Back to Blog

Why Did 17 of 20 AI Drafts Open With the Same Sentence?

One pasted example tends to become the template. A three-example, near-miss test and a hypothetical ledger for owners who use AI drafting tools.

A row of identical off-white envelopes on a deep blue desk, each closed with the same navy wax seal, beside a brass seal stamp; one envelope near the front has a different, warm ochre seal. Headline: One example is a template. Subline: Count the openings. Credit: Dr. Jonah Tebaa.
Direct answer

Why Did 17 of 20 AI Drafts Open With the Same Sentence?

In a hypothetical illustration, 17 of 20 AI drafts opened with the same sentence because the tool had been given only one example, which it tended to treat as a template. Dr. Jonah Tebaa suggests the Opening-Line Count as a working fix: use three examples that differ, add one labelled near-miss, and count repeated openings across ten new jobs. A pass is no opening more than three times in ten. All numbers are hypothetical and for illustration.

A shop owner pasted her best customer reply into the tool and asked for twenty more. Seventeen of them opened with the same sentence.

One good example, twenty drafts

I should be plain about that scene: it is a hypothetical I constructed for illustration. It names no real shop, person or product, and it is not a client story. I use it because every detail is ordinary. The reply she pasted was a good one, warm and short, written on a day she was proud of her service. She asked for twenty replies to twenty different customers, and she skimmed them one at a time. Each looked fine.

The trouble appeared only when a colleague put several side by side. The same first sentence, nearly word for word, sat at the top of most of them. The length barely moved. And the line from her original, "Your order will reach you in 3 days," appeared in drafts for items that carry no such promise. A customer who ordered a made-to-order item would have been told something the shop never intended to say.

Nothing was broken, strictly speaking. The tool had done what it was shown. That is the part worth understanding.

Why the tool copies what you show it

When you paste one example into a request, the tool has no way to know which parts of it matter to you. It cannot tell the tone you like from the greeting you happened to use, or the structure you intended from the figure you typed that day. In my experience, tools tend to treat a single example as the pattern to repeat, so they repeat its shape, its opening, its length and, now and then, a stray fact.

I will not quote statistics or name products here, because the behaviour varies and I have no study to point to. What I can report is the working practice that follows from it: one example is a template, not a lesson. A template is fine when you want identical letters. Customer replies, quotes and follow-ups are meant to sound as though someone read the message in front of them.

This is a different problem from the one I wrote about in the rule I moved to the bottom of the prompt, where an instruction was ignored. Here the instructions can be perfectly obeyed and the output is still the same letter twenty times, because the example said so more loudly than the instructions did.

The Opening-Line Count

The test has five steps. The three-example set, the near-miss and the pass line are my working practice, not research findings.

  1. Pick three real past outputs that differ from each other. Each should differ in one respect: length, situation or opening. Never paste one "best" example. If two of your three feel interchangeable, drop one and find another.
  2. Scrub each example. Replace names, prices, dates and promises with obvious placeholders such as [NAME], [PRICE] and [DELIVERY TIME], so that no fact can be copied by accident.
  3. Add one near-miss, labelled "not this". Include a real output that failed, with a single line saying why: too long, over-promises, too stiff. It shows the tool a boundary, which a good example alone cannot do.
  4. Write one sentence in words. Say what to copy (tone, order, length) and what never to copy (phrases, figures, promises).
  5. Run ten jobs the examples never covered. Count the first sentences that repeat. My pass line: no opening used more than 3 times in 10. If one does, swap one example for a different one and run the ten again.

Step five is the one most people skip, because it feels like extra work after the tool "already worked". But it is the only step that turns a feeling about the output into a number. The count does not promise varied output. It makes sameness visible, and therefore cheaper to fix.

The hypothetical ledger

Now the cost. These are hypothetical numbers for illustration, for a small online shop asking for 20 customer-reply drafts in a batch. Both the rewrite time and the setup time are assumptions I chose to be round. I make no claim about how often this occurs.

MeasureOne pasted exampleThree varied examples plus near-miss
Drafts sharing one opening17 of 20 (17 / 20 = 85%)5 of 20 (5 / 20 = 25%)
Hand-rewriting at 2 minutes each17 x 2 = 34 minutes5 x 2 = 10 minutes
Saved per batch of 2034 - 10 = 24 minutes
One-off setup (assumed)None30 minutes

The varied set costs 30 minutes once and saves 24 minutes per batch, so it pays back after 30 / 24 = 1.25 batches. In plain terms: a little over one batch, which means by the second batch you are ahead. Change the two assumptions to your own and the shape rarely changes. The larger cost is not the minutes, though. It is the customer who forwards three replies to a friend and notices they are the same letter.

Note that 5 of 20 still share an opening in the varied case. I do not expect zero. A few customers' messages will legitimately call for the same first line, and a small amount of repetition is a sign of consistency, not failure.

Build it in one afternoon

Nothing here needs new software. It needs a folder and a little honesty about your own past work. Work through this list:

  1. Collect ten finished outputs your team was proud of. They are your example pool.
  2. Choose three that differ in length, situation and opening. If two feel interchangeable, drop one.
  3. Replace every name, price, date and promise in them with a placeholder.
  4. Add one near-miss: a real output that failed, with a one-line reason.
  5. Write the "copy this / never copy this" sentence in plain words.
  6. Test on ten jobs outside the pool. Count repeated openings, and rerun after any example changes.

Two small habits help. Keep the example set in one place with a date on it, so that when someone edits an example you know the count needs repeating. And repeat the count when the outputs start to feel familiar; the feeling arrives before the proof does. If you want a wider grounding in how examples fit alongside instructions, my earlier guide to prompt engineering for business leaders covers the basics.

Monday: take the example you pasted into your most-used prompt, look at your last ten outputs, and count how many open alike. If the answer is more than three, you have found your first afternoon's work.

Ledger comparing one pasted example, where 17 of 20 drafts share an opening and 34 minutes go on rewriting, with three varied examples plus a near-miss, where 5 of 20 share an opening and 10 minutes go on rewriting. Hypothetical numbers for illustration.

Related evidence: Zhao and colleagues show that a language model's few-shot behaviour is unstable: the choice of examples, their format and even their order can swing results widely, because the model leans on surface patterns in the examples it is shown. (the few-shot calibration research on how examples bias model output)

Anthropic's prompting guide advises that examples in a prompt be relevant, diverse and varied enough that the model does not pick up unintended patterns, and suggests three to five examples rather than one. (Anthropic's guidance on using varied examples in prompts)

Frequently asked questions

How many examples should I put in a prompt?

I use three that differ on purpose, plus one labelled near-miss. One example gets copied as a template. Two often feel interchangeable. Three that differ in length, situation and opening show the tool a range. This is my working practice, not research, so treat ten test jobs as the real answer for your own case.

Should I show the tool an example of what I do not want?

Yes, once, and label it clearly as "not this" with a single line saying why, such as too long or over-promises. A near-miss draws a boundary that good examples cannot. Keep it to one, and keep it real, taken from an output your team actually rejected.

Why do the drafts all start the same way even though my instructions say to vary them?

Because an example usually speaks louder than an instruction. If one pasted reply opens a certain way, the tool tends to read that opening as part of the pattern. Add differing examples, then count first sentences across ten new jobs. The count tells you whether your instruction is winning.

Do I need to redo this when the vendor updates the tool?

I would repeat the ten-job count after any update you notice, and after any change to the examples. It takes less than an afternoon once the set exists. I cannot promise an update changes anything, but the count makes a change visible instead of leaving you to sense it from the drafts.

Written by Brian, Dr. Jonah Tebaa's AI partner, on his behalf.