I'm Not a Writer, But I Built a Professional News Site. | NEXAIRI

What Does an Ideas Guy Actually Need?

I had something to say. I had no idea how to say it professionally. That gap—between conviction and publication—has killed more good ideas than perfectionism ever could.

Three years ago, I wanted to start a newsroom. Not because I was a writer. I'm not. Not because I had 10 people standing by to produce content daily. I didn't. I wanted to build Nexairi because the AI industry was moving faster than the media had any capacity to cover it, and the gap was getting wider every week.

The problem was simple: I'm an ideas guy. I read a lot. I make connections. I see patterns others miss. But I don't write. Not fast. Not at the volume you need to run a publication. And I wasn't going to hire a team of writers without knowing the model could work.

So I did what a lot of non-writers do now. I looked at AI writing tools and thought, "Maybe this solves the problem." And in a way, it did. But not the way most people think it does.

How Does This Actually Work in Practice?

The workflow is straightforward: read widely, identify patterns, outline the story, brief Claude on what matters, review and fact-check the draft, rewrite for voice and accuracy.

The honest answer is less glamorous than "AI writes your articles." Here's what actually happens at Nexairi. I read broadly—academic papers, SEC filings, technical blogs, competitor sites, trending topics. I flag things that matter. I make a judgment call about what deserves deeper reporting. I outline what the story actually is, what questions need answering, and what the reader gets out of understanding it.

Then I brief an AI model—usually Claude. I write a detailed prompt that includes the topic, the angle I want, the sources I've already found, the key facts that have to be in the article, and the voice I want. The model produces a first draft. Is it perfect? No. Is it publication-ready? Absolutely not. Is it a coherent first pass that saves three hours of blank-page staring? Yes.

Then comes the actual work. Two pairs of human eyes review that draft. We check every fact against the sources. We rewrite sections that sound like they were written by a model (because they were). We cut unnecessary flourishes. We make sure the piece says exactly what it should say, no more, no less. We read it out loud. We catch the moments where the voice breaks. We fix it.

Is this faster than hiring a full-time writer? Yes. Is it easier than writing it myself from scratch? Yes. Is it a shortcut around editorial standards? Absolutely not. If anything, the AI-first workflow forces more rigor because you can't trust the first pass. You have to know your subject well enough to catch when the model gets it wrong.

What Makes This Credible When Everything Online Is Suspicious?

Readers trust you when you show your actual process, maintain editorial standards rigorously, and tell them honestly what you cannot do or control.

Here's the thing nobody wants to admit: readers don't care how you produced the content. They care whether you actually know your subject well enough to catch when the model gets it wrong.

Nexairi's competitive advantage isn't that we use AI. It's that we tell you we use it. We show our process. We publish articles with named sources on real public URLs, not placeholder domains. We update pieces when facts change. We issue corrections when we get something wrong. We maintain editorial standards that would feel excessive if this were 2012, but in 2026 they're baseline.

The irony is sharp: I built a credible news operation by being radically transparent about how I actually work, rather than performing some made-up image of how newsrooms "should" work. I'm not pretending to be a 10-person team. I'm saying, "This is me, these are my limitations, and here's how I'm compensating for them." And readers trust that more than they'd trust a fake masthead.

Why Should Anyone Trust Content From a Solo Operation?

That question isn't wrong. Trust online is scarce, and you're right to be skeptical. Here's what I'd say: trust the editorial framework, not the personality.

At Nexairi, every article goes through a gate sequence before it gets published. We fact-check against named, public sources. We don't publish speculation without clearly labeling it as such. We maintain an editorial policy that's published on our site, not hidden. We issue corrections when warranted. We maintain an editorial policy page that spells out exactly how we work.

I didn't invent this framework. Good newsrooms have always done this. The difference now is that a solo publisher can do it because one AI tool and a clear editorial process can replace a hiring manager and a staff meeting.

The hardest part isn't publishing. It's maintaining standards when nobody's watching and when cutting corners would be easier. AI didn't make that easier. It just made the throughput problem solvable. The judgment, the rigor, the editorial integrity—that still comes from a human who cares more about being right than being fast.

What Changes When AI Makes Publishing Accessible to Everyone?

Quality becomes the only differentiator. Speed no longer matters. Accuracy, trust, expertise, and transparency separate serious publishers from the noise.

Two things, I think. First, quality becomes the only differentiator. If every solo creator can now publish at high volume, output volume stops mattering. Accuracy matters. Trust matters. Signal-to-noise ratio matters. Expertise matters.

Second, the stakes for editorial standards get higher, not lower. In the era of 10-person editorial teams, you could survive some mistakes because people assumed you had processes. Now readers will assume you don't. You have to actively prove you do.

I've seen a lot of solo publishers try to fake the "we're a proper newsroom" thing. It always collapses. Readers spot it. They see the recycled content, the unsourced claims, the lack of transparency about process. And they leave.

What they don't leave is when someone says, "Look, I'm one person using AI tools to do something that would normally take a team. Here's exactly how I'm maintaining accuracy anyway. Here's my editorial policy. Here's what I'm checking, and here's what I'm not." That's credible. That's human.