Feedback Cycle
One of the main differences between compiled and interpreted languages is the length of the feedback cycle. In compiled languages, you have to wait quite a bit, whereas for interpreted langauges, you can see the results quickly, especially if it has REPL. Thinking about feedback cycles in other activities, I see that one I’m good at, humor, has a very short feedback cycle. Either they laugh, or they don’t.1 In this way it’s like reinforcement learning, where either you got the reward (+1) or you didn’t (0). This is a stark contrast to writing, which can be seen by breaking down the categories where I do write. For work, though I may get praise for writing something, it’s unclear whether the praise is a result of good writing, or more for writing something at all. In any case, unless there is a major flaw people are unlikely to point out mistakes. For this blog, right now I’m writing into the void, so by definition there isn’t any feedback. I think to solve this, I need to either change my content, my distribution, or both. Even in the future though, it’s unclear to me what good writing might mean. That is, something may attract readers, but that doesn’t necessarily mean it’s good writing. I’m not sure what to do about this, but I think either some changes, or this blog is discontinued.
For me humor is done in a synchronous manner, usually face-to-face