I fully agree: over time, we will view coding LLM agents as a form of compiler, and human effort will shift upstream to the development of specifications, the writing and execution of test cases, and final validation.
I'm not sure about this one, though. Today, my experience with LLMs is that they're already pretty helpful for unit test edge cases I wouldn't have thought of. If you mean broader functional tests then maybe, but I wouldn't say the same for unit tests.
Yes but in my experience, LLMs are prone to cheating and implementing "shortcuts" that look functional but are in fact not really if you look closely. At the moment, you still need a human in the loop to make sure the LLM actually did what it promised and didn't cheat.
I fully agree: over time, we will view coding LLM agents as a form of compiler, and human effort will shift upstream to the development of specifications, the writing and execution of test cases, and final validation.
> the writing and execution of test cases
I'm not sure about this one, though. Today, my experience with LLMs is that they're already pretty helpful for unit test edge cases I wouldn't have thought of. If you mean broader functional tests then maybe, but I wouldn't say the same for unit tests.
Yes but in my experience, LLMs are prone to cheating and implementing "shortcuts" that look functional but are in fact not really if you look closely. At the moment, you still need a human in the loop to make sure the LLM actually did what it promised and didn't cheat.