Sign in to follow this  
Bed

I, Robots

Recommended Posts

There's recently been a lot of news about Automation and its effect on the Economy. 

Just recently, Tesla has produced a convincing Self-Driving capabilities. Demonstration here: Click Here I realize that cars will probably be the first reliable, physical, robotic slaves. Mobile devices might make the calls for you, but cars can deliver things to you, drive you places, catch Pokemon, get paid by strangers to drive them, even drive itself across the country with self-fueling, etc. 

And even more recently, Amazon has a store where you're being analyzed visually and virtually in real time to make purchases. So that the need for cashiers are gone. 

The limiting factor from before for these advancements is the processing power needed to do these things. We have of semi-conductors about 20 atoms in length. Probably half that size in 5 years. I can't help but wonder what kind of future we might get, especially on virtual reality... which will probably be much later in the future due to how much processing is needed. A step above that is true AI. 

I read an article talking about someone named Stephen Hawkings addressing automation (link here: Click here), and how in combination with other issues like global warming, can be the single most critical moment in mankind. Well, I guess I can see how that would turn out. If we can make it pass that time, with the solution being be multi-planetary, we'd have a safeguard for potential destruction of the planet from dangerous scientific research (like black holes or fusion... I mean, discovering the unknown can mean anything) or other stuff like warfare, climate, etc. We'd be pass the point where we can just enjoy life and not worry about work. It'd be mostly smooth sailing from then on. Unless you're afraid of being enslaved by AI. 

Interestingly, Stephen Hawking mentioned that automation had a decent influence in the rise of Trump. Or maybe I'm reading the article wrong and its out of context, or the article is assuming things. I'd like to blame it all on the natural course of things.   

We're far from the point where robots walk around and nobody does anything anymore. Could be another 100 years, so this topic probably is not as relevant, but people are acknowledging that factories have less workers, shopping is more efficient, productivity is high. Jobs will be gone, millions out of work, economy is going to crash. My response to that is... "thats what I said when they first had vending machines..."

For a world where the majority of the lower-middle class is replaced with robots, what would be the right economic model? How will people survive?

The only solution I can think of is just making unemployment benefits live-able, or introduce a base income. Increased government presence, which apparently is not popular. I would say, once everything is too reliant on AI, AI power has to be broken up or money generated from it has to go back to the majority somehow. It would have to be a global initiative to control Automation/AI. 

Edited by Bed

Share this post


Link to post
Share on other sites

Throughout history, there has always been a resistance to improved technology causing automation or improvement. Horse and carriage replaced the need for people to carry goods on their backs. Trains replaced horses (and their riders, those who cared for them, and those who built the carriages). Planes replaced trains. Machine production replaced those that crafted by hand.

Economies are not about employment. They are about production. Most people don't understand this. We could easily achieve full employment by removing any significant method of production or transportation of goods. Why don't we? Because we would become poorer as a result. Keep in mind that even those in poverty today are much wealthier than those in the lower class a hundred years ago.

This new technology is merely repeating what has occurred in the past. Our experience today is no different from what has happened a hundred times throughout history.

 

The "problem" comes from the restructuring period where people who were previously employed must now find new employment. With new technology comes new jobs or the need for jobs tending to the new technology. The jobs that are freed from needing to produce may turn to other pursuits, be they a different production or otherwise (which will ultimately end at research, if our current understanding of advancement is correct).

The question is merely how rapidly the restructuring occurs. If it's slow enough, it will merely be uncomfortable for those affected by the transition. If it happens quickly (or is forced by foolish policies such as a significant rise in the minimum wage), then it becomes painful.

Share this post


Link to post
Share on other sites
17 hours ago, Shattered Rift said:

The "problem" comes from the restructuring period where people who were previously employed must now find new employment. With new technology comes new jobs or the need for jobs tending to the new technology. The jobs that are freed from needing to produce may turn to other pursuits, be they a different production or otherwise (which will ultimately end at research, if our current understanding of advancement is correct).

The question is merely how rapidly the restructuring occurs. If it's slow enough, it will merely be uncomfortable for those affected by the transition. If it happens quickly (or is forced by foolish policies such as a significant rise in the minimum wage), then it becomes painful.

Yes, I take it that you think it might be painful, but not fatal?

At the rate of technology, which appears to exponential at the time, it'll happen more quickly than slowly. How does raising the minimum wage affect automation? I thought it was always about, if the capabilities are out there, and the product is reliable, affordable and more profitable, it would be a no-brainer to use the new technology. For ex. It's risky at best, to invest so much in R&D as a small comp, but R&D is common practice in most manufacturing industries to make the process more efficient, no matter how minuscule the improvement. In the long run, automation beats a $7 minimum wage. Plus you still got china and other countries to build factories. Although minimum wage is a factor, I don't see a strong correlation since automation > any wage in the long term for big business. The issue about minimum wage affecting small business, I think, is not closely related thing... since to get to a level of reliability in automation, you need a lot of money. I've seen small business use tablets and things like that to increase productivity. The tablets and things are made from technology companies, software companies in it with the purpose of automation. start ups try to use the technology to create software for restaurants and stuff, never worked since its not reliable enough. A small business wouldn't have an extensive R&D to offset the rise of minimum wage. What else am I missing?

If we can create an environment where the need for repetitious jobs are eliminated and if we can all just focus on new research, that would be great, I think. If raising the minimum wage does that, should we do it more? But like you said, its how rapidly we can transition. In the case that it will be rapid, given the nature of technology or our current understanding of the rate of improvement, how do we transition? How do we prepare for the case that it will be painful?

I get what you're saying that low-income jobs would just be displaced by other low-income jobs, and it's been happening for over hundreds of years, but you also noted that it would end at research. Do you think there is a plateau where low-income jobs won't be displaced by other low-income jobs, but specifically jobs like research? Would high-income jobs become low-income jobs? What is needed for that transition and restructuring? How can everyone get the education needed for something like researching if low-income jobs are eliminated? Is it, the standard is so low for researching, anyone can do it, because we can hire anyone, anytime, since there'd be almost no competition left. Are we all going to be artists at some point? :P

My understanding from you is that Automation, technology should be forcibly slow to accommodate the slow transition and restructuring capabilities of humans. Are you saying that creating an economy that slows down technological advancement would be the saving grace, and the only path until we get an economy where everyone can easily be a researcher?  

Edited by Bed

Share this post


Link to post
Share on other sites

Yeah, painful but not fatal.

I was coming off of a related discussion on the minimum wage on Facebook, so it bled over into my thoughts here. I'll summarize it by addressing this point...

10 hours ago, Bed said:

How does raising the minimum wage affect automation? I thought it was always about, if the capabilities are out there, and the product is reliable, affordable and more profitable, it would be a no-brainer to use the new technology.

Automation is not strictly "more profitable" than, say, a hand-made approach. It depends entirely on the product and the market. Consider the fast food industry. We have the technology to create a burger-making machine that can consistently crank out ordered burgers. But this machine is still more expensive (between production, maintenance, and life expectancy) than hiring a minimum wage "burger flipping" employee. However, a rise in the minimum wage makes that "burger flipping employee" more expensive, which in turn makes us reevaluate the cost of her replacement by a machine.

There is also, as you pointed out, the general push in technology towards improvement and efficiency. Eventually, that burger-making machine will be more profitable than an employee at the current minimum wage. If the minimum wage is increased, the machine matches that employee in cost-efficiency that much sooner.

R&D is a related but tangential issue. Plenty of businesses (especially small businesses) invest in little to no R&D, but these businesses do reap the rewards when others eventually develop a product. Rather, a rise in the minimum wage causes certain branches of R&D to be more competitive than they previously were, provided that they ultimately develop a viable product.

10 hours ago, Bed said:

If we can create an environment where the need for repetitious jobs are eliminated and if we can all just focus on new research, that would be great, I think. If raising the minimum wage does that, should we do it more?

That's a very long-term game. It also makes the assumption that there will ultimately be insignificant limitations on production. And that will probably require a way to create matter (or instantaneous transportation anywhere in the entirety of space). Like I said, very long-term.

10 hours ago, Bed said:

Do you think there is a plateau where low-income jobs won't be displaced by other low-income jobs, but specifically jobs like research? Would high-income jobs become low-income jobs? What is needed for that transition and restructuring? How can everyone get the education needed for something like researching if low-income jobs are eliminated? Is it, the standard is so low for researching, anyone can do it, because we can hire anyone, anytime, since there'd be almost no competition left. Are we all going to be artists at some point? :P

This... really depends on the situation. Part of the pain of restructuring is that workers develop specialized skill sets that aren't directly transferable. Quite a few middle-aged workers have returned to college and university in recent years to develop a new skill set to allow them to switch industries.

Low-income jobs will never be eliminated... unless the minimum wage was raised absurdly high, I suppose. Say, if it jumped up to $20 an hour in the next ten years, assuming no other significant changes. But inflation is almost always the natural trend in an economy, and it would compensate to a degree. Competition can only be eliminated if most or all industries become oligopolies or monopolies, which is highly unlikely outside of a true socialist or communist economy. Even pure capitalism has a difficult time reaching that point.

10 hours ago, Bed said:

My understanding from you is that Automation, technology should be forcibly slow to accommodate the slow transition and restructuring capabilities of humans. Are you saying that creating an economy that slows down technological advancement would be the saving grace, and the only path until we get an economy where everyone can easily be a researcher?  

You're misunderstanding me. If the market is left to itself, it will naturally progress at a near-optimal pace that is in the interest of the majority of people. However, we do want restrictions in place that encourage competition and discourage monopolies in essential industries. This still leaves the issue of barrier of entry, which is why we haven't seen companies besides Windows and Apple and Linux in the operating system industry. Though that industry has its own interesting issues. Microsoft is very unlikely to ever release another operating system. Google (and others) have created mobile OS's that could potentially see serious longevity as the market trends towards the new portable computer. (Laptops, tablets, and smartphones all have significant issues plaguing them. The market wants a happy medium, but none of the happy mediums quite fill the need.)

Share this post


Link to post
Share on other sites
6 hours ago, Shattered Rift said:

Automation is not strictly "more profitable" than, say, a hand-made approach. It depends entirely on the product and the market. Consider the fast food industry. We have the technology to create a burger-making machine that can consistently crank out ordered burgers. But this machine is still more expensive (between production, maintenance, and life expectancy) than hiring a minimum wage "burger flipping" employee. However, a rise in the minimum wage makes that "burger flipping employee" more expensive, which in turn makes us reevaluate the cost of her replacement by a machine.

 
This reminds me of 3D printing technology that prints meat. Seems like the best it can do so far are cookies and stuff. It's far from efficient though. For burger flipping/preparing, what would be the threshold that people would be replaced with crappy machines? I don't know if something like that is quantified, but for a minimum wage to do that, it could be $20 or $200 per hour.
 
An example that I could think of that is already happening is self-checkout machines. You can say that, maybe the increments of the minimum wage is more sensitive here. A slight increase might remove cashiers, and more self-check out machines will be installed. Already, stores have these machines, as well as vending machines and other automated services. I would speculate that the existence of these machines are not the result of increased minimum wage, they are the result of improved technology & potential for profit. We see cashiers instead of stores entirely filled with self-check out machines because they can break down, customers don't know how to use them, presence, operational stuff. Though, depending on the scale of this, I agree the merit that minimum wage as a driving force of increased automation can be defined, which I have no idea of the numbers.
 
Maybe the natural pace for improvement in technology (due to different demands) is much faster than the demand generated from an increased minimum wage. There isn't much analysis I could find about this subject, so I'm just open to any idea at this point. Or maybe it should be an obvious observation?
 
I've been hearing people blaming Amazon's new store on Seattle's $15 minimum wage too. Amazon is an online store company. They never had cashiers. They are pretty much a software company and have been invested in AI for years to stay competitive. They never really had a physical store. Essentially, they are entering a new market, the physical market you might say, with efficient automation made possible because of the Titan X GPU (maybe). Amazon, Ebay, companies that sell online goods didn't form because of increased minimum wage, it was because the opportunity presented itself through the internet.
 
I also acknowledge that Technology can decrease if the demand is not there, like rocket technology... which is a unique kind of technology. But the Pandora's box of things depend on mostly the computing technology, which the demand is bound to everything these days.
 
6 hours ago, Shattered Rift said:

Competition can only be eliminated if most or all industries become oligopolies or monopolies, which is highly unlikely outside of a true socialist or communist economy. Even pure capitalism has a difficult time reaching that point.

 
I think competition can be eliminated once AI reaches a point that it can develop ideas better than humans. I fear that there will be a single breakthrough of AI that snowballs in enhancements - and it doesn't have to be sentient or anything, it could be just a super computer that develops all possible cases and selects the best ones.
At this point, wouldn't that company that developed the all-purpose AI rise far ahead of any other competitors? Though far in the future, I can't tell. It seems all you need is a lot of fast computers, and software which isn't exactly in low demand right now. intel, nvidia, amd are releasing chips slowly on purpose to get the most profit, but it isn't lagging behind in the real potential I think. It might be even before more simple technology is developed.
 
6 hours ago, Shattered Rift said:

If the market is left to itself, it will naturally progress at a near-optimal pace that is in the interest of the majority of people. However, we do want restrictions in place that encourage competition and discourage monopolies in essential industries.

Hmm, I dont' have enough knowledge to have an opinion on this statement, so I'm just going to leave it at that. :P 

Edited by Bed

Share this post


Link to post
Share on other sites
22 minutes ago, Bed said:

For burger flipping/preparing, what would be the threshold that people would be replaced with crappy machines? I don't know if something like that is quantified, but for a minimum wage to do that, it could be $20 or $200 per hour.

I'm not familiar with the cost of these machines, their maintenance, or their life expectancy, but we do know that minimum wage is largely a flat cost.

Employee Cost = Wages + Taxes + Liability + Hiring + Training

Note that hiring and training are fixed costs (pay once: we'll ignore the possibility of ongoing training). Wages, taxes (which are grouped into wages, but remember that the $10 an hour you earn in Washington State is lower than the full cost to an employer), and liability (which will vary from employee to employee but can be averaged based on the information employers have and will vary by industry).

Machine Cost = Machine + Maintenance

The machine will be purchased outright as a fixed cost, while maintenance requires tech work such as the initial programming (a fixed cost) and maintenance over time (variable but calculated as an average).

Let's just say employees costs $10 an hour. This is low, but we lack sufficient information anyway. For a particular fast food joint, you need 500 hours of labor each week that can be replaced by a machine. That's about 71.4 hours of labor needed each day. The cost is $5,000 a week for this labor or approximately $260,720 a year.

Next, let's assume that a machine costs $10,000 upfront, plus $100 in maintenance each week. That's a cost of $15,200 each year, and let's say the machines crap out after a year and need to be replaced. If we assume the machine is working 84 hours a week (a 12 hour open period), it takes about six machines to cover the same labor, costing a mere $91,200 a year.

Now, obviously these numbers are wrong (and I pulled them out of the air), but it simplifies the general idea of how these calculations occur. The machines clearly cost significantly more than this (either in purchasing or in maintenance or in some other area I've overlooked).

 

The example of self-checkout is an excellent one. These machines are already cheaper than the comparable cost of labor.

37 minutes ago, Bed said:

I would speculate that the existence of these machines are not the result of increased minimum wage, they are the result of improved technology & potential for profit.

It's a consequence of both. Increased minimum wage raises the cost of labor. Improved technology and production methods lower the cost of these machines. As soon as the two costs meet in the middle, restructuring will begin to occur (barring other obstacles such as regulation). But yes, these machines will absolutely increase in frequency as the minimum wage rises. It's already inevitable, and the question is simply how much longer it will take for most (or perhaps all) cashiers to be replaced. (I'm inclined to guess that it will occur in another decade or so when the baby boomers pass on, but that's speculation.)

40 minutes ago, Bed said:

Maybe the natural pace for improvement in technology (due to different demands) is much faster than the demand generated from an increased minimum wage. There isn't much analysis I could find about this subject, so I'm just open to any idea at this point. Or maybe it should be an obvious observation?

This varies by industry, but I would largely agree with it. Some machines (such as computers when they were first developed) are infinitely more efficient than humans. Others are only marginally more efficient. It depends.

43 minutes ago, Bed said:

I've been hearing people blaming Amazon's new store on Seattle's $15 minimum wage too.

I'm not very familiar with the details, but I would be inclined to assume that it has little to do with that. You made a good point about the advent of the Internet. It allowed for competition in the market by removing many of the physical barriers between sellers and consumers.

45 minutes ago, Bed said:

I think competition can be eliminated once AI reaches a point that it can develop ideas better than humans. I fear that there will be a single breakthrough of AI that snowballs in enhancements - and it doesn't have to be sentient or anything, it could be just a super computer that develops all possible cases and selects the best ones.
At this point, wouldn't that company that developed the all-purpose AI rise far ahead of any other competitors?

Hypothetically, that may be true for any market where the AI can find the optimal solution. Many personal service industries will remain in business (though would likely be influenced by such advancements), as they're very difficult to replicate. "Art," for example, is still widely considered to be unique and better suited to human minds.

It depends. A monopoly only remains a monopoly by being able to push competitors out of the market and prevent new competitors from entering.

51 minutes ago, Bed said:

Hmm, I dont' have enough knowledge to have an opinion on this statement, so I'm just going to leave it at that. :P 

It was a pretty factual statement. :P The benefit of capitalism is that, in a competitive market, it creates the best possible economy for the greatest number of people. Goods and services are as cheap as they can be, and the greatest amount of wealth is produced as a result. The problems begins when competition is weakened or monopolies emerge that no longer need to remain competitive and can thus raise their prices... that is, assuming there's a high demand for their particular product.

Share this post


Link to post
Share on other sites
19 hours ago, Shattered Rift said:

I'm not familiar with the cost of these machines, their maintenance, or their life expectancy, but we do know that minimum wage is largely a flat cost.

Employee Cost = Wages + Taxes + Liability + Hiring + Training

Note that hiring and training are fixed costs (pay once: we'll ignore the possibility of ongoing training). Wages, taxes (which are grouped into wages, but remember that the $10 an hour you earn in Washington State is lower than the full cost to an employer), and liability (which will vary from employee to employee but can be averaged based on the information employers have and will vary by industry).

Machine Cost = Machine + Maintenance

The machine will be purchased outright as a fixed cost, while maintenance requires tech work such as the initial programming (a fixed cost) and maintenance over time (variable but calculated as an average).

Let's just say employees costs $10 an hour. This is low, but we lack sufficient information anyway. For a particular fast food joint, you need 500 hours of labor each week that can be replaced by a machine. That's about 71.4 hours of labor needed each day. The cost is $5,000 a week for this labor or approximately $260,720 a year.

Next, let's assume that a machine costs $10,000 upfront, plus $100 in maintenance each week. That's a cost of $15,200 each year, and let's say the machines crap out after a year and need to be replaced. If we assume the machine is working 84 hours a week (a 12 hour open period), it takes about six machines to cover the same labor, costing a mere $91,200 a year.

Now, obviously these numbers are wrong (and I pulled them out of the air), but it simplifies the general idea of how these calculations occur. The machines clearly cost significantly more than this (either in purchasing or in maintenance or in some other area I've overlooked).

Hmm, for all employees to be gone, the cost of the machine would include additional infrastructure costs. 

I guess this isn't just burger flipping, but to replace the entire staff with machines, which should be similar to a vending machine but with additional capabilities like heating.

The typical service provided by a fast food joint is a menu of about 20 or so items. Diners (I think diners are typical) These items are not pre-prepared like vending machines. And humans would still be preparing them too if that was the case, unless u buy a factory to produce pre-made burgers. Unprepared Bulk would have to enter the machine, and the output would have to be a combination of the items within the 20 item menu.

The scale of machines, even if the technology is a little bit better 3d printing from today would be very sophisticated, messy, and take up a lot of space. Either give up your menu and modify it so the machine is capable, or buy a bunch of machines to risk them failing. Also need consider buying spare parts.

I've never worked at a burger joint, but I've seen as much as 5 people work there. You have a person at take-out, a person in the front, 1-2 person in the kitchen, a supervisor. 

If you have machines, you need at least 1 person to clean all the food waste in the machines, clean the diner, monitor the space, traffic control around the machines, resupplying the machine, disposing old meat. This person would also need special operations and maintenance training. Could be a lot of work for 1 person.  

Diner Cashiers/Take-out Cashiers would be replaced with machines, which shouldn't be an absorbent amount of money. Machines on the outside for take-out might need additional maintenance due to it being more vulnerable to weather and/or vandalism. 

I can see cashiers being replaced easily, it's just that the flexibility of talking to your costumer is gone but I think that is insignificant and wont impact the service too much.

The kitchen area however, would have to be transformed into a mini-factory. I've worked in a factory before, and I know what kind of machines are used. They no doubt will require a lot of space and maintenance, and tending to. Especially cooked meat from a cold state ( the fridge would have to change) to a hot state, ready to eat. The heating element messes things up a lot, due to accumulation of food pieces on the burner, and possible fire concerns that might need to be solved. The other option is to just microwave the burger once its prepared. I've seen japanese vending machine do that, pretty neat. It wont be the same burger, but I guess the difference might be unnoticeable. Yum, microwaved raw meat.

Not sure what kind of cost utilities would be for potential food waste, cleaning, power consumption, but I'm guessing that is also not significant.  

I would not do this since it seems more food will be wasted, disasters can occur, machines can break, cleaning will be very difficult, space needed, maintenance, machines will be very expensive due to the sophistication needed to deliver a 20 item menu. This is all assuming that such machines have been developed and tested. 

It would only work if the machines were built in-house from a fast food chain, as costs from independent sellers will be quite high due to it being very specialized, and probably sent from germany or something... of all places...  

Japanese vending machine sounds too good now.

Edited by Bed

Share this post


Link to post
Share on other sites
On 12/5/2016 at 6:10 PM, Bed said:

A step above that is true AI

"AI" has become a nebulous term. We have AI now - this combines things ranging from machine learning to fuzzy logic. Advancing the level of complexity without changing the fundamentals could eventually get you something like Star Trek's "Data" was supposed to be outside of plot holes, able to process and make logic-based decisions from available input, able to act on it's own initiative, and able to come up with original solutions from the available data. Others see AI as being what that Data aspired to become (with the aspiring itself being a plot hole) - a full "person" with emotion, creativity and intelligence, though, which is much like FTL drives. It works in theory, except for a mysterious black box in the center that does something currently impossible. This is the problem with the "technology singularity" being interpreted as an AI uprising. Computers won't spontaneously become true beings when reaching a certain level of complexity, processing power, or getting struck by lightning. They need the as-yet-unknown software, and possibly even new hardware, to go that extra step.

Automation, however, does have the logical extreme where it replaces all labor. Any single task a human can physically do, a machine can be made to do it faster, more efficiently, more consistently, and without breaks. Where machines fall down is flexibility. Assembly lines demonstrate this well - where I can make a wooden toy in 30 minutes with the right tools, 80 different machines synchronized by a computer make it in 10 minutes with 80 others in the assembly line at the same time. We're still about 15 years away from a computer being able to listen to you and respond with anything even close to human accuracy, but there are very few jobs where you actually need to listen to your customers. The customer may not like to hear it, but there are a limited number of options you will ever provide as a business. For most businesses, anything the customer might say to a human is going into a computer interface. Tidy up the interface a bit and turn it around so the customer can use it instead, and you don't need anyone at the register. That's being done now. The only tasks that we can't fully automate humans out of for at least 15 years are: tech support, surgery, about a third of desk jobs, art, politics, and anything where a significant number of people simply won't pay to hear from a machine. Tech support fades as people become familiar enough with increasingly static technology to support each other. Fully automated surgery is just a bit further down the line (20-40 years) unless nanotechnology makes an unexpectedly quick leap and makes surgery obsolete...which still won't begin to be accepted for 15-20 years. Desk jobs vary a bit, but some of the most complex ones are being automated in small chunks. As a programmer, I automate any part of my job I can, and people are being (too) aggressive at it with the "never start with a blinking cursor" movement. It began long ago with templates and has progressed to the point where it's trying to write code for you up to a point - and even that will be obsolete when machine learning advances enough to remove the need. Eventually, we're left with politics, art, and science - or ego, self-actualization and advancement.

It's one prediction I think Star Trek got right. There's no reason to respect money at that point. As soon as you automate just the manual labor industry out of existence, you no longer have sufficient justification to "incentivise" work. You can put in the safety nets and basic standards of living. You can make things free. Things like Kickstarter (that get money because people are interested in a thing) become your model for employment. I could go to work for 5 hours a day because that's my level of interest in it, be fully engaged and productive for those 5 hours, and someone else you couldn't afford to hire in the modern world could be intrigued enough to come in as well. Things like exploring the universe, solving the spread of some new disease, or designing a game that appealed to nostalgia would get intense interest. Things like making a tower with someone's family name emblazoned across it would become laughable. 

The first downside that Star Trek didn't really touch on is that it leaves the wrong people with too much free time. People who want to live the old ways can still do so, and people who actually want to just lie on the beach all day can do so, but there is a certain element of counter-function boredness in humanity. People who become gray-hat hackers are the educated form of this now. They're knowledgeable and skilled enough to cause problems, have the free time to do it, and the limited consequences give them the space to start thinking "because it's there" is justification enough. In a society where no one actually has to work to survive, this could manifest as anything from freeclimbing buildings to disabling critical systems to making a sport of harming other people in ways that the idiots involved can convince themselves aren't "real" harm. There's also more time for faux-activism. One of the reasons Westboro has faded to only sending 2-3 people to even major events is because their source of funding has been curtailed. Take away transportation costs, and we might see their equivalent en mass at 10 funerals a day, with more violence because the people who wish they could be there to "do something about" the faux-activists don't have to be at work. Other fiction (Log Horizon, for example) throws in another category - the people who feel they lack a purpose, but I think the number of people with free time whose skills are interpersonal will solve this. The people who would become psychologists now for a career, plus innumerable others who don't for financial reasons, would be able to address a lot of them, and you'd have something akin to a "job placement" structure naturally be developed to address those who just need something to do. 

The second downside is that there's a lot of potential for stagnation. A lot of the things we do right now to advance are because we hate our jobs, or we have friends who hate theirs, or we think there's profit in making someone else's hated job easier or more fun. When work becomes optional, the incentive to advance pares down to the roadblocks in the way of our aspirations. We would certainly be going to Mars and beyond, and solving those problems, but other areas may fall through the cracks. Or maybe not, it's hard to say when the subject is solving problems that haven't arisen yet.

I don't think, though, that you can dismiss the abolishing of physical labor as being a repeat of the machines the Luddites fought against. With each such technology change, new markets and with them new physical labor positions arose. When you can automate all that with more efficient machines, it might create new markets, but it pre-fills the physical labor positions created with more machines. 

While the technology to do it is looming, the modern Luddites and later the wealthy will slow the implementation. That said, I don't think they'll slow it much. Losing jobs to China resulted in a lot of angry posturing by people dressed in clothes made in China holding signs written with markers made in China on posterboard made in China because those were cheaper. When you replace "made in China" with "made in the USA" without mentioning "by a robot", it will look different because the "job stealing" is close enough for people to take trivial-but-spectacular destructive action, but ultimately it won't be any more effective. Cost will win out over time and we will automate physical labor out of existence. Laws will change quickly after that because "millions of hard working Americans" are scary words to have against you when you're a politician hoping to keep your job. Once you create that situation in the US, other first world countries will follow suit rapidly, and second world will have probably beaten us to it since it fits better with their model anyway. After that, the third world will start benefiting from the surplus, and the friendly third world will have the automation shared with them until they're first world. The unfriendly third world will become briefly violent as the difference becomes more stark, but their unfriendly leadership will quickly run out of resources and become laughable despots like Kim Jong-un with outdated "new" weapons systems and an economy almost completely disconnected from the rest of the world. At some point in the much-further future, political tides of one of the major countries will inevitably shift towards interventionalism and we'll see even those collapse.

I think by 2250, we will be a world without anyone doing physical labor that they don't want to do outside of illegal coercion. But the part I'm glossing over is the intervening years when the newly created job types are filled by robots. It's one thing to replace janitors as they retire with bots, but it's not going to be pretty when it hits retail and fast food where the turnover rate is high.

Share this post


Link to post
Share on other sites
2 hours ago, InuyashaOhki said:

"This is the problem with the "technology singularity" being interpreted as an AI uprising. Computers won't spontaneously become true beings when reaching a certain level of complexity, processing power, or getting struck by lightning. They need the as-yet-unknown software, and possibly even new hardware, to go that extra step.

This kind of thinking reminds me how some people think monkeys spontaneously evolved into humans one day. Magic is still very ingrained into our culture. Sometimes it is fun and I'm all for it, but sometimes it is slowing down progress...

2 hours ago, InuyashaOhki said:

Automation, however, does have the logical extreme where it replaces all labor. Any single task a human can physically do, a machine can be made to do it faster, more efficiently, more consistently, and without breaks.

I think in the case where we create robots in a form similar to us, it might be then that robot slaves will be normal. And then a revolution will take place to give robots, rights. I learned this from watching movies.

Robotic autotomy is hard, especially for physical tasks. It makes sense in factories. Unless you got the dexterity and fidelity of a human arm/hand, it's going to be hard to sew some jeans. I've read that there are technology that can react like muscles. Maybe if that gets to the level of human anatomy or even surpass it, we can get rid of physical jobs. It would be robots that look like us. It may or may not be physically possible depending on if we can develop the necessary materials that can mimic the complexity of a human body. Something like this seems more complex than even True AI.

2 hours ago, InuyashaOhki said:

People who want to live the old ways can still do so, and people who actually want to just lie on the beach all day can do so, but there is a certain element of counter-function boredness in humanity.

These are the things I wonder about, and really interested in what kind of world that it will look like. I'm thinking sports and games culture, especially entertainment will be the THING.

 

Edited by Bed

Share this post


Link to post
Share on other sites

A human shaped robot is only good for appealing to human emotion. The bipedal balance issue is nearly resolved, but it's taken decades and isn't more effective than a much simpler quadraped arrangement. The dexterity of a human is again multitasking. Machines sew jeans now. They die-cut the denim, stiffener, and liner fabric, assemble and sew it inside out, then blow it right side out. Put simply, the human hand is a jack of all trades, master of none. Any single task can be done better by a specialized machine. Think about robotic laproscopy if there's some aspect of sewing you think machines can't do.

Automated warehouses are the space to watch for most labor. If fully automated, Walmart could handle your entire local shopping trip from a kiosk with a sixth or less of the floor space. Nothing remotely shaped like a human will be involved. It will look like an inverted assembly line, with the specialized armature-based tools on a track moving around the warehouse or storeroom. The shelves in the storeroom in Walmart were already electronics-driven in the late 90s (massive hydraulic accordion shelves controlled by an overgrown calculator), and could store a third of what was on the floor in space smaller than the electronics section.

 

There also won't be robot slaves. A Roomba is a robot with the same level of sentience as any that will be build for service. There is no financial benefit to putting feelings into machines, and a society that replaces human labor with machines would have no reason to put it in a labor device either.

Share this post


Link to post
Share on other sites
10 hours ago, InuyashaOhki said:

A human shaped robot is only good for appealing to human emotion.

The human body can be good at physical things. It may be less efficient than a single purpose machine, but there are other things other than efficiency that is desired. It would fill physical jobs that have attached human emotion to it, or nostalgia, mostly assistant type jobs. A robot might look atheistically pleasing, but it doesn't mean it isn't good for physical labor. That is discrimination against pretty robots!

10 hours ago, InuyashaOhki said:

. Put simply, the human hand is a jack of all trades, master of none. Any single task can be done better by a specialized machine. Think about robotic laproscopy if there's some aspect of sewing you think machines can't do.

There are robotic that are better than the human hand to do precise tasks. I don't know much about robotic surgery, is it fully automated or is someone navigating it? Some areas, its better than humans, some not. But to quickly eliminate all human physical jobs, you need a jack of all trade which should be equal to human capabilities or better. Because it is just more desirable to have a machine that does everything. One machine that repeatedly does a task forever is good for high yield production, but a machine that can transition to one job to another would most definitely out-compete humans in every kind of job that exists are going to exist. For sewing, right now you need to stiffen the material to do it because the autotomy can't be done otherwise. I think that's recent, not confident on what it can do. If it can do it and make all sorts of garments, that'd be great. It may mean we don't have to outsource the jobs. This leads me to another thought - How would automation affect outsourcing jobs and China?

10 hours ago, InuyashaOhki said:

There also won't be robot slaves. A Roomba is a robot with the same level of sentience as any that will be build for service. There is no financial benefit to putting feelings into machines, and a society that replaces human labor with machines would have no reason to put it in a labor device either.

I think there's financial benefits. People will buy them. Efficiency is only one thing we desire, we also desire sociality.

Not sure what you mean by labor device. If a society replaced human labor with machines, then yes, the machines will create the machines and humans wouldn't need to put In more machines/devices.

It'd be like a wishing well. I'd like a space ship! Then one appears.

Edited by Bed

Share this post


Link to post
Share on other sites
On 12/11/2016 at 1:45 PM, Bed said:

There are robotic that are better than the human hand to do precise tasks. I don't know much about robotic surgery, is it fully automated or is someone navigating it? Some areas, its better than humans, some not. But to quickly eliminate all human physical jobs, you need a jack of all trade which should be equal to human capabilities or better. Because it is just more desirable to have a machine that does everything. One machine that repeatedly does a task forever is good for high yield production, but a machine that can transition to one job to another would most definitely out-compete humans in every kind of job that exists are going to exist.

With currently approved robotic laparoscopic surgery, the robot matches movements directed by the surgeon who is making decisions based on what he sees. My point in bringing it up, though, is that it's sewing in a far more detail-oriented and tighter space than a human could ever do. It's currently automatically repeating the actions input by the doctor. If you wanted some special stitch done over and over, the same mechanical action would be automated without repeated input.

But since you ask, there are fully automated prototypes. It'll be years before anyone is comfortable enough with them to let it be tried on a human, though, and even then it will be supervised for at least a decade before we see anything resembling an "autodoc" just because we're not all that trusting of computers yet. Robots have flown commercial planes 99.9% of the time for years with far fewer incidents than human pilots, but people are still scared to hear a computer is flying their plane (or driving a car...).

On 12/11/2016 at 1:45 PM, Bed said:

Because it is just more desirable to have a machine that does everything. One machine that repeatedly does a task forever is good for high yield production, but a machine that can transition to one job to another would most definitely out-compete humans in every kind of job that exists are going to exist.

As for "quickly" - first off, nope. None of this is quick. It's going to happen as it becomes profitable to eliminate labor in each instance. Manufacturing is automated when it's cheaper than having someone in Bangladesh do it for a penny a week, yet a lot is being done with automation. The service industry can't outsource overseas, so there is more cost pressure to replace them, especially when talk of major shifts in minimum wage are brought up. Amazon is a prime example (pun intended), with their attempts to automate their warehouses right around the time employees were proving to be more costly than Amazon wanted. 

Second off on "quickly" - also nope. By the time you finish working out the bugs in an individual robot with even a fraction of the dynamic capabilities of a human that's able to fit in a human environment, this shift to labor devices will be long over with. It takes -months- to design, order, and set up a new automated factory that produces a new device now. I have an idea for a novelty item (actually, not hypothetical) that involves fabric, some magnets, layers of foam, and a two-part sturdy handle that the magnets would hold together. If I felt it would be a good investment (it's currently not, or I'd be seeking funding and doing it, but I'm not saying what it is online because it might work in the future), it would be a custom stitching process, a custom foam shaping process, a custom process for making the handles and then assembling the whole thing and sewing it up. If I had an investor, I could have the machines in place in a facility near me within 10 months, with orders filled in 14 months. The level of complexity increases if you're limited in space, but corporations aren't. McDonalds are torn down and rebuilt all the time. Walmart has enough space to build it in the back of every store and then level the front half of the building for more parking. None of them would roll it out nationwide right away, but Walmart's "supercenter" and McDonalds' dual-lane deployments are examples of similar major changes, both of which rolled out right under people's noses. Most people saw one local Walmart replaced, then suddenly realized there were supercenters almost everywhere a few years later. For something as straightforward as fast food, it's so simple that prototyping challenges regularly build fully automated prototype restaurants that function in less than a week. A human can walk from a grill flipping burgers to a wastebin to take out the trash, then to a broom to sweep, but a grilling machine is a controller, 2 metal plates, 4 pistons and 2 mechanical arms off the shelf with some code written to control it, an automated wastebin is available to order already, and I owned an automated broom for a while (Roomba). Unlike creating the multi-function robot, nothing actually has to be invented. It just has to be cost effective enough to write the code and buy the parts.

You also wouldn't transition a machine from one job to another. Even when you have humans, companies since Henry Ford have been single-tasking human laborers for efficiency. They're not doing that as much now, because they're learning that humans are really bad at repetitive tasks, but robots excel at them, making that efficiency one they would naturally gravitate to. 

These specialized robots are still versatile in what they can do, though. You can make one grill your burger to exactly the level of "done" you want, put on the toppings you want, swap out a different kind of meat, add in other foods, etc. A 3D printer can produce everything from a comb to a complex honeycomb lattice to...all the plastic parts you need to make another 3D printer. And a 3D printer is just a set of simple gears, motors, heaters and injectors with a controller. We're very close to being able to 3D print food, human organs and numerous other things with this same simple set of equipment.

On 12/11/2016 at 1:45 PM, Bed said:

For sewing, right now you need to stiffen the material to do it because the autotomy can't be done otherwise.

Stiffening is a material. It's also called "interfacing fabric". It's a stiffer fabric that you use to make shirt collars and waistbands stiff. If you've ever worn a hole in either of those, it's that white material that sort of looks like a very dense mesh. It has absolutely nothing to do with automation. I've got plans for pants I intend to make that are period accurate to the mid 1800s. The process for making them would be -simple- to automate. There is nothing in sewing that is hard for a machine to do.

On 12/11/2016 at 1:45 PM, Bed said:

This leads me to another thought - How would automation affect outsourcing jobs and China?

Well, China is losing it's place as the target of outsourcing just like the US started to 200 years ago. Cost of living, standard of living, and consumer interests are growing there as money flows into China. That drives up the costs of labor, and outsourcing shifts to the next cheapest. Clothing is currently being outsourced more and more to Cambodia than China and even China is starting to outsource labor, despite having the largest labor pool on the planet.

As for how it would affect it, first it would just be a matter of building more automated factories in China. I mentioned earlier that the second world would be quicker adopters than we will be, and China is the industrial giant of the second world right now. As the costs go down from more automation, it would be cheaper to build them in the US and save on shipping of finished products, but China will be past the labor requirement threshold by that point. Eventually, as the costs of more advanced 3D printing come down, you eliminate the warehouse and just store and ship raw materials, with everything made on demand from sushi to a 6 foot tall Furby with custom articulation and circuitry. Over several decades it will eliminate the need for assembly outsourcing and you'll gradually see "made in (country)" disappear off labels as it becomes trivial. But the returning jobs won't go to humans. The human management of the facilities will be the same people who previously flew to Cambodia to oversee sweatshop operations.

On 12/11/2016 at 1:45 PM, Bed said:

Efficiency is only one thing we desire, we also desire sociality.

You wouldn't ask someone to build you a sentient slave. You would ask for a "plastic pal who's fun to be with", to quote Douglas Adams. Even those are likely to just be simulated emotions to give the illusion of empathy with a human without the free will or free thought that makes the difference between complex data processing and a person. A real sentient machine can say "I don't want to be friends with wetware!" which isn't a desirable outcome. More likely, if sentient robots are created, it will be as a directed scientific effort after required labor is ancient history.

On 12/11/2016 at 1:45 PM, Bed said:

Not sure what you mean by labor device.

Any device that performs labor. The welding robots currently making cars, a burger grilling machine, each automated sewing machine in an assembly line, a robotic arm that puts what a customer wants on a tray, or a hex-copter that delivers your Amazon order would all be labor devices. The term distinguishes from laborer, which is a person who performs labor.

 

EDIT: Totally relevant to this discussion.

Edited by InuyashaOhki

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this