Imagine two models have the same architecture, but one model is able to solve some question and the other one is not. The only difference between these two is going to be that the one model has the neurons for it and the other doesn’t. They are just not configured that way. And it’s one of the easiest problems to solve. You just have to train it a bit, and now the other model can also solve it. I think the same is with humans. We all have something we can’t do. It’s just because we don’t have the neural pathway made for it. With a bit of training, you can do it all.

go train your noggin

if its not breaking laws of physics you can do it