Computer algoruthm to search for evidence of design? IS this an idea of merit?

Question

Ok I don't believe ID is science personally and I am not on the side of those who do, but I read about about a guy who is trying to publish his data on a computer algorithm to detect whether something has been designed or if it occurs naturally. I think aside from the intelligent design nonsense this would be a pretty big computer science achievement in the field of AI, anyone else think this idea actually has merit

I completely agree with you.
The whole idea that it is obvious something is designed is not something I agree with because as humans we are conditioned to note things that are distinctly human, but would someone from another planet know a building from a mountain?

Answer 1

I see this as going about the problem backward. The way to test a theory is to make predictions based on the theory and then test to see if those predictions can be observed. If you say "I believe XYZ" and then go out in search of something that appears to support XYZ you have proved nothing.

Answer 2

A leading scientist in the ID movement has come up with a type of flow chart which when followed by answering certain questions involving familiarity and frequency of pattern determines if something is designed or not. This sounds similar to your algorithm computer search which I think would be valid because a basic scientific premise incorporated in ID is that information is a sign of intelligence which makes it an artifact. This is why the DNA code is often used as an example of a highly ordered pattern of information which suggests design.

Answer 3

The trouble is how do you decide whether the algorithm makes any sense, and isn't full of the biases of the programmer. That is, how do you create an algorithm that detects intentional modification/optimization as opposed to optimization that follows the random mutation/selection process. I think it is a not an impossible task, but I do think such an algorithm would have to be tweaked for individual classes of problems to solve. That is, there won't be any silver bullet algorithm to solve all types of queries.

So, no, this is not entirely a waste of time, but I don't think that any such algorithm can ever be trusted to be completely reliable nor can it be expected to work well over a large set of differently posed optimization queries. And, you have to have a way to ensure that the selection criteria is truly objective, and not subject to the whim of the programmer who changes the rules of the game to fit the outcomes they desire.