So What?

Due to the paywall I generally avoid linking to op-eds, columns, or articles in the New York Times. However, since Business Insider has linked to a NYT op-ed on intelligence, I’m going to comment on it. Here’s the meat of the op-ed:

Research has shown that intellectual ability matters for success in many fields — and not just up to a point.

Exhibit A is a landmark study of intellectually precocious youths directed by the Vanderbilt University researchers David Lubinski and Camilla Benbow. They and their colleagues tracked the educational and occupational accomplishments of more than 2,000 people who as part of a youth talent search scored in the top 1 percent on the SAT by the age of 13. (Scores on the SAT correlate so highly with I.Q. that the psychologist Howard Gardner described it as a “thinly disguised” intelligence test.) The remarkable finding of their study is that, compared with the participants who were “only” in the 99.1 percentile for intellectual ability at age 12, those who were in the 99.9 percentile — the profoundly gifted — were between three and five times more likely to go on to earn a doctorate, secure a patent, publish an article in a scientific journal or publish a literary work. A high level of intellectual ability gives you an enormous real-world advantage.

My reaction to that is so what? I don’t think there’s any doubt that highly intelligent people are better able to connect things mentally and do it faster than those who are less intelligent. However, any number of studies have found only a weak correlation between very high intelligence and income.

Furthermore, we’re talking about a rather small number of people. The 99th percentile is around two standard deviations above normal. The 99.99th percentile is around four standard deviations above normal. That’s roughly one in every 4,000 people or 75,000 people in the entire United States. Most of the people you went to school with, have met, or deal with on an everyday basis don’t make the cut.

Most of those who are in the top 1% (or even the top .1%) of income earners don’t make the cut, either.

To my mind there are several key points to remember: there’s a difference between accomplishment and the ease of accomplishment and incomes are, as I said, only weakly correlated with IQ. The highest paid nuclear physicist, whose IQ is almost undoubtedly four standard deviations above normal, probably earns less than the median cardiac surgeon who almost certainly isn’t even above the second standard deviation above normal. The supply of nuclear physicists may be low but the demand is pretty low, too.

Another factor that should be kept in mind. Cognitive development isn’t the only kind. Physical, cognitive, emotional, and social development are all important and, guess what? Individuals with high levels of social development are more likely to succeed physically, cognitively, and emotionally than those with the highest levels of cognitive development and lower levels of physical, emotional, etc.

For me the bottom line is that for the foreseeable future we’re going to be building a society and economy in which half of the people are normal (i. e. within one standard deviation of median) while a quarter are below normal and a quarter are above normal. Allowing our society to transmogrify into one in which the only people who can prosper are those in that top quarter or, worse, in the top .01% is a society doomed to failure.

25 comments… add one
  • michael reynolds

    Going forward the interesting thing to me will be the IQ (using that term very loosely, obviously,) of machines and programs. At present machines can’t replace the smartest people, but can in many cases replace the least intelligent. (In fact, if we didn’t have such easy access to so many people earning next to nothing — Mexican immigrant labor, foreign labor, we’d have replaced still more with machines.)

    I don’t think the robots are working their way up the IQ scale in some steady point-by-point progression, taking out the 81’s today and the 82’s tomorrow, but I think the effect will be that large numbers of people — disproportionately at the bottom of the IQ scale — will be rendered permanently unemployable.

    Is there a way to build a society where people who can be profitably replaced by machines will nevertheless be employed? Maybe, but what would be the point? Would it make sense to take the robots off GM’s assembly line? Should we outlaw Siri to save the jobs of people at the Information booth?

    The robots are coming. And with them come social and economic (and political, ideological and philosophical) dislocations of revolutionary proportion.

    I think this coming change may be more unsettling and dangerous than global warming, and I don’t know that anyone is paying that much attention. As it bites more into the engineers, radiologists and lawyers it will suddenly take on political urgency, but those people are probably better able, by virtue of IQ, to reprogram for different careers. People in the bottom quarter? Not so much.

    Putting on my sci-fi guy hat, I’d guess the societal adaptation will be to redefine the ‘job’ as a privilege rather than an obligation. Those who are lucky enough to hold jobs will work — as will the robots — to support society.

    The Randian fantasy is deader than the dodo. The future is socialist. Either we’ll create pointless make-work jobs for people who can easily be replaced by machines (one kind of socialism) or we’ll subsidize the people who can’t work, (another version of socialism.)

    To paraphrase Orwell, if you want a picture of the future, imagine a lot of people watching TV and eating Cheetos, while a minority telecommute (and watch TV and eat Cheetos.)

  • The problem I have with these sorts of things is that it assumes there is one metric for intelligence. I think how “smart” people are really depends on context and what they are doing.

  • jan

    The problem I have with these sorts of things is that it assumes there is one metric for intelligence. I think how “smart” people are really depends on context and what they are doing.

    A good point, Andy.

    According to Howard Gardners book, Frames of Mind — The Theory of Multiple Intelligences, everyone possesses at least seven intelligences, lending itself to a unique blend of competence factors in each individual, which, IMO, would be difficult to reproduce in an artificial intelligence source.

    Having said that, I listened to a provocative book review, a while back, which looks into the future use of technology, called, The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future.

  • Icepick

    Typo Alert: there’s a difference between accomplishment and the ease of accomplishment and incomes are, as I said, only weakly correlated with income.

  • steve

    Heckman has written on this topic a lot. I would put self-discipline and empathy up there as important abilities that correlate with success. Intelligence should be looked at as a threshold quality. You need a certain minimal amount to be successful. Beyond that, other factors are more important.

    Steve

  • Drew

    “The robots are coming.”

    Get a horse.

    “The Randian fantasy is deader than the dodo. The future is socialist.”

    Fascinating straw man. A fed govet that spends, what, 25% of GDP? is “Randian? Add state and local and I bet govt taxation is 40% of the national income. Randian?

    Bizarre.

  • michael reynolds

    Randian “fantasy.” You know, your political ideology.

  • Drew

    “Randian “fantasy.” You know, your political ideology.”

    No, your bizarre straw man.

    This reminds me of every debate I’ve ever had with a leftist: “so how much should we pay in taxes?”………..(blank stare)……..”more”

  • Icepick

    I would put self-discipline and empathy up there as important abilities that correlate with success.

    Empathy? Really? How many political leaders and financiers are sociopaths or pyschopaths?

  • Icepick

    Also, I’m sure Genghis Khan doesn’t have all those n-children because of his empathy.

  • The robots are coming. And with them come social and economic (and political, ideological and philosophical) dislocations of revolutionary proportion.

    Sure, we’ve seen it before. Will “this time be different” maybe, but you have yet to make a convincing case. And even if this time “is different” why is that you assume those who aren’t book smart/don’t have great SAT scores wont be able to do worth while things? That is, why wont those unemployed people also be viewed as an unemployed resource…an opportunity to do something that makes everyone better off (or at least some better off and others no worse off)?

    I think this coming change may be more unsettling and dangerous than global warming, and I don’t know that anyone is paying that much attention.

    Bah, in the long run we are all dead so who cares about global warming or robots or any of this other bullshit?

    The Randian fantasy is deader than the dodo. The future is socialist. Either we’ll create pointless make-work jobs for people who can easily be replaced by machines (one kind of socialism) or we’ll subsidize the people who can’t work, (another version of socialism.)

    Yeah, because stupid people are so…well stupid that there is nothing they can do. I know, maybe a nice eugenics program.

    I bet you don’t even know what Rand’s philosophy is (i.e. it isn’t mine). But please do continue on with your creativeness which you one of the greatest people on the planet Micheal.

    See where you fall completely on your face Michael is that you think robots can perfectly substitute for human labor in a cost effective way everywhere and that the unemployed labor of humans ends up having zero value. Problem is that robots are like any other form of capital and we will only have so much.

    Is it possible that producing capital goods becomes so cheap that the above will happen? Maybe, but if so you have just joined up with likes of the late Julian Simon…cornucopian and libertarian.

    See you have a stunted and childish view of my philosophy Michael. Libertarianism isn’t about getting rich. It isn’t about getting mine and you can fuck off. It is about liberty and freedom. Many libertarians view the market economy as a good way to promote liberty and freedom. If, however, through technological advancement we get to a point where resources are no longer scarce and the market economy goes away? Fine. It is a means, not an end. Do I think we will get there? Probably not in my life time and probably not in my son’s.

  • Icepick

    And even if this time “is different” why is that you assume those who aren’t book smart/don’t have great SAT scores wont be able to do worth while things? That is, why wont those unemployed people also be viewed as an unemployed resource…an opportunity to do something that makes everyone better off (or at least some better off and others no worse off)?

    Yeah, I’m a definite wealth opportunity.

  • Icepick

    I forgot to add to my last comment:

    @@

  • Yeah, I’m a definite wealth opportunity.

    I know, even when we have big changes the likes Michael is talking about there is a period of dislocation that isn’t fun.

  • Icepick

    Isn’t fun? Wow.

  • Drew

    “Libertarianism isn’t about getting rich. It isn’t about getting mine and you can fuck off. It is about liberty and freedom.”

    You pig. You just don’t care about people……………and you are probably a racist to boot. Probably even want to take the bottom half and turn them into fertilizer….

    When Michael turns over his entire net worth to the poor, I’ll listen. Otherwise, its just self congratulatory gibberish.

  • michael reynolds

    Steve:

    This seems like the key paragraph:

    See where you fall completely on your face Michael is that you think robots can perfectly substitute for human labor in a cost effective way everywhere and that the unemployed labor of humans ends up having zero value. Problem is that robots are like any other form of capital and we will only have so much.

    How many copies of Siri are there? How much does it cost to make more? (Answers: an essentially infinite number, and zero dollars.)

    Sorry, but you’re just wrong. Yes, this time is different. And the result will be a version of socialism. It’s almost not worth debating since all I really have to do is wait.

  • michael reynolds

    To expand a bit, Steve, your notion of robot labor is big steel machines replacing muscle-bound workers. Nah. That’s been done. Yes, we’ll see more of that — manage to shut off the flow of Mexican labor and see how long it takes to develop machines to pick strawberries.

    But that whole thing is old hat. What we’re talking about now is programs. Programs to replace teachers? Do-able. Programs to replace order-takers at the Wendy’s? Do-able. Programs to replace travel agents? Oops, already done. Programs to replace programmers? Do-able.

    And the idea that we’ll somehow employ as many people to create programs to replace programmers is nonsensical. The machines, the programs, the robots are getting smarter. They aren’t just more precise than humans (automated machine tools) or more consistent and tireless than humans (assembly line robots) they’re smarter now, too. Unless you think the old lady working the Information booth knows more than Siri knows.

    We’ve been in a race with machines for a long time. You can take comfort in the fact that we haven’t “lost” yet, but the past isn’t always prologue. Sometimes all you’re seeing is phase one, the early innings. The game is still on. Tied at half-time does not mean it will still be tied by the final play.

  • sam
  • sam, you might have mentioned one of those in the list: writers.

  • sam

    Apparently, none of us are safe.

  • PD Shaw

    I see somwhat of a different pattern in many of the jobs in sam’s link. I see work that is already becoming unproductive for humans either because of labor costs or changes from the internet. For example, is the sports writer to be replaced by robots or does the robot save sports writing for the general reader?

  • How many copies of Siri are there? How much does it cost to make more? (Answers: an essentially infinite number, and zero dollars.)

    Bzzt…wrong, if that were the case then the optimal price is zero. Is the price zero? No? then I highly doubt that the (marginal) cost is zero. Or to put it this way, Siri’s programing is worthless without the hardware where it can do its stuff.

    To expand a bit, Steve, your notion of robot labor is big steel machines replacing muscle-bound workers. Nah. That’s been done. Yes, we’ll see more of that — manage to shut off the flow of Mexican labor and see how long it takes to develop machines to pick strawberries.

    No, I’m talking about artificial intelligence and for it to be useful in a real world sense it also has to be accompanied by robots as well. Things that can move physical things around as well as handle information and in some cases both. So it isn’t just a lever or a crane, but a smart crane. A crane that wont need a human operator and when it encounters a problem not covered by its initial programming can adapt and overcome the problem.

    By the way, we are seeing this in the utility industry, right now we are deploying our “smart” meters. Now they wont be able to solve problems with an artificial intelligence, but they will be a huge improvement over the old meters. One downside is the need for far fewer meter readers.

    But that whole thing is old hat. What we’re talking about now is programs. Programs to replace teachers? Do-able. Programs to replace order-takers at the Wendy’s? Do-able. Programs to replace travel agents? Oops, already done. Programs to replace programmers? Do-able.

    You know what you sound like? The guys who spouted off about that “New Economy” crap and got their asses handed to them by the “Old Economy” when the tech bubble burst., but even worse. They at least acknowledged that processors were a factor in the economy. Of course they left out people though and their consumption spending and failed to recognize a bubble economy.

    Even if you are right though, you’ll still need servers and processors and so forth. The iCloud isn’t really just a cloud hanging around with data in it, it is a set of servers that you’ll be able to access anywhere…huge improvement in information technology? Absolutely, will it be completely devoid of the physical? No. As such, there is a capital aspect to this and your failure to recognize that means your predictions are highly suspect. Oh and cloud computing is not a new idea, been around since the 1960s.

    And the idea that we’ll somehow employ as many people to create programs to replace programmers is nonsensical.

    Never said that.

    The machines, the programs, the robots are getting smarter. They aren’t just more precise than humans (automated machine tools) or more consistent and tireless than humans (assembly line robots) they’re smarter now, too.

    Not yet they aren’t, at least not on a cost effective manner. I know they have come up with computers that are good at Jeopardy, but is that really a thinking computer or is it a computer that is really good at data retrieval? And it sure a Hell isn’t “free” the damn thing has 4 TB of storage, that is only good as the juice is running to power the thing.

    Really Michael, it would be nice if you actually responded to my comments not your version of them.

  • Brett

    @Michael Reynolds

    To expand a bit, Steve, your notion of robot labor is big steel machines replacing muscle-bound workers. Nah. That’s been done. Yes, we’ll see more of that — manage to shut off the flow of Mexican labor and see how long it takes to develop machines to pick strawberries.

    They likely already have those machines – it’s just that it isn’t cost-effective to use them when human labor is cheaper and more versatile.

    That might apply to more “physical manipulation” positions than you realize, Michael. Robots are replacing assembly line workers, but those are positions where they can work under very controlled conditions doing the exact same things over and over again. The newer automation wave seems to be hitting some more difficult tasks (such as driving cars, depending on when Google’s automated car-driving system hits the market), but they’re still far, far short of being able to duplicate humans in this area. Robots like ASIMO are not close to replacing the guy who brings your couch in through the front door, or the person installing your flooring.

    And the idea that we’ll somehow employ as many people to create programs to replace programmers is nonsensical. The machines, the programs, the robots are getting smarter. They aren’t just more precise than humans (automated machine tools) or more consistent and tireless than humans (assembly line robots) they’re smarter now, too. Unless you think the old lady working the Information booth knows more than Siri knows.

    Why would need to employ as many people? Most of the world is hitting the Demographic Transition, which means that the overall pool of Working Age adults is going to start shrinking in the not too distant future.

    On top of that, personal assistant AI like Siri don’t just replace jobs – they could also potentially be a huge boon for workers. Twenty years down the line, everybody could have the equivalent of Siri 10.0, which can supplement and enhance their life and productivity while they work. *

    * Not to speculate too much, but we might also be in shouting distance of physical augmentation by that time period. At that point, the boundary between you and Siri 10.0 become much blurrier.

  • Tully

    I hear echoes of discussions I’ve had lately with pols claiming we don’t need to fund libraries at current levels because the “Library of the Future” will obviate the need for the space and expenditures. (Never mind that librarians have little idea what the “Library of the Future” will look like — they’ve seen that applied tech moves in unpredictable ways with unanticipated effects — and pols have as much clue about that as I do about which horse will win the 2016 Kentucky Derby.)

    Yeah. Just like that “paperless office” I kept hearing about in the early ’90s resulted in less paper being generated and subsequently fewer employees being required overall to handle information.

    Instead what happened was that as the cost of handling information dropped, the amount of information being handled radically increased, along with resulting demand for IT services, along with space and capital requirements for servers and storage and, of course, hardcopy records that couldn’t go poof in a power surge and would be acceptable by the courts and bureaucrats as valid.

    Predicting the future effects of tech on employment is a mug’s game.

Leave a Comment