This seems to be a transition stage for fine motor skill robots with advanced object recognition capabilities. A few more iterations will surely make the human counterparts obsolete, as demonstrated by the rapid automation of manufacturing giants such as FOXCONN.
The most exciting part is that once a few large companies like FOXCONN make a commitment to robotic manufacturing, we may be able to see the exponential decreases in robot costs we have been waiting for as they want cheaper sensors and actuators in larger quantities. Robotics have been (and are) incredibly expensive, but but we are reaching a point where the price will begin to plummet.
This doesn't excite me, it scares me. A rapid transition to robotic manufacturing could mean a very rough time for an entire generation of human workers.
> And who will supposedly buy all these good those factories make? The 1%ers?
Yeah, pretty much. Human wants are unlimited. (To quote Boethius, "If free-handed Plenty should dispense riches from her cornucopia as plentiful as the sands cast up by the storm-wracked sea, or as the stars that shine in heaven - men still would not stop their miserable complaints.")
Have you read Chase's 'Plutonomy' papers from a few years ago? Fearsome reading.
Over time it becomes easier to give more value to more people thanks to progress in technology and greater amounts of money/economic wealth sloshing around.
The people who do something of value, or capture it, derive a huge surplus from the leverage provided by what's out there right now ( billions of people, trillions of dollars )
You can see the difference between Nikola Tesla - an absolute genius - who created the basis for all modern technology/progress and Google.
Even with all the value he created, he couldn't really derive that much personal benefit, as the utility of his inventions were low before network effects set in and there were fewer people with less disposable income to spend money on his inventions. He also got screwed but that's another story.
Compared to Google ( very smart guys yes, geniuses perhaps ) which started less than 14 years ago and has become one of the largest and most profitable companies in the world.
Lady Gaga ( 4 years ) vs. Madonna ( 40 years ) for huge success.
Avatar ( 3 billion ) vs. Jaws ( .5 billion )
You catch my drift. Feedback loops are getting larger and faster - it's like nothing stands still anymore.
> Those factories only exists if they have a market for the goods they produce.
You can worry about the computers taking over, or worry about humanity collapsing under the weight of resource crises. I know which I'd prefer. Someone's going to have to build all those wind farms, and retrieve scarce resources from landfill, and I'd rather it wasn't me.
Maybe if the products being manufactured by robots are only expected to be sold to the independently wealthy. If everyone else is unemployed, they won't be buying many products.
If you're wondering what society might look like in that scenario, Roman society during slavery (the robots would be the slaves) might be a relevant model to help understand the effects on the various economic classes.
Unfortunately I don't know any AI researchers who have done work in the past 5 decades. Would you care to share what would happen?
In all honesty I have no idea. Would they laugh and say that "no way AI needs another 20 years"? Or would they say it's only a few years until everyone working with their hands is out of work? Kudos on the pretentious reply btw.
Where AI has consistently failed is creating a real thinking machine that can reason and adapt. That's not what we're building today though.
The problem is to many peoples jobs are to easy and highly specialized. They don't require strong AI - just good enough "dumb" algorithms, and we're getting really good at those. Doing paperwork by strict specifications. Driving a truck from point A to B. Assemble a circuitboard from a blueprint. As long as you keep a handful of humans in the loop incase something unexpected happens, this is all stuff robots can do today.
Yeah, but here's the thing: it will always be cheaper to design products to be manufactured by special-purpose robots than to design them to be manufactured by humans and use "human-capable" robots to do the work instead.
A perfect example is modern electronics. A large part of the reason for the rise of surface mount electronic components (SMT parts are soldered to the same side of the circuit board that they are placed on) is that they make it much cheaper to assemble electronics by automated pick & place machines than it would have been to keep the older through-hole components (with pins/wires that go through the board and are soldered on the opposite side) and use human-like robots to assemble them.
Likewise, airplanes have autopilots designed into them for that purpose instead of having robotic pilots sitting in seats.
The reality is that in most large volume manufacturing, special purpose automation will win out over general purpose robots.
Maybe not, but it could replace low-wage jobs in developed countries.
Imagine a 24-hour fast-food restaurant that employs several people at minimum wage. A robot could work all 3 shifts around the clock nonstop, not require much space, and be "trained" in new types of food preparation with an instantaneous software upgrade.
My economics professor once said he'd heard that if minimum wage got up to $20/hour, it would be cheaper for McD's to use robots. This is changing the other side of the equation.
Minimum wage in Australia is $15.51/hour + 9% superannuation contribution. This comes out to $16.91/hour or USD17.23/hour. If these things can work 24/7 with the same productivity of a worker on minimum wage they'll pay themselves off in eight months (plus/minus other costs associated with employees and these machines).
The price will naturally come down as they go into mass production. Also keep in mind robots will happily work 24 hour shifts in addition to being more effective per hour.
One decade, yes, but decades? I bet that, if you ordered 10,000 of these, the price would fall significantly; let's say to $25,000.
Now suppose these things manage to run at a 80% duty cycle. That way, they run for around 130 hours a week, replacing 2.5 workers. Those workers make, say, about 2.5 x $1000 = $2500 a year, for a gap of 10:1.
I do not think it will take a decade to close that gap.
Firstly, you can design your factory for robots; get rid of most toilets, lunch facilities, walkways, etc. Also get rid of 90% of your huge HR staff (which is huge due to high personnel turnover). Pay the high-tech workers who do robot maintenance from the salaries of the toilet cleaners, restaurant personnel, etc, you got rid of.
Also get rid of the non-billable hours you have now that you spend training personnel on producing this month's product.
I think there will still be a large margin, especially since these things eat power (1.5kW per robot adds up) and since these robots may work slower than the humans they replace.
Decades, however, is a long, long time. A single decade is a factor of 64 in Moore's law, so power usage will decrease to about 80W (for the robot) + 1500W/64 (for control hardware) ~= 100W in a decade. Alternatively, some of the extra CPU power, could, conceivably, be used to speed up these robots.
is this something we should worry about? If we, as a civilisation (forget countries for a second) hypothetically replace all labourers by robots, then what? Is it any different from using cheap overseas labour? Will we be able to find something else to do for these people? Should we turn to communism in the long term, if the only 'working' people are software&robotics engineers? We obviously can't all be tech workers, can we? And would that even be useful?
There will always be unknown things and things to achieve! There is still nuclear fusion, computer brain interface, genetic engineering, inhabiting other planets. Figuring out how universe works, if there are other realities, etc.
My personal thoughts are that communism is merely the next step after capitalism (after discovering fusion and fabricating nanotechnology), however I don't think/hope it will play out as it had before in Russia.
Correction: automation possess a threat to those who do not work principally with their mind. Knowledge workers won't be in danger of marginalization in the near future
34 comments
[ 3.8 ms ] story [ 66.8 ms ] threadThose factories only exists if they have a market for the goods they produce.
Yeah, pretty much. Human wants are unlimited. (To quote Boethius, "If free-handed Plenty should dispense riches from her cornucopia as plentiful as the sands cast up by the storm-wracked sea, or as the stars that shine in heaven - men still would not stop their miserable complaints.")
Have you read Chase's 'Plutonomy' papers from a few years ago? Fearsome reading.
http://www.principiadiscordia.com/forum/index.php?topic=2512...
It's not that fearsome.
Over time it becomes easier to give more value to more people thanks to progress in technology and greater amounts of money/economic wealth sloshing around.
The people who do something of value, or capture it, derive a huge surplus from the leverage provided by what's out there right now ( billions of people, trillions of dollars )
You can see the difference between Nikola Tesla - an absolute genius - who created the basis for all modern technology/progress and Google.
Even with all the value he created, he couldn't really derive that much personal benefit, as the utility of his inventions were low before network effects set in and there were fewer people with less disposable income to spend money on his inventions. He also got screwed but that's another story.
Compared to Google ( very smart guys yes, geniuses perhaps ) which started less than 14 years ago and has become one of the largest and most profitable companies in the world.
Lady Gaga ( 4 years ) vs. Madonna ( 40 years ) for huge success.
Avatar ( 3 billion ) vs. Jaws ( .5 billion )
You catch my drift. Feedback loops are getting larger and faster - it's like nothing stands still anymore.
But society is getting better.
You can worry about the computers taking over, or worry about humanity collapsing under the weight of resource crises. I know which I'd prefer. Someone's going to have to build all those wind farms, and retrieve scarce resources from landfill, and I'd rather it wasn't me.
http://www.amazon.com/Race-Against-Machine-Accelerating-eboo...
http://www.thelightsinthetunnel.com/
Say this to any AI researcher who has done work in the past 5 decades. See what happens.
In all honesty I have no idea. Would they laugh and say that "no way AI needs another 20 years"? Or would they say it's only a few years until everyone working with their hands is out of work? Kudos on the pretentious reply btw.
However, there is some evidence that it really is different this time. Big data makes a big difference to things like image recognition.
[1] http://en.wikipedia.org/wiki/AI_winter
The problem is to many peoples jobs are to easy and highly specialized. They don't require strong AI - just good enough "dumb" algorithms, and we're getting really good at those. Doing paperwork by strict specifications. Driving a truck from point A to B. Assemble a circuitboard from a blueprint. As long as you keep a handful of humans in the loop incase something unexpected happens, this is all stuff robots can do today.
On the other hand , threatening to replace workers with robots in mass , seem like a powerful negotiation tactic.
I'll believe it when i see it.
A perfect example is modern electronics. A large part of the reason for the rise of surface mount electronic components (SMT parts are soldered to the same side of the circuit board that they are placed on) is that they make it much cheaper to assemble electronics by automated pick & place machines than it would have been to keep the older through-hole components (with pins/wires that go through the board and are soldered on the opposite side) and use human-like robots to assemble them.
Likewise, airplanes have autopilots designed into them for that purpose instead of having robotic pilots sitting in seats.
The reality is that in most large volume manufacturing, special purpose automation will win out over general purpose robots.
Imagine a 24-hour fast-food restaurant that employs several people at minimum wage. A robot could work all 3 shifts around the clock nonstop, not require much space, and be "trained" in new types of food preparation with an instantaneous software upgrade.
My economics professor once said he'd heard that if minimum wage got up to $20/hour, it would be cheaper for McD's to use robots. This is changing the other side of the equation.
Wages in China and India are growing. There are places with lower wages, but the infrastructure for major industry isn't always there.
Now suppose these things manage to run at a 80% duty cycle. That way, they run for around 130 hours a week, replacing 2.5 workers. Those workers make, say, about 2.5 x $1000 = $2500 a year, for a gap of 10:1.
I do not think it will take a decade to close that gap. Firstly, you can design your factory for robots; get rid of most toilets, lunch facilities, walkways, etc. Also get rid of 90% of your huge HR staff (which is huge due to high personnel turnover). Pay the high-tech workers who do robot maintenance from the salaries of the toilet cleaners, restaurant personnel, etc, you got rid of.
Also get rid of the non-billable hours you have now that you spend training personnel on producing this month's product.
I think there will still be a large margin, especially since these things eat power (1.5kW per robot adds up) and since these robots may work slower than the humans they replace.
Decades, however, is a long, long time. A single decade is a factor of 64 in Moore's law, so power usage will decrease to about 80W (for the robot) + 1500W/64 (for control hardware) ~= 100W in a decade. Alternatively, some of the extra CPU power, could, conceivably, be used to speed up these robots.
My personal thoughts are that communism is merely the next step after capitalism (after discovering fusion and fabricating nanotechnology), however I don't think/hope it will play out as it had before in Russia.
Correction: Automation poses a real threat to everyone
http://news.nationalgeographic.com/news/2009/04/090402-robot... http://news.bbc.co.uk/2/hi/health/131697.stm
If we didn't have all the automation we have today, our lives would suck.
The only ones threatedned by automation are those who are unwilling to get a different job.