Today's AI Tip
← All posts

I took the bullets out of my prompt. AI still used bullets.

· View on Instagram · View on Facebook

I took the bullets out of my prompt. AI still used bullets.
A tip worth testing. A common tip says that if you don't want bullets back, don't write your prompt as bullets. Anthropic's Claude guide says removing markdown from a prompt can reduce markdown in the output. So I tested it.

A tip worth testing

A common tip says that if you don't want bullets back, don't write your prompt as bullets. Anthropic's Claude guide says removing markdown from a prompt can reduce markdown in the output. So I tested it.

Source: Anthropic, Claude API docs: Prompting best practices

The plain-paragraph prompt. Same five facts and the same 120-word limit each time. This is the version with no bullets anywhere in the prompt.

The plain-paragraph prompt

Write a message to my team about our new expense rules. Claims are now due by the 5th of each month instead of the 10th, and receipts must be uploaded as photos in the expense app. Late claims get paid in the next month’s cycle. The change starts November 1, and questions go to Priya in Finance. Keep it under 120 words.

Tested in GPT-5.5 (OpenAI API), Oct 2026

Same five facts and the same 120-word limit each time. This is the version with no bullets anywhere in the prompt.

Plain sentences in, bullets out. The model saw a short policy update and reached for a list anyway. This prompt gave bullets in 3 of 4 runs.

Plain sentences in, bullets out

Real output · GPT-5.5 (OpenAI API) · Oct 2026

Team, Starting November 1, our expense rules are changing: - Expense claims are now due by the 5th of each month, instead of the 10th. - Receipts must be uploaded as photos in the expense app. - Late claims will be paid in the next month’s payment cycle. …

The model saw a short policy update and reached for a list anyway. This prompt gave bullets in 3 of 4 runs.

A hint versus an instruction. Your prompt's style is a hint, not an instruction. The bars show how many of four runs came back bulleted in this test.

A hint versus an instruction

Your prompt's style is a hint, not an instruction. The bars show how many of four runs came back bulleted in this test.

Source: Our test, GPT-5.5 (OpenAI API), Oct 2026

The fix: name the format. One change at the end. OpenAI's GPT-5.5 guide tells developers to describe the expected output shape, and this does exactly that.

The fix: name the format

Write a message to my team about our new expense rules. Claims are now due by the 5th of each month instead of the 10th, and receipts must be uploaded as photos in the expense app. Late claims get paid in the next month’s cycle. The change starts November 1, and questions go to Priya in Finance. Keep it under 120 words, written as two short flowing paragraphs.

Tested in GPT-5.5 (OpenAI API), Oct 2026

One change at the end. OpenAI's GPT-5.5 guide tells developers to describe the expected output shape, and this does exactly that.

Source: OpenAI, GPT-5.5 guide

Two paragraphs, four out of four. No bullets in any of the four runs, and all five facts were kept in every run.

Two paragraphs, four out of four

Real output · GPT-5.5 (OpenAI API) · Oct 2026

Hi team, please note our updated expense rules starting November 1. Expense claims are now due by the 5th of each month instead of the 10th, and all receipts must be uploaded as photos in the expense app. Any claims submitted after the 5th will be paid in the next month’s cycle. If you have questions or need help with the new process, please contact Priya in Finance.

No bullets in any of the four runs, and all five facts were kept in every run.

What this test doesn't prove. Four runs, one model, one message: a tendency, not a guarantee. This was GPT-5.5 through the API, not Claude or the ChatGPT app. Still skim the facts before sending.

What this test doesn't prove

Four runs, one model, one message: a tendency, not a guarantee. This was GPT-5.5 through the API, not Claude or the ChatGPT app. Still skim the facts before sending.

Try it: Add 'written as two short paragraphs' to your next message request.

Sources and assumptions

Assumptions:

The short version

Follow @todaysaitipSave this post on Instagram

More posts