I invested a great deal of effort over 30+ years to learn biology, which I started to love in high school when a teacher introduced us to molecular biology. Over time I've come to appreciate that biology is a huge field and people who master one area often know little to nothing about many others.

To be proficient in biology you need to have "Extra" skills: extra ability to work with ambiguity,ability to memorize enormous amounts of descriptive information, and highly abstract representations. Digital biology often loses many aspects of biological reality, and then fails to make useful predictions.

Over the years, I've come to realize I know less and less about biology- that I greatly underestimated the complexity and subtlety of biological processes, and have come to admit that my own intelligence is too limited to work on some problems that I originally thought would be "easy engineering problems".

A great example of the rabbit hole that is modern biology is summed up here: what is the nature of junk DNA? To what extents are digital readouts like ENCODE representative of true biology, rather than just measuring noise? What is the nature of gene and protein evolution?

https://www.cell.com/current-biology/fulltext/S0960-9822(12)... (note that while I disagree strongly with Eddy in many ways, I've come to recognize that I simply don't understand the modern view of evolution outside the perspective of molecular biology (IE, what geneticists like Eddy think).

Also, recently, Demis Hassabis postulated that if he is successful, we will come up with silver bullet cures in 10 years time simply using machine learning. It's amazing how many computer scientists (I call him that rather than a biologist, although he has worked into neuro) make this conclusion.