A New Model for Public Policy
I was just talking to someone who is majoring in International Relations. He told me that IR was a very customizable major that allowed people to focus on different areas, like economics or public policy, “which is basically a lot of history” he said. I wanted to know what he meant by that.
“Well, when you study Public Policy, you study the history of different policies. So you end up learning what worked and what didn’t.”
Somehow, this isn’t enough for me. I like history, don’t get me wrong. I think it is important to study the past, but I think we can do better when it comes to public policy. I want to create something like Pandora.
“We ended up assembling literally hundreds of musical attributes or “genes” into a very large Music Genome. Taken together these genes capture the unique and magical musical identity of a song – everything from melody, harmony and rhythm, to instrumentation, orchestration, arrangement, lyrics, and of course the rich world of singing and vocal harmony. It’s not about what a band looks like, or what genre they supposedly belong to, or about who buys their records – it’s about what each individual song sounds like.”
With this system in place, Pandora is able to play music that you like, even if it is an artist you have never heard of. With each song and artist broken down into bits and pieces, you find yourself enjoying brand new music you normally would not have listened to. And you can refine its system by telling it what songs you like, and what songs you don’t.
So what I want to do is create a model like this for the public policy, within international relations. Break down countries into “genes” and use them to predict optimal public policy decisions. The genes would include:
- Style of government
- Strength of science and technological development
- Economic system
- Media influence
This is a really short list. But having a set of these “genes” for any give time period, we could start feeding it different scenarios and having it predict the best policy. Best would be defined as maximizing certain areas like “GNP” or “International Stability”. By testing historical cases, we could refine the system so that it would get better and better at predicting what a certain course of action would lead to.
Sure, this would leave out a lot of things. Freak accidents. Data unaccounted for by the program. But once those things happened, we could incorporate it into the system and make it even better. I think that with the advances in evolutionary algorithms, this program could rapidly improve itself to a point where it could be of use to decision makers.
Because what do people who decide public policy really do? They take in as much information as they can and try to determine what is the best course of action. I don’t this is implausible at all. And I think it would make governments much more open about their real objectives, because people could see what the maximizers were. Open flow of information make democracies better.
So yeah, a CS major and an IR major need to get together and make it happen.