On Technical Writing with AI
As AI makes written content more plentiful, generating rivers of slop in the process, it will paradoxically make good writers and editors scarcer and therefore more valuable.
All would agree that AI has revolutionized how we generate, process, and express ideas. The explosion of AI-generated content has made bespoke, artisanal human writing all the more precious. Everything I have ever published has been composed without the aid of AI, and I will continue to produce scholarship and these blog posts that way. But I am a professional writer who has worked assiduously to hone my craft over two decades of education and three decades of practice. There are very few people with such expertise, and we should not expect engineers to become great writers.
And there are other domains, such as emails, business letters, and corporate communications, where AI is not only useful but now obligatory, especially when producing content at scale. I never let generative AI produce an email or letter from me afresh. But I now use it as a tool like spellcheck to offer critiques and suggestions on what I have written, and I am amazed each time by how astute its observations are.
The key is to use LLM’s as a research assistant and writing partner, not as a shadow author. To demonstrate this, I asked ChatGPT to find a piece of technical writing about machine learning in investing in order to show how human authorship with LLM assistance could work. It scoured the literature, itself something that would take at least several days and access to scholarly journals that lay behind paywalls.
It returned four choices but had to be steered back on course before finding an appropriate one, below. In fairness to the authors, this was probably one of the best-written academic papers on the topic available, but still opaque to an audience of non-specialists:
Ronen Israel, Bryan Kelly, and Tobias Moskowitz, “Can Machines ‘Learn’ Finance?,” Journal of Investment Management 18, no. 2 (2020): 23–36. (Excerpt pp. 24-25).
“…Of course, looking at multiple models and selecting the top performers in-sample mechanically leads to overfit and poor out-of-sample performance.…”
Look, I am sure they tried very hard to make this as comprehensible as possible, and indeed it is for specialists, but for the rest of us it is, as Shakespeare said, “Greek to me.” So I then asked the LLM to summarize and condense the excerpt, something that would have taken me far longer to do. I then asked the LLM to reveal its sources and reasoning, and checked them to see if they made sense or were hallucinations.
At this point, the LLM had done things I know how to do, which people with less experience might struggle to accomplish. Everything I have done so far are forms of research and prewriting, just like I would do if I were conducting research, taking notes on scholarship, and organizing my thoughts before writing anything. Now, I craft my own version, with wit, rhetoric, logic, and style that are distinctly human.
Investing with Stats and Machine Learning: a Summary of an excerpt of: Ronen Israel, Bryan Kelly, and Tobias Moskowitz, “Can Machines ‘Learn’ Finance?,” Journal of Investment Management 18, no. 2 (2020): 23–36.
by Jacob S. Dorman, Ph.D.
Many machine-learning investment techniques rely on statistical methods; they do not replace them. To the contrary, they use statistics to find complex patterns in mountains of financial data. In other words, they don’t just find the needle in the haystack; they analyze the position of every piece of hay. Using machine learning to expand statistical techniques enables analysis of much larger and more complex datasets in a greater variety of ways. It enables statistics to operate with greater flexibility, allowing researchers to let the AI choose methods and patterns rather than predetermining them.
Frankly, another factor we speak of “machine learning” or “AI” is marketing: those phrases are punchier and have greater commercial cachet than “computer-assisted advanced statistics.”
But applying stats and machine learning to investing is not without peril; searching for models can itself produce misleading results, so researchers use model-tuning techniques known as “regularization” and “cross-validation”—kind of like using various methods to “check the math” of LLM-generated computer code. Regularization limits unnecessary complexity, keeping models from becoming more elaborate than the data supports. Cross-validation checks whether a model that performs well on past data will still be effective on new information. Together, these methods help researchers distinguish useful patterns from those that might seem predictive but are actually just the result of chance events. The goal is to find the “Goldilocks” model: just right.
After all, not all machine learning investment models are created equal. Smaller models require more assumptions, but they are often preferable to large models when there isn’t much data. On the other hand, larger models can find more complex patterns in the data, but they also risk “overfitting,” which happens when a model cannot distinguish between “noise” and meaningful patterns. As a result, researchers must balance the size, sophistication, and flexibility of their models to find genuinely predictive patterns. In other words, models need to analyze a bale of hay and adapt so they don’t go haywire when they try to make sense of a barnful of data. As a result, researchers carefully develop and test many possible models. Humans need to be in the loop—the hay is not going to bale itself!
So what have we learned?
Throughout this process, I used generative AI as a researcher and thought partner, not as an authoritative writer. I delegated the intermediate tasks that AI performs efficiently—finding articles, summarizing them, and generating alternative drafts—while making all substantive editorial decisions myself.
The finished piece differs substantially from the AI drafts. I reorganized the material into a more logical structure, removed repetition, condensed 619 words into 366, and relied on rhetorical techniques that remain difficult for LLMs to produce consistently: extended metaphor, alliteration, subtle humor, rhythm, and carefully sequenced paragraphs. The recurring hay metaphor, for example, grew out of my own memories of stacking hay bales on a friend's farm decades ago—an experience that provided sensory details and emotional texture unavailable to an AI system. I also restored several effective metaphors from the original text that the LLM had discarded because I judged their communicative value to outweigh their complexity.
I revised the passage repeatedly, printed it for another round of editing, and then used Grammarly and Microsoft Word as final quality-control tools. Grammarly was valuable for catching grammatical issues and overlooked typos, but I accepted relatively few of its stylistic suggestions, making far more substantive revisions on my own. Like any tool, AI is most useful when paired with expert human judgment.
This project reinforced my view that AI works best as an accelerator rather than a replacement for experienced editors and writers. It can dramatically speed research, brainstorming, and early drafting, but effective writing still depends on judgment, structure, voice, lived experience, and rhetorical craftsmanship developed over decades. When I ran the finished version through Grammarly’s AI detector, it identified none of the prose as AI-generated—exactly the outcome I was aiming for.
As AI makes written content more plentiful, generating rivers of slop in the process, it will paradoxically make good writers and editors scarcer and therefore more valuable.
