One thing I’ve personally seen is software engineers with an interest in deep learning use it to solve very simple problems that just need a linear statistical

One thing I’ve personally seen is software engineers with an interest in deep learning use it to solve very simple problems that just need a linear statistical model. That’s a risk you take, and one reason “gatekeeping” happens.