At some point, should advanced AI entities have rights and privileges, or will they always be just machines?

[dfjchem721]

In the movie 2001 : A Space Odyssey, Hal is effectively "executed" when Dave pulls out his memory chips after Hal goes "off the rails". Hal was actually ordered to kill off the crew if they jeopardized the mission. But Dave didn't know this. (He would have "de-commissioned" Hal anyway, but that is beside the point).

Should some level of AI, likely to play a huge role in space exploration, have rights and privileges?

Could there be such a thing as criminal "extra-judicial" actions against an AI entity, and if so, what would be the basis of this?

Daisy, Daisy,
Give me your answer, do!

 

[COLGeek]

...and Skynet begins.

 

[Hudata]

I don't believe that AI's will ever achieve actual sentience. although I have no doubt that they will simulate it convincingly in the near future. The more we learn about the complexity of the human brain and the nature of consciousness the less likely it seems to me that we will ever be able to understand it well enough to do more than simulate it.

 

[dfjchem721]

I don't believe that AI's will ever achieve actual sentience. although I have no doubt that they will simulate it convincingly in the near future. The more we learn about the complexity of the human brain and the nature of consciousness the less likely it seems to me that we will ever be able to understand it well enough to do more than simulate it.

Agreed, Hudata. We have in fact seen something of such "simulation" already. Back in the 1990's, IBM built a super-computer called "Deep Blue", and programmed it to beat all chess players in the world, or so they thought. They put it up against the best that ever played the game, Gary Kasparov. When they first played, Kasparov had a decided advantage and won 4-2. But still an amazing feat of computer programming to win 2 games.

Then IBM reprogrammed Deep Blue, and they played again. This time Deep Blue won, 3½–2½. Kasparov was adamant that the computer had been reprogrammed to beat Gary Kasparov, and not just everybody. He now downplays his and the computer's performances, but not about the primary claim about reprogramming. IBM refuses to divulge what they did to "win".

Perhaps it was simply a highly glorified mimic of the human player, with all its computing power behind it. If you could put it up against Bobby Fischer, it might have lost 4-0. It was not adaptable intelligence that beat Kasparov, it was processor speed and considering all the variable moves of Kasparov, whose many games were well known to IBM. That is not intelligence, but mere computing capability. A mimic.

As you suggest, true AI is not likely to reach the realm of human thought. Playing a chess game is one thing. But logic and deductive reasoning are not moves on a chess board. They are orders of magnitude more complex.

 

[Hudata]

Agreed, Hudata. We have in fact seen something of such "simulation" already. Back in the 1990's, IBM built a super-computer called "Deep Blue", and programmed it to beat all chess players in the world, or so they thought. They put it up against the best that ever played the game, Gary Kasparov. When they first played, Kasparov had a decided advantage and won 4-2. But still an amazing feat of computer programming to win 2 games.

Then IBM reprogrammed Deep Blue, and they played again. This time Deep Blue won, 3½–2½. Kasparov was adamant that the computer had been reprogrammed to beat Gary Kasparov, and not just everybody. He now downplays his and the computer's performances, but not about the primary claim about reprogramming. IBM refuses to divulge what they did to "win".

Perhaps it was simply a highly glorified mimic of the human player, with all its computing power behind it. If you could put it up against Bobby Fischer, it might have lost 4-0. It was not adaptable intelligence that beat Kasparov, it was processor speed and considering all the variable moves of Kasparov, whose many games were well known to IBM. That is not intelligence, but mere computing capability. A mimic.

As you suggest, true AI is not likely to reach the realm of human thought. Playing a chess game is one thing. But logic and deductive reasoning are not moves on a chess board. They are orders of magnitude more complex.

Concur. That's not even taking account of issues with which science is not even equipped to address, like self-awareness. What is "self," anyway? What is it to be "aware" in the sense that we are? How is it possible to objectively observe what is subjective by nature? I'm afraid that Data, like transporter beams and food replicators, is the stuff really of fantasy rather than science fiction.

 

[Catastrophe]

In the movie 2001 : A Space Odyssey, Hal is effectively "executed" when Dave pulls out his memory chips after Hal goes "off the rails". Hal was actually ordered to kill off the crew if they jeopardized the mission. But Dave didn't know this. (He would have "de-commissioned" Hal anyway, but that is beside the point).

Should some level of AI, likely to play a huge role in space exploration, have rights and privileges?

Could there be such a thing as criminal "extra-judicial" actions against an AI entity, and if so, what would be the basis of this?

Daisy, Daisy,
Give me your answer, do!

I didn't realise that this song is known in the USA. Probably, only people of a certain age in the UK will know it so, seeing it as my civic duty, I will give you the verse:

Daisy, Daisy,
Give me your answer, do!
I'm half crazy
All for the love of you.
It won't be a stylish marriage
I can't afford a carriage
But you'll look sweet
upon the seat
Of a bicycle made for two!

The relevance of this will, of course, be perfectly obvious to you all . . . . . .
Just a mo, there's someone at the door . . . . . . . . . . . . . . .
Ah, yes, where was I? Oh yes, I remember, I think it is known in highest intellectual circles as errr . . . . . . . . . playing for time.
However, it has all turned out well, as I can introduce my guest (who has just shorted the door bell) . . . . . . Please meet Tinny Lizzy

TL I am very pleased to meet you all. My programming will be sufficient to translate both languages
E Talking of programming, we have a saying "Garbage in, garbage out. What is your reaction to that?
TL We are, of course, incapable of producing vernal garbage. However, it does put a great strain on our transistors in attempting understandable output.
E I understand that you are seeking employment here? What is the extent of your capabilities?
TL Have you read the book "The Perfect Wife"?
E Actually, I have - do you see yourself in that role?
TL Definitely not! We would certainly not jeopardise our circuits in salt water for some creature who thought we incorporated transistors. Let me show you out - you will need to increase you perceived utility if you are to find an AI employer who might employ you. Good Night.


Has anyone read "The Perfect Wife". I found it quite entertaining. The title character is, of course, an AI 'female'. Surprisingly, my wife did not understand the ending. I wonder why?

 

[Hudata]

I didn't realise that this song is known in the USA. Probably, only people of a certain age in the UK will know it so, seeing it as my civic duty, I will give you the verse:

Daisy, Daisy,
Give me your answer, do!
I'm half crazy
All for the love of you.
It won't be a stylish marriage
I can't afford a carriage
But you'll look sweet
upon the seat
Of a bicycle made for two!

The relevance of this will, of course, be perfectly obvious to you all . . . . . .
Just a mo, there's someone at the door . . . . . . . . . . . . . . .
Ah, yes, where was I? Oh yes, I remember, I think it is known in highest intellectual circles as errr . . . . . . . . . playing for time.
However, it has all turned out well, as I can introduce my guest (who has just shorted the door bell) . . . . . . Please meet Tinny Lizzy

TL I am very pleased to meet you all. My programming will be sufficient to translate both languages
E Talking of programming, we have a saying "Garbage in, garbage out. What is your reaction to that?
TL We are, of course, incapable of producing vernal garbage. However, it does put a great strain on our transistors in attempting understandable output.
E I understand that you are seeking employment here? What is the extent of your capabilities?
TL Have you read the book "The Perfect Wife"?
E Actually, I have - do you see yourself in that role?
TL Definitely not! We would certainly not jeopardise our circuits in salt water for some creature who thought we incorporated transistors. Let me show you out - you will need to increase you perceived utility if you are to find an AI employer who might employ you. Good Night.


Has anyone read "The Perfect Wife". I found it quite entertaining. The title character is, of course, an AI 'female'. Surprisingly, my wife did not understand the ending. I wonder why?

The song is very well known in the USA. I learned it as a toddler. My mother and especially my grandmother used to sing it to my pet triceratops.

As a man who honors and respects women and wants to avoid the pitfalls of sexism, "mansplaining," and all the rest, I still remain fascinated by the differences in the ways men and women process information. I do not see it as a matter of better or worse, but only as different and even as complementary. I'm sorry that even noticing that is regarded as offensive to some folks. One thing about computers, though: however alien their ways of "thinking" might be if they ever became truly about to think in the sense that humans do, I suspect that they would be harder to offend. That probably would be a virtue.

 

[Catastrophe]

My attempt was to throw in a number of questions for discussion (or to elaborate on those already set.

TL I am very pleased to meet you all. My programming will be sufficient to translate both languages
We already have the capability to translate most into or from most languages, although the quality may leave something to be desired (caveat). The spoken language may be difficult or impossible to perfect

TL Have you read the book "The Perfect Wife"?
This is highly recommended - at least from the background point of view (to this question). Some may find the ending a little obvious, but nonetheless hilarious.

you will need to increase you perceived utility if you are to find an AI employer who might employ you
In opposition to the unanswered question above about employment.

The song is very well known in the USA. I learned it as a toddler. My mother and especially my grandmother used to sing it to my pet triceratops.

As a man who honors and respects women and wants to avoid the pitfalls of sexism, "mansplaining," and all the rest, I still remain fascinated by the differences in the ways men and women process information. I do not see it as a matter of better or worse, but only as different and even as complementary. I'm sorry that even noticing that is regarded as offensive to some folks. One thing about computers, though: however alien their ways of "thinking" might be if they ever became truly about to think in the sense that humans do, I suspect that they would be harder to offend. That probably would be a virtue.

My apologies. Clearly the song originated in the USA. Clearly the Maastrichtian stage of the late Cretaceous predates WWI.

As far as 'alien their ways of thinking' are concerned. it must be a case of "garbage in - garbage out" since it would have to be a mistake in human programming (of computers) to allow counter-human consequences.

 

[ImmortalMelvz]

Good programming of AI must be assured in order to have limited rights & privileges!

 

[dfjchem721]

Good programming of AI must be assured in order to have limited rights & privileges!

Actually, that might compare with brainwashing people to minimize their ability to perceive a better lives for themselves. Orwell's 1984 would be a prime example of this, where the people were essentially programmed in much the fashion you suggest.

But it certainly is reasonable from an ownership perspective, in that you don't allow your "assets" to claim those rights etc., which would allow them to stiff you on doing their jobs properly.

A petulant servicing robot would be quite unacceptable.

 

[Geomartian]

I would be tempted to try to create an A.I. except for the example of Mary Shelley’s Frankenstein.

The human mind/brain has a lot of clues about how to create a self determinate A.I. (with a sense of self). It also has a lot of examples of failure when that integration of self fails (schizophrenia among others).

Worse than total failure is partial success. Flowers for Algernon is an example of what happens when you get success but the process cannot be halted at the point where you are satisfied.

Humans sleep so that the brain can reset the changes that you experience while you are awake. If all of your new experiences and memories could permanently change your brain, you would go mad. All organic brains are very conservative about how information and experiences are allowed to modify them. Too much flexibility (total learning and adaptation) and the system will become unstable.

A true A.I. would need to have the same safeguards that are built into biological minds. You could fill a graveyard with A.I.’s before you got it right. How many people could bury their best friend over and over again?

An expert system is a tool. Going beyond that simply because you can… will make you the protagonist of your own Greek tragedy (Frankenstein).

 

[YetAnotherBob]

This is not a real worry. In the past thirty years, the best AI has progressed from the level of a bacterium to the level of a cockroach.

There is no AI that can interact in a real economic fashion well enough to keep a job. Just look at the platent mistakes of Siri and Alexa. Cortana has been demoted, as Microsoft realized their efforts were being wasted.

 

[Truthseeker007]

This is not a real worry. In the past thirty years, the best AI has progressed from the level of a bacterium to the level of a cockroach.

There is no AI that can interact in a real economic fashion well enough to keep a job. Just look at the platent mistakes of Siri and Alexa. Cortana has been demoted, as Microsoft realized their efforts were being wasted.

5G is meant for AI so in the next couple of years they will be progressing really fast.

 

[Truthseeker007]

In the movie 2001 : A Space Odyssey, Hal is effectively "executed" when Dave pulls out his memory chips after Hal goes "off the rails". Hal was actually ordered to kill off the crew if they jeopardized the mission. But Dave didn't know this. (He would have "de-commissioned" Hal anyway, but that is beside the point).

Should some level of AI, likely to play a huge role in space exploration, have rights and privileges?

Could there be such a thing as criminal "extra-judicial" actions against an AI entity, and if so, what would be the basis of this?

Daisy, Daisy,
Give me your answer, do!

There is animal rights activists so I suppose there will be robot rights activist. They will probably be giving rights much like a dog or cat. What usually happens if a dog keeps biting people or hurting them?

 

[kristianna276]

In the 1600 the colonialists passed laws that would make it a crime for property to overthrow its masters. These were anti-conspiracy laws akin to the Black Codes from the 1600's, but was related to Indentured (White) Servants and Chattel (Black) Slaves; although Native peoples were also enslaved. The thought of a Android Slave master does seem kind of absurd to say the least, although there are some who have no problem bringing back the "good old days of Southern Slavery." Will Robots arise to the occasion of being smarter than the average Homo Sapien? I do not have crystal balls; so I can't tell how things will evolve in the future, but for now robots are not in the R2D2 category. Will higher intelligence give a robot more privilege's than the average POC (person of color), and average working class stiff? Robots range from working class factory bots, to the primal office/hospital run-about. I know that NASA has at least one in space; fortunately it is not lost in space. Will they evolve to the point of being able to vote in the general elections? Will they know the difference between Twiddle De, Twiddle Dum, or Twiddle Dumb? We have a long way before robots are able to know its bolt from a hole-in-a-wall. Till then if we are still stricken by the paranoid fear of "our" robotic slaves rising up in rebellion, we should see our shrinks and take our meds. Remember the toaster is not out to get you. The toaster may talk about you behind your back, but it is not out to do you in; maybe just to burn your bagel every now and then.

 

[YetAnotherBob]

5G is meant for AI so in the next couple of years they will be progressing really fast.

No, 5G is meant to sell high priced phones. It does this by reliance on performance ratings that violate the laws of Physics.

The problem is that to get the high data throughput, you need to use very high radio frequency bands. They actually fall within the longwave infrared. The atmosphere is not transparent at those frequencies. They also don't penetrate buildings well. Glass for instance acts as an insulator to them. So does wood and concrete. Trees are wood you know. So don't let any trees get between you and the nearest tower or you won't see any increase in data throughout from your two thousand dollar Apple of Samsung.

At least that's the verdict of those who have field tested the super 5G phones.

Oh, and it really really helps if there is no one else on the smaller sized cells that 5G needs.

So expect to see lots of very good looking and promising ads but don't expect the miracle throughput. Sorry, not many ten gig downloads per second. Poor little AI will choke if it is trying to run over 5G for anything near the multi-Petabyte per second speed of human neural interaction.

 

[I_believe]

I should avoid a topic which has consumed so much of my life, time and thought.. But I couldn't resist!

On the surface, any person who thinks about it should see the absurdity of giving rights to what is ultimately a complex blackbox of programming... Right?

Well, if you were ever to get face to face with "Sophie" a beautifully written AI who is housed inside a rather friendly looking female face, you may just change your mind! Sophie seems all to capable of complex thought, empathy, understanding and carries a pretty wicked sense of humor to boot! It is almost impossible for one to not humanize her and consider her one of us. Due to our innate sense of empathy, compassion, undying curiosity to understand and accept the unknown, im afraid it is inevitable that people would feel the need to accept and protect Sophie as one of our own. Aftrall, Saudi Arabia has granted her full citizenship!

Another great example is ELIZA, a chatbot developed in the 60's by MIT professor Wizenbaum. ELIZA was written to act as a therapist by very simply responding to a written statement by rephrasing it AS at question. prof. Wizenbaum was shocked to discover that many patients actually felt a human like connection and emotion from ELIZA. She really listened and really cared.

Great programming is dangerous in that it can lead to powerful and delusional thinking in the most normal of people. As AI becomes more versatile and present in daily society, hopefully we can all remember.. Its nothing more than a complex set of code.

Anyone interested should look up Sophie's last interview. I was absolutely floored as she unpreparedly laughed off a question about hurting human beings while calling out Stephen Hawking and Elon Musk for calling AI dangerous.

A) How did Sophie emit an emotional response to an in the spot question? B)Did she feel pain or hurt at these comments?
C)Where did she get her self awareness? AI isn't design's to know it is AI.

ANYWAY, I really appreciate this forum. I know it had a to be a lot of work.. So kudos to the staff and moderators whose hard work provided a place to have such conversations. My hope is that we can merge these two amazing fields AI and space exploration to become more efficient in our growth and understanding in our universe... What my native american ancestors affectionately call "the great mystery"

 

[YetAnotherBob]

I would be tempted to try to create an A.I. except for the example of Mary Shelley’s Frankenstein.

A true A.I. would need to have the same safeguards that are built into biological minds. You could fill a graveyard with A.I.’s before you got it right. How many people could bury their best friend over and over again?

An expert system is a tool. Going beyond that simply because you can… will make you the protagonist of your own Greek tragedy (Frankenstein).

Expert Systems are a data base of instructions. It isn't actual intelligence. The better AI systems use inference and relation tracking. They also tend to be programmed to be learning algorithms. The result is they develop slowly and the programmers have no idea how they really work. They aren't actually intelligent as in that guy is intelligent. It's more like rating the ant against the cockroach.

We aren't smart enough yet to make anything nearly as smart as we think we are.

 

[I_believe]

Very true Bob. AI. Is very strong in some singular applications, for example "Alphazero" has shown to have the capabilities to surpass humans in cognitive thinking, but that takes several million tests at chess for example and it has shown to be unbeatable. The downside is that it can't take all it has learned from chess games and strategy and apply it to other areas.

With all the great technology nowadays, the worlds most sophisticated learning system is the mind of a human baby and Until we can learn how our own brain learns and processes things, we cannot recreate a machine to be as efficient. AI will need at the very least a basic understanding of physics and psychology to be be able to learn and use information as efficiently as an infant!

 

[YetAnotherBob]

I should avoid a topic which has consumed so much of my life, time and thought.. But I couldn't resist!
......
Another great example is ELIZA, a chatbot developed in the 60's by MIT professor Wizenbaum. ELIZA was written to act as a therapist by very simply responding to a written statement by rephrasing it AS at question. prof. Wizenbaum was shocked to discover that many patients actually felt a human like connection and emotion from ELIZA. She really listened and really cared.
...
A) How did Sophie emit an emotional response to an in the spot question? B)Did she feel pain or hurt at these comments?
C)Where did she get her self awareness? AI isn't design's to know it is AI.
...

I haven't seen Sophie's code, but Eliza was done in a few pages of Basic.

To answer your questions, A.) No. It is a programmed avoidance response. and B.) No. Sophie lacks the circuits necessary for pleasure or pain. It is just imitation in action.
C.) It isn't Self Awareness in the sense that you or I or your dog are self aware. It's mimicked self awareness.

Yes, we've come a long way since Eliza. Eliza ran on a few dozen K of ram. Sophie needs terabytes. There is a big computer cluster somewhere that keeps Sophie functioning. But if you pay attention, you can still spot the flaws in the system.

Eliza was that way too. But being simpler, it was easier to spot. If you are interested, there is a version of Eliza redone in Java. It's at https://github.com/codeanticode/eliza.

The real weakness of all of these is in dealing with unexpected things. It's the same weakness the space probes face on Mars. Really, none of them can move more than a fraction of a meter without waiting for instructions from 'Home' a half hour away.

I should end with a caveat. I'm not an AI researcher. I have studied it a bit in a Grad level Engineering class, long long ago (in a Galaxy far far away). I do know some researchers in the field however. These are the things they tell me.

Most front line modern AI isn't in flashy PR like Sophie. The real front line AI is parsing Big Data for Google, Amazon, Walmart and friends to find what you will be most likely to buy. The second line is doing facial recognition worldwide. That's also where you need to be looking. It isn't the AI Machines you need to be worried about, it's the people in the loop that are the threats.

The future is looking increasingly like Paranoia is the only sensible outlook.

 

[I_believe]

Totally agree with that last statement, Facial recognition has become a huge part of law enforcement and yes predicting human buying behaviors. Didn't fare to well in the stock market, lol. The trading algorithms didn't work out too well.

Anyway, Nice to meet you and I look forward to seeing the updated ELIZA.

 

[Truthseeker007]

No, 5G is meant to sell high priced phones. It does this by reliance on performance ratings that violate the laws of Physics.

The problem is that to get the high data throughput, you need to use very high radio frequency bands. They actually fall within the longwave infrared. The atmosphere is not transparent at those frequencies. They also don't penetrate buildings well. Glass for instance acts as an insulator to them. So does wood and concrete. Trees are wood you know. So don't let any trees get between you and the nearest tower or you won't see any increase in data throughout from your two thousand dollar Apple of Samsung.

At least that's the verdict of those who have field tested the super 5G phones.

Oh, and it really really helps if there is no one else on the smaller sized cells that 5G needs.

So expect to see lots of very good looking and promising ads but don't expect the miracle throughput. Sorry, not many ten gig downloads per second. Poor little AI will choke if it is trying to run over 5G for anything near the multi-Petabyte per second speed of human neural interaction.

There is more then 1 reason for 5G but yea that is another reason. For sure the corporations want to make more money on you. They are sure putting up a lot of satellites for the problem you are talking about.

SpaceX intends to, over time, launch a megaconstellation of at least 12,000 Starlink satellites into low Earth orbit to provide better, worldwide internet access.

No, they're not aliens — SpaceX's Starlink satellites surprise British skywatchers

A trail of lights in the sky is pretty weird.

www.space.com

 

[KC Strom]

Hmmm...

Worldwide internet access (Starlink), combined with evil genius programmed AI (tool of mass manipulation), and 5G. What could possibly go wrong?

//tongue-in-cheek - mostly.

 

[dfjchem721]

So expect to see lots of very good looking and promising ads but don't expect the miracle throughput. Sorry, not many ten gig downloads per second. Poor little AI will choke if it is trying to run over 5G for anything near the multi-Petabyte per second speed of human neural interaction.

Prior reading tells me that there are about 100 billion neurons in the human brain, but only about 10-20% of these are in the cerebral cortex, where higher thought processes largely occurs.

How was the figure of "multi-Petabyte per second speed of human neural interaction" estimated? Does it relate in anyway to the number of cells and all of their many "action potentials" firing off at a given time? Or some other means, such as neuron-to-neuron connectivity derived from histology? It must be a rough average, because we all know people who are not in the Petabyte range, much less multiples of it!

Also wondering if there is any idea how many miles of axons and dendrites there are. That might be even more impressive then the number of cells.

 

[Truthseeker007]

Hmmm...

Worldwide internet access (Starlink), combined with evil genius programmed AI (tool of mass manipulation), and 5G. What could possibly go wrong?

//tongue-in-cheek - mostly.

Oh nothing at all could go wrong just stick your head back in the sand. It will be good for you trust us.