Expiration date on the economy as we know it

[Jarvis323]

Ion thrusters would be useless in such a scenario. Again, you do not understand the logistics of your proposals.

That may be so, but I never claimed that I did. This is physics forums. So how do we get our 10 billion solid gold mansions down to Earth most effectively?

 

[phinds]

That may be so, but I never claimed that I did. This is physics forums. So how do we get our 10 billion solid gold mansions down to Earth most effectively?

No, it's your proposal so it's up to you to figure that out. If you are interested in such topics I'd suggest a new thread but I can tell you that using today's technology, you just don't. It's cheaper to mine gold here. Future goals are much better but they are just goals, not yet reality.

 

[russ_watters]

AI in this context just matters in the sense that input from human beings is not a limiting factor. In space, the limitation is energy and natural resources, which are quite vast. It really doesn't take much more than the technology we already have (only a little bit more sophistication) to begin exploiting it...

It all sounds bizarre, and I was trying to be a little comical. But the core of what I am saying isn't science fiction any more than laser beams or televisions are.

To me, you are making several very large leaps that cross the boundary between reasonable speculation and sci-f/fantasy. I don't think any of your three premises are anywhere close to accurate/realistic.

 

[russ_watters]

Well, now I have to switch gears and agree w/ jarvis. I agree w/ you that nothing is totoally free, but if we have (and I think we will have) intelligent robots in space that maintain themselves and/or each other, and that are capable of building and using mining machinery and building and using return rockets (this is a bit far off in the future but doable) then the cost to us on Earth becomes more about where we PUT all the hardware that is returned along w/ the stuff we value, not what it costs to get it here.

Why wouldn't we use such robots on Earth first?

 

[phinds]

Why wouldn't we use such robots on Earth first?

Oh, I wouldn't disagree w/ that, I'm just trying to stay within the OP's context.

 

[russ_watters]

Oh, I wouldn't disagree w/ that, I'm just trying to stay within the OP's context.

I think missing context is precisely the problem withe the thread...

 

[russ_watters]

Ok...

What such questions tend to lack in their fanciful speculation is perspective (context). So I'll take a few steps back and try to provide some.

Part 1: What Do Humans Want from Technology?
Fundamentally, what separates humans from animals is our nearly unlimited ability to use tools, thanks to our big brains and opposable thumbs. The entire history of human development is based on using tools to reduce our workload and increase productivity -- as well as accomplish tasks we are not capable of doing naked.

[stops in middle of writing post to make a sandwich]

The qualitative improvements may be hard to quantify (by definition), but the quantitative ones are not:

I could probably travel 30 miles in a day, with a small pack, on foot. On a horse, maybe 100 miles, with a larger pack (never tried either of those things, so I don't know - I'm just making it up). Later this month, I'll be driving 450 miles in about 9 hours including stops, carrying probaby only a couple hundred pounds of gear but could do thousands if I wanted. An orderds of magnitude improvement. Even planning the trip; no more spending an half an hour with a map, planning and measuring the route and then hand-calculating the time; google maps does it faster than I can type in the request.

We build buildings 100 floors high with cranes capable of lifting dozens of tons. If I want a bathroom floormat, I don't weave one myself in 3 days, I drive to Kohls where one that came off a machine in China at a rate of 10 per minute is sitting there waiting for me.

If I want to communicate with my sister in Boston I don't write a letter and then hire a Pony Express rider to run it up there in a few days, I just Skype her.

The point is, all of our tools make us better and faster at things we want to do and enable things we otherwise wouldn't even be capable of without them. And that progress - while uneven - has produced exponential improvement.

But what hasn't change much at all is the human's workload. Over the past century, most people have worked between 35 and 60 hours a week, always.

Why?

Because despite some people's protestations to the contrary - they are fooling themselves - we all want "stuff", because "stuff" is just awesome. I could choose to work fewer hours at an easier job, but then I wouldn't be able to afford air conditioning, Netflix, a good car, a telescope permanently mounted on my deck, etc. Any one of those "things" I own would make a medeval king drool, and the combination of them is just fantastic.

The point of all of this is that automation reduces our workload allowing us to do more, better. The net result is a near constant workload for humans, for a long time in the past, which will almost certainly continue for a long time in the future.

But what about robots/AI? Can't they make human labor obsolete?

 

[Michael Price]

But what about robots/AI? Can't they make human labor obsolete?

Super-AI is different from labour augmentation, because we are talking about the complete replacement of humans in the productive economy. (You don't see chimps doing assembly line jobs, do we?) More jobs will be created, but not for humans. Who, or what, would employ a human when a robot is cheaper, more reliable and more capable. Humans have to live on investment income, from the robo-economy, as pensioners.

 

[russ_watters]

Part 2: AI/Robots

I'll just say it: I think AI is a myth. It's bigfoot. It doesn't and won't ever exist, except that it already does.

[finishes sandwich, puts plate in diswasher]

Bigfoot is probably a bear. AI is just my refrigerator ordering milk for me when I run out (I don't own one, it's just an example).

"But", you say, "isn't AI 'human replacement intelligence'"? I reply: I've been running spreadsheets on a PC for 25 years -- they are much, much better at math than I am.

"But", you say, "isn't AI also technology successfully masquerading as psople?" I reply: who cares? No, seriously. Why do I care if I can't tell if the voice on the other end of the line is a human or computer? Neither have ever provided me good customer service.

Then, robots; We have them. They're awesome. They recently built me a car.

"But", you say, "they don't look/talk/walk like people!" I reply: Why would I want that? The human body is a terribly designed machine. That's nothing more than an unnecessary aesthetic preference. I don't need a robot that looks like an English butler in a 3-piece suit and a bowler hat to make me a sandwich. Honestly, once the novelty wears off, I'll just find him creepy. I'll be perfectly happy when I get my sandwich robot from Kohls/Amazon in a few years. It'll be about the size of a microwave oven and cost $59.95. Once a week, I'll load in a loaf of bread, jar of mayo and slabs of turkey and cheese. Then, at the press of a button (or on a timer), it'll spit out a perfectly made turkey sandwich for me. Awesome.

A turkey sandwich robot is an incremental improvement only, but that's my point. That's how technology has worked for the past few millenia, and it will likely continue for the forseeable future. Humans will not suddenly find themselves obsolete, we'll just have technology do more and more for us while we do the same amount.

 

[phinds]

Humans have to live on investment income, from the robo-economy, as pensioners.

Where do they get the investment income if they have never worked?

 

[russ_watters]

Part 3: But what if We Could?

Ok, I'll bite. But what else is missing here is a hand-waving over what are major - perhaps insurmountable - technological challenges requiring multiple technological revolutions to achieve:

1. Spacecraft Propulsion.
You can't handwave this away. It's expensive/inefficient. It may always be. The idea that technology can multiply itself to overcome this is a non-starter because it needs a start. Human-replacement robots can't do everything on their own. They need infrastructure and tools. You can't send a single robot into space and have him build a civilization. You need to start with a small city of them, with factories and power plants and other infrastructure. Even with fully-autonomous robots, getting all of that into space would cost - pun kind of intended - an astronomical amount of money.

2. Human-Replacement Robots
Human-replacement robots are a bad idea, but assuming we want them, I'm not so sure we can build them. We're way, way off in terms of energy and automation machinery. Humans are complicated and poorly designed, and making a quality bad robot is no easy thing. So even though we have robots rolling around on wheels or tracks exploring Mars and vacuuming my floor, people are struggling to build a robot that can walk over 2x4 like a human without tripping. It's stupid, but whatever. I think we are many decades away from making robots who can play basketball in the NBA.

But let's say human-replacement robots become possible. Now what?

3. Everything Else
Everything else still needs to be invented. I suppose we could send an army of human-looking robots to an asteroid with pickaxes and wheelbarrows, but it would be a lot better if we invented purpose-built robotic mining equipment. I suspect we've made some progress on that front already for earthbound equipment, but we'd need to invent some to operate on an asteroid/the moon. That's just one example. We're talking about re-inventing *everything* in an industrial city and then launching it into space.

So a more direct answer to the OP's question: when a human-replacement robot gets invented, then we can start conversation of what to do about it. We're many decades from that point, and society will look very different then anyway, so having that conversation is skipping over the more immediate conversation of what we should be doing for the next several decades.

"But", you ask, "won't it be too late by then?" No. This isn't "iRobot", where a company builds an entire army of millions of them in secret and then one day flips a switch and they are all online and PO'd. Decades before we have true human-replacement robots, we'll have almost human replacement robots and not quite human replacement robots. They will be fabulously expensive, have limited lifespans, and limited but ever-increasing capabilities. In fact, with such qualifiers in place, we already do have them. And we'll have decades to figure out what to do with them.

Just like every other tool we've developed over the past few millennia.

 

[Michael Price]

Where do they get the investment income if they have never worked?

Basic minimum income - it has to happen. Human-run governments will chase unemployed human votes by guaranteeing handouts funded from corporate tax (on the robo-economy). It can become self-funding (c.f. Norway's Sovereign Fund), if managed with restraint.

 

[phinds]

Basic minimum income - it has to happen.

Nothing HAS to happen.

 

[Vanadium 50]

Fifty years ago we sent 12 people to the moon. We haven't been back since. Yet we are expected to believe that in "a few decades" one will have a pocket device that can fabricate a trillion houses in space and "few", the number of decades, is so small we need to worry about it night now.

 

[Jarvis323]

Fifty years ago we sent 12 people to the moon. We haven't been back since. Yet we are expected to believe that in "a few decades" one will have a pocket device that can fabricate a trillion houses in space and "few", the number of decades, is so small we need to worry about it night now.

No t much more than 50 years before we went to the Moon, we made the first working plane. Once it is known that something extraordinary is now possible, people race to be the first to do it.

 

[Michael Price]

Fifty years ago we sent 12 people to the moon. We haven't been back since. Yet we are expected to believe that in "a few decades" one will have a pocket device that can fabricate a trillion houses in space and "few", the number of decades, is so small we need to worry about it night now.

We haven't been back to the Moon because there wasn't, and still isn't, an economic case for space exploration. Going to the Moon was a publicity stunt. But AI and robotics is being driven by commercial considerations, Moore's Law and such like.
Economic reality is being ignored in this thread. Pie in the sky does not happen.
For a technology to change to the world it has to be:
1) technically possible
2) economically viable
AI and robotics is only going to change the world precisely because it is NOT free.
Let's take a contemporary example. Solar energy. Sunlight is free, but solar energy is not. Solar energy or photovoltaics (and other renewables) is revolutionizing the world not because it is free (it isn't) but because it is cheap - cheaper than fossil fuels in many situations and getting cheaper.

 

[Michael Price]

[AI is] Just like every other tool we've developed over the past few millennia.

AI is not like every other tool. There is a qualitative difference between augmenting human labour and replacing humans with something cheaper and better.

 

[PAllen]

More and more it keeps ocuring to me that the modern economies and styles of government that we have now may have a fast approaching expiration date. For example, with AI and autonomous robotics, we could soon have exponentially fast growing industrial complexes in space. And the amount of precious resources available in asteroids dwarf those that we have access to on Earth, not to mention mining them doesn't harm our environment. So who dominates in space and robtics+AI will control the planet? Will there soon be a time when making a living wage through time-value transactions will be obsolete? So what will we do if that is the case? Should this be a hot button political issue?

Just pointing out that this is very similar to predictions by Karl Marx circa 1850 for the "not too distant future". How well did that prediction fare?

 

[Jarvis323]

Just pointing out that this is very similar to predictions by Karl Marx circa 1850 for the "not too distant future". How well did that prediction fare?

Comparisons with apparently similar historical events is something worth doing. But in the end, you have to analyze a unique situation on its own merit.

 

[Vanadium 50]

And how is this relevant?

I am coming to the agreement that this discussion is more about science fiction than science.

 

[MikeeMiracle]

I think it's telling that most scientists are primarily concerned with making changes due to climate change.

I think we need to focus on this problems which are right here and right now rather then worry about an issue which to me seems at least 100 years away.

 

[pinball1970]

And how is this relevant?

I am coming to the agreement that this discussion is more about science fiction than science.

 

[gleem]

Autonomous robots exist today so it is not scifi. Mars lander Curiosity is one application AI let it operate when in-communicato with Earth. There could and will be much more sophisticated devices sent In the probe being sent to Europa to explore the ice layer an autonomous device called Valkyrie will be deployed to drill below the ice layer. For probes sent to Jupiter for example it take about 54 minute to send a signal to a probe. The probe must have capabilities to respond to changes that might occur between signals sent and received. It seems to me that the AI available on Earth for say, autonomous cars is even more sophisticated.

When it comes to the commercialization of space the age old paradigm of cost to benefit will be employed. What innovations will reduce cost. Look at SpaceX which reduced cost of a launch to only $60 M. The almost perfect vacuum of outer space will have definite manufacturing advantages for some processes as well as availability of of continuous solar power.

 

[bhobba]

More and more it keeps ocuring to me that the modern economies and styles of government that we have now may have a fast approaching expiration date.

During my student days I remember when I turned up to the subject Methods Of Mathematical Economics I was the only student. The lecturer almost canceled the subject but the others I could have done instead didn't really appeal and my academic adviser said it was probably a good idea to do some kind of humanities subject so is there some way I could still do it?. He said OK and I used to see him for an hour once a week in his room where he gave me reading, problems to do and assignments - usually computer simulations.

The above was just a pre-amble. I always remember what he said on the first day. Our current models are all basically a crock when compared with what actually happens. For example it is well known department of treasury predictions are never right - they always make assumptions such as people will put their savings into the bank to get safe interest from which they can collect tax. Instead they often put it in the share market to avoid just that. Nice math and some interesting insights but they never really actually work. With the advent of more powerful computer and complex models, our ability to predict short term economics was getting better, medium term had some value as a guide, but long term - still utterly useless. I do not think things have changed - predicting long term what our economy will look like is just an intellectual exercise not to be taken seriously.

I sometimes do it eg discuss with people the idea with more automation maybe in the future we all will have a basic wage and only a sort of technical elite will have a job maintaining the automation and of course a political elite doing whatever politicians do. Regardless of how such deliberations pan out I am in no doubt its chances of being correct are zero.

Thanks
Bill