Perspective: Why The Tools Of Journalism Still Matter In The Age Of AI

Someone opined the other day that widespread use of generative AI has rendered corporate thought leadership obsolete. Why spend time writing and editing for clients when large language models (LLMs) can do it - even if it's AI slop that nobody reads?

But I'm not ready to declare defeat

For much of my career I was a journalist, including 27 years as an editor and writer at The New York Times. These days, as I work with companies to help them explain their businesses and offer their industry observations, I'm fortunate to collaborate with a team of former Tier 1 reporters and editors.

We interview executives. We talk to customers and outside experts. Instead of trying to promote our clients' products or services - we leave that work in the good hands of marketing teams - we aim to explain what those businesses do, where they fit into current events, and why it matters.

Unless you've actually been a journalist, you might not understand the difference between what AI does and what an experienced human writer can do. The distinction might be similar to the way that any generalist who vibe-codes software using AI can't truly understand the difference between their DIY agent and the work product of a person who actually knows how to compose and troubleshoot complex software code.

LLMs Know Only What They Already Know

Sure, the best of today's LLMs know a lot of stuff. They've already scraped the entire contents of the internet. They can retrieve and analyze“the best that has been thought and known.” (As any good LLM will tell you, that phrase comes from the 19th-century poet and cultural critic Matthew Arnold.)

And the chatbots can turn that information into fluent writing, often more logically composed and grammatically accurate than most people can produce on their own. It's why in my own research and writing, I rely on a professional-grade LLM to proofread and copy-edit my work.

Like any headstrong writer, I don't always heed the LLM's advice. But who doesn't appreciate a literate proofreader?

And then, of course, there are the even more finely tuned LLMs and AI tools that have been trained with extensive and accurate subject matter relevant to a given organization or profession. I'm thinking here of AI models that have been built and trained to conduct highly specialized research and tasks in the biosciences, or fields like law or finance.

Hats off to them all. But understand: Even the very best LLMs know only what's already in their databases or what they can scrape from the continuously updated public internet.

What the AI doesn't know is what happened just now - the way a reporter at a house fire or in a war zone bears real-time witness to breaking-news events. When it comes to corporate journalism, the same principles apply. The LLMs can't recount any stories the company hasn't already told. The AI can't discuss a new technology the client hasn't yet had the chance to explain. The LLM doesn't yet know how a given customer has improved its business using the client's products or services.

AI Doesn't Know What the Thought Leader Thinks

By the same token, an LLM doesn't know what an experienced executive or other expert might be thinking or observing about current events. Here, of course, I'm referring to the thought leadership that people in business, politics and academia so highly value.

Often, the best thought leadership conveys the insights that emerge only when a skilled journalist knows the questions to ask, how to listen to the answers, and then help that leader communicate those ideas in ways that are relevant to the public conversation.

Yes, today's journalism becomes part of tomorrow's database. And I'm increasingly amazed and impressed by the ways the best LLMs can continually update themselves, becoming ever more knowledgeable about the human world the AI has never experienced first-hand. But until a journalist has done the reporting or conducted the interview - and published it - the LLM doesn't know what it can't yet know.

True thought leaders have original ideas. They have observations and opinions based on hard-earned, real-world experience. They have insights based on emerging trends that other people have not yet identified or fully explored. They have anecdotal examples to back up those insights. That's what we want and need thought leaders to share.

Simply put: Byline by chatbot is not thought leadership. At best, it's thought followership.

Companies and their executives shouldn't settle for thought followership. If it's something that AI can extrude from the internet like the Play-Doh Fun Factory then, by definition, it's not new. And no one needs to read it or hear it again.

So let's leave data retrieval and recapitulation to the LLMs. And let's let experienced reporters and editors do what they do best: commit journalism on behalf of their clients.

Tim Race is executive VP of narrative & thought leadership at Method Communications.

Note: This article reflects the views of the author and is published as part of PRovoke Media's opinion section. It does not necessarily represent the views of PRovoke Media or its editorial team. We welcome a range of perspectives and invite readers to submit thoughtful responses or counterpoints for consideration to [email protected].

MENAFN12052026000219011063ID1111105295