Chad Perrin: SOB

17 December 2008

How is an economy like technology?

Filed under: Cognition,Geek,Liberty — apotheon @ 01:49

In a comment at his own Weblog (Chip’s Quips), my friend Sterling had this to say:

Yes — the myth still circulates that technology will eliminate the need for brains. The reality is that technology simply enables brains to work at a higher level while spending less time concerned with repetitive details.

This was part of a conversation related to the fact that every few years there’s a new push to try to eliminate programmers from the business of programming by creating some kind of programming environment that can be operated by MBAs. He’s right, of course — as is Scott Westfall’s Eliminating the Programmer, which is what spawned the discussion at Chip’s Quips in the first place.

Something that struck me about Sterling’s comment, however, is the parallel that can be drawn between:

  1. technology’s effect on technology markets when the “work smarter, not harder” breed of technologists are given the liberty to employ their technological skills as they see fit
  2. wealth’s effect on economic markets in general when the “work smarter, not harder” breed of wealth producers are given the liberty to employ their wealth as they see fit

In each case, the producer — whether it be a producer of new technologies (such as engineers and programmers) or a more general producer of wealth (such as an entrepreneur) — is better able to develop paths to more positive outcomes for everyone on average than external regulators. This is because the real producers, the innovators in their respective fields, tend to operate at a higher level of abstraction than outsiders who try to command the field via statistical studies.

Statistics is a field that can only produce meaningfully accurate results when you measure extremely narrow metrics and know how to interpret your results appropriately. One of the biggest and most common errors made by mediocre (or worse) statistical analysists is that of failing to eliminate or account for subtle variables that skew results. Another is assuming correlation implies causation. These errors make up the lion’s share of the reason that the software industry is full of middle managers who do stupid things like try to measure programmer productivity in lines of code produced (aka lines spent), version control check-ins, and so on, over a given period of time.

They also make up a lot of the reason that Congress does stupid shit like pass a $700B bailout bill that is likely to cause more harm than good. Keynesian economists, and Keynesian leaning econometricians, as well as people with no economics knowledge to speak of at all (such as congresscritters), have a tendency to measure a very limited number of economic factors, make unwarranted assumptions about what these measurements mean, then generalize from there while ignoring all the other factors that pertain to the situation to try to come up with ham-handed “fixes” for misapprehended problems. The end result is that government’s involvement in the economy ends up being an attempt to fix a problem using the same techniques that created it in the first place — of trying to make two nearly identical wrongs become a right, in short — at least 98% of the time.

At least, that’s the case if we assume that what middle managers actually want is more, and better, productivity, and what congresscritters and Keynesians actually want is greater general wealth production and across the board improvement of quality of life. If we assume these are the respective goals of these people, they fail because their methods are so often simple (and narrow) minded.

Another possibility, though, is that their actual goals are not what we might naively assume. Perhaps the real goal of the middle manager is to have documented, demonstrable criteria for making decisions that can be used to justify his or her continued employment. Perhaps the real goal of the Keynesian with influence over economic policy is to maintain and extend that influence, regardless of the actual effects on the market. Perhaps the real goal of your district’s congresscritter is to get reelected and indirectly improve its own economic standing — and to hell with the rest of us.

Meanwhile, if you look at the individual innovative technologist or entrepreneur, he or she is likely more concerned with making the core functions of his or her work easier, of abstracting away the drudgery, so that more of that effort can be put into solving new problems, and so that efficiency of producing either new technology in specific or new economic wealth in general is increased. As this new technology and wealth is distributed through free exchange to those who need it, more and more people are freed from having to focus their efforts on their current level of needs, and can focus them on a new, higher level — because that previous, lower level has already been met. That is, after all, what “achievement” and “success” are really all about: satisfying the current want or need, freeing one up to seek the next, higher level want or need.

The innovators who spawn the means of satisfying others’ wants and needs in the first place are in turn served by a higher level of innovator. At each higher level of the resulting hierarchy of innovators and producers, you find someone who creates the abstractions that ease the efforts of those on the next tier down, generally because they themselves wanted to break free of the drudgery of that lower level of the innovation and production hierarchy. The ability to do so is maximized by leaving such innovators and producers control over their own resources so they can be put to work by these people in the most effective manner possible toward further innovation and wealth production.

Standing outside the hierarchy of technological innovation or an economic market, punishing people for failing to use resources in a manner that fits your statistical model or rewarding them for gaming the system, in no way improves the situation. In fact, it retards true productivity.

The reason all this came to mind upon seeing Sterling’s comment about technoogy eliminating the need for brains was the recognition of a very simple fact:

The fastest way to eliminate the need for specialized knowledge workers from the set of requirements for a given task is to let them innovate in the realm of their own productivity. Truly innovative workers tend to create tools and techniques that abstract away the drudgery of their jobs, freeing them up to focus more on the work that’s harder to abstract away. The best way to help them do that is to stay the hell out of their way.

In short, keep your damned metrics-inspired “regulation” off my programming and wealth if you actually want me to be more, rather than less, productive.


  1. Well, the motives of a programmer or entrepreneur may not always be pure either — job security and the accumulation of personal wealth probably figure high on their respective lists — but that doesn’t alter the fact that regulation doesn’t help them to be more productive for the common good.

    Comment by Sterling Camden — 17 December 2008 @ 01:59

  2. Of course they will generally work for their own best interests. The point is that, in doing so, they produce new technologies and techniques that improve technology and increase wealth not only for them, but in general. That all assumes a free market, naturally — which, in turn, means that there isn’t heavy-handed regulation mucking up the works.

    Comment by apotheon — 17 December 2008 @ 02:36

  3. […] How is an economy like technology?This may help explain why so many technologists have libertarian leanings.Tags: none […]

    Pingback by Chipping the web: December 17th -- Chip’s Quips — 17 December 2008 @ 03:01

  4. Good points. Managers and legislators could be turning to proxies (such as your lines of code example) because they’re unable to see the whole picture, or grasp how exactly the process for creating valuable output works. And instead of working to improve their understanding of the process, they turn to shortcuts, proxies, rules of thumb, in order to simplify the decision making process, thereby oversimplifying the situation and losing the nuance and complexity which will impact whether their decisions will lead to outcomes which are successful or unsuccessful. In a way though, you can see that most anywhere. Similar proxies exist for how parents scrutinize their children, how teachers keep track of students’ progress, how people choose their significant others, and so on and so forth. The problem with legislators though is that their ability to wreak havoc is amplified by their positions, and the amount of taxpayer money at their disposal.

    Comment by Chris | Tech — 23 December 2008 @ 02:18

  5. I liked that statement by your friend; gives technologists a good definition and makes them stand tall. :)

    Comment by Wer — 23 December 2008 @ 04:05

  6. […] decided to take some implicit advice from my own writings, and work on things from a higher level of abstraction. It would be nice to have more stuff figured […]

    Pingback by Chad Perrin: SOB » notes for building cities — 26 January 2009 @ 05:19

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

All original content Copyright Chad Perrin: Distributed under the terms of the Open Works License