Say exactly what you mean: The one skill AI can't help with

AI made words cheap. Clarity will be the moat.

👋 Hi, I’m Jeff. I write weekly about lessons I’ve learned leading product teams and working to lead well in chaotic environments, drawing on 25+ years of building.

1:1 exec coaching: I work with a small number of senior product leaders one-on-one as a coach. These are typically six-month engagements. If that’s you or someone you know, let’s talk.

⏱ Read time: 3 minutes

---

Have you ever been in a meeting at work and noticed what you might call “corporate speak?” I’m not immune. For example, I’ve caught myself front-loading too much context before getting to the point someone was after. In a world where AI-generated text is increasingly everywhere, anti-corporate speak, or saying what you mean, will stand out more than ever.

For me, corporate speak is when between 30-50% of a conversation is padding. This is language that’s not clearly saying what’s happening or what should happen. It’s often the result of someone’s fear of offending someone else in the room, or from a lack of the speaker’s clarity of thought.

When you do run across someone who consistently speaks directly, you realize how resonant that is compared to everyday corporate speak.

It’s like working underwater and looking out at the sky through that water vs. lifting your head out of the water and suddenly seeing the sky more clearly.

My favorite example of getting to this clarity is still from Ben Horowitz’s book, The Hard Thing About Hard Things. “Hard Things” is a fast read and is useful for anyone interested in learning about management, and more specifically, how tech companies work. The book was originally published in March, 2014 but is worth a revisit from time to time.

The most useful part of the book is infused throughout. It’s the drop-dead simple way Horowitz has of communicating. You get a sense for this in his take on specific stories. But you also see it in-between the stories in terms of how basic thoughts are conveyed: Use. As little. Words. As possible. Say exactly what you mean.

Take this example from the story about Horowitz trying to hire a VP of Sales into his struggling company. He has someone he thinks is fantastic in mind, but this person is a little odd. So board members plus Chairman and co-founder Mark Andreessen are taking issue. The potential hire doesn’t look exactly right. He went to a small school, and he has an off way about him. Horowitz listens and then makes a short, clear case:

“I agree with every single one of those issues. However, Mark Cranney is a sales savant. He has mastered sales to a level that far exceeds anybody that I have ever known. If he didn’t have the things wrong with him that you enumerated, he wouldn’t be willing to join a company that just traded at thirty-five cents per share; he’d be the CEO of IBM.”

Simple. Direct. This candidate has some issues, but he is exactly what we need right now and let’s get real, we are not exactly playing from a position of strength at this time.

Imagine instead how that statement could get mangled by someone who is either not clear about exactly what he (or she) thinks or not willing to just state it:

“Hmmmmm. You have some good points, there. Really. You do. This guy has some positive points, but we’re all agreed we need the best possible hire. I guess I need to take another look at this situation with that in mind and report back to the board in 30 days.”

That statement is a recipe for delay and then losing the candidate.

THAT, in the end, is the most valuable message of The Hard Thing About Hard Things. It’s a book about communicating with as little wasted language as possible.

Take the time to know what you think and then say EXACTLY that.

Doing so saves time and builds trust, both of which most startups are short on and desperately need.

This is even more important in 2026. In 2014, when this book came out, the enemy of clarity might have been fear or fuzziness of thought.

Both of those are still there. But now, AI can generate a thousand fluent, reasonable-sounding words in seconds. The result is more language than ever and less clarity than ever — vagueness with a clean coat of paint. AI can produce the words for you. It can’t know what you think, or make sure you say exactly that.

Try trimming your words this week. Before your next meeting, write the one sentence you’d say if you had only ten seconds. Say that first thing.

Let me know how it turns out.