So I'm currently doing a data science program. I'm torn on what my capstone should be. This is a definite consideration. I could also do stuff with race, or with alleles and intelligence. Or use it as an excuse to do something for my employer. I'd really like to create something where you give it your dna and the dna of another person, and it determines the average predicted intelligence of your offspring and also the bottom and top 20 percentiles. That could be way over my head. But I heard recently that climate scientists weren't accounting for things like cosmic rays that can increase cloud nucleation, and every factor of solar variance, and processes that have delay. One of the nice things about some of these new data science tools is that make it stupid easy to include all of these variable. You just feed all the variables in and let it train, and you can see which factors actually matter and it takes care of combinitory effects and all that jazz with no real effort on the part of the person using it. I could do an all data inclusive neural net and see what comes out of it. It's in some ways less scientific because the results are less interpretable than doing everything manually and doesn't explain the why, but if you train once with a ton of cosmic factors included, and once without them, and you find that CO2 is way less predictive when you included the additional factors then at least it tells you they made an oopsie and should have controlled for some of those things.