Skip navigation

Monthly Archives: October 2009

OK, so we should be all familiar with Paley’s Watchmaker argument. That is, you see a watch in a field, then you know someone made it. Ergo, you see some critter, you know someone made that as well. Praise be to God, and so on.

We all know that evolution through natural selection etc can adequately explain the diversity of life. So that must mean that Paley’s argument is wrong, right? So if we see a watch in a field, that watch must have evolved!

Well, that’s the basis of memetics: take things that we know have been designed, and claim that they have instead evolved. In the paper that started memetics off, Viruses of the Mind, Dawkins does this quite explicitly when he uses computer viruses.

This strikes me as completely bizarre. As Dawkins is fond of saying, evolution through natural selection etc. gives us the appearance of design, without requiring a designer. However, to take his example, computer viruses do have designers, for they are computer programs and such programs need human authors. Therefore we do not need clever theories to explain their properties (beyond the mechanistic properties of those programs), we can just point to some hacker and say “there’s your cause.”

Advertisements

This is a common defence of the patent system, but I believe the reasoning behind it is erroneous.

I can understand why such a rationale would be put forward. After all, the converse, that a system of state-backed monopolies would help the formation of… um… monopolies and therefore not help the little guy, would not be a good selling point. The rhetoric around this has become quite sophisticated, given that there has been a lot of time and effort expended into making the arguments as emotionally appealing as possible, which can make it feel somewhat hard to counter. I shall give a fairly crude example of one such argument, nonetheless. One should understand that my counter arguments apply to more sophisticated versions of the same argument, however.

2 man start up spends a few years generating brilliant product x. It uses new and clever ways of doing things. A big company comes along, steal the idea, puts a 300 man team on it for a few months and gets to market first.

Link

As you might be able to gather if you click the link, I have already responded. There I gave some reasons why such a rationale would not normally apply.

So, there are some problems with this statement as it applies to the real world.

Most forms of technological development are incremental. While there are revolutionary, “game changing” inventions, most build upon previous work. Most of the game changers are themselves accretions of other technologies — think of the internal combustion engine, a synthesis of atomisers (from perfume), swamp gas detectors (a useless invention) and pistons from steam. A new entrant with some fabulous invention is therefore quite likely to infringe in some way on existing inventions. Furthermore, incremental inventions themselves open up new opportunities. High-pressure steam engines were a requirement for some of the more important developments of the industrial revolution, and they were held back by at least ten years by patents. This is analogous to saying that land ownership is available to all at no cost because of homesteading, which is rendered null and void when all land is owned. If one is a new entrant in a field that is swamped with patents, then one is at the whim of the incumbents, and so patents help the big guy.

There is the issue I brought up in my response above. That is, the large incumbent company is seriously outnumbered by the small challengers, and so it would be prohibitive for them to imitate all the challengers’ inventions. They must therefore wait for a particular invention to be successful before it’s worth imitating, at which point it is too late, because unless they are a heavily entrenched monopolist (for instance, they have been granted lots of patents), the new entrant has already carved out a market for themselves.

In reality, small companies do not benefit from patents. Preventing competition does not help the competitor who has the advantage. It helps the incumbent, who has run out of ideas.

Scheme’s DO is generally treated with some disdain. Yet I often find myself recreating it with named LET.

So,


(let next ([i 0] [res '()])
  (if (= i 4)
    res
    (next (+ i 1) (cons i res))))

Is equivalent to:

(do ([i 0 (+ i 1)]
     [res '() (cons i res)])
  [(= i 4) res])