Happy Labor Day!

Arnold Kling writes:

Meanwhile, a relatively small set of people will be able to figure out the artificial-intelligence economy and be able to position themselves to reap its rewards.

From another direction, here is a mention of “cartels of mutually satisfied mediocrities.”


In relation to Lanier's piece, Arnold Kling says read it but "I do not think all of it makes sense". In relation to AK's post, I say yes read it but I think it makes little sense. In my view, we should look at what is happening with labor in the world market economy as a problem of a fundamental change in the relative scarcities of different types of skills as a result of the integration of over 3 billion people into the world economy rather than as a consequence of technological change. Whereas AK thinks it is a problem on the demand side, I point the problem on the supply side as the critical one (let me repeat it again: the U.S. economy should not be analyzed as a closed economy). To make sense of my idea, however, I must define clearly what those types of skills are and how we can measure their relative scarcities. I have yet to make significant progress in addressing these two issues. Indeed, as long as we aggregate labor into just one category, or into two or three categories reported in some data, we are not going to make any progress.

Kling is very generous in his treatment of Lanier, creating an interpretation in which he actually said something that makes sense.

However, I find it hard to take seriously someone who states, among other things, that the internet is a "failed technology". I guess the internet did fail if it was designed to eliminate the gap between rich and poor and solve most other major problems in the world. But it looks to me like the internet has been a huge benefit a lot of people, and not only those rich ones who are lucky enough to control the servers.

I agree with you. As much as TV has increased the value of sports (and therefore the services of some professional sportswomen/men), Internet has increased the value of ideas (and therefore the services of some professional thinkers and communicators). Not surprisingly, despite the large increase in demand triggered by technology, even in the new world economy of over 6 billion people, supplies have struggled to meet the new high demands, particularly the high quality end of these demands, and therefore the earnings of many suppliers have increased albeit very unequally.

Today America must deplore the big waste of labor that government at all levels has become. How many people are the federal, state, and local governments hiring to do stupid and often illegal things? We can list some examples, like this one


Heh, that's funny: I just finished registering for the Stanford AI class.

I wonder how many signed up for the advanced option? Should be interesting!

RE: cartels of mutually satisfied mediocrities

So, the second most certain city in which people who actually deliver what they say are most punished is Athens? What does that tell us to expect about a Greek bailout?

I've really appreciated the posts on MR lately about ZMP workers, The Great Stagnation, information technology, the impact of free trade on U.S. labour markets, etc. Had an interesting conversation with a friend about this stuff a couple weeks ago after the London riots (which didn't seem to have any real cause other than a general sense of hopelessness and boredom).

What if all of our economic and technological development has been taking us to somewhere that isn't really providing happiness? That is, if machines do most of what humans once did, even if it's feasible to pay for welfare for everybody other than a few kingpins that run the system... what if this isn't really producing happiness? How do you account for people's need to be a part of something greater than themselves when calculating the costs and benefits of different policies?

As a liberal progressive, I've always assumed that a general "working towards a better humanity" was enough of a reason for living for most people. For me, who gets paid to work on something really interesting and cool and someday useful (magnetic fusion energy), it certainly is enough. But if you are someone whose use to society is, say, packing Thermos bottles into their retail packaging, and your job is about to be replaced by an ABB Frida robot... what's left? Sitting around drinking and instant-messaging your friends from a tablet computer? I guess conventional economics would say that that's what you want, since that's what you'll spend your money on... but I don't see it leading to long term stable societies. More like the London riots, across the world.

I don't mean to sound arrogant here. This is not a matter of "haha, I can do physics, but these lumpenproles here who are fit only to be a cog on an assembly line, sucks to be them..." I really am having doubts in my usual optimism about the future lately. If robots, AI, and cheap overseas labour can replace 50% (or 75%) of what people in the First World do for work... what's left? A handful of high-end jobs, and the rest of the people either selling each other insurance and coffee, or collecting welfare? Even if we all have the latest gadgets. Freedom from labour was supposed to give us happiness. What if it doesn't?

Well the ZMP stuff is a load of crap to make the people who are doing okay feel better. You raise some good points though, Economics as a discipline has this view of labour where if you can produce something with less labour then before it's all the better for everyone it's always sort of assumed the labour can be redeployed elsewhere which in the real-world obviously isn't true. Economics doesn't at all take into account that people actually like having meaningful work. Work is not purely an economic transaction. The more technological advances that occur , the more disruptive they are, you start getting this labour market where only the very best-of-the-best can participate.
But then it's expected that everyone should be happy to go along with it - even though a growing number of people won't even reap any benefit from new technology. From my perspective, for instance, you can go on and on about say new mobile tablets and how great they are, but it does f@#k all for me, I can't afford any of them period and even fairly moderate cost reductions won't help me out. In fact these new gadgets become widespread enough that it becomes something you NEED to have to operate in society and if you can't afford one well the cost of basically living in society goes up - more people get pushed out.
It's probably not really too wise to argue against technological advances but at the same time you deprive people of meaningful work I think you can expect a lot more riots and violence in the future - and this is definitely not something that enters into any economic modelling.

Congratulations, Mr. Liberal Progressive. All that time you were tearfully orating about a 'living wage,' the 'right' to a job, cradle-to-grave social services, you were lowering the marginal productivity of American workers. All that time you were screaming 'racism!' at anyone daring to question the 'right' to travel and the blessings of diversity, you were setting blue-collar workers on a race to the bottom for the benefit of Corporate America. Your naive faith in global institutions is exploited ruthlessly by net-exporters who have no qualms about asking what is best for THEM and THEIR NATION and who do not give a second thought to the 'global' good.

Now that all the monetary inflation necessary to enact the delusional, progressivist social engineering schemes has borrowed all from future productivity that can be borrowed; now that it's clear an increasingly 'diverse' workforce is not going to pay nearly enough taxes to fund Baby Boomer retirements; now that government bonds have completely distorted the capital markets, you pull your head out of the sand long enough to blink in the bright sunlight of reality and exclaim, "What?! Nobody told me...!"

Wow, that's a lot of anger. Here are five (somewhat scattered) points:

1. I'm not American. Even worse, I'm an anti-nationalist. As distasteful as this must seem to you, the drastic rise in living standards of hundreds of millions of people overseas actually means something. It sucks that it seems to require that the workers of the first world see times get worse.... but globally, things are getting better. I know, I know, I'm supposed to care about certain people more because I happen to be born on the same continent as them... this has never made much sense to me.

2. Exactly what do you think is the alternative? Somehow stop technological progress, and return to the 1950s when basically the entire non-USA world was either pre-industrial or bombed out of existence? Those hundreds of millions of new workers do exist and tariff walls can't keep them from working forever.

3. What is this stuff about "diverse" workforces and the "right to travel"? I didn't even say anything about that. Can you be more specific as to what you mean here?

4. So the real point of my original comment was to speculate that "meaningful work" is a requirement for a stable population. (That is, since the decline of religion as a controlling force. Wasn't a lot of meaningful work in 1450 but people had other reasons to get up in the morning.) And it doesn't look like meaningful work for uneducated people is going to be plentifully available any time soon...

5. I hardly think it's fair to blame the decline of tariffs and the opening up of global free trade on liberal progressives. I suppose it could be the result of a sort of alliance or meeting-of-minds between anti-nationalist/anti-racist progressives, and pro-free-trade pro-business libertarians. That's an interesting thought, actually.

That Gambetta paper on L-doers is fabulous.

"A feeling of familiarity develops among L-doers: L-prone people recognise other L-prone people as familiar, as ‘friends’."

AI replacing jobs is the 'fusion energy' of the computer crowd - it's always something that's just 20 years away, and it's been that way for 30 years.

When I was in college in the 80's, there were constant dorm-room debates over whether there would be any jobs left for anyone by the year 2000. The consensus among many of my peers was that 80% of all existing jobs would be done by robots and computers by 2000-2010. I thought that was crazy then, and I think it's crazy now.

Most lay people have no idea just how difficult the AI problem is, and how little progress we've actually made towards significant artificial intelligence. In the last 30 years, we've seen some improvements in natural language (but it's not perfect), and we've managed to make bipedal robots that can walk and do simple tasks in a controlled environment. There are a few other limited areas where AI has made inroads, such as facial recognition or some autonomous decision-making for drone aircraft and such. But there is no general purpose AI capable of solving problems the way a human would.

Robots are most prevalent in factories, and that's actually my specialty. I write software for machine control, factory automation, and MES (software that analyzes factory inputs and outputs and improves efficiency). The companies that are buying this type of software aren't really eliminating employees with it - they're using it to fine-tune production processes to eake out efficiency gains of a percent or two here and there. The robots in factories are incredibly dumb - they are basically programmed step-by-step by humans. Some of them have sensors that can detect things like a misaligned part and correct it, but mostly they just do exactly as they are told.

There's no AI controlling these things. Even the programming languages for them are very old - many of these devices are programmed using ladder logic, which is a remnant language from the days when everything in a factory was controlled with relays and transistors. We're still running factories with a language designed in the 1960's and 1970's.

We've known for a long time that assembly-line workers were doomed to be replaced by automated processes. But once you get past that and into jobs that require decision-making and judgment, replacing humans with AI is about as far away now as it was when I was in college.

Hey - don't ruin Arnold Kling and Bryan Caplan's wet fantasy here.

Yes, exactly. Here's an experiment you can do at home -- try dragging a few photos into Google's 'Image Search', for like this for example. The matches make it pretty clear that the image processing algorithms are still often completely clueless at recognizing objects. Which is not surprising, because AI has been lousy at that for going on 50 years now. And this image search is from the company whose AI is going to power the real-time vision systems used by autonomous cars?

Even as 'robots' go, humans are pretty hard to beat. There are no capital costs to speak of, operating costs (especially for basic models) are quite low, and they're relatively easy to program and reprogram (by non-programmers using natural language) to do even fairly complex and intricate tasks. Their mobility and balance and the coordination between their vision systems and 'robotic arms' is excellent. And so on...

Happy Labor Day All You ZMP Workers!

The mythical ZMP workers that is

Of course. We need the myth of ZMP to make us feel good and to anaesthesesize the public from doing anything to correct it.

The sci fi utopias Greg Olynik refers to above presumed a much flatter distribution of wealth than the one we actually have.

"Greg Olynik"? Really? You drunk, holmes.

Presumably the artificial-intelligence economy can figure itself out.

The Lanier as interpreted by Kling thesis seems to be something like a version of Karl Marx. At a very broad level of generalization, Marx believed that capitalist development of industrial technology would lead to high level of total economic production, enought for everybody to have a high standard of living. However, he also thought that capitalism would lead to an unsatisfactory and worsening distribution of income which he felt was (a) normatively wrong and (b) politically unstable. Seems to me that Lanier/Kling are saying much the same thing except not expressing an opinion on the politically unstable part.

Comments for this post are closed