This is Noah Gift's Coding Blog. I only talk about coding and technical stuff here, and that is mostly Python, although I will mix in some other languages, and talk about Artificial Intelligence.

Tuesday, November 18, 2008

Religion: An API To The Brain?

I have been thinking a lot about publicly exposed API's to the brain, in the context of AI programming. It seems like one of the more interesting APIs to the brain is religion. At PyWorks, there was a talk on AI, in which, hunter/prey behavior was simulated. I wonder if it would be possible to simulate religious behavior in much the same way? The specific behavior would be a pastor/flock behavior.

It seems like if we could simulate religious behavior, then it might allow us to gain insights into how to create artificial life that follows that same API. How could that look in Python?


from brain.religion import PublicApi
from brain import emotion, rationality, memory

p = PublicApi(denomination="MyChurch", scale=10, **kw)



One of the more fascinating parts would be to study the actual brain scans of deeply religious and non religious people while they talk about politics, and other hot issues. This research could help to identify what regions of the brain are firing, and why, and thus help software engineers create a similar API with a future artificial life form.

2 comments:

Guido van Rossum said...

Brain scans are so primitive when it comes to higher-level thinking processes. I will (once again) compare them to connecting an oscilloscope to a computer's memory. It may be a great tool when debugging parts of the BIOS, but it's not so great when your compiler is generating faulty object code.

We should reverse-engineer the software that's running on the brain!

Noah Gift said...

In future, there may be a device that could detect higher-level thinking processes in a more accurate manner. I almost want to write a science fiction book about it. What if we could detect when a thought was purely based out of emotion? A person could where this device when buying a house, or a car, and it could alert the user to the fact that they were potentially making a poor decision.

In terms of reverse-engineering the brain, I do think religion could be one of the more fascinating things to start with. Creating AI simulations of religious behavior has the potential to be more rewarding than hunter/prey, due to the fact that animals don't have a religion that we can observe.

Obviously, hunter and prey evolve to get better at what they do, or they die. What about religious behavior, in particular, could modeled, and then mimicked in software? I guess I see a benefit, in both creating a software implementation of the religious API, and in both understanding how to tap into the actual religious API that appears to exist in humans.

I think a true understanding of the brain, and concurrent application of that understanding to the field of Artificial Intelligence, could be this centuries next revolution. At least I hope.