Comment by nabla9

The problem with numerical programming is that all data is the same type. If you access wrong row, multiply across wrong axis, function uses wrong parameters, it's semantically wrong, but it can still work and system can learn something.