Few-shot / Zero-shot
Zero-shot - you ask without examples. Few-shot - you give the model 2-5 examples of “input → output”, and it copies the pattern. The simplest way to improve quality.
Zero-shot - you pose a problem without examples and hope that the model understands. Few-shot - you attach several examples in the format “here is the input, here is the correct output,” and the model copies the pattern.
Few-shot is the most underrated technique among marketers. Do you want AI to write headlines in your style? Don't explain the style in words - show 3-5 of your best headlines. The model captures the tone more accurately than any description.
My rule: if the zero-shot gives something wrong, don’t rewrite the instructions ten times, add examples. Two or three good examples usually solve the problem faster than a paragraph of explanation.
Related terms
Where is it understood in practice?
Need to set this up on your project?
I analyze metrics, calculate unit economics and collect funnels on real budgets. 30 minutes on call - free.