The big issues in the dockworkers strike are 1) pay and 2) automation. International Longshoremen's Association president Harold Daggett has mainly been talking about automation, which drives most economists crazy. American ports are not efficient by global standards, and we all pay a price for that. But it is true that automating American ports would eliminate thousands of jobs. Should we care?
Consider that Daggett once fumed about EZ Pass eliminating the jobs of toll booth workers:
Take EZ Pass. The first time they come out with EZ Pass, one lane, and cars were going through and everybody sitting in their car goes, ‘What's that all about, I'm going to get one of them.’ Today, all those union jobs are gone, and it's all EZ Pass. People don't realize it, everybody's got three cars, everybody got an EZ Pass on the window, and they go through like it's nothing, and they get billed in the mail. They didn't care about that union worker working in the booth. . . . Someone needs to go to Congress and say, ‘Whoa, time out,’ this world is going too fast for us. Machines have got to stop.
And for that I have zero sympathy. Sitting in lines at tool booths is awful, so E-Z Pass has improved my life. If your plan for your own future is all about imposing suffering on others, count me out.
On the other hand, let's imagine a future in which machines can do pretty much all work. What happens then?
I have no idea. So I tend to think we should be careful about these decisions.
Who bears the cost of inefficient ports? Mainly working and middle-class people, since we spend a lor more of our incomes on stuff than the rich do. American manufacturing companies also pay, since they use a lot of imported parts and materials. So this isn't a transfer from the rich to the poor, more a general tax on all of us.
I think what I would like to see here is the port workers getting a nice raise in exchange for accepting automation, but I can't manage particularly strong feelings about it.
1 comment:
On the other hand, let's imagine a future in which machines can do pretty much all work. What happens then?
I have no idea. So I tend to think we should be careful about these decisions.
So putting aside the fact that people have been contemplating the possibilities of a "post-scarcity" civilization since at least the 1960s, most famously with Star Trek...
...it's still a hugely moot point in a lot of cases, because people vastly overestimate how much can actually be automated, or will be able to be automated in the foreseeable future.
Case in point, clothing manufacturing. We have all sorts of automation for spinning thread and weaving thread on auto-looms and similar things - but one thing we've NEVER been able to automate is the actual sewing of garments.
Robots can spin and weave thread into bolts of cloth, and even do some degree of cutting, but sewing of said cloth into finished garments is beyond robot capabilities. Every single piece of clothing sold on the face of the earth is sewn by a human operating a sewing machine - most of them brutally exploited women and minors in Asian sweatshops, depressingly.
It turns out, sewing fabric is exceedingly complex, and requires a high degree of technical skill which no robot can even begin to approach. Robots are great with rigid materials - but they're utterly awful with materials that bend, stretch, flow, etc. The problem is so difficult, in fact, that people have tried to bypass it entirely by looking into the possibility of somehow 3D-printing clothes as an alternative.
Lots of other things have the same sort of problems. Any task which is uniform enough, and has few enough variables to grapple with, can potentially be automated. But lots of very basic societal tasks involve a considerable degree of variation in their performance, which demands a human level of flexibility (both figurative AND literal) which robots simply lack.
Housecleaning is another great example. Something as simple as loading and unloading a dishwasher is tremendously difficult to automate, in part because of the variability of the physical objects being manipulated, but also because of the space and form-factor constraints of the system which are designed to accommodate humans, rather than machines.
Plumbing can't be automated. Electrical wiring can't be automated. Most facets of construction, in general, can't be automated.
Automotive repair can't be automated. Computer repair can't be automated. Heck, basically any kind of repair requires a human.
Anything that involves social interactions. Anything that involves working with animals. Anything that involves human judgement can't be automated.
You can automate cars on roads because road systems are designed around concrete rules and standardized parameters. But you can't really automate off-roading because robots aren't able to intuit solutions to the complex physics required to drive a truck up a densely forested, boulder-strewn, intermittently muddy mountainside the way we humans can.
So many things require "going with your gut". And a great many of those things also require the ability to quickly change your mind on the fly. You have to be able to anticipate possible failure states before you attempt something; and you have to be able to recognize when you approaching said failure state BEFORE you actually reach it; and you have to be able to rapidly iterate on new alternative solutions to attempt, based on all that new information you just gleaned seconds before.
Post a Comment