936 replies to this topic

[adamh]

Oh sure, there always is some social disruption when something new and game changing came out. For example, when smart phones became popular, phone booths started disappearing, they are very hard to find now. I think people would look at having a robot the same as having a car, an expensive purchase they can use for a long time. But why should it be expensive? The only expensive part comes from trying to look and act like a human. We already have chatgpt and whatever the next iteration is. So you would pay $10 grand for a chatgpc that looks human and can walk? Why?

 

As for them taking jobs, you will not be able to make a robot that can do everything a human can for at least 100 years, imo. The physical part is very difficult, climbing a ladder, walking, running. Even harder is seeing a situation and deciding what needs to be done. It takes robot vision which is getting better but is not too good yet. To look at something, see a bolt is missing, look for and pick up the right bolt, and install the bolt, that takes a lot of brainpower that is not easily appreciated. 

 

To look for and pick up and object takes more than cameras for eyes. The brain has to recognize the object which may be at different angles. That alone takes a lot of computing power. Then it has to manouver its arm into position and pick the object up. We aren't quite there yet except in some dedicated environments where everything is in the right place. And to grab something flying by in the air like a ball, that is way way more difficult. It has to calculate the speed, direction, the path its going to take, gravity effects on the flight of the ball, wind and so on. To catch it, it has to do all that and to calculate exactly where to put its hand and do it in time. It will years before we have that

 

We don't need that. What would you use a robot for? Going to the store and picking up groceries? They deliver for free. A tennis partner? Can't do it yet though it can throw balls at you. Cut the lawn? You need a mower and teach the machine to mow. Better to have a mower that is programmable. It can do some house keeping, we already have roomba.

 

What will people want a robot for once they realize its not going to look or act exactly like a human nor can it do many human things? Someone to talk to? That can work, we already have chatgpt as I keep pointing out

 

In factories, it will be an extention of a programmable machine. It will have additional capabilities like examining parts for flaws, operating a machine and so on that a human used to do. It might be built into the machine, be a part of it or be like a supervisor checking on the whole factory operation. It will take time since many things can not be automated or its too expensive.

 

Certain things will become cheaper right away. Computers are used now in trouble shooting cars, doctors use more and more diagnostic programs and equipment. AI is just another iteration of that

[mag1]

adamh, thank you for posting!

 

I am especially happy that we have been able to be grown ups about our disagreements and have tolerated diversity of opinions.

The internet too often becomes an echo chamber -- it is invigorating to still have forums which accept a range of perspectives.

 

Admittedly, I am rapidly cycling in a manic depressive like way on the topic of AI.

I am very unclear how long this seemingly very unstable situation can be maintained without large scale civilizational disruption occurring.

While proclaiming the end times or chanting the End is Nigh is typically used for comical effect, I really cannot see myself beyond what is

now approaching.

 

It feels like the Singularity. I know that many do not like misusing Singularity for anything but superintelligence, though

Civilizational Singularity will likely be a bad hair day when it arrives -- probably best to just call in sick when it is finally here.

When they start rolling out millions of humanoids at a reasonable consumer level price, then things could happen FAST!

 

Consider with these humanoids how they would be able to use all existing human designed tools.

I had thought of the easy ones like the lawnmowers, the brooms, the ironing boards, the stoves, etc..

 

 

It then occurred to me that they would also be able to drive our cars. That is quite startling!

You see if you have to turn everything into a robot then it means it could take perhaps decades to

make the transition to robotransport. But ... if you could simply sit a humanoid into the drivers seat

of existing cars, then the changeover might happen over the course of 3-4 years. As soon as the 

transition is complete, the 40,000 fatalities and tens of thousands of other serous injuries would no

longer happen. I found this idea quite stunning! These robots will be multipurpose -- they could do

anything a human could do but probably better. The idea that we need to hardwire a robotic transport 

computer into everyone's car might simply be unnecessary. Getting humans out of the driver's seats

of cars will be such a victory for safety! It is quite strange that the robot videos to date have not shown

a humanoid driving around -- that would be hilarious!

 

There are likely a wide range of other human chores that I am not thinking of right now that will also greatly enhance our quality of life

when done by robots. 

  

[Advocatus Diaboli]

"It then occurred to me that they would also be able to drive our cars."

 

Not sure if a robot is needed to drive a car--well, unless you want the car-driving robot to unload the groceries when you get back from the market. The "robot" is the car itself. For example, Elon Musk is going to have Tesla integrate his AI, Grok, into Tesla vehicles.

Edited by Advocatus Diaboli, 24 May 2024 - 01:58 AM.

[adamh]

"It then occurred to me that they would also be able to drive our cars."

 

Not sure if a robot is needed to drive a car--well, unless you want the car-driving robot to unload the groceries when you get back from the market. The "robot" is the car itself. For example, Elon Musk is going to have Tesla integrate his AI, Grok, into Tesla vehicles.

 

Good point I think people are too hung up on what they saw in scifi movies and expect robots to basically be like humans in a robot disguise. Every movie I saw with robots, they were always humanoid which makes no sense. Walking on legs instead of wheels is very hard, they do have robots that can do it but their whole build and purpose is to walk. They need a dedicated computer to do that.

 

The demand is for lifelike bots that can talk, can think, can do things and make decisions. They want too much. Things like vision are complicated. Teslas have the latest tech but they get into collisions in self driving mode. Are you sure you want a bot driving the car?

[adamh]

adamh, thank you for posting!

 

I am especially happy that we have been able to be grown ups about our disagreements and have tolerated diversity of opinions.

The internet too often becomes an echo chamber -- it is invigorating to still have forums which accept a range of perspectives.

 

Admittedly, I am rapidly cycling in a manic depressive like way on the topic of AI.

I am very unclear how long this seemingly very unstable situation can be maintained without large scale civilizational disruption occurring.

While proclaiming the end times or chanting the End is Nigh is typically used for comical effect, I really cannot see myself beyond what is

now approaching.

 

 

No need to panic, the world has seen much greater upheaval than this. Think about the panic in 1999 over the year 2k situation that turned out to be not so much. People were talking like civilization will collapse. Didnt happen

 

"Consider with these humanoids how they would be able to use all existing human designed tools.

I had thought of the easy ones like the lawnmowers, the brooms, the ironing boards, the stoves, etc.."

 

The stove will be easy perhaps, the broom takes a lot of physical skill and training

 

 

"There are likely a wide range of other human chores that I am not thinking of right now that will also greatly enhance our quality of life

when done by robots."

 

No doubt

[Mind]

I used to think self-driving cars would be cool, because I thought I would purchase the car and operate it as I see fit. I would be the one telling it where to go and when.

 

But that is not what is happening. Self-driving cars will be like the bus or the train. The government and/or corporations with own and operate the cars. They will decide where you can go, how fast, and when. You will lose the skill of driving and then become dependent upon the government even more. IMO.

[Mind]

Elon Musk asks a similar question to what has been discussed in this thread. If AI can do everything you can do and do it better and you have nothing left to contribute to society or the future, "does your life have meaning?" If you are "on vacation" for the rest of your life, sleeping in, doing drugs, watching porn, or AI-generated super-entertainment all the time, is that good?

 

If you are a non-contributing life form on the planet, will you be eliminated?

 

This is new territory. On the planet right now there are millions of species. Even though humans are of the highest intelligence, we still need "the rest" of nature in order to survive. AI only needs energy - essentially. Once AI and robots have secured the energy they need, why keep humans around? After all, many governments and environmental NGOs around the world already constantly screech about how awful humans are, like a virus, destroying the planet, obliterating "the climate".

 

Do you think you are going to be allowed to travel wherever you want on your own mega-yacht? Live in a mansion? Play in extremely energy-intensive virtual reality all day long? Have a sexy robot partner? Based upon current trends, you will be getting a tiny ration of food/energy/housing/travel. You might have the most material wealth right now. It might go downhill soon.

[mag1]

Thank you for replying Advocatus Diaboli!

 

The thinking for the humanoid to drive the car is that self-driving cars goes from something that might be years on the horizon

to something has potential for medium universal uptake. Further, robots could then be recognized as being able to replace nearly

any human action in bricks and mortar reality.

 

This is of importance because otherwise we would need to create a robotic replacement for every tool/activity and eventually

people would give up because of all of the expense involved. With robot world, there is only ever one item to purchase -- the robot itself.

Once this is recognized the roll out of robots would be greatly accelerated.

 

This could have profound benefits for our societies. As noted earlier, car accidents alone cause a colossal amount of mortality and

serious injuries every year. There are many hundreds of thousands of serious brain injuries each year caused by car accidents.

Life would be immensely better if we could have universal roll-out of auto-drive sooner than later. With humanoids this then becomes

highly plausible. The robots could simply jump in the driver's seat and away they go. Safety greatly amplifies when all the transport

grid is done by humanoids. At the price points that we are seeing for humanoids and the potential for roborental, the economics

already seem feasible for a rapid transition to roboworld.

 

One could also imagine that robots could be very helpful personal care workers for people with severe disability. This would be an

enormous win for the disability community.

 

Thinking of humanoids then as essentially universally capable offers us an extremely positive future.

[mag1]

Lately I have been thinking about the alpha Go discussion that the thread had almost 10 years ago. At that time I was quite concerned

about the social implications of such powerful AI. Some on thread (presumably from Asia) thought it odd that I would have such concerns.

Yet, it has occurred to me that it was exactly around this time ~2016 that Asian nations (in particular South Korea) entered into a decline

into near extinction level total fertility rates. Currently, we are at ~0.70 and there seems no bottom in sight. Leadership in South Korea

have recognized this as a social crisis. What we see here is that Asia has this broadly positive interpretation of AI. Once AI defeated Go

masters, this was seen as the time for humans to step aside and allow our robot overloads to take over and run a more efficient society.

Humans had served their purpose.

 

Most in the West did not see the triumph of Go as the end of our species. Fertility continued on mostly as it had been. Now however, we

see the potential roll-out of universally capable of humanoids. Westerners deeply care about work and the need to feel that they are 

meaningfully contributing. Humanoids could easily take this sense of mission away. This certainly suggests to me that this could be the

Go moment for the West. This might be the trigger for fertility collapse in the West just as we saw fertility collapse in Asia starting with

alphaGO.

[mag1]

I will not out members of this thread, though I really think we all need to feel some love.

We need to get away from all of this "people are only air breathers", "useless eaters" etc. etc. type thinking and embrace more of an unconditional love perspective.

That might be the only way we as a species can survive.

 

You are loved for who you are -- not what you are.

Love yourself.

 

Indeed now that humanoids and AGI feel to be close, we could all soon be "air breathers".

All of this there are these important people and then there are these technoserfs really needs to stop.

There could soon be a very rapid social decompression in which there is one class -- technoserfs.

The arguing for exceptionalism could soon seem very implausible.

When you have an AI that knows everything that is known and a humanoid that can be rapidly programmed

by a planet full of amateur AI reinforcers, the social hierarchy quite reasonably might soon look like a pancake.

 

That truly will be stunning! You have all of these humanoids out there and then people start teaching the robots to

do all sorts of interesting tasks-- such as ironing, fixing the plumbing, dusting etc. etc. and all of this training is then

uploaded back onto the robotic companies servers ... the training effect could be profound and rapid.

 

I am not sure if enough time has passed since my last outburst (think it was about 10 years ago on thread) when I proclaimed

Vive le free lunchers libre!

(I really felt that one at the time and still do, though I got quite a bit of pushback and at least one wtf!)

Nevertheless, I will repeat -- Vive le air breathers libre!

(This might also result in some push back)

 

While it is true that previously I had a decidedly anti-natalist perspective as there seemed to be a great deal of uncritical reproduction going on out there,

I think now that we have entered this extinction level era, we need to be more caring and loving.

We need to care more about people because our planetary life boat is now starting to thin out.

So, there is a self-interest perspective to work.

Yet, it needs to be genuine.

 

Interestingly, humanoids could move us to this more loving era, by doing the dangerous and unpleasant jobs for us.

Humans are simply too valuable to be put at risk of harm.

Jobs should not expose workers to physical, mental or esteem harm.

We need to up our conception of human rights now that we have humanoids that can substitute for humans in such capacities.

Also when humans are simply not able to reliably perform tasks such as driving cars, humanoids/robotics need to be developed for such tasks.

 

We need to step up and pay up and show our humanity.

There have been nations that have tried the more abysmal route of wanting higher fertility and then not paying the bills when the stork arrives.

This has only given us a permanent example of very flawed population policy that if anything has made our fertility crisis even worse.

Edited by mag1, 28 May 2024 - 04:02 AM.

[Mind]

I will not out members of this thread, though I really think we all need to feel some love.

We need to get away from all of this "people are only air breathers", "useless eaters" etc. etc. type thinking and embrace more of an unconditional love perspective.

That might be the only way we as a species can survive.

 

You are loved for who you are -- not what you are.

Love yourself.

 

Indeed now that humanoids and AGI feel to be close, we could all soon be "air breathers".

All of this there are these important people and then there are these technoserfs really needs to stop.

There could soon be a very rapid social decompression in which there is one class -- technoserfs.

The arguing for exceptionalism could soon seem very implausible.

When you have an AI that knows everything that is known and a humanoid that can be rapidly programmed

by a planet full of amateur AI reinforcers, the social hierarchy quite reasonably might soon look like a pancake.

 

That truly will be stunning! You have all of these humanoids out there and then people start teaching the robots to

do all sorts of interesting tasks-- such as ironing, fixing the plumbing, dusting etc. etc. and all of this training is then

uploaded back onto the robotic companies servers ... the training effect could be profound and rapid.

 

I am not sure if enough time has passed since my last outburst (think it was about 10 years ago on thread) when I proclaimed

Vive le free lunchers libre!

(I really felt that one at the time and still do, though I got quite a bit of pushback and at least one wtf!)

Nevertheless, I will repeat -- Vive le air breathers libre!

(This might also result in some push back)

 

While it is true that previously I had a decidedly anti-natalist perspective as there seemed to be a great deal of uncritical reproduction going on out there,

I think now that we have entered this extinction level era, we need to be more caring and loving.

We need to care more about people because our planetary life boat is now starting to thin out.

So, there is a self-interest perspective to work.

Yet, it needs to be genuine.

 

Interestingly, humanoids could move us to this more loving era, by doing the dangerous and unpleasant jobs for us.

Humans are simply too valuable to be put at risk of harm.

Jobs should not expose workers to physical, mental or esteem harm.

We need to up our conception of human rights now that we have humanoids that can substitute for humans in such capacities.

Also when humans are simply not able to reliably perform tasks such as driving cars, humanoids/robotics need to be developed for such tasks.

 

We need to step up and pay up and show our humanity.

There have been nations that have tried the more abysmal route of wanting higher fertility and then not paying the bills when the stork arrives.

This has only given us a permanent example of very flawed population policy that if anything has made our fertility crisis even worse.

 

Just to be clear, I am NOT saying that humans have no value. Just the opposite.

 

I only shared the current NGO/ government/media opinion about humans - which is (essentially) that humans are a virus and are destroying the planet. If that opinion continues to gain momentum, then AI/robotics might be used to eliminate people instead of bring a better future for all.

[mag1]

Mind, sorry I missed the antiphrasis in your previous comments.

 

Yes, I was not sure about that -- I actually read it straight as written -- mouth breathers and useless eaters etc..

I have had enough engineering adjacent life experience to not automatically assume that such talk was intended

in a way other than literal. Basically, I had to eventually escape from that environment because it was not for me.

At some point in my life I Iooked in the mirror and accepted that I was human and that this was how I wanted to

interpret reality. I suppose that I could have fantastically more money than I have now if I went with a more 

pro-robot life view, though I am happy with how things have turned out for me. I am not sure that I could be

personally happy -- even with wads of money -- in techland as we are staring so directly into a social abyss.

The time frame might roughly be on the scale of a year or two before there could be a Civilizational Singularity.

What exactly meaning would money even have at that point? We can all live gated communities in bunkers

patrolled by armed robotic dogs?

 

You do not even need this look in the mirror and escape approach anymore -- Genetics can answer these questions

for you (through the use of polygenic scores). It is now so helpful for people to find out where they belong in the universe

and not just go through life and use a trial and error approach. My genetics is easily understood and it points me

strongly towards where I have large comparative advantages.

 

 

The engineering types just could not stop laughing when I mentioned     ---- studies, or facilitates and I also found it 

best to stay away from deontological. Through time I have more fully appreciated their perspective that complex systems

(such as society etc.) need to be clearly specified before you can make any progress towards completely controlling them.

If you never get around to operationally specifying anything, you can have massive libraries of research that has no actual

value. In time it mostly has to be all discarded. Yet,with true science based upon objective standards you can start with 

1+1= 10 and before you know it you are within sight of AGI. Finally we are moving towards a precise genetic specification

of human biology and psychology which should lead to profoundly better functioning societies.

 

My genetics did not conform well to engineering culture. Not long ago my full genome sequence read out in the single digit

percentiles for autism spectrum, this makes me more of an anti-autist and perhaps largely barred from engineering. However,

the sequence actually did give me more of an autistic adjacent readout. With respect to our current situation it might be very

useful to think of creating engineering schools that specifically selected at a genetic level or otherwise for those who were

non-autistics. The problem that should be all too clear is that we have this rapidly emerging AI and those right by the levers

of power are so much on the spectrum. The counter-balance that was added to prevent run away AI risk was easily removed

when they were not seen as team players. If engineering could be shifted so that the leaders had high technical competency

and also were more towards the anti-autist end of the spectrum perhaps our species would have a higher chance of surviving

the approaching Singularity.

 

Edited by mag1, 31 May 2024 - 02:31 AM.

[adamh]

Mind, are you back to watching horror scifi movies again? "Useless eaters" and ai will somehow get rid of us? I thought you sobered up and decided that would not happen? I have asked many times how computers, which is essentially all ai is, a computer with advanced programming, will destroy humans or how they would get the motive to do so?

 

Yet, each time I ask, you run in the opposite direction and talk about something else. No one has come to your rescue and filled in the blanks but here you are again with this 'evil robot' nonsense. How do they become evil, how do they avoid needing humans to repair them and provide energy? If a computer 'wanted' to kill humans, it means it was programmed to do so. Have you ever seen a computer do something it was not programmed to do? You talk as if humans take resources away from computers, which frankly makes no sense.

 

I don't expect you to back up the latest paranoia with facts but perhaps you can see that this fear is irrational?

 

And the old trope about if humans get too much leisure time then society will collapse and everyone starts to kill everyone else. You probably saw one of those movies too. It hasn't ever happened and the wealthy in society tend to be the most healthy and happy. They also do not have to work so that little theory falls apart when examined in the light of day. 

 

Mag1, you do try to back up your ideas with facts rather than just speculation, and thats a good thing. But its easy to be influenced by hysteria and group fears. You speak of a singularity which comes from a scifi movie I believe. What we will have in the future is a society in which there is no need to compete. People compete for power, money, and status. With ubi there is no need to struggle to make money, decisions will be made with help from ai and will be logical.

 

Games and recreation in general will become major industries. Some people will decide to spend much of their life in ai created reality with limited interaction with the outside world. Most will simply dive in to play a game or to experience life in a different setting, perhaps in nature or in a fantasy world then come back to normal life. People will explore the planet, the solar system and beyond using artificial reality systems which have already started to come out

 

Homelessness will become a thing of the past and the sick or insane will get treatment. The low birthrate problems will be solved and a stable earth population will be maintained.

 

But, the next 5 years will be hard due to wars, disease, bad economic and political policies. After that comes the good part

[Mind]

Mind, are you back to watching horror scifi movies again? "Useless eaters" and ai will somehow get rid of us? I thought you sobered up and decided that would not happen? I have asked many times how computers, which is essentially all ai is, a computer with advanced programming, will destroy humans or how they would get the motive to do so?

 

Yet, each time I ask, you run in the opposite direction and talk about something else. No one has come to your rescue and filled in the blanks but here you are again with this 'evil robot' nonsense. How do they become evil, how do they avoid needing humans to repair them and provide energy? If a computer 'wanted' to kill humans, it means it was programmed to do so. Have you ever seen a computer do something it was not programmed to do? You talk as if humans take resources away from computers, which frankly makes no sense.

 

I don't expect you to back up the latest paranoia with facts but perhaps you can see that this fear is irrational?

 

And the old trope about if humans get too much leisure time then society will collapse and everyone starts to kill everyone else. You probably saw one of those movies too. It hasn't ever happened and the wealthy in society tend to be the most healthy and happy. They also do not have to work so that little theory falls apart when examined in the light of day. 

 

Mag1, you do try to back up your ideas with facts rather than just speculation, and thats a good thing. But its easy to be influenced by hysteria and group fears. You speak of a singularity which comes from a scifi movie I believe. What we will have in the future is a society in which there is no need to compete. People compete for power, money, and status. With ubi there is no need to struggle to make money, decisions will be made with help from ai and will be logical.

 

Games and recreation in general will become major industries. Some people will decide to spend much of their life in ai created reality with limited interaction with the outside world. Most will simply dive in to play a game or to experience life in a different setting, perhaps in nature or in a fantasy world then come back to normal life. People will explore the planet, the solar system and beyond using artificial reality systems which have already started to come out

 

Homelessness will become a thing of the past and the sick or insane will get treatment. The low birthrate problems will be solved and a stable earth population will be maintained.

 

But, the next 5 years will be hard due to wars, disease, bad economic and political policies. After that comes the good part

 

For the 1000th time!!!!!

 

I am NOT saying humans are useless eaters.

 

I am NOT saying AI/robots are guaranteed to kill everyone.

 

Environmentalists NGOs and and some governments ARE saying people are awful for the planet.

 

Many AI researchers worry about how AI will be used - which is a legitimate concern.

[adamh]

Well, Mind, maybe you didn't say humans were "guaranteed" to be eliminated by robots but you have consistently forecast some sort of mad max type dystopian future as a result of ai. What I'm trying to get you to discuss is your frequent claims that robots, in other words computers with ai, are going to cause great harm or eliminate humans. 

 

Do you think computers can become "evil" and program themselves to do harm?

How do you think this would come about?

 

Just because computers now can talk, some of them, does not mean they have all human like qualities including irrational hatred, desire to destroy and so on. Thats anthromorphizing, we do it a lot, we like to think animals are like humans but can't talk, or think the weather or our car is out to get us. But we know better, most of us

 

When you look closely at an irrational fear you always find that its not as bad as you imagined and in many cases it turns out there is nothing to fear at all. You might want to drop the robots turning evil meme since most can see through it. You still have some arguing points over what effect more leisure will have on society, best to stick to that since its harder to prove.

[mag1]

adamh, I think that it is reasonable to be very worried about present capabilities of AI.

The thread, I think rightly, has interpreted the question about possible harms from the point of view of harms to humanity (i.e., the social stability of human civilization).

 

The tech community seem to have more of a take that we will only need to worry when AGI is immediately imminent.

I do not think this perspective is correct.

 

We have already seen the fertility collapse that has occurred (and is occurring) in Asia.

I expressed apprehension when AlphaGo soundly defeated a Go expert in 2016.

This match was widely followed in Asia.

Starting from that time, we can trace the emergence of near extinction level fertility in Asia.

 

The West largely ignored the Go match.

The loss did not force any deep contemplation about human relevancy.

 

However, we might now have our own Go/No Go moment.

AI humanoids might create a similar fertility crisis in Western nations. 

Perhaps over the next year we could have highly functioning human robots that could replace a substantial fraction of the workforce

and this might provoke us to think of our place in the universe and whether the next generation will even be given the chance

to make a place for themselves in such a universe.

 

 

It would then not take a leap of the imagination to realize that the next generation of the robots will be able to do everything better

than all humans. Fertility would then really collapse. There is no great counter-argument that I could give that would

refute the logic of humanoid supremacy. If such a fertility crisis were to start, I am not sure how we could counter-act it.

Asia has yet to be able to do so. It would be a Go / No Go type moment.

 

I do not think deep concern about this emerging potential social consequence of AI is unreasonable.

 

Edited by mag1, 19 June 2024 - 02:33 AM.

[adamh]

Someone lost a go match and that made the country infertile? No, because fertility is down all over the world. Its down in usa, down in europe, down in china... I'm not sure if there is any country where its up. In many if not most places, the birth rate is below the replacement rate. 

 

I think its because the planet is overcrowded. I think there is something in the genome that reacts to over crowding and reduces the fertility rate. I think you will find that in areas that are thinly populated you will find larger families. Many years ago large families were commonplace. Infant death was high and having a lot of kids meant some would probably survive and be able to work. Large families are less and less popular for financial reasons among others. Some people look at what is going on in their country and in the world and don't want to bring kids into it. Its that or maybe they lost a go match.

[Advocatus Diaboli]

Re post #916

 

The answer, of course, lies within Aldous Huxley's "Brave New World". Artificial wombs, childhood indoctrination, and the separation of people into classes, or castes, according to mental ability. Add in a little "soma" to keep the common folk happy and in check, and you have a solution--plenty of peons for menial work, like building those robots that can't build and repair themselves, and caring for an ageing population, among other functions.

 

As far as putting trust in an AI to answer and act on what I believe most people would consider to be a no-brainer such as: 

 

"If killing a cat could have prevented Hitler from killing millions of people, would it have been ok to kill the cat?"

 

Here is the first part of how Perplexity AI answers the question:

 

"This is an extremely complex ethical dilemma without a clear right or wrong answer. There are several important considerations:" (My emphasis)

 

"extremely complex ethical dilemma", say what? "without a clear right or wrong answer", say what? Keep that AI away from me and out of the hands of any government leaders that run the country and that may rely on AI in their decision making.

 

 

[Mind]

Every musical artist can be impersonated with perfect fidelity. AI can generate de novo music and voice. Artists are worried of course.

 

The positive: everyone can "create" great music now.

 

The negative: Human artists are done. No more gravy train. No more huge record deals. (of course, human artists will still perform, and make a tiny bit of money, but the big corps are going to pump out AI-generated music by the billions of tracks and scrape up billions of pennies along the way.)

[adamh]

Its just another tool. You think ai art is so great and will put real artists out of business? I think most of it looks like crap. It can imitate great artists but thats it. Artists will use ai to further their art. Da vinci used students and other artists to do less important parts of his work. That was ai in the old days. Todays artists will lay out the theme of the work and let ai fill in the less important parts or to show how a technique looks

 

Advocatus diaboli, that answer sounds exactly like ai. Humans don't talk so foolishly. People think ai is highly intelligent but its just a jazzed up computer, a piece of machinery. A book can contain a lot of information but is it smart?

Edited by adamh, 21 June 2024 - 10:04 PM.

[mag1]

Late 2025 for GPT 5 with PhD level intelligence?

This is going to be fantastically exciting!

 

Most people have maxed out their cognitive ability at not much more than high school level and then coasted.

You can read a newspaper then you're good.

GPT5 will allow people to be constantly challenged at their proximal level of development and that will not have to mean 

some population determined outcome.

 

We could see robots and high level GPT in about a year very very awesome!

Of course, the downside might be that fertility goes poof. 

[mag1]

One of the insights that the thread has picked up on is that we might soon witness a robo-invasion. Oids offer such enticing economic opportunities that there is this near frenzy happening 

to bring the tech to market. From what I understand 2025 could be the year of mass rollout. The market for such a product could be truly massive. Global capitalism needs to constantly find

something that people want.

 

For quite a while I have been somewhat blah -- I did not really want anything -- give me money  -- meh? - mo money mo problems. But robo-world! Yeah!

this could be super-fantastic! It would be a whole new world. It would be very exciting. People are just too busy all the time -- they need to work all the time -- they need to be on their

cell phone; there is just no time. With an oid, life could be all now and all the time it'll be great! People can do what they want to do and I can do what I want with my oids: Everyone is happy..

Edited by mag1, 16 July 2024 - 01:28 AM.

[adamh]

One of the insights that the thread has picked up on is that we might soon witness a robo-invasion. Oids offer such enticing economic opportunities that there is this near frenzy happening 

to bring the tech to market. From what I understand 2025 could be the year of mass rollout. The market for such a product could be truly massive. Global capitalism needs to constantly find

something that people want.

 

For quite a while I have been somewhat blah -- I did not really want anything -- give me money  -- meh? - mo money mo problems. But robo-world! Yeah!

this could be super-fantastic! It would be a whole new world. It would be very exciting. People are just too busy all the time -- they need to work all the time -- they need to be on their

cell phone; there is just no time. With an oid, life could be all now and all the time it'll be great! People can do what they want to do and I can do what I want with my oids: Everyone is happy..

 

Yes this is the logical result. People will have more free time, they will not have to work, they will be able to do all the things they wanted to but were stopped by economic limitations. 

 

Another cool thing that is being developed is the exoskeleton. Originally developed to help factory workers move heavy items and carry them around. It can be adapted to give a paralyzed person mobility. It substitutes for your muscles or just helps a little depending on the setting.

[adamh]

Mind and others often claim that UBI would cause people to go crazy or have many problems. However, when its been tried it has worked well. For example, alaska gives its residents money every year. It can be up to $10,000 per family and is given with no strings attached. People said unemployment will go up if you have ubi but that has not turned out to be the case. Unemployment actually went down, poverty went down and overall health improved. Pretty much the opposite that the nay sayers have been telling us. 

 

https://m.kuow.org/s...c-fund-dividend

 

With robots being able to do more and more jobs, we are moving closer to the day of ubi large enough that working becomes optional. That will be good because at the same time, there will be less jobs available because machines will be doing them. Machines don't require a salary, just maintenance, energy and repair. Sort of like your car, your car does a lot of work for you, makes your life easier but instead of paying it a salary, you just use it. 

 

There will be a period of adjustment, similar to how the automobile replaced the horse and carriage. Horse breeders and carriage makers found work in factories. When the robot replaces the human laborer, there will be some disruption but the monthly check will make it good. Throughout history there have been examples like that. The telephone replaced the telegraph which replaced messengers. The horse replaced a lot of human labor and the internal combustion engine replaced a lot more. No more pony express.

 

In 10 years the benefits will be clear and only a few luddites will still moan about the changes taking place. There are those who still say we should go back to the horse and buggy, ditch electronics and other advances but progress goes on. Social security will probably be folded into ubi along with welfare, food stamps, vouchers and so on. They say ss will go broke soon or reduce payouts. We need the boost from robots right away

[Mind]

A couple of recent articles about the advancement of AI.

 

It isn't quite "intelligent" yet. It can't do de novo creative activities. It still relies upon the database of human data.

 

As people use AI more, they lose "intelligence" and might be more susceptible to neuro-degenerative diseases. I see this all the time where I work. Young people have little internal problem solving skill, very little memory of facts, hardly any innate creativity, etc... As long as they have their phone/AI they can work and answer questions. Without it, their IQ is embarrassingly low.

[Advocatus Diaboli]

re: post #925

 

"It isn't quite "intelligent" yet. It can't do de novo creative activities. It still relies upon the database of human data.".

 

"It can't do de novo creative activities."

 

Such as?

[adamh]

 

As people use AI more, they lose "intelligence" and might be more susceptible to neuro-degenerative diseases. I see this all the time where I work. Young people have little internal problem solving skill, very little memory of facts, hardly any innate creativity, etc... As long as they have their phone/AI they can work and answer questions. Without it, their IQ is embarrassingly low.

That argument has been made throughout the ages. While its true that not using an ability leads to not being good at it. If you don't read you lose the ability to read, if you don't do math problems you will never be good at that. Fortunately the education system has been able to make most people able to read and write and do simple arithmetic. 

 

Many years ago when writing was invented they said it was useful but it made people lazy. Instead of memorizing long stanzas of epic poems and accounts, they could simply read them from paper. No need to memorize. People much like yourself sounded the alarm saying this will cause the public to become stupid since learning and memorizing was considered very scientific. And while having a good memory is certainly a worthwhile skill to have, not doing memorization does not seem to have made anyone stupid. Or should we go back to the old ways? 

 

Whether computers can become intelligent may be a matter of semantics. Humans learn from trial and error and by following procedures that work. We learn the process to solve a math problem and follow the steps. We use logic to solve problems and computers can do this too. The things the computer can't do well right now are becoming fewer and fewer. They don't understand emotion but can mimic it. 

 

Robots are coming and are already here. We might as well get use to it and make the best of the situation

[mag1]

The excitement builds!

 

Many times after you let some enthusiasm of the moment cool down on a shelf for a while, when you return with a more clear perspective you tend to reevaluate to a less

enthused stance. With the humanoids/scooterbots, this would seem to be untrue. If anything my excitement has intensified in my month away from thread.

 

During the last month I became aware of a scooterbot implementation that could bring down the cost of bots to ~$2,000. That is a truly stunning price point. Up till then I 

was locked into a price of $20,000 with the humanoid robots. For many that would not realistically fit in their budgets at a consumer level. With those type of prices we started

hearing how this might work out to ~$30 per hour roboworkers. Yet, there is very minimal advantage in spending almost $20,000 just to have a cool bipedal. Moving to a

scooterbot would give you all of the benefits of a humanoid (without bipedalism) for only $2,000! OK, scooterbots it is!

 

 

At ~$2000 there simply will be a massive massive consumer market. Typically, one might frown upon governments trying to

centrally plan consumer markets, yet with this bot market it is so obvious and near term and some level of coordination by government could be helpful.

 

The range of capability of a domestic scooterbot essentially might be human equivalent -- every task that a human could complete, a scooterbot could do to perfection.

A near term goal of artificial general human equivalence (AGHE) no longer seems implausible. With all of the robots reporting back what they have learned to some central computer,

such AGHE might occur quite rapidly. We could have a global scale reinforcement and selection for many many bot behaviors. Basically, download the task app that you need.

 

As I noted earlier on thread this rapid commericalization of these robots has been brewing for many decades. The basic robotic hardware has been on the shelf for quite a long time. The

main limiter was the lack of brain power. As soon as we had the GPTs, Frankenstein's brain could be grafted into the bots and here we are in roboworld. This is just really really stunning.

There is widespread talk of moving to mass production for various bots (over the scale of months) and the beta versions are already moving out into the field for initial testing.

2025 simply could be a massive year in transforming life into a robot saturated landscape. 

 

I am somewhat surprised by the speed at which this rollout appears ready to happen in comparison to my other great passion on thread -- the robobuggies. 10 years after the start of

the robobuggy global takeover, we still do not see many of them around. It has been a somewhat less than total victory. My assessment had been modified to expect that bricks and

mortar would likely always be like that. You can write software and change the world immediately (e.g., GPT) --  You can make hardware and not change the world

even 10 years later. The Bricks and Mortar world simply has too many people who can block innovation. Admittedly with robobuggies, my perception was that the entire urban retail 

landscape would turn into moonscape if it were to scale. Robobuggies then are an example where the Luddites won.

 

Why will the Luddites probably not win with scooterbots? Scooterbots will offer such overwhelming value and convenience for billions of people. Scooterbots will allow everyone on the planet

more time to do the things they want in their lives and less time devoted to chores that give them little if any joy. Cooking, cleaning, shopping, collecting garbage bins .... these activities give me

no particular joy-- I do them because I have to. I am forced to do these labors against my will. Scooterbots will free me and everyone else of these nuisances. It will be great! Perhaps the secret

sauce here is that all of these chores do not actually create any labor dollars for others. It all amounts to dragging everyone through unpaid drudgery. Scotterbots will be such a boost!

So there really is not an obvious victim with this technology. We can live again! Happy New Year to everyone for 2025! It is going to be fantastic!

 

Other big news today is that Strawberry has been released (what has previously been referred to as Q*). This LLM has enhanced reasoning ability up to that of PhD level -- clearly things are now

starting to intensify. It now feels as if we are in that exponentially filling lake and there is only one or two doublings until it overflows -- This possibly could happen within the next year or two. There is

actually a very short interval with an exponentially filling lake in which it is somewhat filled but not overflowing -- that is probably where we are right now. Also over the last month or two a scientist

LLM has been demonstrated that can think up interesting experiments and write them up. One can only imagine to expect if we were to go to bed one night and closed the loop on this scientist bot 

and just let it ask interesting questions, run its own experiments in a lab and then iteratively continue with such research. Knee high green goo at planetary scale?  Enjoy the calm before the storm everyone.

 

Edited by mag1, 12 September 2024 - 11:59 PM.

[adamh]

Very good points, mag1. It is indeed foolish to insist that a 'bot look just like a human and walk on two legs. They also want human faces and voices but why? Most future bots will look more like r2d2 which I recall was a short squatty bot that ran on wheels. They should be built to do the jobs we need done. There will be specialized delivery bots, amazon will boot all their delivery people to the curb and use robots. Your lawn mower will be self programmed and not need supervision. A bot will fix dinner and take out the garbage, watch the house against burglars or fire or other emergencies. 

 

I saw a short video of a robot on the Ukraine battlefield. It was on wheels, not very big, it was armored on top against small arms fire. It rolled into position where the enemy was and then blew them and itself into bits. In the future it will deliver mail and hopefully not blow up.

 

The super humanlike robots will be the personal care bots and sex bots. They will become our friends and confidents. Bad guys will hack them to steal our secrets. Poor people will have mechanical looking bots and the rich will indulge in super lifelike bots. I saw a video of human like bots doing gymnastics, quite amazing. There was a robot show in china recently in which they put real human models inside costumes and gave them makeup. They tried to pass them off as super advanced bots but I'm not sure anyone was fooled

 

The day will come when we can't tell the difference.

[mag1]

Adam, i just think this is stunning! The beta versions that they are rolling out for home testing look very very shaky. It seems almost that they feel obligated to launch a product now because there is just so much buzz for them. My guess is that we will see truly massive improvement in these bots over the next year. That would seem to be true because: 1. they have a teleoperation functionalty and 2. domestic applications are fairly static and standardized. If there is some task that needs to be done, then they can go to teleoperation mode to get it done. All of these movements could be stored and then they could used as training allowing for fully autonomous task performance. Another consideration is that in a home environment most task are highly routinized. The AI training for self-driving cars has been quite challenging because roads are not highly predictable environments -- every time you drive down the same route there are many new features. Yet, with domestic tasks if you cut the lawn once you have pretty much got the task down. There is minimal change in the environment from one cutting to the next. This would apply to most features of household chores -- people largely can go onto autopilot when they are doing these tasks -- so could the robot. Further, the suburban environment is highly standardized; so doing a task in one house is probably pretty much the same as doing this same task in another house. This suggests that once the task is perfected in one environment it will largely be perfected in all environments. With self-driving cars they have needed to have buildings filled with humans who must annotate the video from self-driving cars to help the AI "see" important features of the drive. The training requirement for domestic robots is probably going to be much less.

 

It might be a good idea for the robot companies to give guidance on the type of domestic tasks that have been perfected so people would not feel disappointed when they order one. Also they might want to have instructional videos on what they need to show the robot to complete various tasks. So for lawn mowing, a video might demonstrate how a robot owner should show the robot where the power cord might be, where the mower is, the cutting pattern for the grass, etc..

 

This is super super exciting! What we are seeing emerge is a technology that is more of the world. Over the last century we have become increasingly sedentary -- we stare at screens almost all day long. The mental health crisis that we are experiencing is then largely inevitable. I have been out most weekends over the last few weeks doing yard cleanup and it has made me feel great! If any one out there is feeling in the dumps -- then go out and do some yard work, you will so much better for getting away from screens and actually moving around a bit. Robots will allow us to stop staring at screens and interact more in a human way -- we could talk to our robots and they can engage us in physical activity perhaps activities such as fencing etc. that might help us keep nimble and in shape. This next year is going to be super amazing!

 

One other feature of the scooterbots is they have an 8 hour charge. Bipedal robots are quite inefficient so they only have about a 2 hour charge. The scooters easily outcompete them. I suppose there might be clever ways to allow scooterbots to move between floors --perhaps they will use pulleys are have an attachment for the stairs. While the current AI robot technology is fairly non-functional, it clearly seems likely that we will see a rapid evolution over the nearish term.

Edited by mag1, 25 September 2024 - 02:22 AM.