On: The current state of LLM-driven development

[edfletcher_t137]

The first two points directly contradict each other, too. Learning a tool should have the outcome that one is productive with it. If getting to "productive" is non-trivial, then learning the tool is non-trivial.