Get Ready for Killer Robot Technology.

[barbos]

Multi-purpose robots are not an efficient way to do anything - a purpose built robot will almost always be a better option.

Shit, our entire educational system is designed around taking multi-purpose humans, and getting them specialised.

But in case of humans all this specialization is all in the brain. Robots are not going to have problems with uploading different brains for different specialization. You can probably try to make a specialized and dumb robot which makes t-shirts only. But I doubt it would be any good. t-shirts are much harder than welding car frames. T-shirt robot would have to be much "smarter" and will naturally be capable to do other kind of works.

The guy who performs heart transplants doesn't also mop the operating theatre floor, and for the exact same reasons, the robot that performs heart transplants won't also be used to mop the floor of the operating theatre.

Diverse abilities are only good when there is not enough specialist work to fill a robot's (or a human specialist's) time. If there are only enough cardiac surgery patients to fill half of a surgeon's working hours though, it still makes more sense to have half as many surgeons working, than it does to have the surgeon spend half his time operating, and the other half cleaning.
In a world of 7.5 billion humans, there is enough work to occupy at least one robot 24x7x365 doing pretty much anything - even the most infrequently required tasks start to be needed continuously when you multiply the frequency of need by 7.5 billion.

Containerised shipping is INCREDIBLY cheap, even with a big labour element in the cost - although that has fallen dramatically in recent decades, and big container ships now circumnavigate the world with only a handful of crew.

Automated (self-driving) trucks will cut shipping costs even more dramatically in a few years; the cost of truck drivers is a large fraction of the cost of shipping an item from, say, London to Sydney, despite the vast majority of that trip being by sea - the cost per tonne from Harwich Dock to the docks at Port Botany is somewhat less than the cost of trucking the goods from London to Harwich, and then from Port Botany into Sydney.

As transport becomes more automated, and therefore cheaper, the benefit of having a handful of factories for any given product, manufacturing in bulk goods for shipment worldwide, will only increase; manufacturing locally can't compete, as the economies of scale when making a million units in a 24x7x365 mega factory will always outweigh the transport cost savings for a local factory making a few dozen units a month.

Interestingly, as the automation of production progresses, factories will move away from the cheap labour of the developing world, and back to the high-tech and reliable infrastructure of the first world. But this move won't bring many jobs with it - a tiny number of engineers are needed to maintain the factories, but not a single worker is needed in the manufacturing process.

In such a scenario, it is crazy to expect people to find work; and it is inhuman to give the whole (or even majority) of the proceeds of the manufacture of goods to the owners of the means of production. Most people will be unemployable; and society will be able to afford to keep them in comfort regardless. If society chooses not to do so, civil unrest is the inevitable result. Our entire economic system will need to be re-modelled to handle the idea that huge swathes of humanity cannot work, and should not be expected to.

This should already be happening.

I agree. Shipping is cheap but with full automation there would be no point in shipping anything other than raw materials like iron and stuff.

 

[bilby]

But in case of humans all this specialization is all in the brain. Robots are not going to have problems with uploading different brains for different specialization. You can probably try to make a specialized and dumb robot which makes t-shirts only. But I doubt it would be any good. t-shirts are much harder than welding car frames. T-shirt robot would have to be much "smarter" and will naturally be capable to do other kind of works.

The guy who performs heart transplants doesn't also mop the operating theatre floor, and for the exact same reasons, the robot that performs heart transplants won't also be used to mop the floor of the operating theatre.

Diverse abilities are only good when there is not enough specialist work to fill a robot's (or a human specialist's) time. If there are only enough cardiac surgery patients to fill half of a surgeon's working hours though, it still makes more sense to have half as many surgeons working, than it does to have the surgeon spend half his time operating, and the other half cleaning.
In a world of 7.5 billion humans, there is enough work to occupy at least one robot 24x7x365 doing pretty much anything - even the most infrequently required tasks start to be needed continuously when you multiply the frequency of need by 7.5 billion.

Containerised shipping is INCREDIBLY cheap, even with a big labour element in the cost - although that has fallen dramatically in recent decades, and big container ships now circumnavigate the world with only a handful of crew.

Automated (self-driving) trucks will cut shipping costs even more dramatically in a few years; the cost of truck drivers is a large fraction of the cost of shipping an item from, say, London to Sydney, despite the vast majority of that trip being by sea - the cost per tonne from Harwich Dock to the docks at Port Botany is somewhat less than the cost of trucking the goods from London to Harwich, and then from Port Botany into Sydney.

As transport becomes more automated, and therefore cheaper, the benefit of having a handful of factories for any given product, manufacturing in bulk goods for shipment worldwide, will only increase; manufacturing locally can't compete, as the economies of scale when making a million units in a 24x7x365 mega factory will always outweigh the transport cost savings for a local factory making a few dozen units a month.

Interestingly, as the automation of production progresses, factories will move away from the cheap labour of the developing world, and back to the high-tech and reliable infrastructure of the first world. But this move won't bring many jobs with it - a tiny number of engineers are needed to maintain the factories, but not a single worker is needed in the manufacturing process.

In such a scenario, it is crazy to expect people to find work; and it is inhuman to give the whole (or even majority) of the proceeds of the manufacture of goods to the owners of the means of production. Most people will be unemployable; and society will be able to afford to keep them in comfort regardless. If society chooses not to do so, civil unrest is the inevitable result. Our entire economic system will need to be re-modelled to handle the idea that huge swathes of humanity cannot work, and should not be expected to.

This should already be happening.

I agree. Shipping is cheap but with full automation there would be no point in shipping anything other than raw materials like iron and stuff.

Of course there would. Shipping is cheap; robots are expensive. Expensive capital equipment (like robots) needs to be used as much as possible in order to pay for itself. A $90,000 dollar robot that turns $1 of materials into a $10 T-shirt needs to make 10,000 T-shirts before it starts making a profit. Having one in your home is pointless - you don't need 10,000 T-shirts. Having one in your town or city might make a small profit. Having one producing T-shirts for the entire USA will make a much larger profit; and having one that makes T-Shirts for the whole world will make a humongous profit.

A robot that can make T-shirts only is going to be cheaper than one that can make T-shirts and jeans. Sure, you could even make a robot with general purpose limbs and manipulators that can make any kind of clothing, or even do other tasks, but that just adds to the cost. it is far cheaper (per item produced) to build a specialist robot that makes T-shirts only; and a second specialist robot that makes jeans only, than it is to build a robot that can do both, and that switches between the two jobs - as long as there is enough demand so that the two machines never have an idle moment.

And with cheap transportation, and a world population in the billions, specialisation will ALWAYS beat generalisation.

In WWII, car factories were converted to make tanks, or aircraft. That change was possible, because all three products require similar machinery and skills to make. But each plant switched from cars EITHER to tanks, OR to aircraft; because making both tanks and aircraft in the same factory is inefficient. It's better, cheaper, faster more efficient to make one thing, over and over again, using specialised plant where possible, than it is to make lots of different products using multi-purpose plant.

The ONLY situation where the flexibility to make lots of different products is useful is when demand for your product falls short of supply, to the point where your factory/machine/robot sits idle; which can only happen in situations such as prototyping, where the total number of products required worldwide is very small; or where the cost of transportation to global markets is so high that only local demand is worth filling.

If one person in a thousand wants your product; and even those people only buy one a year, that's still seven million units per annum required worldwide. Why waste money on a multi-purpose robot/factory/machine, if both it and a cheaper single-purpose robot/factory/machine can churn only out 100 units an hour? You need seven or eight cheaper single-purpose machines for that task; using seven or eight more expensive multi-purpose robots for the job is a waste, as they will never need to do anything other than make that one product, to keep up with demand.

In fact, you have it exactly backwards; Shipping inexpensive commodity raw materials is the last thing you want to do, because the proportion of the cost to the buyer that is transportation is a larger proportion of the total. All else being equal, goods should be made close to the supply of raw materials, and then the finished goods shipped to markets worldwide. If you can make 1,000 widgets, with a total weight of 900kg from a tonne of iron (10% goes to scrap, and can be fed back in to the start of the process, or maybe it just gets wasted), it is very obvious that it is 10% cheaper to ship widgets than it is to ship the iron needed to make them. Putting the factory next to the iron works is a better choice than putting the factory next door to the customer.

 

[barbos]

But in case of humans all this specialization is all in the brain. Robots are not going to have problems with uploading different brains for different specialization. You can probably try to make a specialized and dumb robot which makes t-shirts only. But I doubt it would be any good. t-shirts are much harder than welding car frames. T-shirt robot would have to be much "smarter" and will naturally be capable to do other kind of works.

I agree. Shipping is cheap but with full automation there would be no point in shipping anything other than raw materials like iron and stuff.

Of course there would. Shipping is cheap; robots are expensive. Expensive capital equipment (like robots) needs to be used as much as possible in order to pay for itself. A $90,000 dollar robot that turns $1 of materials into a $10 T-shirt needs to make 10,000 T-shirts before it starts making a profit. Having one in your home is pointless - you don't need 10,000 T-shirts. Having one in your town or city might make a small profit. Having one producing T-shirts for the entire USA will make a much larger profit; and having one that makes T-Shirts for the whole world will make a humongous profit.

A robot that can make T-shirts only is going to be cheaper than one that can make T-shirts and jeans. Sure, you could even make a robot with general purpose limbs and manipulators that can make any kind of clothing, or even do other tasks, but that just adds to the cost. it is far cheaper (per item produced) to build a specialist robot that makes T-shirts only; and a second specialist robot that makes jeans only, than it is to build a robot that can do both, and that switches between the two jobs - as long as there is enough demand so that the two machines never have an idle moment.

And with cheap transportation, and a world population in the billions, specialisation will ALWAYS beat generalisation.

In WWII, car factories were converted to make tanks, or aircraft. That change was possible, because all three products require similar machinery and skills to make. But each plant switched from cars EITHER to tanks, OR to aircraft; because making both tanks and aircraft in the same factory is inefficient. It's better, cheaper, faster more efficient to make one thing, over and over again, using specialised plant where possible, than it is to make lots of different products using multi-purpose plant.

The ONLY situation where the flexibility to make lots of different products is useful is when demand for your product falls short of supply, to the point where your factory/machine/robot sits idle; which can only happen in situations such as prototyping, where the total number of products required worldwide is very small; or where the cost of transportation to global markets is so high that only local demand is worth filling.

If one person in a thousand wants your product; and even those people only buy one a year, that's still seven million units per annum required worldwide. Why waste money on a multi-purpose robot/factory/machine, if both it and a cheaper single-purpose robot/factory/machine can churn only out 100 units an hour? You need seven or eight cheaper single-purpose machines for that task; using seven or eight more expensive multi-purpose robots for the job is a waste, as they will never need to do anything other than make that one product, to keep up with demand.

In fact, you have it exactly backwards; Shipping inexpensive commodity raw materials is the last thing you want to do, because the proportion of the cost to the buyer that is transportation is a larger proportion of the total. All else being equal, goods should be made close to the supply of raw materials, and then the finished goods shipped to markets worldwide. If you can make 1,000 widgets, with a total weight of 900kg from a tonne of iron (10% goes to scrap, and can be fed back in to the start of the process, or maybe it just gets wasted), it is very obvious that it is 10% cheaper to ship widgets than it is to ship the iron needed to make them. Putting the factory next to the iron works is a better choice than putting the factory next door to the customer.

Usable t-shirt robot would have to be smart enough to make any kind of clothing. And who says robot have to cost $90k?
Slower robot does not have to be expensive. Slow 3D printers are very cheap nowdays. And slow t-shirt robot is not that different from 3D printer, as far as mechanics concerned. Problem is with software, it has to be smart enough to deal with fabrics and such.

As for the shipping, if iron was used only for nails then making nails in place and then shipping would make sense. But the problem is, iron is used for a lot of stuff and that stuff used a lot of other raw materials, You can't have a world where only final products are being shipped (long distance)
In any case, I believe mindless shipping as in the case of US made cotton will reduce. Bad for Bangladesh but that's inevitable.

 

[Crazy Eddie]

You can probably try to make a specialized and dumb robot which makes t-shirts only.

They already have those.

t-shirts are much harder than welding car frames.

LOL no they're not.

T-shirt robot would have to be much "smarter" and will naturally be capable to do other kind of works.

No, it would not. The T-shirt robot would be equipped with highly specialized tools and equipment that are optimized for the task of constructing and assembling garments and printing designs on fabric. Those same tools that make it a superior T-shirt robot are inferior for just about any other task imaginable; you don't use a needle and thread to weld a car frame, you don't need ultra-fine high precision motor control to mop floors, you don't need high temperature thermal printing equipment to change a light bulb.

The generalist tools employed by humans -- e.g. hands and fingers -- are inferior to the specialized tools we invented to make those jobs easier. The t-shirt robot is, if anything, a MORE specialized tool that is superior to a human in every respect.

I agree. Shipping is cheap but with full automation there would be no point in shipping anything other than raw materials like iron and stuff.

With full automation you would ship raw materials to the place where large-scale automation is being implemented. In other words, the factory that has ten thousand t-shirt making robots can produce more volume and fill more orders than the place that only has two. Throw in economy of scale and the fact that even automated shipping is still cheaper in bulk, local manufacturing makes no economic sense.

Usable t-shirt robot would have to be smart enough to make any kind of clothing.

Real textile machines do not have this feature.

Problem is with software, it has to be smart enough to deal with fabrics and such.

It's not a matter of "smarts", it's a matter of tooling. The ability of a machine to manipulate its workspace and accomplish a task is a function of the tolerances of its parts, range of motion of its components, operating time and precision. Software is actually the EASY part of that equation, as there is a very huge industry devoted to designing very clever algorithms that allow these machines to more efficiently manufacture products.

But even a dumb CNC machine can cut metal more precisely and more reliably than a very smart human, which is actually the whole reason people use CNC machines. A very well-programmed CNC machine will do an even better job than a very smart android, to the point that it makes no sense to program the android to do the job at all.

 

[barbos]

They already have those.

t-shirts are much harder than welding car frames.

LOL no they're not.

T-shirt robot would have to be much "smarter" and will naturally be capable to do other kind of works.

No, it would not. The T-shirt robot would be equipped with highly specialized tools and equipment that are optimized for the task of constructing and assembling garments and printing designs on fabric. Those same tools that make it a superior T-shirt robot are inferior for just about any other task imaginable; you don't use a needle and thread to weld a car frame, you don't need ultra-fine high precision motor control to mop floors, you don't need high temperature thermal printing equipment to change a light bulb.

The generalist tools employed by humans -- e.g. hands and fingers -- are inferior to the specialized tools we invented to make those jobs easier. The t-shirt robot is, if anything, a MORE specialized tool that is superior to a human in every respect.

I agree. Shipping is cheap but with full automation there would be no point in shipping anything other than raw materials like iron and stuff.

With full automation you would ship raw materials to the place where large-scale automation is being implemented. In other words, the factory that has ten thousand t-shirt making robots can produce more volume and fill more orders than the place that only has two. Throw in economy of scale and the fact that even automated shipping is still cheaper in bulk, local manufacturing makes no economic sense.

Usable t-shirt robot would have to be smart enough to make any kind of clothing.

Real textile machines do not have this feature.

Problem is with software, it has to be smart enough to deal with fabrics and such.

It's not a matter of "smarts", it's a matter of tooling. The ability of a machine to manipulate its workspace and accomplish a task is a function of the tolerances of its parts, range of motion of its components, operating time and precision. Software is actually the EASY part of that equation, as there is a very huge industry devoted to designing very clever algorithms that allow these machines to more efficiently manufacture products.

But even a dumb CNC machine can cut metal more precisely and more reliably than a very smart human, which is actually the whole reason people use CNC machines. A very well-programmed CNC machine will do an even better job than a very smart android, to the point that it makes no sense to program the android to do the job at all.

I presume you found t-shirt machine? If so you proved my point, because nobody uses it for the reason I stated.

 

[Crazy Eddie]

I presume you found t-shirt machine? If so you proved my point, because nobody uses it for the reason I stated.

You're exactly right, nobody uses textile processing machines to weld sheet metal. Because that would be idiotic.

I'm not sure how that proves your point, exactly, since it is yet another example of you making a prediction that turns out to be a) several years out of date and b) totally wrong in any case.

 

[barbos]

I presume you found t-shirt machine? If so you proved my point, because nobody uses it for the reason I stated.

You're exactly right, nobody uses textile processing machines to weld sheet metal. Because that would be idiotic.

I'm not sure how that proves your point, exactly, since it is yet another example of you making a prediction that turns out to be a) several years out of date and b) totally wrong in any case.

100% t-shirts are not made by robots, they are made by humans in 3rd world countries.

 

[Crazy Eddie]

You're exactly right, nobody uses textile processing machines to weld sheet metal. Because that would be idiotic.

I'm not sure how that proves your point, exactly, since it is yet another example of you making a prediction that turns out to be a) several years out of date and b) totally wrong in any case.

100% t-shirts are not made by robots, they are made by humans in 3rd world countries.

Well, I work next door to a T-shirt company that has machines that cuts and sews the cloth to order based on computerized templates, graphic design software and some very clever automated systems. So there's that.

OTOH, China and South Korea -- where the majority of cheap textiles are manufactured -- are not actually "third world countries" by what has become the conventional definition of the term. A simple google search also turns up the names of six different companies that manufacture t-shirts domestically in the United States without depending on seamstresses or printing equipment overseas.

Basically:

100% t-shirts are not made by robots, they are made by humans in 3rd world countries.

^ EVERY WORD OF THIS IS WRONG

 

[barbos]

100% t-shirts are not made by robots, they are made by humans in 3rd world countries.

Well, I work next door to a T-shirt company that has machines that cuts and sews the cloth to order based on computerized templates, graphic design software and some very clever automated systems. So there's that.

OTOH, China and South Korea -- where the majority of cheap textiles are manufactured -- are not actually "third world countries" by what has become the conventional definition of the term. A simple google search also turns up the names of six different companies that manufacture t-shirts domestically in the United States without depending on seamstresses or printing equipment overseas.

Basically:

100% t-shirts are not made by robots, they are made by humans in 3rd world countries.

^ EVERY WORD OF THIS IS WRONG

Go to walmart and find a $5 t-shirt which was not made in a 3rd world country.
And your machine which google search gave you, we don't know whether or not it actually works as you think it works. Youtube videos of other t-shirt "machines" are utterly unimpressive, they all look like stuff from the 80s.

There are no robots which can completely replace/exclude humans from that industry. That's why Bangladesh exists.

 

[Crazy Eddie]

Go to walmart and find a $5 t-shirt which was not made in a 3rd world country.

Tags on the clothes I'm wearing now:
Shirt: "Made in Taiwan."
Sweater: "Made in Korea."
Jacket: "Made in Vietnam"
Hat: "Made in China

Which of the above are "third world countries?"

There are no robots which can completely replace/exclude humans from that industry. That's why Bangladesh exists.

Interestingly enough, I can't find a single tag on anything I own that says "Made in Bangladesh." Which sort of makes sense because I'm pretty sure "making T-shirts for walmart" is NOT the reason why "Bangladesh exists."

You are a moron.

 

[barbos]

Tags on the clothes I'm wearing now:
Shirt: "Made in Taiwan."
Sweater: "Made in Korea."
Jacket: "Made in Vietnam"
Hat: "Made in China

Which of the above are "third world countries?"

Technically all of them. And I asked specifically about $5 t-shirt.

There are no robots which can completely replace/exclude humans from that industry. That's why Bangladesh exists.

Interestingly enough, I can't find a single tag on anything I own that says "Made in Bangladesh." Which sort of makes sense because I'm pretty sure "making T-shirts for walmart" is NOT the reason why "Bangladesh exists."

What do you think they do in Bangladesh, economics wise that is?

You are a moron.

I don't think it's me.

 

[bilby]

Technically all of them. And I asked specifically about $5 t-shirt.

There are no robots which can completely replace/exclude humans from that industry. That's why Bangladesh exists.

Interestingly enough, I can't find a single tag on anything I own that says "Made in Bangladesh." Which sort of makes sense because I'm pretty sure "making T-shirts for walmart" is NOT the reason why "Bangladesh exists."

What do you think they do in Bangladesh, economics wise that is?

Most Bangladeshis work in agriculture; primarily growing rice and jute. Most of the rice is consumed domestically; Jute is the export upon which modern Bangladesh was founded.

Sadly, modern synthetics have led to a dramatic downturn in world jute demand; It was a strategic material in the age of sail, and was prized by the East India Company, who had a huge market selling cordage to the Royal Navy (as well as for use in their own vessels) but now it is of little value. Textiles (including the odd T-shirt) and shipbuilding are replacing jute as the main drivers of the economy; but most Bangladeshis still grow crops for a living.

 

[barbos]

Technically all of them. And I asked specifically about $5 t-shirt.

There are no robots which can completely replace/exclude humans from that industry. That's why Bangladesh exists.

Interestingly enough, I can't find a single tag on anything I own that says "Made in Bangladesh." Which sort of makes sense because I'm pretty sure "making T-shirts for walmart" is NOT the reason why "Bangladesh exists."

What do you think they do in Bangladesh, economics wise that is?

Most Bangladeshis work in agriculture; primarily growing rice and jute. Most of the rice is consumed domestically; Jute is the export upon which modern Bangladesh was founded.

Sadly, modern synthetics have led to a dramatic downturn in world jute demand; It was a strategic material in the age of sail, and was prized by the East India Company, who had a huge market selling cordage to the Royal Navy (as well as for use in their own vessels) but now it is of little value. Textiles (including the odd T-shirt) and shipbuilding are replacing jute as the main drivers of the economy; but most Bangladeshis still grow crops for a living.

It's all true, but as far as first world concerned, Bangladesh is just a sweat shop for walmart&co, the only source of hard cash for them is making clothes. They can't export agriculture because they barely have enough for themselves.

 

[bilby]

Technically all of them. And I asked specifically about $5 t-shirt.

There are no robots which can completely replace/exclude humans from that industry. That's why Bangladesh exists.

Interestingly enough, I can't find a single tag on anything I own that says "Made in Bangladesh." Which sort of makes sense because I'm pretty sure "making T-shirts for walmart" is NOT the reason why "Bangladesh exists."

What do you think they do in Bangladesh, economics wise that is?

Most Bangladeshis work in agriculture; primarily growing rice and jute. Most of the rice is consumed domestically; Jute is the export upon which modern Bangladesh was founded.

Sadly, modern synthetics have led to a dramatic downturn in world jute demand; It was a strategic material in the age of sail, and was prized by the East India Company, who had a huge market selling cordage to the Royal Navy (as well as for use in their own vessels) but now it is of little value. Textiles (including the odd T-shirt) and shipbuilding are replacing jute as the main drivers of the economy; but most Bangladeshis still grow crops for a living.

It's all true, but as far as first world concerned, Bangladesh is just a sweat shop for walmart&co, the only source of hard cash for them is making clothes. They can't export agriculture because they barely have enough for themselves.

They grow more than half of the world supply of jute. Most of it is exported.

The Ganges delta (Bangladesh plus the Indian province of West Bengal) between them make 85% of the world supply; most of the rest is grown in China.

 

[barbos]

They grow more than half of the world supply of jute. Most of it is exported.

The Ganges delta (Bangladesh plus the Indian province of West Bengal) between them make 85% of the world supply; most of the rest is grown in China.

OK, I guess they will survive robots, that's good.
But general trend would be reduction of sweat shop manufacturing.