Rough Drafts logo

Rough Drafts

Subscribe
Archives
August 15, 2024

8 generative AI prompts to help you write better

Last week I wrote about how I use AI for writing. This week I’d like to share some detailed tips and prompts that have worked for me and other writers.

The number one piece of advice I have for writers working with AI chatbots is to treat them like interns or research assistants you are giving assignments to.

In other words, don’t expect the AI to do the work for you — certainly not at the same level that you do. Treat it as a helper and collaborator. Give it clear instructions, and know that you will have to revise, correct, and improve much of what it does for you.

It helps to think of AI as a colleague that is pretty clever and has access to an unbelievably large quantity of source material, but which is utterly lacking in common sense or any notion of originality. 

You know that coworker who is kind of a know-it-all and also a bit of a bullshit artist? The one who will make up answers and pretend to know things even if they don’t? That’s your chatbot coworker.

If you understand these limitations and can work around them, AI chatbots can be powerful collaborators. 

Important note: If you’re writing for clients or publications, make sure they know you are using AI in your writing process. Some clients won’t care; others will have policies against AI use that you need to respect. Some publications, like The Hill and Entrepreneur, now have policies against AI content, while others like Harvard Business Review, require disclosures of any AI use. 

Tip 1: Give the chatbot a clear, detailed assignment

Whether you’re using the chatbot for research or help with writing, put as much specificity into the prompt as you can.

You may not need to create a detailed assignment brief, but it wouldn’t hurt. The pointers I shared on how to start a writing project with a team are equally valid when your “team” is you plus a chatbot.

As with an assignment brief, it’s useful if your prompt includes a few specifics:

  • A working title
  • A short description of the content you’d like to create — just a sentence or two describing the idea is enough, or a full abstract if you’ve got that.
  • The purpose for which you’re writing this piece. What’s the goal you hope to accomplish?
  • The target word count you’re aiming for
  • The target publication, and/or any details on the audience that you’re hoping to reach
  • Any notes on the tone or style you’re hoping for (formal or casual, technical or general-interest, use of “I, we, you, they” pronouns)
  • Any supporting notes and resources, such as transcripts, white papers, blog posts, or news story links.
  • You can also list any stats or data points you’d like the story to include.
⚠️
Warning: Don’t ever put anything confidential into a chatbot prompt or into a customized chatbot, like a custom GPT on ChatGPT. Chatbots can leak information, through attacks known as prompt injection and other means.

Tip 2: Revise your prompt iteratively

If you don’t get what you want, revise your prompt and try again. Suppose the chatbot delivers something that seems surprisingly articulate, but it’s kind of verbose and not very original (the default mode for both ChatGPT and Claude). Try revising and resubmitting your prompt, adding specific instructions aimed at correcting what went wrong.

Some examples of instructions you might add to your revised prompts:

  • Limit your responses to five sentences or less.
  • Don’t use more than three sentences in each paragraph.
  • Vary sentence and paragraph length to provide variety and musicality to the writing.
  • Use direct, conversational language, with short sentences and varied sentence structures.
  • Eliminate fluff from your response. Get right to the point.
  • Avoid jargon in your response.
  • Avoid the following words: delve, showcasing, underscore, comprehensive, crucial, intricate, pivotal. (These are 7 words that suggest a text was written with AI. I would add a few others to that list: transformative, ensure, critical.)
  • Provide examples, preferably with a human dimension, in order to give your writing a more emotional, relatable aspect.
  • Provide sources for each argument, with links.

Tip 3: Use the AI as a debate partner

You can use AI to preview objections to what you’ve written, so you can strengthen your arguments. Try this prompt:

  • You are a researcher. Please review the following text and provide 3-5 arguments against it, with sources, representing opposing perspectives.

Once you have the response, you can review the arguments and sources, and decide if you want to build preemptive objections into your piece. 

Tip 4: Ask for additional questions to prepare for an interview

Before an interview or sourcing call with an expert, write a list of questions you hope to ask. Then provide the chatbot with your list of questions, plus some background on the interview subject, and ask it to provide additional questions you could ask.

The prompt:

  • I am going to interview someone who is a (title) overseeing (responsibilities), as preparation for a (project outcome). I have written some interview questions, which I'm pasting below. What other questions should I consider asking this person in order to make my interview more complete?

Tip 5: Use the chatbot to check for common knowledge

Chatbots’ “knowledge” is built out of what’s already on the internet. This is why AI-generated text is not particularly original. You can use this to your advantage, though. Suppose you’re writing a piece that answers a common question that potential customers or readers might have.

The prompt:

  • Please give me a range of short answers summarizing possible responses to (the question or problem you're answering in your piece).

Once you have those answers, compare them to what you’ve written. If what you’re writing shows up in the chatbot’s answers, you’re not saying anything original. Revise it to ensure that you’re writing something truly new, something that does NOT show up in the chatbot’s response. 

This tip comes from a webinar led by PR pro Michael Smart last year. 

Tip 6: Ask it to highlight errors in your text

Writer Alicia Bonner shared this tip with me: “I love asking ChatGPT to bold errors in any text block before I publish it.”

The prompt:

  • You are a copyeditor. Please use boldface type to highlight any errors in the text below, and give me a suggestion on how to correct each error.

Alicia also shared that she loves using it to write cover letter drafts, something that I also found useful. Cover letters are formulaic and unlikely to be read by actual humans, so that’s a perfect use case for a robot writer.

Alicia added, “You’re absolutely right, sometimes the language pattern seems good and then you read it a third time and you’re like, hang on, that makes no sense.” 

As I said before: Check the output — and revise, revise, revise.

Tip 7: Ask AI for advice on revising

PR professional Suzanne Collier answered a LinkedIn post of mine with a great tip. Suzanne writes, “I find it especially helpful for research, refining content, identifying gaps, and uncovering different ways to present information so that it is clear and engaging.”

Suzanne specifically recommends prompts like the following:

  • Help me make this more engaging, descriptive, persuasive, etc.
  • Provide options for shortening/lengthening this text.
  • Take this text and change the voice to be more empathetic (or more direct, or other directions on tone).

Tip 8: Check everything

As I wrote last week, generative AI platforms are dream machines: Their job is to make up stuff. While they often produce factual statements, there’s no fact-checking involved in the process, so anything truthful is basically an accident. 

It’s up to you to check everything and make sure that the AI isn’t bamboozling you. That includes clicking through on any links it provides to verify that A) the link is real, and B) whatever it links to actually says what the AI summary says it does. You don’t want to be those lawyers.

You can try using one chatbot to check the work of another chatbot. For instance, you might take a paragraph from ChatGPT and feed it into Claude with this prompt:

  • You are a fact-checker. Please alert me to any factual errors or inconsistencies in the following text.

I used this prompt for the first two paragraphs in this section, and Claude's answer was pretty good. Its summary: "Overall, while the language is somewhat colloquial and uses some generalizations, the core message about the need for human verification of AI-generated content is accurate and consistent with current understanding of AI limitations."

💡
This prompt can surface potential issues, but I wouldn't rely on any chatbot for fact-checking, ever. Always check any facts yourself.

Bonus tip: The AI writing tool in WordPress

Automattic, the maker of WordPress, is rolling out an AI tool to make your writing more concise.

That's the reverse of the way most AI tools write. It's also the opposite of what most SEO tools will tell you, which is why the web is so clogged with looooong posts. 

Since WordPress powers an enormous proportion of the web — over 40% of all sites — the impact could be huge.

For now, it’s only available to users of WordPress.com. I’m looking forward to it coming to self-hosted WordPress.org sites so I can test it out. For more details, see the story by Paul Sawers in TechCrunch. 

Don't miss what's next. Subscribe to Rough Drafts:
dylan.tweney.com Bluesky LinkedIn
Powered by Buttondown, the easiest way to start and grow your newsletter.