Don’t just prompt AI to “edit” a thought leadership article. Prompt it so that it becomes an intellectual sparring partner that challenges your article’s premise and analysis.
Most people using AI to improve their thought leadership content are probably only scratching the surface of what it can do for them. That has more to do with the person prompting the AI than with the AI itself.
I should know: I was once in that boat.
I probably spent the past year experimenting with how AI can help make my thought leadership articles logically tighter and better reasoned. I wasn’t getting the output I was looking for, and just assumed the AI was failing me. It turns out I was failing it.
But I think I finally figured it out! Here’s what I discovered so that you, hopefully, can save yourself valuable time and energy by not having to wander through the AI desert.
It’s not enough to ask AI to help you “edit” a thought leadership article
For a long time, whenever I asked AI to help improve my article, I asked it to help me “edit the article” or “improve the draft.” What I got back was fine, but superficial. I got back line edits, improved word choices, and smoothed sentences or reordered paragraphs.
The AI was useful the way a copy editor would be: tinkering with words, sentences, and paragraphs, and improving the language, but not necessarily my arguments, the strength of my logic, or the analysis in my articles. Yes, my writing was getting smoother, but my thinking, as communicated in my articles, wasn’t getting any sharper.
I finally figured out why.
When I asked AI to “edit” or “improve my draft,” it went into copyediting mode, optimizing for fluency, good transitions, and a consistent tone. It didn’t go deeper than that, so it couldn’t critique and improve my thinking, which is what I wanted it to do.
I realized that if I wanted AI to do more than improve my writing, I had to be more specific with my prompts.
Tell AI how you want it to help you “edit” a thought leadership article
In hindsight, it seems obvious, but it dawned on me that if I wanted AI to challenge my thinking in my articles and tell me where my reasoning was weak, I had to-wait for it-ask it to challenge my thinking in my articles and point out where my reasoning was weak.
So I did! And when I did, I received much more actionable feedback that helped me improve how I was making my points.
For example, I began asking AI:
“What’s the strongest counterargument that I haven’t addressed?”
“Read this as someone who disagrees with my conclusion. What’s the weakest link in my article? What’s the weakest point that I made that you would seize upon?”
“Are there logical fallacies or inconsistencies in my article?”
I would also ask the AI to give it to me straight, such as “Don’t be polite. Be skeptical. Tell me what’s wrong with this article.”
No matter the AI model I was using at the time, when I began asking it these kinds of questions, I got better results and the kind of feedback I was looking for all along on weaknesses in my arguments, weaknesses in the logical flow of my articles, and in the analysis that I applied.
Prompt AI to unlock its sparring partner abilities
The value of these more specific prompts is that they created a thoughtful, creative intellectual sparring partner I didn’t have before.
In the process of writing thought leadership, most of us don’t regularly have an intellectual sparring partner. Instead, we have to come up with a way to challenge the arguments we’re making ourselves, which we already think are solid (otherwise we wouldn’t have made them in the first place).
When challenging our thinking and looking for weaknesses in our arguments, we’re limited by our knowledge and imagination. Having a colleague review a draft of a thought leadership article is a good way to challenge our thinking and identify weaknesses in the article’s logic. But that’s not a luxury we all have at any given time.
AI, however, is a 24/7/365 sparring partner with extensive knowledge of the law or industry we’re discussing, along with expertise in structuring arguments (and picking them apart). With AI, we have an always-available, worthy adversary that can make our thought leadership articles better by prompting us to rethink the arguments we’re making and how we present them.
When AI offers counterarguments, points out weaknesses in our analysis, and generally critiques an article, we now know what we must do to strengthen it. Before, we couldn’t see those shortcomings. But through more specific prompts, AI is giving us a list of issues to fix to improve our article.
On a related point, that’s why the copy edits that AI gives us when we ask it to “edit” our articles can be counterproductive. An article might have clearer language thanks to AI’s suggestions, but those suggestions are unlikely to make our underlying arguments stronger-they’ll just sound better.
Three steps to unlocking AI as an intellectual sparring partner through better prompts
Once you understand that AI can be a sparring partner that improves the quality of your thought leadership articles, here are three pointers for unlocking those sparring partner abilities when you prompt it.
First, name the failure mode you want to avoid. That is, in addition to telling it what to do, be sure to tell it what NOT to do (i.e., the failure mode).
For example, if you want it to give you “the strongest counterarguments a skeptic of your thesis” would give you, tell it to “not to be polite, not to hedge, and not to give me any deference as the author of the article.”
If you don’t make that clear, your AI platform might go easy on you and your article, not giving you the tough love you need to improve your article and the thinking reflected in it.
Second, ask AI to review your article from a particular point of view. Not just as an expert in the field, or as a partner who has practiced in an area of law for 50 years, but as a person who disagrees with your thesis and theme, or someone who is skeptical of the entire premise of your article.
By asking AI to role-play as someone on the opposite side of the ideological spectrum from you on a particular topic, you’ll likely receive more critical feedback that will help you craft a stronger article.
Third, and finally, ask for a handful of objections, not a long list. The way that AI typically works is that if you ask it for more things, it’ll give you more things, but they’ll be watered down.
So ask for the top one, two, or three biggest problems with your draft. That forces the AI to focus on fewer problems, which should yield more substantive criticism. If you ask it for the seven biggest problems, you’ll get a list that likely runs out of steam as you get down to the second half.
Unlocking what could be an indispensable use of AI
The light bulb went on for me regarding this use of AI when I was working on a thought leadership article for a client. The client wanted to talk about a novel theory of law by connecting a couple of dots. I thought I connected those dots well.
The AI platform I was working with disagreed, pushed back, and pointed out flaws in the logical progression from dot A to dot B to dot C that gave me pause. My reaction was, “Wow, this is actually spot on.”
And of course, I double-checked the case and statutory citations that the AI suggested. Those suggestions and others undoubtedly improved the arguments and, by extension, the entire article. They tightened up the logic and eliminated a flaw that, had it not been addressed, would likely have been picked up on by my client, or worse, experts in the field if my client missed it too.
The big picture takeaway for me and hopefully for you regarding these prompts is that you can use AI to do more than “edit” your thought leadership articles. You can use it to stress-test your articles by challenging your arguments, your analysis, and the articles’ entire premise.
There are plenty of reasons NOT to use AI in the process of developing a thought leadership article, such as its default robotic writing style. But using AI as a sparring partner is an effective, non-controversial application.
If you use AI to challenge the thinking, logical reasoning, and analytical skills displayed in your thought leadership articles, I’m willing to bet you’ll be impressed with the results-and you’ll create a higher quality library of thought leadership that more persuasively positions you as an authority regarding the areas of law you practice and the industries you serve.
Want to build a thought leadership strategy that works? Take our free Thought Leadership Opportunity Finder and discover exactly where your content strategy has gaps — and what to do about it. Or, learn more about how LFES works with attorneys to produce consistent, high-quality thought leadership without the time investment.