And a lot more jobs are soon to be automated out of existence.
That’s a good thing. For example, the job of swinging a pick at the coal face in the mine has been automated out of existence. Do you really want it back?
Whether it’s a good thing, and how good, depends on what happens to the people who had those jobs.
Imagine that there are a thousand people doing a horrible job for $20k/year. Now a machine of negligible cost comes along that can do the same job as one of them for $1k/year. Then, oversimplifying in some obvious ways, we have the following:
Option 1: Machines replace humans, workers take the benefits. Each of these people switches from doing the job for $20k/year to doing nothing for $19k/year. They’re better off, their employer is exactly the same as before. This is better for some people and not worse for anyone. Of course it will never happen.
Option 2: Machines replace humans, owners take the benefits. Each of these people switches from doing the job for $20k/year to doing nothing for $0k/year. Maybe they can find other jobs, maybe not. Their employer is just $19k/year per worker better off. This is better for some people (the owners of the business) but much worse for the employees, at least in the short term. It is quite likely to happen.
(In this case, what probably happens next—at least if there is competition—is that the company lowers its prices somewhat. So now the business owners win and their customers win. In the long run these lower prices may lead to new jobs.)
Option 3: Something in between. A union negotiates a special deal, or the government steps in in the hope of reducing unemployment and disaffection, or something. Exactly what happens will vary but it’s probably better for workers than 2 and worse than 1, better for owners than 1 and worse than 2, and probably a non-negligible fraction of the benefits get eaten up by administrative costs.
In the long run, all of these are probably better than leaving things as they are. In the short run—say, a few decades—option 2 (which is the most likely of the three, I think) means a thousand people out of work, and quite a lot of them may be unable to find other jobs. This may well be a bigger loss of net utility than the business owners’ gain in wealth.
If machines end up taking everyone’s jobs, that could be glorious (if it leads to lives of comfortable leisure for all) or terrible (if it leads to lives of comfortable leisure for people who are already wealthy enough not to need to work, and starvation for everyone else).
So: yes, a lot of jobs are pretty terrible and an optimal world without those jobs is much better than an optimal world with them. But we don’t have the option of either sort of optimal world, we only get worlds designed by Moloch, and the Moloch-world without those jobs may be even worse than the Moloch-world with them.
Then, oversimplifying in some obvious ways, we have the following:
A bit too much oversimplification for my taste. In particular, there is a rather important bit missing: what happens to the price of the products (or services) that these people produced. Let’s say they made widgets. In a market-based economy the price of the widgets would go down considerably and this will lead to a lot of extra consumer surplus for all buyers of widgets. This could easily be the most important effect and the greatest source of utility.
And because of that your option 2 is actually very unlikely to happen, other than in a monopoly-like situation, even in the short run.
In the long run, all of these are probably better than leaving things as they are.
There we go :-)
we only get worlds designed by Moloch
I don’t know about that. Moloch has to wrestle with the invisible hand of Adam Smith :-D and the world we are currently living in isn’t half bad, is it?
In particular, there is a rather important bit missing: what happens to the price of the products (or services) that these people produced.
Yeah, in retrospect I should have said something like
(In this case, what probably happens next—at least if there is competition—is that the company lowers its prices somewhat. So now the business owners win and their customers win. In the long run these lower prices may lead to new jobs.)
No, what happens is that the original producer plays a stock short on the new guy, does an M & A on them, then liquidates the new company while trapping any intellectual property and patents in the safe.
Then we lose competition, price pressure, and innovation.
Arguably, some people are already not economically viable as employees. But I don’t know why most of people wouldn’t be able to adapt, just like their no-longer-economically-viable peasant ancestors did.
Until now, humans always had multiple competitive advantages in sensing, intelligence, and motor control, and an integrated system for all. That competitive advantage made them the best adaptable machines available.
That advantage is going away, particularly for the less intelligent and less educated.
Horses used to compete for real work in the economy, and win. Their population dwindled in the US as they were competed out of the marketplace by machines. Horse genocide. Their domesticated population has been coming back, but now as pets more than workers, and they’re still not near the numbers they used to be.
How many human pets do people want? How many people want to be a pet?
When other things get smarter, cheaper, better, and you don’t, eventually you lose. And machines have advantages to employers that people don’t.
When other things get smarter, cheaper, better, and you don’t, eventually you lose.
That’s the basic luddite proposition and the problem is that the entire history of mankind says that this is not the way it works. And if you are going to pronounce But This Time It Will Be Different, you need stronger arguments.
That’s a good thing. For example, the job of swinging a pick at the coal face in the mine has been automated out of existence. Do you really want it back?
Whether it’s a good thing, and how good, depends on what happens to the people who had those jobs.
Imagine that there are a thousand people doing a horrible job for $20k/year. Now a machine of negligible cost comes along that can do the same job as one of them for $1k/year. Then, oversimplifying in some obvious ways, we have the following:
Option 1: Machines replace humans, workers take the benefits. Each of these people switches from doing the job for $20k/year to doing nothing for $19k/year. They’re better off, their employer is exactly the same as before. This is better for some people and not worse for anyone. Of course it will never happen.
Option 2: Machines replace humans, owners take the benefits. Each of these people switches from doing the job for $20k/year to doing nothing for $0k/year. Maybe they can find other jobs, maybe not. Their employer is just $19k/year per worker better off. This is better for some people (the owners of the business) but much worse for the employees, at least in the short term. It is quite likely to happen.
(In this case, what probably happens next—at least if there is competition—is that the company lowers its prices somewhat. So now the business owners win and their customers win. In the long run these lower prices may lead to new jobs.)
Option 3: Something in between. A union negotiates a special deal, or the government steps in in the hope of reducing unemployment and disaffection, or something. Exactly what happens will vary but it’s probably better for workers than 2 and worse than 1, better for owners than 1 and worse than 2, and probably a non-negligible fraction of the benefits get eaten up by administrative costs.
In the long run, all of these are probably better than leaving things as they are. In the short run—say, a few decades—option 2 (which is the most likely of the three, I think) means a thousand people out of work, and quite a lot of them may be unable to find other jobs. This may well be a bigger loss of net utility than the business owners’ gain in wealth.
If machines end up taking everyone’s jobs, that could be glorious (if it leads to lives of comfortable leisure for all) or terrible (if it leads to lives of comfortable leisure for people who are already wealthy enough not to need to work, and starvation for everyone else).
So: yes, a lot of jobs are pretty terrible and an optimal world without those jobs is much better than an optimal world with them. But we don’t have the option of either sort of optimal world, we only get worlds designed by Moloch, and the Moloch-world without those jobs may be even worse than the Moloch-world with them.
A bit too much oversimplification for my taste. In particular, there is a rather important bit missing: what happens to the price of the products (or services) that these people produced. Let’s say they made widgets. In a market-based economy the price of the widgets would go down considerably and this will lead to a lot of extra consumer surplus for all buyers of widgets. This could easily be the most important effect and the greatest source of utility.
And because of that your option 2 is actually very unlikely to happen, other than in a monopoly-like situation, even in the short run.
There we go :-)
I don’t know about that. Moloch has to wrestle with the invisible hand of Adam Smith :-D and the world we are currently living in isn’t half bad, is it?
Yeah, in retrospect I should have said something like
… Oh, wait. I did.
No, what happens is that the original producer plays a stock short on the new guy, does an M & A on them, then liquidates the new company while trapping any intellectual property and patents in the safe.
Then we lose competition, price pressure, and innovation.
And even more folks are out of work.
Good point. I expressed myself poorly.
A lot of people are soon to be automated out of economic viability as employees.
Arguably, some people are already not economically viable as employees. But I don’t know why most of people wouldn’t be able to adapt, just like their no-longer-economically-viable peasant ancestors did.
Until now, humans always had multiple competitive advantages in sensing, intelligence, and motor control, and an integrated system for all. That competitive advantage made them the best adaptable machines available.
That advantage is going away, particularly for the less intelligent and less educated.
Horses used to compete for real work in the economy, and win. Their population dwindled in the US as they were competed out of the marketplace by machines. Horse genocide. Their domesticated population has been coming back, but now as pets more than workers, and they’re still not near the numbers they used to be.
How many human pets do people want? How many people want to be a pet?
When other things get smarter, cheaper, better, and you don’t, eventually you lose. And machines have advantages to employers that people don’t.
That’s the basic luddite proposition and the problem is that the entire history of mankind says that this is not the way it works. And if you are going to pronounce But This Time It Will Be Different, you need stronger arguments.
Wrong. This proposition has never been tested before. Things were not a threat to be smarter, cheaper, better before.
Humans have dominated the world through their intelligence. It’s the most powerful factor of production.
You need a stronger argument than “intelligence is just the same as all other factors of production”.
“Cheaper and better” existed before, so you are focusing on “smarter”. So what is that which is smarter than humans? Point it out to me.
Computers are already better than us in a ton of intelligence tasks, and that list is only going to get longer and longer.