AI, Is Concern Warranted?

[Ophiolite]

I am amazed no one has mentioned Azimov's Three Laws of Robotics.

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws

If we incorporate code in any developing AI, to the effect above, we have gone a long way to removing most concerns.

 

[Radrook]

When dealing with AI we cannot completely use the human perspective.

For starters, AI/computers doesn't have to deal with hormones or body chemicals which could induce emotions, pleasure, and affect decision-making.

If you remove those components in human physiology, including the sensation of pain, we'll probably behave like robots too.

We would simply look at our world purely from an information or logical standpoint. Efficiency and Order will be the rule..




Could Jesus be talking about computers here?

Luke 19:40
"I tell you," he replied, "if they keep quiet, the stones will cry out."

The most abundant mineral in stones (rocks) are silicates. It is from silicates we get silicon. Silicon forms the core physical component of information processing in computers.

Of course, it took thousands of years before that promise is fulfilled if that is the case. But time is usually seen differently from a divine perspective.

I totally agree about the hormones which influence our thoughts via helping to produce emotions after they are released into the bloodstream. Recreation of an android that would be the exact representation of us would indeed seem to require that it have analogous chemicals in its system that would trigger similar reactions. Also, perhaps that material; employed in the android creation would also have to be analogous to flesh and not metal as a certain expert on the subject suggested in his video.

About Jesus and the stones scripture which you suggest might be prophetic in relation to computers? Well, I never imagined that he meant it that way but since he has access to the future developments I would not be surprised if indeed he had that in mind as well. Very interesting thought!

 

[Radrook]

I am amazed no one has mentioned Azimov's Three Laws of Robotics.

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
2. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws

If we incorporate code in any developing AI, to the effect above, we have gone a long way to removing most concerns.

Certainly! That was in the back of my mind all along and I was waiting for someone to bring it up. Thanks for finally specifically mentioning it. It does make plenty of common sense to include safeguards against the type of malfunction which would endanger us and it's really hard to conceive that the designers of androids would mindlessly neglect such an obvious necessity. Asimiov's Three Laws were alluded to in the film Aliens.

 

[AV1611VET]

About Jesus and the stones scripture which you suggest might be prophetic in relation to computers? Well, I never imagined that he meant it that way but since he has access to the future developments I would not be surprised if indeed he had that in mind as well. Very interesting thought!

Computers run on silicon.

Silicon is mined from sand.

And what did Jesus say about building a house on the sand?

 

[Radrook]

Computers run on silicon.

Silicon is mined from sand.

And what did Jesus say about building a house on the sand?

You have very interesting insights!

 

[AV1611VET]

You have very interesting insights!

Amen! Thank you!

They say blind people have heightened senses of hearing and smelling.

I like to think that God made me scientifically challenged, but insightful.

 

[Radrook]

Amen! Thank you!

They say blind people have heightened senses of hearing and smelling.

I like to think that God made me scientifically challenged, but insightful.

James 1:17 - Every good gift and every perfect gift is from above, and cometh down from the Father of lights, with whom is no variableness, neither shadow of turning.


1 Corinthians 12:1 - 12:31

 

[durangodawood]

Computers run on silicon.

Silicon is mined from sand.

And what did Jesus say about building a house on the sand?

If youre going to be that literal, then lets all note that a computer is not a house.

 

[RDKirk]

Not any closer? Are you kidding?


How far are we from Nanobots?

Nanosilver based products have already been rapidly commercialized worldwide. So it’s pretty much safe to conclude that we are not far away from Nanobots or Nano robots. In fact, we’re pretty close.

LOL. A graph about computer technology that starts at 1970?

You realize that there weren't even pocket calculators in 1970, right? So if you start the graph far back enough, even the most moderate increase looks tremendous. That's an old, old statistics trick I was hippped to in Media 101...back in 1970, before calculators.

That article says:

"Nanotechnology is an emerging field of science dealing with substances or materials in the ‘nano’ or nanometer range. "

That's not just micro-robots. That definition includes electricians and photographers. Again, another trick. Use a hugely vague definition as though it were proof of a specific point.

 

[AV1611VET]

If youre going to be that literal, then lets all note that a computer is not a house.

But it's part of an infrastructure built on sand.

 

[RDKirk]

Certainly! That was in the back of my mind all along and I was waiting for someone to bring it up. Thanks for finally specifically mentioning it. It does make plenty of common sense to include safeguards against the type of malfunction which would endanger us and it's really hard to conceive that the designers of androids would mindlessly neglect such an obvious necessity. Asimiov's Three Laws were alluded to in the film Aliens.

The Three Laws works for Asimov's stories because he posited only one company literally in the entire universe building robots, so they could enforce a company policy. (Asimov's "Robots of Dawn" illustrated why there was only one company in the entire universe building robots.)

 

[durangodawood]

But it's part of an infrastructure built on sand.

I opened my computer and sand fell out!

 

[TagliatelliMonster]

I regularly play chess against computers on the Internet and the computers SEEM to be analyzing the position in the same way I am, visually. However, that isn't really the case at all. The computer is merely dealing with numerical representations of the position of the pieces on a numerical representation of the board and calculating outcomes based on the dynamic interrelations of those values. In short, it doesn't see pieces or a board visually. It is merely reacting to mathematical inputs and responding by providing a solution via mathematical output. But to the human player it seems otherwise since the mimicry of human thought is so convincing that it is very hard to believe that the machine isn't actually consciously pondering as we are.

Talking about computers and chess... My dad is a chess enthousiast and has loads of specialised chess software which use cutting edge "chess engines", backed by a database of millions of chess games, even going back all the way to documented games of chessmasters of the 1600s. Next to the actual "AI" calculations those engines perform during a game, these engines also cross reference the current game against that entire database. These engines also automatically scale as you add more CPU horsepower to the hardware.

So the dificulty level of playing against the CPU will also be dependent on how fat the computer is that is running the software.

Most of the time, my dad wins. On max level. He is really smart.
But, more importantly (in his opinion, anyway), he is very human.

He always says that "it's still a computer. It can not think like a human. All it takes for me to beat it, is to think like a human and not play chess like a predictive calculator"

The way he wins against the pc, is (in his own words) by setting up a "human trap" that a machine simply can't see. I'm not sure what that means exactly, but it sounds smart.

It's pretty much in line with the way Garry Kasparov approached his matches against Deep Blue (the IBM supercomputer that played a few chess matches against Kasparov in the late 90s).

If memory serves me right, in total it was about 50/50 for wins vs losses/draws.

But think about that for a second....
A supercomputer, which was designed to play chess, didn't manage to whipe the floor consistently with a human. In fact, the supercomputer lost half the time.

This computer was an absolute beast that calculated the outcome of millions of potential chess moves in a split second.

Off course, that was 20 years ago (wow, I'm getting old...) and I probably wouldn't even last for much longer then a minute against that computer... But the fact that a human actually was able to achieve such a result... It's quite remarkable imo.

 

[timewerx]

I totally agree about the hormones which influence our thoughts via helping to produce emotions after they are released into the bloodstream. Recreation of an android that would be the exact representation of us would indeed seem to require that it have analogous chemicals in its system that would trigger similar reactions. Also, perhaps that material; employed in the android creation would also have to be analogous to flesh and not metal as a certain expert on the subject suggested in his video.

I don't think it would wise to model something from a corrupted creation such as we.

The whole point of automation (which I assume is one of the major reasons for AI) is to offset human labor but more importantly, to improve efficiency of certain tasks.

If machine intelligence is to be deliberately held back by processes which makes us inefficient, wouldn't that be absurd?

The only reason I would give AI "flesh" is to test it but only in a virtual or simulated environment. They would be rated on how well they overcome the limitations of the "flesh".

However, once the AI have passed such test, they will be taken out of the virtual world, be given unconstrained hardware and be allowed to exist in the real world.

It's possible, we are AI ourselves, being tested in a simulated universe.

 

[Radrook]

LOL. A graph about computer technology that starts at 1970?

You realize that there weren't even pocket calculators in 1970, right? So if you start the graph far back enough, even the most moderate increase looks tremendous. That's an old, old statistics trick I was hippped to in Media 101...back in 1970, before calculators.

That article says:

"Nanotechnology is an emerging field of science dealing with substances or materials in the ‘nano’ or nanometer range. "

That's not just micro-robots. That definition includes electricians and photographers. Again, another trick. Use a hugely vague definition as though it were proof of a specific point.

Well, I see no reason why the authors would maliciously go out of their way to employ the deceptive strategies you describe. I perceive them as being merely striving to be informative by providing a timeline instead of striving to trick the reader.

BTW
You provide absolutely no reasons why attaining what is predicted on the chart is impossible.

 

[Radrook]

Oh, I think the concern is much less than it is for nanobots.

So you consider the concern for nanobots more warranted than concern for an AI?
Sorry! I just found that you did address this question an it was no. I'm not sure that I totally agree however. There is what we might term unjustifiably intense concern and a certain concern which is justifiable. I think that a certain amount of concern is indeed warrantied. However, I agree that an unjustifiable irrational amount of concern isn't warranted. So I guess it's a matter of degree.

 

[Greg J.]

Well, IBM can arrange atoms so they say "IBM" and "Star Trek," but the finest artificial intelligence can't read a children's book and explain it.

 

[Radrook]

I don't think it would wise to model something from a corrupted creation such as we.

The whole point of automation (which I assume is one of the major reasons for AI) is to offset human labor but more importantly, to improve efficiency of certain tasks.

If machine intelligence is to be deliberately held back by processes which makes us inefficient, wouldn't that be absurd?

The only reason I would give AI "flesh" is to test it but only in a virtual or simulated environment. They would be rated on how well they overcome the limitations of the "flesh".

However, once the AI have passed such test, they will be taken out of the virtual world, be given unconstrained hardware and be allowed to exist in the real world.

It's possible, we are AI ourselves, being tested in a simulated universe.

Very good points! True, efficiency to perform tasks is indeed a very important factor in the field of robotics and creating an android with full human limitations would be counterproductive in that sense. For example, providing a robot with emotional components would make it susceptible to all the irrationality that emotions trigger in humans. Such emotions lead to violence, inconsistencies of policies, unnecessary expenditure of energy in useless irrelevant worrying. One Isaac Asimov sci fi short story had a robot commit suicide when it was suddenly endowed with emotional capacities.

So the ideal would seem to create an android which is gioverned by laws as Asimove delineated but which is nevertheless capable of simulating the emotions of human compassion and all other characteristics which are conisdered admirable.

Actually, providing such a creation with a capacity iof self awareness might be considered a cruelty from a certain perspective since we might endow it with the capacity to suffer the horrible realization that it is a freak of nature.

That is the theme of the short story "I have No Mouth but I Must Scream" where the computer realizes its own freakish nature and turns on mankind as an act of revenge.

Here is a very interesting video on robotic consciousness:

 

[TagliatelliMonster]

Well, IBM can arrange atoms so they say "IBM" and "Star Trek," but the finest artificial intelligence can't read a children's book and explain it.

Hmm. That's not entirely true actually...

Sure, we still have a long way to go, but AI engines specialising in understanding and interpreting "natural language" has come a long way already. We are seeing some application of this pop up in the real world also.

Speech recognition is getting better and better.
Then there are also the automated translation applications, which are also slowly but surely become better at interpreting metaphores, sayings, figures of speech and other such symbolic language. That is off course what the real challenge is in translating into another language. It is not sufficient to simply look up the "other language" words one by one and paste them together. More often then not, that just results in nonsense.

There's also this project from Microsoft, which translates spoken language in real time over skype.
You speak english to the computer and at the other side, it comes out in chinese.

 

[Greg J.]

Which has nothing to do with recognizing abstract concepts intended by an author by understanding the culture for which it was written, such that the knowledge can be used anticipate questions people would ask and what information is important (and what is not) so it can be explained to them (e.g., thematic elements of the book, the use of color, idioms, mood, and photos) before they say anything.