What You'll Learn
- Relying on AI to generate writing for you takes away from your ability to provide a unique, human perspective.
- Search engines — including AI-based ones — prefer original, valuable content, which favors writing it yourself.
- AI can still be a powerful tool for advancing your thinking if used carefully.
I'm getting straight to the point: No, you shouldn't use AI to do your writing for you — even if that just means a first draft that you'll later edit.
Why? Because in doing so, you skip a critical opportunity to frame the piece in human terms, from a human perspective and in a human context. And everything that comes afterwards in the writing process will be limited by what the AI first proposed.
Now, this is a no-judgment zone. If you turn to AI to produce a first draft of your written content, I don't blame you. We’re all tight on time. And writing is hard.
But stick with me as we set the chatbot aside, open that blank text document, and have a go at doing it the old fashioned way.
Writing Is Thinking
In June of last year, a preprint paper out of MIT found that those who used an LLM while writing an academic essay showed significantly different brain activity compared to those who went without AI assistance.
Participants who relied on a generative AI tool to help them write their essay "performed worse than their counterparts" who relied on their own thinking "at all levels: neural, linguistic, scoring."
The paper's authors stress that these conclusions are "preliminary," that the results may vary depending on the context, that the paper itself has yet to be peer-reviewed, and that more research is needed in order to fully understand the impacts of LLMs on human cognition. Still, writing is a key component of our thought processes.
"Writing compels us to think — not in the chaotic, non-linear way our minds typically wander, but in a structured, intentional manner," an editorial in the journal Nature Reviews Bioengineering states. "By writing it down, we can sort years of research, data and analysis into an actual story, thereby identifying our main message and the influence of our work."
So if you need to process your thoughts on a topic, a good way to do that is to just write it out. Take the time and effort to jot down what you're thinking, what you're feeling, and where your mind is leading you. It doesn't have to be good. It doesn't have to be a proper piece of content to start. It can just be a loosely organized list of bullet points on a page. From there, you can sort out your ideas and craft them into something more coherent.
This very blog post sprung from a collection of fragmented thoughts that only came together once I committed them to a text document.
Writing from scratch may be hard at first, especially if you aren't used to working this particular mental muscle. But as you write, you will begin to make connections and think of new directions and possibilities.
If You Want to Stand Out, Write It Yourself
AI-generated content is an amalgamation of other people's thinking. Yes, you can use it to draft that fundraising appeal or grant letter of inquiry. But since AI is based on content that's already out in the world, its output is, by definition, derivative content.
For instance, a 2024 paper in Science points out that, while AI can produce polished content, its use tends to reduce the creative diversity of work overall. All this is leading to a sort of "cultural stagnation," as Ahmed Elgammal, a professor of computer science at Rutgers University, argues at The Conversation:
"[The derivative nature of LLMs] has broader implications, because modern culture is increasingly influenced by exactly these kinds of pipelines. Images are summarized into text. Text is turned into images. Content is ranked, filtered and regenerated as it moves between words, images and videos. New articles on the web are now more likely to be written by AI than humans. Even when humans remain in the loop, they are often choosing from AI-generated options rather than starting from scratch."
AI's effect on cultural works isn't just hypothetical. In fact, evidence suggests that it's already influencing the way we communicate with one another.
In 2024, Florida State University (FSU) researchers identified a selection of what it termed "focal words" — words that ChatGPT tended to overuse that began showing up more frequently in scientific literature after ChatGPT's introduction. Meanwhile, more recent peer-reviewed research, also out of FSU, found a similar phenomenon occurring in everyday speech. If AI's influence is showing up in these areas, it stands to reason that AI is influencing communication in all forms.
An oft-repeated writing "rule" is to avoid clichés. By doing so, you push yourself toward more original thinking and more stylistically interesting prose. That alone may be reason enough to avoid using LLMs to generate content. LLMs — by their very nature — fall back on common tropes, relatively limited word choice, and frequently used writing formulas.
While formulaic content is fine for certain purposes, it's not great if you want to stand out. If every fundraising appeal and every grant application sounds the same, none of them will break through the noise.
You Know Your Audience
If you're engaging with your audience or your community on a regular, basis, you probably have a pretty good sense for what their needs are. What sorts of content will resonate with them. What messaging will work and what will fall flat.
If every fundraising appeal and every grant application sounds the same, none of them will break through the noise.
That isn't necessarily the case when you rely on AI-generated content that pulls from a wide range of content aimed at audiences that may be significantly different from your own.
While an increasing number of enterprise-grade AI tools can pull from contextual information that you provide — brand guidelines, content samples, and so on — generative AI creates content based on what it "thinks" you want it to say, not on what you actually should say.
You can always revise a piece of AI-generated content, sure, but depending on how close or far it is from what you want, you may not necessarily save time compared with doing it yourself.
Search Engines Value Unique Content
This doesn't just apply to human readers; it also applies to search engines. As our Wes Holing points out, search engines still prioritize valuable, unique, high-quality content. And for the reasons I outline above, that favors human-produced work.
"Copy generated by ChatGPT, Claude, Gemini, or any other generative AI service can only draw from what has already been published," Holing writes. "Google’s E-E-A-T search guidelines and more restrictive spam policies mean that unoriginal content lacks a competitive edge and is less likely to rank or be included in generative summaries."
But what about AI-powered search features? Turns out, the same holds true there. Holing writes:
"[T]he most significant recent addition is the generative AI summary. Website owners have since worked to revise their content to appear within this prime piece of search real estate, but the good news is that many of the traditional SEO strategies still remain relevant."
In other words, it's a good idea to start with original content, not to just tweak an AI-generated piece and call it a day.
How to Thoughtfully Use AI in the Writing Process
I may be arguing against using AI to do your writing for you, but I'm not saying that you should never use AI in your writing process. In fact, if you use AI only to get answers or generate content, you're thinking way too narrowly about it.
Here are a few ideas for how AI could help augment your writing and research without doing your thinking for you.
Just be sure to exercise a little caution — LLMs don't have the critical thinking or media literacy skills that humans do. They can make mistakes and hallucinate things. They also may not surface what you're looking for on the first try. But with some care, they can provide a jumping-off point for your writing.
To Help You Identify Topics or Themes to Write About
Are you stuck on what to say? Use AI to help you identify topics to write about, and then further develop, iterate, and brainstorm your own ideas from there. You can ask it for a list of broad topics, or you can go a step further and use it to get an idea of what people online want to know about a topic.
I would shy away from asking it to produce a clearly defined thesis for a piece of writing.The idea here isn't to use AI to replace your thinking or to rely solely on the topics it produces; it's just to help you get past any initial writers' block.
Example Prompt
I'm writing a blog post about using AI at your nonprofit, but I don't know where to start. What sorts of questions are people asking online about the topic?
As a Template Generator for Formulaic Writing
If you've never written a SWOT analysis or a concept note for a grant application, you may be forgiven for not knowing how to format them. You can ask an AI chatbot to provide you with a basic outline that you can follow and populate yourself.
Example Prompt
Give me a template for a concept note for a nonprofit grant application. I don't need you to produce the content; I just need to know what sections to include.
As a Virtual Reference Desk
Have you ever gotten help from the reference desk at a library? They won't do your research for you, but they can point you toward resources to check out. A well crafted AI query could do the same with online research. In this example, you're not looking for answers so much as you're looking for a starting point for further research.
Personally, I don't really use AI much, but when I've asked it about a topic, I'm less interested in what the chatbot itself says and am more interested in the sources that it surfaces. So consider AI a way to begin your search for sources. And don't just stop with what it tells you. Be curious. Dig deeper. Ask follow-up questions.
Example Prompt
I work at a nonprofit and don't know much about grant writing. Can you point me toward some resources I can read to learn more about the topic?
As a Mental "Sparring Partner"
This is an idea I've seen floating around the internet — and among fellow TechSoup staffers — and I think it's a good one. Rather than have AI do your thinking for you, do your own thinking, then turn to AI to help advance your thinking. AI can be useful for poking holes in your reasoning and finding gaps in your logic.
For example, one of my colleagues uses a chatbot to "interview" him about a topic to help him think through a problem in more detail, challenge assumptions, and suggest additional pathways of thought and action.
Example Prompt
I want to write a piece explaining the grant writing process. What key questions should I ask myself about this topic, and what resources and arguments should I consider during the writing process?
Educator Mike Kentz discusses the concept in more detail on Substack. Kentz also links to a list of additional prompts to try.
A Bicycle for the Mind, If You Let It Be One
Machines simply can't replace the heart and soul that powers our work.
Steve Jobs once called the personal computer a "bicycle for our minds." A bicycle doesn't replace human ability, but through a drivetrain of cogs, chainrings, and pedals, it greatly amplifies human effort. Jobs posited that the personal computer does the same thing for how we work, play, create, and communicate.
AI can serve a similar purpose — not just act as an answer machine or content generator.
I'm not the only one who sees it that way. As David Brooks writes for The Atlantic (paid subscription or trial access required), "The people who are going to make a difference are not the ones who seek relaxation and passively use AI to work less. They are the ones who will seek improvement and actively wrestle with AI to develop their own mental capabilities and accomplish more."
So should you use AI in your writing process? With a little thought and care, sure. But don't let it write a single word on your behalf.
Because at its core, the act of writing is humans conveying information to other humans. There is intrinsic value in that. And human perspective, experience, and insight is key in our work as nonprofits. We're not here just to make a line go up. We're not here just to optimize every last aspect of our lives. We do this work because we care about making an impact in our communities. And that takes empathy. Compassion. Critical thinking. A humane view of the world.
Machines simply can't replace the heart and soul that powers our work, and they shouldn't replace the means by which we communicate, either.
Explore More AI Topics
- Explore how AI-powered search has changed SEO — and how it hasn't.
- Learn 4 Tips for Better ChatGPT Prompts.
- You need an AI use policy at your nonprofit. Here's what to know.
- Read how a member of TechSoup's development team uses AI when writing grants.
Thumbnail photo: Shutterstock