When the mechanized revolution hits, how will humans deal with it?

That fear has been around for centuries.

Look at how educational requirements have increased over time, and you’ll see that the real problem is that education is treated as a luxury here, which will cause the US to implode while other nations that are less into tax cuts and for-profit education systems will leave us behind.

The goal of technology is a post-scarcity economy, though those who profit from scarcity will seek to impede technology as much as they can. The world will lose one major superpower, with all the ripple effects that implies/entails.

Apparently most consumers don’t seem to care. Very few consider where or how the object is made. 12 year old Chinese slaves? No problem just keep the price below $9.98 at Wally World. Robots? What’s that? Just make it cheap.
The population will need to decrease to keep pace with the loss of jobs. It had better or we will have a lot of young Jihadist, Crusader, Missionaries, etc. roaming the Earth looking for something to do.

Sounds like a good work opportunity for robot programmers, robot maintenance technicians, robot refurbishment facilities, robotic efficiency consultants, etc.

I would mourn the loss of the job of the half-drunken oil change tech who forgets to replace the air filter half the time about as much as I mourn the loss of the job of seven year olds having to run behind the first automatic loom to clear the cotton blockage (losing fingers or whole limbs in the process).

It also bears mentioning that as more cheap products are available and more things are being automated, a market for handmade and artisan crafts is starting to flourish.

Yes, you can buy plastic bowls in a three-pack for a dollar, but I have friends who currently make their living throwing custom pottery for eager clients.

When random mundania can be automated, it frees time up for people to learn what they WANT to do, instead of what they HAVE to do.

I’m more than happy to sit at a loom all day. I know many other women who would find it torturous. Good thing we don’t all have to weave our own clothing, huh?

It’s going to be the most liberating accomplishment in human history so long as we manage our resources well. When people are freed from daily drudgery just think what will happen with the arts, music, scientific study…Game changer does not even begin to describe this. People will wonder why the even worried about their shitty factory job, especially once we take the mechanization off world.

Humanity is going to have too many other problems to deal with, like a rapidly falling average life expectancy, and rapidly rising death rate from antibiotic resistant superbugs, There was a time when the life expectancy of a man was 45. How would we deal with that now? Mechanization? Something will have to do the required work in the face of all the death and disease!

How cool will that be‽ I’ll be able to get a robot to shop for me.

@Seek Why do you think I learned to be a CNC Machinist? At the very least, computers are incapable of “judgement calls”. Most can’t even see what they are doing and thus need human supervision. The machines I usually run aren’t smart enough to know that a the cutter grabbed the part and ripped it out of the fixture. Hell, they don’t even know where the fixture is unless I tell it, and in the interest of preventing personnel casualties and/or equipment damage in the six-figure range, rest assured that I make sure I tell it correctly.

Of course, there are some people that lack the aptitude required to learn those skills required to tell the machine where to go and what to do when it gets there. Those people will be at a disadvantage, just as you can’t get an office job (or answer a modern phone) without at least some computer skills.

In that regard, I don’t see things changing much more than they have since computers became a thing. We can pretty accurately extrapolate how the future will go by looking at our current “tech gap” that separates those who know the machines and those who don’t.

@ARE_you_kidding_me When people are freed from daily drudgery just think what will happen with the arts, music, scientific study…Game changer does not even begin to describe this.
How do these newly liberated people who spend their time painting, weaving, writing poems, making music, buy bread, milk, clothes, etc.? Are they going to have a stable of very wealthy clients who want their wares? If their job is taken away by a machine, unemployment will eventually run dry.

@jerv At the very least, computers are incapable of “judgement calls”. Most can’t even see what they are doing and thus need human supervision. The machines I usually run aren’t smart enough to know that a the cutter grabbed the part and ripped it out of the fixture. Hell, they don’t even know where the fixture is unless I tell it, and in the interest of preventing personnel casualties and/or equipment damage in the six-figure range, rest assured that I make sure I tell it correctly.
Y0ou do not ever see machines being able to make calls like that 80–95% of the time in the future? If humans never believe machines can, then I guess they will never try to program them to do it. Look at what computers do now than when they took up whole rooms of not floors of a building and could basically only count cards are crunch numbers. The garden variety cell phone is a basic microcomputer. Back when phones were attached to cords, if no one thought a phone could be more than that, they never would have tried to make it more, but people did, even then I am sure some thought it science fiction to have a phone tell you where all the sushi restaurants within 8 blocks of your location are, and how they rate by customers. That doesn’t even include banking by phone; the brute beast phones attached to cords I am sure was never envisioned to do any of that. As processors get faster and smaller along with memory, programs could be made for a machine to ”watch” another machine and detect if there is a malfunction, alignment issue and either alert the humans or even make the corrections to the other machine. As processors get leaner and meaner they may be able to process thousands of possible scenarios and solutions to each in less than a second and make the right call based off algorithms wrote into their programming. Machines of today may not be that smart, but you honestly think they will never be that smart?

Computers are smaller and faster, but no smarter. They really cannot do much more than they could back in WWII. Now, some of their peripherals are a bit more advanced; a 1920*1080 screen with 32-bit color is a step up over a single row of LEDs or some punch cards/tape, but really it’s just the same bit-pushing, only massively sped up and with a little more parallelism.

Sure, phones and computers have merged, and we also put radio in our phones to cut the cord. My opinion of that is perfectly summed up in Ecclesiastes 1:9. I mean, how long did chocolate and peanut butter exist separately before we combined the two into a tasty, tasty treat? We had bank records too; usually a ledger at the teller’s desk. Turn that ledger into a database and you have the same thing, only smaller and with better search ability. Hook it to a phone line so it can talk to your radio-telephone with a computer translating your finger-presses on the screen into commands to alter the database in a manner specified by whoever programmed your phone and the bank computers, and you have old ideas combined in new ways.

“As processors get faster and smaller along with memory, programs could be made for a machine to ”watch” another machine and detect if there is a malfunction, alignment issue and either alert the humans or even make the corrections to the other machine.”

I have $een $uch $y$tem$. For instance, the Handtmann SC I used to run would, in accordance with it’s programming, begin certain operations by calling up a probe capable of measuring to 0.00002”, moving a couple inches from where it thinks the part/fixture is, measuring with a few taps, and writing the results of the probe’s position readings into the appropriate work offset. That probe was also pretty damn expensive, but when you consider what the machine itself cost, another $30–50 grand is relatively negligible.

Of course, it could (and occasionally did) fuck up pretty bad if the programmer or the operator screwed up. The procedures we had for “first article” programs made acute paranoia seem like blind trust. So despite the expense and sophistication, it was still rather limited.

On the other hand, the limited capabilities of computers can actually be exploited to a certain extent, at least if the programmer isn’t stupid. It’s pretty simple to a program along the lines of ”IF all is well THEN do your thing ELSE stop and alert human operator”. That SC had a multi-ton part about the size of a van that could travel along the X-axis at speed high enough to turn a person to jelly. Why did that never happen? You could not walk within five feet of the working area without breaking a light beam; if anything broke that beam, it would assume that there was a human in the danger zone, stop the X-axis as quickly as possible (usually with a thunk not unlike hitting a brick wall), and refuse to move until the beam was unblocked and someone manually reset it.

But it had issues that, while they erred on the side of human safety, pretty much precluded the possibility of even running unsupervised for more than a few minutes at a time. It didn’t care whether it was a foot, a dropped blueprint, a moth, or a flying metal chip that broke the beam.

I suppose that it’s theoretically possible to keep adding peripherals and increasingly complex programming to get around that, but there is a point of diminishing returns that I don’t see us ever surpassing, at least not in a general purpose machine. I think that the human mind will remain the most flexible system there is, even if it’s not the fastest or most reliable.

@Hypocrisy_Central , The answer to the question of how people will be able to afford to live is very simple. Even with all the work being done by machines, humans play a vital economic role as consumers. A company can’t make a profit unless it sells its merchandise and it can’t sell its merchandise if nobody can afford it. One way or another, money will be transferred from companies to consumers, most likely as a tax distribution. The companies will no doubt complain, but they will have no choice. Zero sales does not make for a very good bottom line.

That still leaves the question of how people will live their lives once the retirement age effectively becomes 0. @Seek has suggested that people will devote themselves to creating handmade items. I suppose that may come to be. Humans are very adaptable. We did not evolve to live in large cities, so we may be able to adapt to a fully automated economy, but I am not entirely convinced. How do you find meaning in life if everything is handed to you.

There are other interesting issues. Suppose that machines could be made more efficient by having what appear to be human characteristics. Perhaps a machine could come up with better results if it has an aesthetic sense or felt frustration at not achieving its goal. We can argue whether machines could actually have feelings, but how would you respond to a machine that gives the external appearance of having them? Would there be a danger of having them take over?

@jerv I have $een $uch $y$tem$. For instance, the Handtmann SC I used to run would, in accordance with it’s programming, begin certain operations by calling up a probe capable of measuring to 0.00002”, ….] That probe was also pretty damn expensive, but when you consider what the machine itself cost, another $30–50 grand is relatively negligible.
Is that a failure on the machine or the lack of knowledge of the humans? Many things started out much more expensive and did less. I remember when the Apple IIe came out and by today’s standard it would be quite dumb, but it cost a mint. Same with the Bata Max, the thing was as large as luggage and my phone has better camera sensitivity than it had, and it cost a mint compared to the camcorders of today which are better and smaller. That is before you even get into flat screen TVs, the early rear projection things are a joke when matched up to today’s flat screen and for its time, cost more in comparative value. If they (smart machines) became more common, like digital watches and all that, the price will come down. Just as battery life has increased, we have way more power over the simple carbon battery, programs will be smarter, technicians and programmers will find ways to take the existing technology further as they have with other things.

@LostInParadise Even with all the work being done by machines, humans play a vital economic role as consumers. A company can’t make a profit unless it sells its merchandise and it can’t sell its merchandise if nobody can afford it.
Agreed, but if machines displace enough people, where do these now freed workers, new consumers have the money to be consummers?

@Hypocrisy_Central , They will get the money from the government, which in turn gets it from companies. The companies will be willing to fork over tax money to the government to give to consumers, because that is the only way that the companies will be able to sell their goods. The companies either share their profits or go out of business. Sharing the profits would seem to be the preferred route.

Well, @LostInParadise – some people will devote themselves to the arts. Generally, I think, people who are already artistically inclined will find it more profitable to do so as more things become automated. Sure, anyone can buy a mass-produced thing from Ikea, but plenty of people already choose to have the same sort of thing, but more unique, from antique shops and handmade sources. I think this will increase as time goes by.

Other people will, as mentioned above, find other outlets more fulfilling: robot design, etc. There will always have to be people to, at the very least, tell the robot farm equipment what to plant where and to maintain the power sources and clear equipment jams.

I do not see why robots could not tell the farm equipment what to plant, maintain power sources and clear equipment jams.

Uh, because someone would have to tell the robot what to tell the farm equipment.

@Hypocrisy_Central That sensor is the refined, low-cost one. And aside from price dropping tenfold, they haven’t changed for pretty much in the last 30 years or so, largely for the same reason that modern tires are still circular. Sure, there are laser-based CMM probes available for roughly the same cost, but they are an order of magnitude less accurate (±0.001” accuracy doesn’t cut it when working to aerospace tolerances!) and far less robust. While the former will likely change in time, the latter won’t get around the inherent limitations of optics, which precludes it’s use in that particular application; it’s only suitable for use in inspection rooms, and would get wrecked in a heartbeat if you tried to use it on the shop floor by putting it in a tool holder and using it the same way as the contact probe. As for non-laser things, the only sort of electromagnetic waves that could achieve the needed resolution would also pose EMF risks that would render them unsuitable for use outside of lead-lined rooms devoid of humans.

And it still doesn’t get around the software issue. The only reason a CMM can even do what it does is because it’s programmed a certain way. While it can somewhat make up for a precisely placed/oriented part being a few millimeters off-center or rotated a couple of degrees either way, programming in in a manner that could accommodate just slapping a part on the table any old way would wind up being slower than a trained operator, and would still require a human programmer.

Betamax vs digital imagery is a little different as one is analog and one is digital. Modern electronics are easily capable of handling more colors than the human eye can discern, which is why we stopped at 32-bit color (~16.7 million colors) for consumer-grade imaging, but it also loses a bit of detail because digital imagery is always pixelated. Maybe the pixels are tiny, but zooming in on a digital image will result in a tile mosaic whereas zooming in on analog images will result in actual magnification. Why do you think digital cameras consider optical zoom such a strong selling point?

As I said earlier, modern computers can’t really do anything the Apple ][ couldn’t do if properly programmed. The only real improvement is that modern computers can do the same things faster and cheaper. If one really wanted to, they could get an Apple ][ to run Photoshop, but it would take at least a few hours to redraw the screen every time you moved anything by even a single pixel.

@Seek , You tell the robot the rules for what to plant one time as part of a computer program. If it is a potato farm then it plants potatoes. If you want to be more sophisticated to take into consideration things like the prices of various commodities and local weather conditions, that too can be programmed. Computer programs are already running which automatically buy and sell stocks. Determining what to plant should be considerably easier.

@LostInParadise You seem to think it’s far easier than it really is.

When you walk, are you aware of EVERY single thing you do? How does each one of the many hundreds of individual muscles in your body move? How do you synchronize them all? Do you consciously alter those hundreds of commands many times every second in response to stimulus from sources such as your inner ear? Speaking of the ear, don’t forget that your brain is also analyzing every minute pressure change (in other words, sound) around you. And every ray of light that strikes your eyes for that matter.

Odds are that you don’t even notice because you have instincts. Computers totally, utterly, and completely lack those instincts. Things you can do without thinking would require at least a few hundred lines of code. If you want to do anything more than blink or breath, the complexity of the code increases dramatically.

The programs I run on a mill are generally simple as it’s mostly simple cartesian coordinates with a few simple commands like “Spindle on at 3000 RPM clockwise” (written M3 S3000), but aside from the spindle, coolant, and 3–5 axis servos, there isn’t much to them that needs controlling; they are much simpler than even one human finger.

So we return to @Seek‘s point; who programs the robot and the machine?

^ How do you synchronize them all? Do you consciously alter those hundreds of commands many times every second in response to stimulus from sources such as your inner ear? Speaking of the ear, don’t forget that your brain is also analyzing every minute pressure change (in other words, sound) around you. And every ray of light that strikes your eyes for that matter.
Isn’t that similar to many vehicles made now? The ”brain”, the ECU or whatever the particular unit is called by which manufacturer, tells the vehicle how to adjust to load, terrain, speed, etc. All this has taken the fun out of working on a vehicle. I had a vehicle once where the ”brain” went out of whack because I disconnected it from the battery to replace the water pump, I could never get it to run right because if I did this, the brain did that. If I did this to counter the brain doing that, it did something else to offset what I did. It was the same problem the son of a friend had and this was before vehicles really got brainy like today.

Computers totally, utterly, and completely lack those instincts. Things you can do without thinking would require at least a few hundred lines of code.
A few hundred or a few thousand, the fact is, it can be done even if it has to have a command a mile long. The code tells the machine is it this? No? Then do that, even if it has to do that hundreds of times to get to the end, it can. As processors get faster it can get through the myriad of options quicker, and quicker.

So we return to @Seek‘s point; who programs the robot and the machine?
We know for now, humans program the machines, making that are a rhetorical question. One human can program 20 machines that in turn can do the work of 1,500 humans, so while the programmer, and other techs that keep the robots going, if one of those robotic machines can paint a house, once the techs set it up, their job is done until time to take the machine apart and pack it up to set it up at the next building. If the machine is made and program to till the soil, fertilize and then plant whatever, if it has enough variables to draw on such as what to do if it encounters rocks, existing pipes under the ground, trees, etc. you can now plant hundreds of acres and you only need maybe one person to oversee the machine, the rest of the humans are sent to Walmart.

@jerv , We do not yet have robots that can walk on two legs, but we are getting to the point where people will own self-driving cars. We also have 3D printing machines that can create rather intricate designs. There has been a lot of work designing machine learning algorithms that create programs on their own. I think that the singularity, if it ever arrives, is a long ways off, but I do not discount the possibility.

Yes we do

Creative artificial intelligence is a completely different conversation, and still science fiction.

We’re not taking about cylons. We’re talking about industrial tools.

^ We’re not taking about cylons. We’re talking about industrial tools.
Someone builds a machine that someone else assembles at the job site. The parameters and dimensions of the building is programmed into the machine’s brain. The program has very detailed instructions of what to do when it comes to windows, what type of windows, vents, doors, other openings, if there are people or animals in the spray path, etc. One man hits the switch, the machine goes to do its job, the 3–5 other men who would have been doing the painting are not needed and told to go shopping. If, and ONLY IF, the machine runs into something new and not in its programming, it alerts ONE PERSON, the tech or programmer who finds out what it is and makes the adjustment so the machine can complete its job. After it is done, THEN some humans can return to take it apart and cart it off to another job site. The actual painting has replaced 3–5 living souls.

Yes. And the industrial revolution replaced an entire industry of unmarried women over thirty, who spent their days at spinning wheels and manual looms.

Now we have so much clothing available for so little money we had to develop machines just to keep up with the laundry.

What’s your point?

“Isn’t that similar to many vehicles made now?”

Yes and no. Cars are relatively simple. They really only oversee a few dozen details at a time. Yet even that degree of parallel computing is moderately expensive despite having been around for years, and in some cases is still far inferior to what a trained human can do.

As for your anecdote, I think you figured out why I distrust cars from the early-90s. While OBDII (mandated as of 1996) is pretty good about just throwing a code and doing what it has to do to limp you home, and pre-OBD cars (generally up to about the late-80s) are too dumb to care, there was a period where the computers were smart enough to cause problems but too dumb not to.

Also, it sounds like you may have borked the Crank Position Sensor on that one; a fault that a human mechanic could easily check, but the car’s ECU cannot check at all. That would throw both the ignition and injector timings off, causing it to run poorly in a way that the computer would try to correct but lack the adjustment range to adjust properly. Also, pulling the battery generally costs you a bit of power and MPG until it re-learns. Then again, it’s possible that the engine had other issues that the computer managed to correct but forgot, and your attempts to fix it confused it as it attempted to re-learn the proper settings.

Regarding programming, that only goes so far. Reality is full of unforeseen circumstances, and while a default behavior is usually a good failsafe (that Handtmann’s default state was “Turn off all pumps and servos, slam on all the brakes, and lock all the doors and panels”), there are times where even computers can make human-like mistakes. More to the point though, their perfect memory is balanced by a lack of instinct, so even those capable of some learning are limited compared to a human. And those that can learn tend to do so in ways that make little/no sense to humans. (At least neurotypical ones; there is a reason high-functioning autistic people tend to gravitate towards computers, feeding hte stereotypes about computer geeks having poor social skills.)

Most CAD/CAM programs that turn a blueprint of a curved part into G-codes make enough small linear moves that, although the programs were basically raw-text files with no formatting information, they were still often a few megabytes; too big to fit in the storage of older machines, but attempting to “dribble-feed” it from an external computer would… well, cause other problems. Newer machines (like that Handtmann) are more capable, as would be expected from something a decade or three newer, but exploiting that capability to make more intricate parts not only costs a bit more, but also requires an intermediate step (CAD/CAM software) to turn a CAD model into G-code (that’s what the CAM part does). And that program still needs double-checking and tweaking by humans before it even goes out to the floor to be tested. And that testing is only entrusted to the most skilled operators, usually with supervision from the engineer who “wrote” the program. And even then, 98% of the time, edits needed to be made for reasons no computer could foresee because reality doesn’t give a shit about how things should be.

The funny part there is that while the procedure I outlined above was for a modern machine cutting highly detailed part, running far simpler programs on older machines generally only differed in that simple programs (those not involving any curves other than circles or circular arcs) were far smaller, could be written fairly quickly by hand, and the operator was the one that turned the blueprint into a program. That last point means that anyone capable of writing a program was generally the one to test it and edit as needed, cutting manpower requirements.

In both cases, putting the machine in Single Block mode and keeping one eye on the tool, one on the Distance To Go display, and one on the next line while having one hand on the Feed Rate Override, the other on the Spindle Speed override, and one finger on the Cycle Start button was pretty much standard practice for a new program. Running the first part of a new setup using a proven program was nearly as demanding, and even replacing a worn/broken tool required a skilled operator to verify the correct offset values were entered had a certain procedure. I saw what happened when someone trusted the machine after replacing a cutter once. Long story short, they dumped a couple of million into attempting (unsuccessfully) to repair the damage, and the incident gave the rest of us a cool story about the guy who got fired.

And I’ve typed enough that my fingers hurt :P

@ARE_you_kidding_me That was then, this is now. Of course, they still don’t move fluidly or quickly, and are absurdly easy to knock over. Also note that they have fewer servos than humans have muscles and less articulation than we do; that simplifies matters greatly when it comes to programming them.

@jerv G00gle owns Boston dynamics now, we may soon have to be subservient to our new robot masters

@jerv Yes and no. Cars are relatively simple. They really only oversee a few dozen details at a time. Yet even that degree of parallel computing is moderately expensive despite having been around for years, and in some cases is still far inferior to what a trained human can do.
Sure, I don’t think there will ever be a machine that can perform an appendectomy, at least not for centuries if we last that long. Going on your premise that vehicles are fairly simple and that is why they can adjust power to whatever wheel, close grill vents, or tweak their spoilers, etc. If there was a fairly straight-forward machine to build dog houses, once the human set the machine up, you can have 5–8 or more of them building dog houses instead of humans. Not only that, the machine doesn’t take lunch hours, call in sick, has to heal from injuries, gets bored, and it can work 12 hours and still be as fresh as when it was turned on. If there was two shifts, you only need two humans as opposed to 10–16, of which are pulling a wage, while the machine works for free.

@jerv, I appreciate your detailed answer based on personal experience. I still think it is remarkable that cars can drive themselves. Maybe if I were more knowledgeable about the engineering involved, I would be less impressed.

I found this article, which says that it is projected that nearly half of all current jobs will be replaced by machines in the next few decades. The article distinguishes between men’s and women’s jobs, but I think that it is best to keep gender out of it. What they call women’s work is jobs in the service sector. It makes sense that jobs requiring one on one human interaction are far less likely to be replaced by machines. Is this the future of employment?

I think it is an open question whether the general public will be able to shift to the engineering and programming jobs less susceptible to being replaced by machines. People are much better educated than they were 100 years ago. Is there a limit to the skills that the average person can acquire?

@LostInParadise, I think the opposite: If you were more knowledgeable about the engineering and mechanics involved you would be more impressed. As @jerv suggests throughout his discourse, computers don’t think like we do. (As a matter of fact, so far they do not “think” at all.) Sure, they make far faster calculations, but they don’t got rhythm – no soul, no flow, and no instinct whatsoever. So while a computer’s superfast calculator can very quickly determine that an action needs to be taken to move the vehicle, change its course or speed, or adapt to changing environmental conditions (and traffic) and can quickly send out appropriate commands to a series of servo motors to move one or many actuators in three dimensions, analyze the result and send more commands to continue the movement, stop the movement, move faster or slower, or move in a different direction – or engage whole new systems – that’s all that it does. It never gets a “feel” for driving; I don’t suppose that even when AI becomes a real thing that a computer will ever enjoy a nice drive in the country.

In addition, robot-operated cars have to manage traffic with human drivers, who are not always as attentive to the rules of the road that computers are programmed to obey. So the humans drive too fast (or too slow), cut in front of the robot drivers in the “safe space” that the computer has judged to be necessary but the human driver doesn’t seem to mind (computers are generally instructed through their programming not to tailgate, but humans can – and usually do – choose to override that safety advice), and while a computer-driven car will come to a full and complete stop at a stop sign, and wait for crossing and oncoming traffic to stop as well, the human drivers in the oncoming and crossing vehicles often fail to stop as completely as they did when they took their driving exam, so the robot could conceivably sit at its stop sign all day while it waits for that traffic to stop, as it “knows” is required. But that’s not acceptable, either, so the robot has to be programmed – specifically programmed, that is – to drive with “fuzzy rules” like humans do. That is, the robot controlling the vehicle has to be programmed to “not drive perfectly” – because humans don’t, and the humans (so far) rule the roads. (Except in India, where the cows do.)

Then, machines always fail. That is, they always fail “eventually”. Either the sensor that provides information to the computer becomes clogged with debris (that another part of the machine, built for this purpose, cannot clear), or a wire breaks (or a wireless signal is blocked by some kind of electrical interference), or a bushing or bearing on the mechanical train somewhere gives out, jams or wears past a point of tolerance – “something happens”. So the computer has to be programmed – using specific instructions, because it hasn’t got “sense” to pull over and stop – or to “know” when that’s not a valid option, either – what to do in case of all kinds of breakages, outages, interference and malfunctions. All of the “normal” vicissitudes of “what can happen while I drive a car?” have to be thought of in advance and programmed into some sort of decision matrix that the controller can work through – very quickly – and come to some sort of default action/s to safely stop the vehicle, get it out of a moving traffic flow if possible, and allow the passengers to disembark.

That’s a hell of a lot of engineering. (And in the case of a vehicle moving at speed in populated areas with other vehicles, there has to be various kinds of redundancy built into the system, too, so that if a major control element goes haywire, there’s something in back of that to take over and perform the “get out of traffic and stop the vehicle” function – at the very least.)

And it will all have to be done at a cost that will make the vehicle sell for an attractive price point, make the manufacturer a profit that will encourage him to make and sell more vehicles, and fit into a package that consumers will find stylish and useful. That’s engineering.

@Hypocrisy_Central You would need some highly educated people though. People like me and my geeky friends who not only paid attention to their K-12 education, but had the curiosity and ambition to go beyond that, whether through college, trade school and/or self-teaching. Not everyone has that.

For instance, my trade doesn’t interest enough people to keep up with demand once you adjust for the number of people leaving the trade, often by retiring. It’s not that only MENSA-level intellects can handle it, just that it’s not something that all people can do, and is complex enough that specialized learning is required.

While low-level jobs are outsourced for cost considerations, skilled positions are usually outsourced due to a lack of qualified applicants. Qualification comes through education.
As an aside, machinists have more job security than you’d think considering how much manufacturing has been outsourced. Those that lack both the education and the desire to learn will find themselves lacking a job.

@CWOTUS What was that about fail? We call failures crashes. And you are correct at pointing out that, while it may be technically possible to add enough geegaws and cruft to mitigate/prevent foreseeable problems, it will be a while before we can do so at reasonable cost, and still not protect well against the unforeseen well enough to avoid the need for human oversight.

When driving, there are times where the logical thing to do is not the correct things to do. How many times has someone crashed into Google’s self-driving car as it drove perfectly competently? And considering how much it takes just to get it to do what it does, can you extrapolate what it would take to go beyond that by using an ability that we haven’t even figured out how to code, and do it in real-time?

Even with current technology, I think that second ability alone (doing it real-time) would require a computer at least the size of a car, which in turn would require a cooling system and power supply that just would not be feasible, especially not if you have enough hardware redundancy for me to feel comfortable trusting my life to the machine.

Of course, we could potentially remove humans from that loop if we did a total overhaul that included making it illegal to drive in any manner other than computer control, though even that would require more technology than I see in the next couple of decades, though the possibility of machine failure is still high enough that humans would still need to do some overseeing.

@LostInParadise I think this thread might be of note to you. I go into a little detail (too much for one person’s taste) about the effects automation will have on the labor force and our economy.

However, I feel that the projections in that article are at least naive, if not outright uninformedly ignorant. In the 1960s, they predicted that the year 2000 would be like the Jetsons, and that we would travel to the Moon (possibly even Mars) like it was a trip to Disney World. In the 1980s, Back to the Future predicted that we would not only have hoverboards in 2015, but that they wouldn’t burst into flame. And where the fuck is my flying car?

Maybe humans put too much stock in their supposed expertise. Here is a site that tells how to outperform the so called experts using nothing more than a calculator, though I would guess that at least some of the equations were arrived at using statistical methods or perhaps machine learning.

@LostInParadise Specify every cue and their weightings though. That is where computers fail, especially with things other than numbers.

Take driving for instance. See that red Nissan over there on the cross street approaching the intersection you’re rolling up to? What is his BAC? If he runs that light and you (or your autopilot) slam on the brakes, will the vehicle behind you stop in time to avoid a collision? Google says “Not always.”, and has the body shop bills to prove it. Which is more important anyways; the braking power of the vehicle behind you, or their reaction time?

Humanity’s greatest strength is adaptability. We are not strong, but we can use prybars, pullies, jacks, and other things to get around our weakness. We are not well insulated enough to survive in the snow, but we can use “prosthetic fur”, whether it be the hide of an animal we killed or a jacket made of materials we synthesized. We cannot outrun a cheetah, but we can make ridable machines that can. You get the point.

The equations you seen to think will allow computers to outperform humanity are limited by their inflexibility and naivete. At best they are akin to idiot-savants. It’s not like they do anything humans don’t do; they just do so without emotional bias. But the GIGO principle affects computers far more than it affects humans, thanks to their literalism, so it only truly works if you are damn sure the equation is right. David X. Li thought that he had a good replacement for the Black-Scholes equation, so I think the first thing we’d need to do is calculate how large a grain of salt we’d need to take any given SPR with.

@jerv I think your self assured futuristic optimism is going to have one hell of a time dealing with the reality of a falling average life expectancy.
Before antibiotics, the average life expectancy of a man was 45 years, and that wasn’t all that long ago. And the new reality of a world devoid of antibiotics is coming back around, I’m afraid, very much faster than anyone would/could imagine. But leave it to people like you who believe in Man’s continued superiority of the elements, to rattle on about how Man will ultimately win out against all logical, realistic odds to the contrary.

@jerv, It still cracks me up that all those self-professed wine experts with all their fancy talk and high salaries can be replaced with a simple linear equation.

Google’s main pitch for self-driving cars is that they are safer than cars powered by humans. Their safety record to date is pretty impressive. An added bonus is that by being networked they can respond to traffic conditions and get you to where you want to go more quickly.

@jerv I am not disputing that, but one of those geeky robotics hounds will replace 10 or more high school diploma, some college people. They will not be needed as some machine operated by the one geeky guy will be doing the job they (the high schooled diploma, some college) people would have been doing.

@Hypocrisy_Central Theoretically, yes. In reality, I have my doubts due to the rampant anti-intellectualism in this country, and similar resistance to change in some other regressive (if not repressive) regimes.

I cannot help but have GURPS:Transhuman Space snippets pop up in my head as I think about this question, but I wouldn’t even know where to begin on a bullet-point synopsis of how they handle this issue. Suffice it to say that sci-fi is often more speculative or extrapolatory that outright fantasy. At worst, it’s food for thought.

@jerv Theoretically, yes. In reality, I have my doubts due to the rampant anti-intellectualism in this country, and similar resistance to change in some other regressive (if not repressive) regimes.
Even if people in this nation went to college for no other reason than to run up a debt, shag like crazy, or party hardy, or all of the above, there will still be enough people coming out of MIT and similar colleges to fuel the move to robotics or a mechanized society. The greed on both sides of the fence may even help it. That may make the future ”haves and have nots” those who are tech savvy as oppose to those who are not.

Other regimes might not embrace it as much as here, they have fewer ”creature comforts” and gadgets to start with, and their people may not be as addicted to them as here. If they had the means to develop it, and in an effective way, I believe they would just to enter the ”First World Club” and be able to feel they are on par with the USA, Canada, Germany, etc. of this world.

First, a little setup….

Technology cannot advance in a vacuum. It needs a base, and it builds upon itself. Even the idea of bashing prey with a rock evolved from us bashing prey with our hands and looking for a way to improve the concept. As we get into the more advanced things, the more prerequisites there are. For instance, modern computers are great, but they rely on transistors which are basically miniaturized vacuum tubes (different operating principles, but they fulfill the same function) as well as the material science to do the whole semi-conductor thing. Oh, and electricity, with all that that implies.

I bring this up because those that can’t meet the prerequisites cannot advance. Haiti and Ethiopia are unlikely to come up with any major breakthroughs like self-sustaining fusion or room-temperature superconductors. If they could, they already would have by now.

There is irony there in that the means to make products cheap are themselves quite expensive. While a Handtmann SC may make a wing spar faster and cheaper than doing it by hand the way the Wright Brothers did, Handtmanns are not cheap, and the industrial power required to run it… well, it takes money to make money.

Put it all together, and the net result is that while the divide between the Haves and Have-nots will grow, the distribution will not be terribly different than it is now, with the possible exception of the US losing it’s superpower status and being left (at best) second-tier. China and the EU will be near/at the top, Africa at the bottom, India divided in the unique fashion that it is now only moreso, and the US… well, depending on how our internal pressures work out, we’ll either be in the top three or balkanized and too busy licking our wounds to keep up with the big dogs.

Now the tie-in…

”[T]here will still be enough people coming out of MIT and similar colleges to fuel the move to robotics or a mechanized society.”

I disagree. We are already on the decline in that regard while the rest of the world is either level or increasing.

“The greed on both sides of the fence may even help it.”

The greed on one side of the fence will not allow that to happen, and the other side really cannot do much about it… at least nothing peaceful. Do you see the US doing the free college thing, or do you think it more likely that the fat cats raking in the dough by raising tuitions to the point of inaccessibility will continue to buy lawmakers to make the Have/Have-not divide larger? While I doubt that is their intent, that is a direct result of their greedy actions.

If the greedy do anything to stop that, then they will sacrifice short-term profit for sustainability. Recent history shows that that will never happen. Not so recent history has plenty of precedent for it not happening to corroborate that. Only the sorts of reforms that will come at the expense of the Haves will ever bring it about, and rest assured that they will fight it to the bitter end. And if that’s what it takes, then the Have-nots will bitterly end the Haves. Just ask Nicolae Ceaușescu.

“If they had the means to develop it, and in an effective way, I believe they would…”

So do I. I just don’t see them having the means unless gifted the means by a benevolent superpower. Of course, rightly or wrongly, that would raise serious questions about sovereignty, so it’s possible that the gifts would be refused just to avoid having the people feel that the government gave their country to the benefactor as a colony.

No matter how you look at it, there are major social and economic issues involved. If you think the invention of the internet and the resulting ability to communicate to billions nearly instantaneously changed the world, wait until the fun actually gets started. Pull up a chair and enjoy the show.

When the mechanized revolution hits, how will humans deal with it?

The same way the manual laborers dealt with the industrial revolution – they adapted to the new standard. They learned how to use machines.

@Jackiavelli That would be nice.

Then again, it’s interesting how the age where one is considered “adult” has risen as educational requirements have risen. Back in the days of yore, if you were strong enough to work and smart enough to work unsupervised, you were an adult. Early/mid-20th century, you needed to be literate and able to do basic math to get most jobs, but that was easily doable by age 16. Late-20th century, just everyday life required more skill than most jobs a couple generations prior, making K-12 education (age 18) a minimum. Now, with so many jobs requiring degrees, some have said, “24 is the new 18.” as one won’t have enough education for many jobs until their mid-20s.

Ponder the implications of how raising the age of adulthood affects society.

I don’t think professional sports, like football, will have the same draw with robot players as compared with real people players. Not enough injuries!

There will be no way to have a robot do child protective investigations, so if you want to be immune to the robot revolution, train to become a CPS worker.

@kritiper As I said in a thread a while back regarding transhumanism, I think it possible that professional sports may divide in a manner similar to auto racing. Given that we already went through that issue with Oscar Pistorius and his prosthetic limbs, I think we would do well to start thinking about the issue now so that we’ll already have ways to deal with things like elective prosthetics becoming a thing.

@jca Honestly, the same is true of much of anything dealing with people. While we currently have automated systems that can handle things like paying your bill over the phone or checking out groceries without human interaction, those tasks are trivial compared to handling the sort of situations that normally flag the caller/customer as having circumstances not covered by their programming.

Given how slowly even things like speech recognition have evolved, I don’t see it being too big an issue within our lifetime. And our descendants will grow up with a better knowledge of which jobs are humanocentric and which will be done by robots than we have in much the same way that kids today grew up with iPhones and Facebook which seem to elude many people of our generation.