On Secrets

I have a B.A. in Economics. I should know game theory beyond the minimax algorithm. But I don't.

If I did understand games I could lay out a solid argument for the thesis that secrets are necessary and beneficial for society, and maybe even provide some rigorously proven stylized example game. Instead, I'll have to suffice with common sense persuasion.

Like virtually every academic, I believe in scientific method of publishing and testing ideas and observations, in order to extract the best knowledge available to mankind. If this works for testing ideas about nature, why shouldn't it be sort of universal? Why shouldn't companies, governments and individuals be open about behaviour as well? If no one had any secrets, society would definitely be more rational. And rationality would provide enormous benefits to us all.

Unfortunately, this is absolutely true, absolutely desirable and yet absolutely unachievable. We'll never get there. Probability of someone cheating, no matter how small, will make this an impossible utopia. As long as someone can cheat AND win the honest player will not only lose, but will lose as a fool.

Now, secrets are the most efficient instrument to protect yourself against both known and unknown malevolent adversaries. When information is not open, you can effectively be 100% sure that this information will not be used against you. And even if there is a negligible probability of someone actually using your secrets against you, there is not necessarily any direct benefits in sharing your secrets either.

An economically relevant matter of sharing secrets is the popularity of Apache or BSD licenses. These form the cornerstone of commercial open source, largely because they do not impose copyleft. Question is, are individual developers just useful idiots if they put their efforts in non-copyleft open source projects?

So far, experience suggests that at least the most successful projects have passed the "useful idiots" danger zone and established themselves as public goods supported by benevolent corporate citizens. For the rest, I don't know. One can argue that even closed-source startups usually end up doing market research for the big players, so what is the difference?