Airs - Ian Lance Taylor

A couple of times I’ve alluded to an idea I first saw from Peter Singer, that the structure of society creates a great deal of the value which flows to the wealthy. Our legal system and our stable system of property ownership is crucial for building successful companies. Yet the value gained from companies flows largely to company management and partially to dividends to shareholders. None of goes to the operations which support the company: the courts, the police, the schools which train the workers, etc. Instead, we tax profits, and use the tax money to pay for the infrastructure.

This naturally suggests that we could arrange matters differently. Instead of using taxes to pay for the courts, companies could pay the courts directly. And likewise for other social services. This would match my understanding of how libertarians think society should work. It would bring the benefits of free competition to the court system: the best courts would get the most money.

Of course, there is an obvious so-called moral hazard: if you choose whether or not to pay the fire brigade, then if you chose not to pay, the fire brigade has a pretty strong incentive to burn down your house as an example to encourage others to pay. And if you pay the court system, then the judges have a very strong incentive to issue findings which favor whoever pays them the most.

Also, if different court systems compete, then how do you pick which court system to use when two parties in a dispute pay for different ones? While this could be negotiated, in extreme cases the winner will be the most powerful party, which will normally be the richest party.

So as far as I can tell such a system degenerates reasonably quickly into control by the wealthiest. Of course they would normally hide their power in a velvet glove, but it would be there when needed.

Is that any different than our system? I think there is a difference, which is that we can vote people out of power. Voting does not require a great deal of commitment, it just requires making a decision. So we have a way of derailing powerful people, a way which would not be available if societal infrastructure were privatized.

Of course there are many very powerful people in our society who are not elected. But even those people can be controlled by the masses, via the court system. It doesn’t happen all that often–the lawsuits which followed the stock market crash are over now–but it does happen, and it does constrain people to obey the rules to some degree.

The opposite of privatizing societal infrastructure is nationalizing private property. The U.S. does this in a very limited way via eminent domain. It has been done effectively on a larger scale in countries like South Africa. But too much privatization leads to a state socialist system like the U.S.S.R., which we could see was a disaster.

So there is some sort of balance to strike, or perhaps a third way to find, in the ideal society.

The New York Times Sunday Magazine ran a short piece by Jim Holt on the idea of universal consciousness. The idea is that consciousness can be found everywhere in the universe, even in rocks.

This is based on the premise that “physical properties alone cannot account for subjectivity. (How could the ineffable experience of tasting a strawberry ever arise from the equations of physics?)” The argument is logical: if we accept that premise, and we accept materialism, then we pretty much have to conclude that consciousness must be in everything.

Now, if this were a sane world, that would be a nice reductio ad absurdum, demonstrating that the premise is false. Yet somehow sensible people are so attached to that premise that they are willing to believe in conscious rocks rather than give it up.

Astonishing stuff. Some people need to apply a little more reasonable thought and a little less ineffability.

Larry Niven’s classic science fiction short story “All the Myriad Ways” points out a problem with the Many-Worlds Interpretation of quantum mechanics: if there is no collapse of the quantum wave function, then all possibilities actually happen. That would seem to imply that we make all possible personal choices, including the really bad ones. This makes a mockery of our notions of free will, even worse than determinism: if we make all choices, then we really don’t exercise our will at all.

There is an obvious flaw in this argument: there is no particular reason to think that uncertainty at the quantum level leads to our making different choices. Most quantum events actually have no effect at the macroscopic level at which we live. If one entire molecule of oxygen tunnels across a room–an event which will occur at a vanishingly smal probability–nothing will change at the macroscopic level. And even if a bunch of molecules in my brain switch places, that doesn’t mean that I will change my mind about something.

Our actions are constrained by who we are. Quantum activity could conceivably make us into different people. It can’t make us be the same person who made a different decision.

Still, quantum mechanics in general, and the Many Worlds Interpretation in particular, pose some real philosophical problems. Unfortunately they don’t pose the problems addressed by science fiction novels about parallel worlds. The problems are more along the lines of what it is like to live in a world in which low probability events are not only possible but actually happen all the time.

Iain M. Banks, in his novel “The Algebraist,” argued that the only universal religion is one in which we assume that our universe is actually being simulated by an incomprehensibly large computer, and that our only meaningful goal is to try to reach the creator by proving that it really is a simulation. That seems as good a reason as any for why we never see low probability events in macroscopic universe: because the computer doesn’t bother to simulate them.

Can we ever have a science of ethics? That is, some way to determine how people ought to live, and some way to demonstrate its truth? And can there ever be some way to persuade people to live according to its guidelines? Philosophers and prophets have proclaimed the truth many times, and in fact their truths tend to be broadly similar. But clearly all people don’t follow these proclamations.

That said, ethics has changed over the millenia. Slavery is an obvious example: it was once normal, and is now considered to be wrong. Although slavery has not disappeared, the people who engage it now believe they are doing something wrong, which was certainly not the case 200 years ago. I think of this as an expansion in what it means to be human, or to be taken seriously: once only the people in the tribe mattered, now all humans matter.

Could we ever backslide from our current position on slavery? Clearly anything could happen if civilization collapses, but, short of that, I think not. I don’t think there are fashions in ethics, I think there are long term movements.

Back to a science of ethics. I think the steps are to define what a good society looks like, and then figure out the rules to get there. Of course this is an iterative process. An unachievable good society is useless, as are rules that nobody will follow. We don’t know the good society, though we know plenty of bad ones, so we may need to experiment.

How can we ethically run an experiment on developing ethics? It’s impossible today, but it is now possible to imagine such an experiment. It could be run in a detailed simulation on a computer. This would require something close to artificial intelligence, and it would require many different interacting agents. We could apply certain rules and see if the resulting society seems good. We could let the agents develop their own society. We could compare the results to real societies as a cross-check.

I don’t know whether this would lead to anything useful. But it’s definitely worth trying. It could be one of the best uses of artificial intelligence.

The only people who seriously study ethics today are philosophers. In general I think philosophy is the study of things which we don’t understand. Once philosophy covered all knowledge; now it just covers a small subset. The goal of philosophy should always be to eliminate itself, by understanding the subject sufficiently that it can turn into its own discipline.

Will there ever be an end to philosophy? That is a good question to ask a philosopher.

As we all know, Plato argued that the objects we see in the world are just instantiations and specializations of abstract template classes. Any given chair is just an instance of the abstract class Chair. The existence of the abstract class is what permits us to identify the chair as a Chair. The very fact that we can talk about chairs, and that we can tell whether an object is a chair or not, means that Chair exists.

Plato also applied this to attributes and relations. For example, the fact that one chair is bigger than another is an instantiation of Bigger. I once questioned a philosophy professor about this. She pointed out that before there were any people, we still tend to think that one star was bigger than another. Therefore, the concept of Bigger must exist independently of people, and therefore it must really exist in some sort of abstract space.

I don’t really believe in any of this, myself. But it is true that we predisposed by nature to see certain types of objects (e.g., faces), and thus we are born with some sort of abstract concepts in our heads. (Plato (or Socrates) of course argued that pretty much everything we know we know from birth (in the Meno dialogue). I don’t believe that, but it seems to be true that we do know some things from birth.)

Anyhow, this is all background for this: does the fact that we naturally carry abstract concepts in our heads mean that object oriented programming is more natural than functional programming? That would be consonant with my earlier arguments about how parallel programming is somehow inimical to our thought processes.