When I first founded Intellerate Consulting, my initial projects involved strategic writing and editing for clients in the research and policy sectors.
Two years later, generative AI has rapidly changed the playing field for this kind of work, prompting me to rethink my business model and expand my portfolio of services.
Yet after a few months of experimenting with AI, I don’t plan to surrender my writing and editing projects to the machines.
Neither should you.
In today’s newsletter, I’ll describe why this is the case.
My journey from AI resistor…
For a long time, I resisted the pull of AI and regularly rebuffed suggestions that I try ChatGPT.
My initial AI avoidance was rooted principally in ethical concerns.
I was uncomfortable using AI models that are opaque about their training sources and algorithms.
I worried about how these models, which are trained on existing content, can perpetuate existing power structures and biases. (For a humorous example, check out Dr. Samantha Pillay’s attempts to use AI to create an image of a man scrubbing a toilet.)
I was troubled by the massive environmental costs of powering AI.
As someone who cares a lot about digital privacy, I hated the idea of donating my intellectual property to Big Tech companies for training new AI models.
… to cautious adopter
I still have these concerns. But as a business owner, I need to understand my competition.
Moreover, I have a commitment to my clients to produce the best possible work, on time, and within budget. It would be irresponsible for me to ignore a potential resource.
For these reasons, I began to experiment with using AI on selected tasks. I’ve found that there are many things that AI can do quickly and do well, such as:
- Reducing texts to meet word/character count limits
- Generating clear and snappy titles
- Rearranging relatively formulaic text and tables
On the other hand, my forays into using AI for text summaries and even basic proofreading ranged from underwhelming to hilariously useless. (More on that in a future newsletter.)
While AI models continue to improve, I remain convinced that AI is woefully ill-equipped to do the strategic writing and editing necessary for organizations to connect with an audience and ultimately make an impact on our complex world.
Rather than add to the Internet’s ever-growing list of AI bloopers, I’d like to step back and describe what strategic writing and editing actually entails—and why these demands are fundamentally human.
Judgment, not optimization
Many of my projects involve strategic writing and editing: I start with a client’s insider knowledge (or draft text), and I create a clear, persuasive document that makes sense to the people that matter, such as policy-makers, grant reviewers, or funders.
In many respects, this work is an act of translation: I craft a text that bridges the gap between internal expertise and an external audience.
To be effective, one needs to keep the audience in mind at all times. And while we usually refer to “the audience” as a single entity, in reality we’re talking about a diverse group of stakeholders, with a wide range of preferences, priorities, and areas of expertise.
Here are some questions that drive my own process:
- How might readers interpret a particular word, phrase, or idea?
- What kinds of experiences and biases will readers bring to the text?
- How can I ensure that the text is accessible to generalists and yet enlightening to subject-matter specialists?
- What is an effective way to address a sensitive topic in a publication aimed at an international group of decision-makers?
- Can the reader follow the narrative logic from beginning to end, or will they get distracted, confused, and frustrated?
These questions illustrate how strategic editing goes beyond mechanical text processing and generation.
Each decision requires that the writer/editor consider the text from multiple vantage points and use human judgment to balance diverse priorities.
Done well, the final product will crystallize complexity without oversimplifying.
And if the text needs to address a controversial issue, it does so with clarity and directness while avoiding sanitized and shallow prose—hallmarks of AI-generated text.
Effective written communication requires distinctly human capabilities: empathy to understand diverse reader perspectives, deep knowledge of organizational cultures, and the kind of judgment that only comes from years of experience navigating complex challenges.