> And I thought: mate, you were already so far up the abstraction chain you didn’t even realize you were teetering on top of a wobbly Jenga tower.
But AI is different. If I program in a high level language like Python, sure I don't know what's going on under the hood. But you get a 'feel' for it because the same code usually reproduces the same results. Does AI reproduce the exact same results when I ask the same thing? That I don't know.
I was thinking the wobbly Jenga tower thing was unfair. The stack TypeScript runs on is fairly stable and buildable on. LLM output is much more random.
But AI is different. If I program in a high level language like Python, sure I don't know what's going on under the hood. But you get a 'feel' for it because the same code usually reproduces the same results. Does AI reproduce the exact same results when I ask the same thing? That I don't know.