AI Can Help You Write Thought Leadership. It Can't Write It For You, Yet

Every investment firm and financial advisor we work with asks us some version of the same question: Should they be using artificial intelligence to write thought leadership?

The honest answer is yes, but carefully.

AI lacks judgment, and its predictive large language models lead to some inaccuracies. But the main problem is voice. As more firms lean on the same AI tools, their writing starts to converge, with similar sentence rhythms, similar structures, similar phrasing. That's the opposite of what thought leadership is supposed to do, which is to make a firm or an advisor sound like themselves, not like everyone else using the same digital assistant.

However, AI is genuinely good at a handful of things. It's fast at tasks like transcribing and simple regurgitation, like summarizing. It's a capable research assistant, pulling together data and context to support an argument someone has already decided to make. And it's useful for stress-testing an idea, poking holes in an argument before a reader does.

So, the real question isn't AI or no AI. It's where, specifically, in the process AI earns its place. Here's how we think about it, stage by stage.

1. Capture

Most good thought leadership starts as an idea trapped in someone's head. Busy advisors, analysts, and portfolio managers lack time to write the idea down and structure it into a narrative.

Traditionally, writers have helped by sitting down with the thought leader for an interview or reading through their notes or presentations. Today, additional steps are easier. Anyone can record their thoughts, and AI can produce a transcript in moments.

Also, Macro Editorial can provide an AI interviewer to ask a few questions to draw out that big idea, record it, and then transcribe it. Yes, that means talking to a robot, but you can do it from anywhere, at any time.

Either way, the goal of this stage is the same: get the person's actual thinking down in their own words, before anyone starts shaping it into a piece.

2. Research

Once there's a point of view to build a case around, AI becomes a genuinely efficient research assistant, pulling together supporting data, precedent, and context. It's not deciding what the argument is; it is helping build the evidence for an argument a human has already made.

AI has become much more transparent with its sourcing, and one of our writers can determine if the sources are credible.

3. Draft

This is the one stage we don't hand to AI.

We have yet to see AI craft a draft that sounds like an authentic human, while also making a cogent argument with credible source material.

Our writers apply judgment, honed from decades of experience covering financial markets and collaborating with compliance departments. We also see writing as a calling – we love it!

4. Refine

Once there's a draft, AI earns its keep again. The latest AI models are surprisingly good at fact-checking specific claims, checking drafts against source material or by searching the Internet, as well as finding typos. Used this way, AI acts less like a writer and more like a colleague whose job is to find the holes before a reader does.

AI can also shift text in a desired direction, such as making it punchier or more formal. In this way, AI is a decent editor, as long as a human editor reviews all drafts.

5. Repurpose

Once a piece is finished and fact-checked, AI can help with promotion and complementary content. It can turn one article into a handful of social posts, a shorter companion piece, or an outline for a follow-up, quickly and with plenty of options to choose from. This is low-risk work, because the substantive thinking and proofing already happened upstream, at the draft and refine stages.

The takeaway

AI isn't going to write your next piece of thought leadership, at least not well and error-free. But used deliberately, at the right stages, it can make the process behind good writing faster and sharper without cutting corners on the writing itself.

The firms and advisors who earn their readers' trust over the next few years won't be the ones who used AI the most. They'll be the ones who used it well.