> If I have any doubt that someone might not complete a task, or at least accurately explain why it's proving difficult, with at least 95% certainty, I won't assign them the task
It gets hard to compare AI to humans. You can ask the AI to do things you would never ask a human to do, like retry 1000 times until it works, or assign 20 agents to the same problem with slightly different prompts. Or re-do the entire thing with different aesthetics.
No doubt, I'm just saying that working with a coding agent is not even remotely similar to being a team lead. If a member of your team can't complete a task and can't accurately explain what the difficulty is, you're in trouble.
It gets hard to compare AI to humans. You can ask the AI to do things you would never ask a human to do, like retry 1000 times until it works, or assign 20 agents to the same problem with slightly different prompts. Or re-do the entire thing with different aesthetics.