When Not to Use AI

AI is a tool, and like any tool it is wrong for some jobs. Reaching for it by default leads to features that are slower, costlier, and less reliable than a simple rule would have been.

Skip it when the rule is clear

If a task has a definite right answer that a simple condition can produce, write the condition. A model adds cost and uncertainty where none is needed.

Be careful where mistakes are costly

• Avoid AI for actions that are hard to reverse without review.
• Do not use it where a wrong answer has legal or safety weight.
• Question it when users expect exact, consistent results.

Use it where judgement helps

AI earns its place on fuzzy tasks: summarising, drafting, suggesting. Match the tool to the problem, and do not let the hype pick for you.