Part 4: Use AI as a research assistant (Bakhtin and model collapse: How to use AI with expressive writing without generating AI slop)
This post is part of a series. See Part 3: From Bakhtin’s heteroglossia to AI model collapse for the previous section.
Part 4: Use AI as a research assistant
The first technique is to use AI as a research assistant. Consider this approach:
- Write out a draft of a blog post using your own thoughts, voice, and words. No AI at this point. The blog post could be 500 words or 3,000, depending on how much you have to say.
- Plug your post into Gemini Deep Research (or similar) and ask AI to research the topic of your post for you. Read the lengthy report.
- From the research report, identify several sources mentioned that you think are significant in the discussion.
- Read the sources and pull out some interesting quotes or passages.
- Now make a second pass through your essay and interweave/reference the quotes as desired into the body of your content.
- Follow up external quotes with summaries and analyses of these other voices.
- Ask AI to verify that your interpretations of the source are correct.
You could also task AI to do steps 3 through 7 in varying degrees (there’s a whole spectrum of possible AI implementation here), but if you use too much AI, you’ll be abdicating your own essay and voice to the algorithm. You have to keep a hold of it yourself if you want the writing to be meaningful to you. Don’t give away your words, in short. Just use AI to enrich, not usurp, your voice.
Using AI as a research assistant can help you identify relevant voices among a sea of competing conversations. We don’t have endless time to research the topics we write about, reading books and articles end to end looking for the relevant chapters and passages—though doing so would of course be more enriching.
In using AI as a research assistant, you aren’t diluting your own voice because you’re using AI to help identify and summarize other research and voices, bringing and weaving those other perspectives into your own. And AI is really good at summaries.
Summary is one format in which we don’t expect much personal voice. This use of AI aligns best with the way we use AI with technical documentation in the enterprise, such as using AI to summarize all the technical changes in an API release. Summary doesn’t tend to be an offensive use of AI.
Next section
Continue on to the next section, Part 5: Use AI for verification.
About Tom Johnson
I'm an API technical writer based in the Seattle area. On this blog, I write about topics related to technical writing and communication — such as software documentation, API documentation, AI, information architecture, content strategy, writing processes, plain language, tech comm careers, and more. Check out my API documentation course if you're looking for more info about documenting APIs. Or see my posts on AI and AI course section for more on the latest in AI and tech comm.
If you're a technical writer and want to keep on top of the latest trends in the tech comm, be sure to subscribe to email updates below. You can also learn more about me or contact me. Finally, note that the opinions I express on my blog are my own points of view, not that of my employer.