IMO it's not about the tests as an outcome it's about the process to get to a solution.
I use tests a discovery tool. I often don't know how the solution will look like but I most certainly always know how to start to write a test for it. It's dictated by my requirements/objectives.
My tests evolve as I discover the solution. A running tests suite is a side effect of this process.
On the other hand, if I don't have clear objectives/requirement and the goal is just gather knowledge, I don't bother with tests unless they happen to be the most convenient workbench compared to a simple script or a throwaway app.
SoFi needs more test cases, IMO, I'm writing tests for @lenster.lens almost every day and making sure nothing should break 🤞
If you can write a useful test, you should (regardless of industry). If you’re writing a test to check a box for your manager or inflate a coverage perecentage somewhere, your probably shouldn’t bother. Unfortunately, the latter seems much more common.
Haha! Sometimes you just have to accept that you will lose some money discovering bugs! You can’t test what you don’t know needs testing right!