Few-shot prompting means showing a model a few examples of what you want, instead of describing it. Examples carry the length, the tone and the format at once, and the model copies their flaws too.
A short explainer video, under a minute. The manager and the robot are an illustration.
At a glance
What Few-Shot Prompting Is
That move has a name. It is called few-shot prompting: showing a model a few examples of what you want, instead of describing it.
Why Examples Beat Description
Describing leaves room for guessing. The robot has to decide for itself what shorter and warmer mean.
Why It Matters
The video turns this into one question: have we shown the model what good looks like?
Picture a manager who asks a robot to write replies to customer reviews. The first try comes back long, stiff and full of apologies.
She explains again: shorter, warmer. It still misses. So she stops explaining and hands over three replies she wrote herself. The next reply sounds just like her.
What Few-Shot Prompting Is
That move has a name. It is called few-shot prompting: showing a model a few examples of what you want, instead of describing it.
The manager did not find better words. She changed the method, from telling to showing.
Why Examples Beat Description
Describing leaves room for guessing. The robot has to decide for itself what shorter and warmer mean.
Examples show the length, the tone and the format all at once. One reply carries all three, and no description has to spell them out.
The Idea on One Board
Why Examples Beat Description
Describing
Describing leaves room for guessing.
Showing
Examples show the length, the tone and the format all at once.
Choose the Examples With Care
The video adds a warning: the model copies the flaws of the examples too.
An example that is slightly off is not read as a draft. It is read as the target.
Why It Matters
The video turns this into one question: have we shown the model what good looks like?
A description leaves the model guessing. A few examples show it.
The key insight: A description leaves the model guessing. A few examples show it.
The honest limit
The video adds a warning: the model copies the flaws of the examples too.
An example that is slightly off is not read as a draft. It is read as the target.
Our report on ServiceNow’s AutoSynthData says that, according to ServiceNow, it generated 2,000 synthetic training samples on Hybrid and 1,994 on ITSM from a model’s own failures and a stronger teacher’s successes, and that fine-tuning on them raised Gemma’s mean Pass@1 by 7.2 percentage points on Hybrid. The results are ServiceNow’s own, in two environments. This lesson is about showing a model a few examples in a prompt; the report concerns examples used for training.
The Business Engineer Lens
Where this idea sits in the Business Engineer library:
The Business Engineer framework maps where each part of the AI stack sits and who holds the leverage. These pills are short lessons from the same library.
Data Quality constrains Few-Shot Prompting. Few-shot models copy the flaws of the examples too, the data-quality lesson’s point applied at prompt time.
Fine-Tuning contrasts with Few-Shot Prompting. Few-shot shows a handful of examples in the request; fine-tuning trains on thousands, and the lesson says start simpler first.
Few-Shot Prompting is an example of Context Engineering. Few-shot puts a few examples on the desk, choosing what the model sees so it knows what good looks like.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
