This is actually quite common situation and it seems that the major cause is that tech is now saturated with "normies" (liberal arts college dropouts, SJW schizoids, etc) instead of math or biology majors like it used to be at the times of Bell labs or Xerox PARC or early Google.
Abstract bullshit, misapplication of statistical methods is another cause. We are drowning in the sea of screaming narcissistic virtue-signaling bullshit.
So, go back to concrete mathematics, grounded in reality modeling, lightweight high level prototyping languages and functional programming. This, like universal principles, will never fail you.
Biology?! I'm not even sure if math is in place here. I would put instead 'CS and engineering'.
People who do things that never existed before are rarely these who are good at math. These people are called engineers. They have different skillset and background and very different attitude towards solving problems.
Don't get me wrong, being computer engineer myself I have huge respect for both biology and math. I took a biology class recently and I do read math books.
> go back to concrete mathematics
I would say: go back to making things aka engineering
You get back the joy of building things by building things. Something that worked for me when I was in similar situation was working on some Arduiono projects. It's not only programming, you are making sometehing in the physical world as well. The projects are usually simple and small, you don't get burned out. The actual programming is pretty simple - you don't need design patterns, some havy framework or functional programming to make few LEDs blink or a mottor to spin back and forth.
Building things involves problem solving. The life sciences often involve problem solving without producing an artifact. I think it is the challenge of solving hard problems that unites both pursuits.
Abstract bullshit, misapplication of statistical methods is another cause. We are drowning in the sea of screaming narcissistic virtue-signaling bullshit.
So, go back to concrete mathematics, grounded in reality modeling, lightweight high level prototyping languages and functional programming. This, like universal principles, will never fail you.