[Red Badger]

It just looks for patterns and tries to correlate those patterns with known areas of genetic responsibility.

Much more efficient at it than a human is............

[Robert DeLong]

But it can only work with what the human has provided to it has the inputted knowledge base. If that required assumptions, which is what I gathered from the article that assumptions were necessary, and those assumptions are off, then it may take years to come to the realization that the assumptions were off.

Like I said, it's an interesting endeavor. But it will still require a human brain to analyze what the computer spits out for accuracy. Much like the disclaimer of AI, or at least the ones I have casually used, that instructs the user that they should verify AI's results. 😋

However, that said it may still lead to a quicker improvements to the application. But it could also muddy the water just as easily.

The problem with using a machine, is that it tends to degrade the human mind. It also tends to create a dependency upon the machine.