I think PG had an article about this: if you have any process with a) a distribution of skill which b) scales with addition of resources, market price at the top of the distribution should explode because talent becomes a license to print money. Banking only got to that first because they have a very simple model: decisions for many types of assets take equal difficulty to get right regardless of how much money is stake (below a ceiling in the hundreds of millions or billions). If you can reliably make the average + 2% trading $2,000 of Coca Cola at a time (and you can't, but work with me here), then you can reliably make the average + 2% trading $200 million of Coca Cola. Comparing your wages to that of someone who trades hours for dollars no longer makes sense.
Quants had a similar compensation scheme for quite a while now. I think other classes programmers are going to hit on it, too: there are many things you can do which demonstrably improve revenues or cut costs and which scale to the moon.
Can I again toot the horn for A/B testing, metrics, etc etc? My homework for tomorrow is writing a consulting proposal. They want to know how much purchasing my services stands to make them. Consider a hypothetical company with 9 figures of revenue a year which does no A/B testing. Pick a number for what you think you could do for a first cut of conversion optimization. 2%? 5%? 10%? Do the multiplication.
You can create staggering amounts of wealth these days with programming and a few other tricks, and you can capture quite a bit of it.
The relevant bit: To get rich you need to get yourself in a situation with two things, measurement and leverage. You need to be in a position where your performance can be measured, or there is no way to get paid more by doing more. And you have to have leverage, in the sense that the decisions you make have a big effect.]
There is nothing wrong with the model PG puts forward, as a start, but it is a simple model and it needs explanation. At the least, you should make some attempt to explain why things change over time. Consider radio in the USA, up until the mid 1950s. It was eventually replaced by TV, but in the 30s, 40s and 50s it was the main medium of mass communication. The nation lived by its radio. The potential audience for a radio star was greater then than it is now. So why do top stars, like Rush Limbaugh and Howard Stern earn such large fees now, when the market is so fragmented?
Changes in the tax code are important to consider. As a thought experiment, imagine if the top income tax rate was still 91%. (During the 30s, 40s, 50s and early 60s, it was 91% on income over $100,000, about a $1 million in today's money.) How many stars would ask for income in excess of $1 million if they knew it would be taxed at 91%? They would probably instead ask for various perks, which was the trend in the US during the mid-20th century. Asking for perks instead of cash remains the trend in those nations with strongly progressive tax rates, for instance, Sweden.
So why do top stars, like Rush Limbaugh and Howard Stern earn such large fees now, when the market is so fragmented?
I did a double-take here, because it sounds fairly obvious to me: monopsonies can get away with paying you next-to-nothing because your best alternative to "next-to-nothing" is "nothing". If there are only a few radio networks (and/or they have a gentlemen's agreement with respect to hiring talent), then regardless of skill the commercial value of your ability is zero unless you sign with them. Rush Limbaugh would be making $20k a year if there was only one radio station in town. There are now 50 and the one that gets to sign him makes a substantial amount of money. (He tapped an unlikely niche market for news and opinion: half the country.) The market now favors him rather than favoring them.
Edited to add: By the way, it appears that Google et al have a gentleman's agreement with respect to hiring talent.
Making wealth is not the only way to get rich. For most of human history it has not even been the most common. Until a few centuries ago, the main sources of wealth were mines, slaves and serfs, land, and cattle, and the only ways to acquire these rapidly were by inheritance, marriage, conquest, or confiscation. Naturally wealth had a bad reputation.
Strongly implying that in the modern world most large fortunes come from making wealth, but presenting no actual evidence. From observation, most large fortunes come from trickery: either in markets or by being hired as a CEO and negotiating high, non-measured compensation by abusing conflicts of interest. I haven't studied the question carefully, though. Has anyone?
He also strawmans the pure-pie model. The real pie model looks something like this. A very large group of people collaborate to create something of value (even in software, there are library authors, hardware makers, ISP maintainers...). Some are part of the same company. Some are tied in via contracts, shaped by negotiators. Once the good is made, the wealth is divided up among the team. (Simultaneously, the team works on the next hunk of wealth -- it's a pipeline, not an alternator.) It is meaningful to look at the distribution phase in isolation, with the caveat that how the distribution phase impacts the production phase is a factor in how good the distribution model is.
Quants had a similar compensation scheme for quite a while now. I think other classes programmers are going to hit on it, too: there are many things you can do which demonstrably improve revenues or cut costs and which scale to the moon.
Can I again toot the horn for A/B testing, metrics, etc etc? My homework for tomorrow is writing a consulting proposal. They want to know how much purchasing my services stands to make them. Consider a hypothetical company with 9 figures of revenue a year which does no A/B testing. Pick a number for what you think you could do for a first cut of conversion optimization. 2%? 5%? 10%? Do the multiplication.
You can create staggering amounts of wealth these days with programming and a few other tricks, and you can capture quite a bit of it.
[Edited to add: PG link: http://www.paulgraham.com/wealth.html My favorite essay of his, incidentally.
The relevant bit: To get rich you need to get yourself in a situation with two things, measurement and leverage. You need to be in a position where your performance can be measured, or there is no way to get paid more by doing more. And you have to have leverage, in the sense that the decisions you make have a big effect.]