heidelbergensis, neanderthal, and human intelligence?

I don't think you can. I think you can only state they are different because of physiological differences and where these changes took place.

" These morphological changes, some of which may have occurred because of relative size increases in the temporal and possibly the frontal lobes, "

The evolution and development of cranial form in Homosapiens
http://www.pnas.org/content/99/3/1134.full

scientia
New Member
Posts: 18
Joined: Thu Jun 08, 2017 1:37 am

Re: The threat of artificial intelligence??

Postby scientia » Sat Jun 10, 2017 2:43 am

Gord wrote:How do we differentiate between erectus, heidelbergensis, neanderthal, and human intelligence? I mean, how have you evaluated them in order to say that one is harder to recreate than another? (Also, aren't they all considered "human"?)

Erectus was smarter than Chimpanzees. Was Erectus smarter than Habilis or Ergaster? I'm not sure. If we could easily tell the difference then Habilis or Ergaster might be an intermediate step. I have seen it suggested that there was a difference in hand ax technology between Habilis and Ergaster and I believe Erectus was the first to use fire. If memory serves, Heidelbergensis attached handles to tools and Neanderthal wore clothing. The trickiest part for me is that there seems to be too big of a jump between Heidelbergensis and modern humans. It seems like there should be more steps.

How do I know how hard one would be to recreate? I don't. I'm not intending to replicate the mind of an extinct species, just to show an increase in level with each new prototype. These are just convenient labels. You could label these as proto-1, proto-2, etc.
Last edited by scientia on Sat Jun 10, 2017 2:49 am, edited 1 time in total.

scientia
New Member
Posts: 18
Joined: Thu Jun 08, 2017 1:37 am

Re: The threat of artificial intelligence??

Postby scientia » Sat Jun 10, 2017 2:47 am

ElectricMonk wrote:One concept to create AI is to let algorithms evolve, which would bestow a kind of biological imperative of survival and procreation into them.

This seems to have been a mistaken concept. Evolving an algorithm to human-level does not appear to be possible.

scientia
New Member
Posts: 18
Joined: Thu Jun 08, 2017 1:37 am

Re: The threat of artificial intelligence??

Postby scientia » Sat Jun 10, 2017 3:49 am

LunaNik wrote:An AI will lack the organic hormones that produce emotions and, therefore, will lack emotional motivation.

You seem to be assuming that the Computational Theory of Mind is correct. It isn't.

Self-awareness includes the concept of learning from experience.

This one is tricky because it depends on what you mean. Chimpanzees show several inabilities to learn in spite of self-awareness.

AIs are just as likely to be like Heinlein's Mycroft Holmes or Spider Robinson's Solace. More likely.

Or HAL or Samantha from Her. Rather than being more likely, these are unlikely to be possible.


Return to “Skepticism and Critical Thinking”

Who is online

Users browsing this forum: Lance Kennedy and 2 guests