Outsourcing your writing to AI tools, like CoPilot and ChatGPT, is undeniably tempting for anyone who finds drafting documents to be the most tedious or challenging part of their day. But for those who spent their career honing their writing craft, and for anyone who spends the majority of their day drafting and editing content, these tools can feel like an insult. However, I have slowly come to terms with these technologies and would encourage any professional writer to do the same.
The biggest shift in my mindset was thinking of AI and LLM software as a tool to assist me, not replace me. To use one analogy: Even the most skilled carpenter uses a table saw and a power drill; it helps them do their job faster and more effectively, but still relies on the carpenter’s judgment, experience and knowledge. Or, to put it another way: Would you write a proposal with spell-check disabled, or drop into the group chat with your phone’s autocorrect turned off?
Balancing human versus robot input
Experts refer to the ways AI can be incorporated into work on a spectrum of four terms:
- Human-only. This one’s self-explanatory and reflects how most every writer worked until recently.
- AI in the loop. The human drives the process, but an AI tool provides assistance.
- Human in the loop. AI does the heavy lifting, but a human checks in through the process.
- AI-only. The one that makes folks like me nervous, this working style lets an AI tool operate independently.
My preferred use case, and the one I would recommend to other professional writers, is AI in the loop. Use AI when it makes sense to help out, but your years of experience, accumulated sense of style and finely tuned flair for prose should lead the way. With that in mind, here’s where AI can help writers during their workday:
Brainstorming and outlining
It almost always feels easier to edit or react to existing work than start from a completely blank page. Using an AI tool to provide a basic outline can be a helpful jumping-off point – especially if you start by dropping in notes or background material for the tool to build on.
When beginning a challenging draft, I’ve dropped presentation slides into AI tools and asking how I could summarize the material as written content. Referencing this generated summary helped me focus on the key points, providing a launchpad for drafting the full story.
Un-stick a stuck brain
You know the moment: Thinking, “There has to be a word or a phrase or an expression that I can use here.” In days past, I’d Google things like, “word or expression for saying work expands to fill time available” or “other ways to say enhance processing power,” which rarely yields useful results. The conversational nature of LLMs – which, remember, were trained on real-world published writing – offers a better way to get past those moments.
Get help with time-consuming tasks
AI proponents often advise thinking of the software as an intern. It might not produce client-ready work, and you’ll definitely need to tidy it up and fix it, but at least the virtual intern did the boring and time-consuming parts for you. For writers, this could mean using AI to search through data and look for patterns. Examples would be poring through reports or transcripts, or pulling statistics out of a trove of data.
Recently, I prompted an AI tool to review interview transcripts to identify quotes on specific topics and themes, then had it digest an Excel sheet of survey data to pick out interesting trends. In both cases, I went back to the source and double-checked everything lined up with the interviewee’s words and the survey numbers, but this allowed me to skip some of the time-consuming admin work up front.
Proceed with caution
Let’s go back to the carpenter analogy from the beginning. Our hypothetical carpenter should still check their work against the blueprints and plans as they build. They should follow safety rules along the way, and make sure what they are building aligns with construction codes.
For a writer using AI, the principles are no different. Always verify anything generated by AI to be sure it’s accurate and truthful to your source materials and goals. Pay close attention to your employer’s policies around AI usage, as well as those of your clients. Finally, give everything a once-over to be sure that your content is polished to the same professional standards you – and your clients – demand of all work.
Embracing the shift
Not that long ago, I would have rolled my eyes at writers using AI in their day-to-day work. Yet carefully integrating AI in my day helps get rid of the sticky, tricky and, let’s face it, dull parts of drafting content. A few years in, it’s starting to feel like using the dishwasher versus hand-washing every plate. Sure, I could do it all without any automation, but why not make the day a little bit easier?
Jake Holmes is an editorial writer at Stratacomm, where he develops written materials from start to finish—conducting research and interviews, crafting key messages, and advising on pitching and publication strategies. With more than six years of agency experience and nearly a decade as an automotive journalist, he offers deep insight into storytelling and media engagement.
