About a year ago a man came to Sam Harper with an offer. He could take Sam’s entire newsletter workflow and automate it. The whole thing, written in under five minutes.
Understand who that offer was aimed at. Sam runs a local media company in Waco, and the newsletter is the company. Finding the stories, choosing between them, writing them up, sending them out three mornings a week. That is where his hours go. Somebody had just offered to hand all of those hours back.
Sam’s reaction was the honest one. “Part of me is like, maybe it can. Maybe it actually can. Great, I would love that. Save me so much time.”
So he said show me.
What he saw
He watched the demo. Then he turned it down.
His explanation is one sentence and it is worth reading twice. “It’s the human element that people connect with.”
He unpacked it a little further for me. A newsletter that assembles itself is a dump of information. Anybody can produce that. The packaging of the information is not much harder, because “any AI can do that just with one click of a button or one prompt.” What the machine had produced was the thing that looks like the product from a distance, without the thing that made people subscribe.
His own description of what The Waco Buzz feels like when it works: “You’re getting the news like you’re talking to a friend.”
I want to sit on this for a second, because I get told constantly that the AI question in content is a quality question. It is not. The demo probably produced perfectly clean copy. The question Sam answered was about whether anyone would still recognize him inside it, and the answer turned out to be no.
The idea he is building his workshop around
Sam is running a workshop at YouXAI called Building Out Loud, and when I asked what he wanted to cover he went straight to the fight everybody is having online.
“Depending on what circles you’re in or where you’re going online, there’s a measure of, hey, we don’t want everything to just be so AI generated, or overly where you can just tell. It’s like, man, there’s no person behind this.”
His position on that fight is that both extremes are going to keep getting louder and that neither one is useful to a person with work to ship. What he wants to give people is the third thing. He had a name for it that stuck with me immediately.
“What’s kind of that beautiful radical middle, if I can say it like that.”
That phrase does more work than it looks like. Radical, because holding a middle position in this argument takes more nerve right now than picking a side. Beautiful, because he means it as the better outcome rather than a compromise everyone settles for.
The practical version of the middle is a question he asks about his own work, and it is the one I would take to your own desk. “What part do I wanna make sure I maintain? What part do I always want to make sure my fingerprints are still on?”
The alternative he names is specific enough to be a warning. “Parking your brain at the door.” “Handing the account over to an AI to run the whole thing.”
The split he actually runs
Here is where the theory turns into a working system, and it is the most useful thing in the interview.
Sam opened by telling me what he is not. “I am not doing investigative journalism. There are other amazing players in Waco who have been doing that for a long time.” What he does instead is one word. “I’m a curator of information.”
That word choice sets up the whole division of labor. The raw material is already public, scattered across timelines and feeds. His problem was never access. His problem was that the feeds he was pulling from are built to surface the wrong things. His description of that is blunt: “large algorithms that push car wrecks and negative experiences,” because the algorithm knows people engage with that. His job is filtering all of it down to the stories his readers should actually see.
Before AI, that filtering ran on screenshots. He described it plainly. He would see something, grab a screenshot so it would not be forgotten, and hope the pile held. “That’s a very limited system that inevitably is gonna have stuff get missed.”
Now the machine does the sweeping. “Leveraging AI helps on a, hey, make sure that these things are constantly being refreshed and looked at, and I’m getting these stories dropped directly to me.”
And the part it does not touch is the part that took him ten years in Waco to earn. He decides what matters. He knows what will resonate here, which is a thing you cannot acquire by reading about a place. When I said that sounded like an executive assistant, he agreed with it flatly.
Sweep with the machine. Choose with the human. That is a rule anybody can apply to their own work this week, and it does not require a single new tool.
The workflow he plans to walk through in the workshop follows the same shape across the whole creative process: ideation and even pre-ideation, editing and feedback, then distribution and placement, and then the business model underneath all of it. His aim for the room is to save people a lot of hours in the places where hours are being wasted, so that the hours left go to whatever they are actually good at.
The bet he made before any of this: own the audience
The other half of Sam’s story has nothing to do with AI, and it may be the most valuable thing in the interview for anyone building anything.
Two years ago he made a deliberate choice to go email first. Then he did something that sounds insane in 2026. He posted nothing on social media for a full year.
His reasoning was that he needed to find out whether the thing was wanted at all. “I’ve gotta build a bit of a product market fit of is there a desire for this and what I’m trying to deliver?” So he spent the year proving that, and treated social as an addition to the flywheel once the flywheel existed.
Underneath that decision is the distinction he keeps coming back to, rented audience versus owned audience. He is careful about it, and he does not dismiss the alternative. Building a business on Instagram or a Facebook page is “totally doable, and you can make tons of money.” His objection is about who holds the switch.
“At any point a lever could get pulled, the algorithm changes, and all of your revenue could be gone overnight. You have no control over it.”
Email gives him three things the feed does not. The audience is his. The attention is undivided. And the permission was granted deliberately, which brings an obligation with it that he takes seriously.
Earning the morning
That obligation shows up as a schedule. Tuesday, Friday and Sunday mornings, every week, on time.
He explained why the timing carries weight, and it is not about deliverability. It is about a habit. “When I get up and I’m getting my coffee, I’m looking forward to getting this email.”
Then he made the comparison that landed hardest for me. It has become part of your morning “in the very same way it used to be 20 years ago of I had my coffee and I walked out to my sidewalk to get my newspaper. No one does that anymore. But now I’ve got my coffee and I’ve got my phone next to me.”
The newspaper on the sidewalk was an appointment. Somebody physically brought it, on time, every day, and if it was late you noticed. Almost nothing in a modern feed is an appointment. Sam rebuilt one.
His statement of the job that comes with it: “I wanna make sure that I’m earning that part of your day with every single newsletter that I send out.”
Why the big players are moving this way too
Sam laid out the industry picture and it is worth repeating, because it reframes what he is doing in Waco as the front edge of something rather than a local curiosity.
The New York Times, he told me, has announced a local newsletter in the Twin Cities. 6AM City has gone hard into AI and now runs in hundreds of American cities. Axios has more than 30 local newsletters running newsletter first.
Then the number. Morning Brew, which he watched get started while he was in Michigan, was acquired a handful of years ago for $75 million and has around 4 million subscribers. Started by two college students.
His read on why: “It’s the power of, hey, I have an audience that has brand equity built over time through consistency, and ultimately, it’s a distribution channel.”
I asked the room to hold on that for a second during the interview and I will do it here too. Morning Brew is a newsletter. That is the whole asset.
The title change
The detail I keep thinking about is a small one. Sam changed his own title on LinkedIn from publisher to head of community.
His reason: “At the end of the day, I’m not just doing this for ad dollars. I’m doing this for people who are here.”
Eight or nine months ago he started putting on his own events, and he says that is what unlocked the flywheel. The logic behind it is one line, and it is the same insight that makes his newsletter work in the first place.
“Everyone else who’s reading this are also my neighbors. So why should we not do fun stuff together?”
A publisher has readers. A head of community has neighbors. He runs the same business either way, and the second word changes every decision inside it, including the one about how many ads to stack.
What this means for you
If you are a creator, a business owner, or somebody who posts on behalf of a company, Sam’s split is the thing to steal. Find the part of your work that is sweeping, gathering, monitoring and assembling, and hand that to the machine. Find the part that is judgment, taste and knowing your own people, and keep your hands on it. Then write down which is which, because most people have never made the list.
And if you have been building your whole audience on somebody else’s platform, ask yourself what happens the day the lever gets pulled. Sam’s answer to that question cost him a year of silence on social media and bought him 15,000 inboxes he owns.
His workshop, Building Out Loud, runs at YouXAI on September 12 at the Performing Arts Community Center in Waco. It is a working session about building an audience that believes in what you are building.
Register at youxai.live. Tickets are $79 through August 31 and $99 after that.
See you in Waco,
Fernando
P.S. Sam’s last point in our conversation is the one I expect to be right the fastest. He thinks every employee is being activated as a creator, whether their company has thought about it or not, because they already have the profiles and they are already the best advocate their product has.