How do you use AI tools in your daily work, and when do you refuse to trust an LLM's answer?
Describe your use of AI tools by task, each with its check: reading cited lines, running code, confirming APIs against the docs for your version. Refuse any answer you cannot check cheaply or that is costly to get wrong, keep restricted data out of unapproved tools, and back it with one real catch.
Answer
Short answer
Answer by task, not by tool. Say which jobs you hand to AI tools, such as explaining unfamiliar code, drafting tests or writing a one-off query, and the check you run on each kind of output before you rely on it. Then give your rule for refusing an answer: if you cannot check it cheaply, or being wrong would be expensive or irreversible, you treat it as a lead and verify it from a primary source. Finish with what you never paste into these tools, and one real time the rule caught something.
Was this answer helpful?
Likely follow-ups
- Which of your verification checks would you drop under deadline pressure, and what is the risk of dropping it?
- A teammate merges assistant-written code they cannot explain. What do you say to them?
- How do you check an answer about a library when the documentation for your pinned version is thin?
- Your team has no policy on what can be pasted into these tools. What do you do before one exists?
Related questions
- What is the piece of work you are proudest of, and why that one?mediumAlso on judgement and behavioural5 min
- You are asked to bring a real piece of your own work and talk through it. What do you choose, and what do you have to be ready to admit?mediumAlso on behavioural5 min
- You are asked to design a product that does not exist, and the setup is missing details you would need in real life. Do you point that out?mediumAlso on behavioural5 min
- Tell me about a decision you had to make in minutes, with information you knew was incomplete.mediumAlso on judgement6 min