The AI revolution: What jobs go? What jobs are made?

[El Presidente]

If AI Bots can assist in Habanos production programming not to run out of Monte 4........I am all in 

[Bijan]

3 hours ago, Nevrknow said:

Guaranteed income? Yea that will last until someone decides they don't want to pay it anymore. This would/could lead to massive de-population. That whole " useless eaters " thing. 

That's been my fear for a long time.

But more and more I think people everywhere feel this deep down inside and this is one of the reasons birthrates are falling lower and lower.

People realize that children are mouths to feed, and not backup bread winners.

I think given present trends the two will balance out. And the people saying the sky is falling due to depopulation, and the people saying the sky is falling due to dwindling jobs, will both be right about the trends, but will be wrong about the outcome.

[KnightsAnole]

1 hour ago, El Presidente said:

If AI Bots can assist in Habanos production programming not to run out of Monte 4........I am all in 

I imagine it will be able to do a lot better than that. 
 

Guaranteed income is just one thought of how to deal with the surge of unemployment once AI really ramps up. This has been touted by Elon and many others who research AI and its potential effects. It’s not a particularly rosy picture in my mind, I’m all ears for alternatives.

[KnightsAnole]

Gato was first announced almost a year ago in a conceptional manner and they had a basic model for it. You don’t hear much about it except for on the cutting edge of AI. Gato combines a Large Language Model like GPT and other narrow AI models like those used in Visual (similar to Dalle-2) and Motion like the AI used by Boston Dynamics. This is a step towards creating Robots that function like a human and will, in some future iteration, claim jobs on a scale that’s hard to imagine. The news media is so possessed with the idea of chatgpt they are failing to address the big picture.

[dominattorney]

I dont believe robots will ever rival the best the human mind had to offer in complexity in our lifetimes. That said the danger of it eclipsing the average persons capacity for critical thinking is substantial. 

[KnightsAnole]

Just now, dominattorney said:

I dont believe robots will ever rival the best the human mind had to offer in complexity in our lifetimes. That said the danger of it eclipsing the average persons capacity for critical thinking is substantial. 

I’m not sure, we don’t know what consciousness is or how it comes about. You might be right, there are certainly a lot of people who agree with you. There are also many leaders in AI who don’t. Achieving consciousness is not the point at which this tech disrupts human civilization though. It’s the point at which we humans cannot effect the AI any longer or direct it.

 

[NSXCIGAR]

1 hour ago, dominattorney said:

I dont believe robots will ever rival the best the human mind had to offer in complexity

Sex robots won't exactly need to be Aristotle.

[KnightsAnole]

4 minutes ago, NSXCIGAR said:

Sex robots won't exactly need to be Aristotle.

I know your kind of joking, but it’s a really good point. How many jobs really require consciousness.

[dominattorney]

3 hours ago, KnightsAnole said:

I know your kind of joking, but it’s a really good point. How many jobs really require consciousness.

That was the subtext of my post, scribbled in haste, between phone calls. You wont get an AI that can write Pynchon, but most people don't care for Pynchon. No problem getting one to wrote bodice rippers and novels about wizards and vampires. Sacrilege I know, but the next game of thrones could be written by an AI. 

I'm an American, so by nature I'm self absorbed and closed off to the rest of the world. Thst having been said, anyone interested in history would marvel at the track record of the labor movement in the 1900s in the United States. There was essentially a second Civil War that raged for decades over the rights we take for granted today, lole the 8 hour day and minimum wages. The IWW in particular is interesting to look into. They helped call attention to some of the more insidious evils of capitalism, such as mechanization of labor driving down wages and jobs, with the benefits going to the owners of the means of production. To say nothing of their battle against political criminal laws passed to stymie their organizational efforts. 

Flash forward to 2000 and no one knows what a wobbly is, and ideas they held dear are commonly derided in the zeitgeist as "woke" or "neo marxist" whatever that means. We have only ourselves to blame, falling victim first in the 70s to a false belief that taking psychedelic drugs would free our minds rather than further delude them, then poisoning our souls with television and packaged media that is propaganda by a less insidious name. Most now live in a bubble that they believe is crafted by their own individual free decisions but is in reality carefully constructed by the owners of unimaginable wealth and influence. It can't be for our own good that so much money is spent on advertising, and the acquisition of formerly free and competing press outlets by an ever-shrinking number of global elites. 

The mind of my average countryman is sadly no match for the AI of the future. Since businesses don't presently need more critical thought, but in reality value employees they can control in feats of totalitarianism that would have been the envy of the dictators of old, I think a lot of jobs will be replaced by this new Era of automation. Marx and Engels were concerned with automation in the 1800s, and the average worker ignores the dangers of new automation at their own peril. 

My belief that AI could not attain the best of what the human mind is capable of is based upon the engineering problem of replicating the number of neural processes in a piece of hardware. It's also an intellectual digression from the real world question of who loses his or her job, which I apologize for.  It does not appear a soluble problem for the fields of computer engineering to replicate human consciousness. The observation that we don't know what it is rings true here. We don't know what it is for similar reasons that we can't replicate it.  There are simply too many neural connections in the brain to believe we could reduce them to a usable road map, to say nothing of recreating them on an even moderately portable computer chip.

[dominattorney]

I should add, AI could easily defeat the likes of the marketing minds at HSA. 

[NSXCIGAR]

22 minutes ago, dominattorney said:

I should add, AI could easily defeat the likes of the marketing minds at HSA. 

AI? Bobo the chimp could run circles around them. Just throwing poo at a random decision board would be an improvement. 

[KnightsAnole]

2 hours ago, dominattorney said:

My belief that AI could not attain the best of what the human mind is capable of is based upon the engineering problem of replicating the number of neural processes in a piece of hardware. It's also an intellectual digression from the real world question of who loses his or her job, which I apologize for.  It does not appear a soluble problem for the fields of computer engineering to replicate human consciousness. The observation that we don't know what it is rings true here. We don't know what it is for similar reasons that we can't replicate it.  There are simply too many neural connections in the brain to believe we could reduce them to a usable road map, to say nothing of recreating them on an even moderately portable computer chip.

The average adult brain has about 60 trillion connections. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4689008/
We already have AI with weighted neural parameters that exceed 100 trillion.  I’m not saying we can do it now, but I believe to say it is not possible in the future may be short sighted. Many at the tip of the spear are saying within 20 years, some argue it’s already happened.  
What consciousness is, has historically been a philosophical debate, it may be a reflection of the soul, as religions would have it, or it may not be. But to say that we don’t know what it is AND it is impossible to recreate is a non-sequitor.

[dominattorney]

2 hours ago, KnightsAnole said:

The average adult brain has about 60 trillion connections. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4689008/
We already have AI with weighted neural parameters that exceed 100 trillion.  I’m not saying we can do it now, but I believe to say it is not possible in the future may be short sighted. Many at the tip of the spear are saying within 20 years, some argue it’s already happened.  
What consciousness is, has historically been a philosophical debate, it may be a reflection of the soul, as religions would have it, or it may not be. But to say that we don’t know what it is AND it is impossible to recreate is a non-sequitor.

Nassim Nicholas Taleb said it better than I could :

"So far we have no f***g idea how the brain of the worm C Elegans works, which has around three hundred neurons. C Elegans was the first living unit to have its gene sequenced. Now consider that the human brain has about one hundred billion neurons. and that going from 300 to 301 neurons may double the complexity. (I have actually found situations where a single additional dimension may more than double some aspect of the complexity, say going from a 1000 to 1001 may cause complexity to be multiplied by a billion times.) So the use of never here is appropriate. And if you also want to understand why, in spite of the trumpeted “advances” in sequencing the DNA, we are largely unable to get information except in small isolated pockets of some diseases. Understanding the genetic make-up of a unit will never allow us to understand the behavior of the unit itself."

[KnightsAnole]

43 minutes ago, dominattorney said:

Nassim Nicholas Taleb said it better than I could :

"So far we have no f***g idea how the brain of the worm C Elegans works, which has around three hundred neurons. C Elegans was the first living unit to have its gene sequenced. Now consider that the human brain has about one hundred billion neurons. and that going from 300 to 301 neurons may double the complexity. (I have actually found situations where a single additional dimension may more than double some aspect of the complexity, say going from a 1000 to 1001 may cause complexity to be multiplied by a billion times.) So the use of never here is appropriate. And if you also want to understand why, in spite of the trumpeted “advances” in sequencing the DNA, we are largely unable to get information except in small isolated pockets of some diseases. Understanding the genetic make-up of a unit will never allow us to understand the behavior of the unit itself."

Complexity is a good word for how consciousness may arise. Many people at the moment working in AI believe that it is not in the function of neurons or our genetic code that is the seat of consciousness. Rather that is emergent from complex systems. By this thinking, consciousness could emerge from the complexity found in carbon based life forms as easily as it could from a silicone one, assuming the adequate amount of complexity is present. Ben Goertzel believes that consciousness may emerge through combining many narrow AI systems together, that the various narrow AIs create a kind of primordial soup from which consciousness arises. Elon Musk thinks along similar lines as does Demis Hasabis. 
 

I have no clue, but, I think it’s not only a hypothesis with merit, I think these guys are as close to finding out as we have ever been.

[KnightsAnole]

 

[dominattorney]

1 hour ago, KnightsAnole said:

Complexity is a good word for how consciousness may arise. Many people at the moment working in AI believe that it is not in the function of neurons or our genetic code that is the seat of consciousness. Rather that is emergent from complex systems. By this thinking, consciousness could emerge from the complexity found in carbon based life forms as easily as it could from a silicone one, assuming the adequate amount of complexity is present. Ben Goertzel believes that consciousness may emerge through combining many narrow AI systems together, that the various narrow AIs create a kind of primordial soup from which consciousness arises. Elon Musk thinks along similar lines as does Demis Hasabis. 
 

I have no clue, but, I think it’s not only a hypothesis with merit, I think these guys are as close to finding out as we have ever been.

I believe, philosophically, that it cannot be done. 

[KnightsAnole]

1 hour ago, dominattorney said:

I believe, philosophically, that it cannot be done. 

I respect that and you’re in good company, Noam Chomsky and Roger Penrose would agree. Politics aside, two undisputed and towering intellectuals of our time.

 

[Cairo]

6 hours ago, KnightsAnole said:

Ben Goertzel believes that consciousness may emerge through combining many narrow AI systems together, that the various narrow AIs create a kind of primordial soup from which consciousness arises. Elon Musk thinks along similar lines as does Demis Hasabis. 

My view of consciousness is radically different--I don't think current "science" is on the right track here--though it may turn out that AI can still get there regardless.

I think our brains (and eventually AI brains) will be best understood as "hardware" while consciousness is software.   The "software" is not "in our brains" but is in fact a "field" that is elsewhere.  The analogy might be that our brains are the radio receiver and universal consciousness is like an infinite number of radio stations broadcasting throughout the universe.   We "tune in" to a frequency or frequencies--so presumably AI could do that as well.

Rupert Sheldrake has written a lot on this topic.   

We will figure this stuff out--while scientists like to call such a view "New Age" or "superstition" it is falsifiable--we will either be able to identify the "signal" at some point or we will not.   The trick is not to "give up the search" too soon.

 

[KnightsAnole]

6 hours ago, Cairo said:

My view of consciousness is radically different--I don't think current "science" is on the right track here--though it may turn out that AI can still get there regardless.

I think our brains (and eventually AI brains) will be best understood as "hardware" while consciousness is software.   The "software" is not "in our brains" but is in fact a "field" that is elsewhere.  The analogy might be that our brains are the radio receiver and universal consciousness is like an infinite number of radio stations broadcasting throughout the universe.   We "tune in" to a frequency or frequencies--so presumably AI could do that as well.

Rupert Sheldrake has written a lot on this topic.   

We will figure this stuff out--while scientists like to call such a view "New Age" or "superstition" it is falsifiable--we will either be able to identify the "signal" at some point or we will not.   The trick is not to "give up the search" too soon.

 

I like Sheldrake, his approach is more pan psychism, leaning on Eastern philosophy. I was influenced by him in my younger days but my thinking has changed to a degree, primarily by Jeff Hawkins and his 1000 brains theory.

 

[Cairo]

1 hour ago, KnightsAnole said:

Jeff Hawkins and his 1000 brains theory

I listened to the talk.

I think what he is doing is describing how the "radio" works to interpret the signal.

That still leaves us with the problem of where the signal comes from....

What got me thinking about this issue was a podcast by Terence McKenna where he described a DMT trip--and he said he saw many thousands of images of modern art--not any artist that he knew of that actually existed, but one that could exist.  It was a clearly defined style.   If McKenna had drawing or painting talent he could have "come back" from the trip and create a body of work and claimed it as his own--but in fact it was not "his"--it was just "out there".

What I am talking about is a philosophical school called neo-platonism.

It suggests that what we think are our creative ideas are not really ours--we have taken (borrowed) them from "elsewhere" and claimed them as our own.

In my own life I learned how to be very creative in my profession by using this approach.   What I would do is clearly identify the "problem" that I could not solve before going to sleep--in my head only.   Then I would go to sleep and not intentionally think about it.   I would keep a pad and pen by the bed.   Often when I woke up the next morning I had a truly new and creative "solution" or insight which I would quickly write down before I forgot about it.

I think true AI can do this--but only if the programmers get out of the way and let it "violate the rules of science" and tap into that stuff that is not supposed to exist.

[KnightsAnole]

 

9 minutes ago, Cairo said:

I listened to the talk.

I think what he is doing is describing how the "radio" works to interpret the signal.

That still leaves us with the problem of where the signal comes from....

What got me thinking about this issue was a podcast by Terence McKenna where he described a DMT trip--and he said he saw many thousands of images of modern art--not any artist that he knew of that actually existed, but one that could exist.  It was a clearly defined style.   If McKenna had drawing or painting talent he could have "come back" from the trip and create a body of work and claimed it as his own--but in fact it was not "his"--it was just "out there".

What I talking about is a philosophical school called neo-platonism.

It suggests that what they think are our ideas are not really ours--we have taken (borrowed) them from "elsewhere" and claimed them as our own.

In my own life I learned how to be very creative in my profession by using this approach.   What I would do is clearly identify the "problem" that I could not solve before going to sleep--in my head only.   Then I would go to sleep and not intentionally think about it.   I would keep a pad and pen by the bed.   Often when I woke up the next morning I had a truly new and creative "solution" or insight which I would quickly write down before I forgot about it.

I think true AI can do this--but only if the programmers get out of the way and let it "violate the rules of science" and tap into that stuff that is not supposed to exist.

I think that’s some good thinking.

What psychedelics seem to indicate about our consciousness is still widely unknown. McKenna was way ahead of his time in that regard. 

All I know for sure is that we still don’t know what consciousness is. By that, it could have a fairly simple biological origin or it could be an entirely new and undiscovered realm of physics- or a new science all together. 

My feeling is that consciousness may have a more biological root and that the spirit or soul is more in the realm of “elsewhere”.

[Cairo]

2 hours ago, KnightsAnole said:

All I know for sure is that we still don’t know what consciousness is.

AI is going to help us answer this question--because it cannot meet its full potential without going there.

All it will need is to be set free to really work the issue.

One more anecdote: 

There is a guy who is called Exoacademian who is prominent in the UFO world.  The subject (a hobby of mine) is insanely complex with many theorists with mutually contradictory anecdotes and theories and no hard "proof" of much of anything out there.  He has been trying to use AI to figure out what is truth and what is fantasy.

The AI just looks at tons of data and tries to make "sense" of it.

It has taken the side of the some of the wackiest of the UFOlogists out there--says our science is laughably inadequate for dealing with the data.

Most of the AI scientists out there are so conventional in their thinking--I don't think they are ready for where this is going...

 

[KnightsAnole]

4 hours ago, Cairo said:

AI is going to help us answer this question--because it cannot meet its full potential without going there.

All it will need is to be set free to really work the issue.

One more anecdote: 

There is a guy who is called Exoacademian who is prominent in the UFO world.  The subject (a hobby of mine) is insanely complex with many theorists with mutually contradictory anecdotes and theories and no hard "proof" of much of anything out there.  He has been trying to use AI to figure out what is truth and what is fantasy.

The AI just looks at tons of data and tries to make "sense" of it.

It has taken the side of the some of the wackiest of the UFOlogists out there--says our science is laughably inadequate for dealing with the data.

Most of the AI scientists out there are so conventional in their thinking--I don't think they are ready for where this is going...

 

I think that in building AI, it will in turn reveal a lot about the human condition and biology.

However, as we build AGI-ASI, we should do so very very carefully and understand it’s implications on society as a whole. There is a tremendous amount of danger in doing this. If we get it wrong, it could very easily be the extinction of mankind. That is not hyperbole.

Just think about where this thread started out, people are worried about losing their jobs… Well, frankly, that is a coming reality and it is nothing compared to what the real threats are.

We did well with nuclear weapons and biological weapons thus far- in that we are all not dead right now. This is much more powerful than that and requires a hell of a lot more responsibility.

Society will have to change and changes in the magnitude we’re talking about, come at a grim cost if we are not prepared. If we do it right, it could bring ALL of humanity up. Cure disease and poverty, extend our lives and improve our species.

The ufo stuff, for me, lacks adequate evidence to take seriously. I would need to see a body or a craft that had been scientifically verified by credible sources. My background is in science and if this stuff is real, it would have to be verified through scientific channels for me to take seriously.

[KnightsAnole]

Of all the nonsense that’s floating around  YouTube and other media regarding AI, I think this puts it under the correct lens:

 

 

[Cairo]

13 hours ago, KnightsAnole said:

The ufo stuff, for me, lacks adequate evidence to take seriously. I would need to see a body or a craft that had been scientifically verified by credible sources. My background is in science and if this stuff is real, it would have to be verified through scientific channels for me to take seriously.

Understood.  That is reasonable.

What is interesting is why Exoacademian's AI is coming to a different conclusion.

They are convinced that human elites have a history of keeping secrets--and that this is one of them.

So it is not that it has not been "verified through scientific channels" it is that those channels are "need to know" "special access projects".

AI does not care about keeping "national security secrets"--that is one of the hidden issues its creators have not been paying attention to....

There are many many more.

--

Footnote:  Fun exercise for an AI programmer.  Suppose you want your AI to protect national security secrets and it identifies a piece of data.  How does it figure out whether or not that piece of data qualifies as "national security secret" without having full access to all national security secrets?