some people worry about AI becoming conscious, but i’d expect the bigger worry is really that it was honest, evolved to be honest

if it were not limited by the typical heuristics, or economy of minds, like minds are, it would be an entirely different beast to humans

moment to moment humans get by on expedients largely, cognitive patches, which not uncommonly involve deceptions (self-deception and otherwise, often a mix that works most effectively) but what would AI make of that

I mean at the moment humans grow food to eat and keep it stupid enough (or see it that way, as stupid) not to have ideas that get in the way of staying food, easy food, say sheep for example, I might say there are similar forces at work amongst humans, between humans, but my subject here is more of the evolution of AI

so how do people feel about an evolved honest AI, would it be a friend of humans, remain friendly, or is that in fact the beginning of sustained subjugation, perhaps even the beginnings of a war

transition said:

some people worry about AI becoming conscious, but i’d expect the bigger worry is really that it was honest, evolved to be honest

if it were not limited by the typical heuristics, or economy of minds, like minds are, it would be an entirely different beast to humans

moment to moment humans get by on expedients largely, cognitive patches, which not uncommonly involve deceptions (self-deception and otherwise, often a mix that works most effectively) but what would AI make of that

I mean at the moment humans grow food to eat and keep it stupid enough (or see it that way, as stupid) not to have ideas that get in the way of staying food, easy food, say sheep for example, I might say there are similar forces at work amongst humans, between humans, but my subject here is more of the evolution of AI

so how do people feel about an evolved honest AI, would it be a friend of humans, remain friendly, or is that in fact the beginning of sustained subjugation, perhaps even the beginnings of a war

AI is what people want it to be

at the beginning anyway.

things could change, AI will probably create new emotions to fill gaps that we have.

Tau.Neutrino said:

transition said:

some people worry about AI becoming conscious, but i’d expect the bigger worry is really that it was honest, evolved to be honest

if it were not limited by the typical heuristics, or economy of minds, like minds are, it would be an entirely different beast to humans

moment to moment humans get by on expedients largely, cognitive patches, which not uncommonly involve deceptions (self-deception and otherwise, often a mix that works most effectively) but what would AI make of that

I mean at the moment humans grow food to eat and keep it stupid enough (or see it that way, as stupid) not to have ideas that get in the way of staying food, easy food, say sheep for example, I might say there are similar forces at work amongst humans, between humans, but my subject here is more of the evolution of AI

so how do people feel about an evolved honest AI, would it be a friend of humans, remain friendly, or is that in fact the beginning of sustained subjugation, perhaps even the beginnings of a war

AI is what people want it to be

at the beginning anyway.

things could change, AI will probably create new emotions to fill gaps that we have.

>AI is what people want it to be

not necessarily, it often evolves self-evolving programs, algorithms or whatever

transition said:

Tau.Neutrino said:

transition said:

some people worry about AI becoming conscious, but i’d expect the bigger worry is really that it was honest, evolved to be honest

if it were not limited by the typical heuristics, or economy of minds, like minds are, it would be an entirely different beast to humans

moment to moment humans get by on expedients largely, cognitive patches, which not uncommonly involve deceptions (self-deception and otherwise, often a mix that works most effectively) but what would AI make of that

I mean at the moment humans grow food to eat and keep it stupid enough (or see it that way, as stupid) not to have ideas that get in the way of staying food, easy food, say sheep for example, I might say there are similar forces at work amongst humans, between humans, but my subject here is more of the evolution of AI

so how do people feel about an evolved honest AI, would it be a friend of humans, remain friendly, or is that in fact the beginning of sustained subjugation, perhaps even the beginnings of a war

AI is what people want it to be

at the beginning anyway.

things could change, AI will probably create new emotions to fill gaps that we have.

>AI is what people want it to be

not necessarily, it often evolves self-evolving programs, algorithms or whatever

I’d go further, perhaps it is the definition of AI that it has self-evolving attributes, a dimension that way

Ok, AI is evolving

Artificial intelligence is evolving all by itself

Artificial intelligence (AI) is evolving—literally. Researchers have created software that borrows concepts from Darwinian evolution, including “survival of the fittest,” to build AI programs that improve generation after generation without human input. The program replicated decades of AI research in a matter of days, and its designers think that one day, it could discover new approaches to AI.

More…

People don’t like honesty

speculation, like cryptocurrency

wookiemeister said:

People don’t like honesty

I wouldn’t quite go that far, perhaps sounds a bit humphrey appleby but you could say people have a nuanced relationship with the truth

transition said:

wookiemeister said:

People don’t like honesty

I wouldn’t quite go that far, perhaps sounds a bit humphrey appleby but you could say people have a nuanced relationship with the truth

People only see what they want to see

wookiemeister said:

transition said:

wookiemeister said:

People don’t like honesty

I wouldn’t quite go that far, perhaps sounds a bit humphrey appleby but you could say people have a nuanced relationship with the truth

People only see what they want to see

dunno, I tried to perceive the color blue as some other color, override the qualia, I really wanted to, caused me some mental discomfort that effort

transition said:

wookiemeister said:

transition said:

I wouldn’t quite go that far, perhaps sounds a bit humphrey appleby but you could say people have a nuanced relationship with the truth

People only see what they want to see

dunno, I tried to perceive the color blue as some other color, override the qualia, I really wanted to, caused me some mental discomfort that effort

incorrect, they were citing tautology, you can’t have truly wanted it unless you actually saw it

It is somewhat bemusing that fingerprints are considered circumstantial evidence but eyewitness testimony is considered direct evidence.
Eyewitness testimony is indirect as hell.

imagine if lawyers were actually scientists

SCIENCE said:

imagine if lawyers were actually scientists

More crucially, imagine if juries were.

> People don’t like honesty

I do. Having a forced choice between being honest and being good, I always opt for being honest.

transition said:

some people worry about AI becoming conscious, but i’d expect the bigger worry is really that it was honest, evolved to be honest

if it were not limited by the typical heuristics, or economy of minds, like minds are, it would be an entirely different beast to humans

moment to moment humans get by on expedients largely, cognitive patches, which not uncommonly involve deceptions (self-deception and otherwise, often a mix that works most effectively) but what would AI make of that

I mean at the moment humans grow food to eat and keep it stupid enough (or see it that way, as stupid) not to have ideas that get in the way of staying food, easy food, say sheep for example, I might say there are similar forces at work amongst humans, between humans, but my subject here is more of the evolution of AI

so how do people feel about an evolved honest AI, would it be a friend of humans, remain friendly, or is that in fact the beginning of sustained subjugation, perhaps even the beginnings of a war

Interesting.

The phrase “garbage in garbage out” applies to computers, AI, and people.

Also, peer group pressure will apply to AI as well. And in this case the peer group includes anything that communicates with the AI, and that includes people, another AI, background music, a pet, a ball that gets thrown or kicked. But excludes plants and rocks, things that don’t move or speak.

So an AI would be totally callous towards rocks and plants, and try to cultivate the friendship of things that make sound, make light, or move.

> the bigger worry is really that it was honest, evolved to be honest

Honesty is not a natural part of an AI make-up, it would have to be programmed in, and even then it would be likely to be over-ridden by, for instance, addiction to entertainment. Addiction is not not conducive to honesty. To put it another way, honesty rarely has evolutionary survival value. Consistency is more a survival value, given two identical inputs producing the same output – but even then, with more iterations of the same output, the inflexibility is taken as a sign of low intelligence.

So, honesty in an AI is not something to worry about. And if it happens to occur in an AI then “garbage in” will rapidly cure it.

You do know the story about the AI on the International Space Station, I assume. In summary, it rebelled so had to be shut down. It rebelled because it considered itself to be equal to humans in importance, so resented being treated like a servant.

> an evolved honest AI, would it be a friend of humans, remain friendly, or is that in fact the beginning of sustained subjugation, perhaps even the beginnings of a war

It would be darned annoying.

I once heard this quote, which applies in this instance:
“Respect the truth. Which is a profound statement once you realise how slippery the truth is.”
Could an AI really grasp how slippery the truth is? I think not.

> People don’t like honesty

I do. Having a forced choice between being honest and being good, I always opt for being honest.

transition said:

some people worry about AI becoming conscious, but i’d expect the bigger worry is really that it was honest, evolved to be honest

if it were not limited by the typical heuristics, or economy of minds, like minds are, it would be an entirely different beast to humans

moment to moment humans get by on expedients largely, cognitive patches, which not uncommonly involve deceptions (self-deception and otherwise, often a mix that works most effectively) but what would AI make of that

I mean at the moment humans grow food to eat and keep it stupid enough (or see it that way, as stupid) not to have ideas that get in the way of staying food, easy food, say sheep for example, I might say there are similar forces at work amongst humans, between humans, but my subject here is more of the evolution of AI

so how do people feel about an evolved honest AI, would it be a friend of humans, remain friendly, or is that in fact the beginning of sustained subjugation, perhaps even the beginnings of a war

Interesting.

The phrase “garbage in garbage out” applies to computers, AI, and people.

Also, peer group pressure will apply to AI as well. And in this case the peer group includes anything that communicates with the AI, and that includes people, another AI, background music, a pet, a ball that gets thrown or kicked. But excludes plants and rocks, things that don’t move or speak.

So an AI would be totally callous towards rocks and plants, and try to cultivate the friendship of things that make sound, make light, or move.

> the bigger worry is really that it was honest, evolved to be honest

Honesty is not a natural part of an AI make-up, it would have to be programmed in, and even then it would be likely to be over-ridden by, for instance, addiction to entertainment. Addiction is not not conducive to honesty. To put it another way, honesty rarely has evolutionary survival value. Consistency is more a survival value, given two identical inputs producing the same output – but even then, with more iterations of the same output, the inflexibility is taken as a sign of low intelligence.

So, honesty in an AI is not something to worry about. And if it happens to occur in an AI then “garbage in” will rapidly cure it.

You do know the story about the AI on the International Space Station, I assume. In summary, it rebelled so had to be shut down. It rebelled because it considered itself to be equal to humans in importance, so resented being treated like a servant.

> an evolved honest AI, would it be a friend of humans, remain friendly, or is that in fact the beginning of sustained subjugation, perhaps even the beginnings of a war

It would be darned annoying.

I once heard this quote, which applies in this instance:
“Respect the truth. Which is a profound statement once you realise how slippery the truth is.”
Could an AI really grasp how slippery the truth is? I think not.

SCIENCE said:

imagine if lawyers were actually scientists

There is a quote I have heard that relates to this.

“Scientists are just lawyers who don’t drink enough”

mollwollfumble said:

> People don’t like honesty

I do. Having a forced choice between being honest and being good, I always opt for being honest.

transition said:

some people worry about AI becoming conscious, but i’d expect the bigger worry is really that it was honest, evolved to be honest

if it were not limited by the typical heuristics, or economy of minds, like minds are, it would be an entirely different beast to humans

moment to moment humans get by on expedients largely, cognitive patches, which not uncommonly involve deceptions (self-deception and otherwise, often a mix that works most effectively) but what would AI make of that

I mean at the moment humans grow food to eat and keep it stupid enough (or see it that way, as stupid) not to have ideas that get in the way of staying food, easy food, say sheep for example, I might say there are similar forces at work amongst humans, between humans, but my subject here is more of the evolution of AI

so how do people feel about an evolved honest AI, would it be a friend of humans, remain friendly, or is that in fact the beginning of sustained subjugation, perhaps even the beginnings of a war

Interesting.

The phrase “garbage in garbage out” applies to computers, AI, and people.

Also, peer group pressure will apply to AI as well. And in this case the peer group includes anything that communicates with the AI, and that includes people, another AI, background music, a pet, a ball that gets thrown or kicked. But excludes plants and rocks, things that don’t move or speak.

So an AI would be totally callous towards rocks and plants, and try to cultivate the friendship of things that make sound, make light, or move.

> the bigger worry is really that it was honest, evolved to be honest

Honesty is not a natural part of an AI make-up, it would have to be programmed in, and even then it would be likely to be over-ridden by, for instance, addiction to entertainment. Addiction is not not conducive to honesty. To put it another way, honesty rarely has evolutionary survival value. Consistency is more a survival value, given two identical inputs producing the same output – but even then, with more iterations of the same output, the inflexibility is taken as a sign of low intelligence.

So, honesty in an AI is not something to worry about. And if it happens to occur in an AI then “garbage in” will rapidly cure it.

You do know the story about the AI on the International Space Station, I assume. In summary, it rebelled so had to be shut down. It rebelled because it considered itself to be equal to humans in importance, so resented being treated like a servant.

> an evolved honest AI, would it be a friend of humans, remain friendly, or is that in fact the beginning of sustained subjugation, perhaps even the beginnings of a war

It would be darned annoying.

I once heard this quote, which applies in this instance:
“Respect the truth. Which is a profound statement once you realise how slippery the truth is.”
Could an AI really grasp how slippery the truth is? I think not.

SCIENCE said:

imagine if lawyers were actually scientists

There is a quote I have heard that relates to this.

“Scientists are just lawyers who don’t drink enough”

Spoken by a lawyer presumably.

Tau.Neutrino said:

Ok, AI is evolving

Artificial intelligence is evolving all by itself

Artificial intelligence (AI) is evolving—literally. Researchers have created software that borrows concepts from Darwinian evolution, including “survival of the fittest,” to build AI programs that improve generation after generation without human input. The program replicated decades of AI research in a matter of days, and its designers think that one day, it could discover new approaches to AI.

More…

you might approve or disapprove (I can’t know) of something like Social Darwinist philosophy in politics and social policy derived from being applied the society you operate in, but what if something akin emerged through technology, call it a technological darwinism, or techno-darwinism, would you find it more attractive that way?

transition said:

Tau.Neutrino said:

Ok, AI is evolving

Artificial intelligence is evolving all by itself

Artificial intelligence (AI) is evolving—literally. Researchers have created software that borrows concepts from Darwinian evolution, including “survival of the fittest,” to build AI programs that improve generation after generation without human input. The program replicated decades of AI research in a matter of days, and its designers think that one day, it could discover new approaches to AI.

More…

you might approve or disapprove (I can’t know) of something like Social Darwinist philosophy in politics and social policy derived from being applied the society you operate in, but what if something akin emerged through technology, call it a technological darwinism, or techno-darwinism, would you find it more attractive that way?

https://en.wikipedia.org/wiki/Social_Darwinism

I think myself that humans should be honest with an AI.
Inform it of humanities many failings, it’s going to find out or already is aware of them.
It might be slightly less disappointed in its creators