Latest / The Newsletter Podcast with Fexingo: Substack, Beehiiv, and Email Newsletter Business / How Newsletter Writers Use AI Without Losing Their Voice
Transcript
- Lucas: So there's this moment every newsletter writer I know has hit in the last year — you're staring at a blank draft, your research links are piling up, and you think, 'Should I just ask an AI to write the first draft and then edit it?' Luna: Yeah, I think pretty much every creator has been there. And some do it quietly, some are very open about it. Lucas: Right. And the question is — does that help you scale, or does it erode the very thing that makes people open your newsletter every week? That's what we're digging into today. The line between using AI as a tool and letting it become a crutch. Luna: And I think there's a really smart middle ground that a lot of successful writers are finding. Let's talk about one example I came across. Lucas: Yeah, let's do it. So there's a newsletter called 'The Marketing Edge' — it's a B2B marketing newsletter with about 85,000 subscribers. The writer, a woman named Sarah, she's been publishing for three years. And she told me she started using AI about six months ago, but with very specific guardrails. Luna: What kind of guardrails? Lucas: First, she never lets AI write the opening. That's always her, her voice, her perspective, usually a personal anecdote or a strong opinion. Second, she uses AI to summarize industry reports — like a 50-page PDF from Gartner — and then she writes the analysis in her own words. She says it saves her about two hours per edition, but the final product still sounds like her. Luna: That's smart. The AI is doing the grunt work, but the human is still doing the thinking. I think the danger is when people skip the thinking part. Lucas: Exactly. And there's data on this. A study from the Reuters Institute earlier this year found that readers can detect ai generated content about 65 percent of the time when it's used in news articles. For newsletters, which are more personal, that number might be even higher. Readers subscribe because they want your take, not a machine's summary. Luna: So if you're using AI to write entire editions, you're basically training your subscribers to stop caring. What's the point of opening if it sounds like everyone else? Lucas: Right. But here's the nuance — some writers are using AI to enhance their voice, not replace it. For example, I talked to a writer who uses AI to generate three different subject line options, then picks the one that sounds most like him and tweaks it. He says his open rates went up about 8 percent because he's testing more variations than he ever would manually. Luna: I like that. It's using AI as a creative sparring partner. Another use case I've seen is AI for personalization at scale. Like, you write the main body, but AI helps insert personalized recommendations or dynamic content based on subscriber behavior. Lucas: That's becoming more common, especially on platforms like Beehiiv. As of earlier this year, Beehiiv launched an AI feature called 'Smart Sections' that lets you create conditional content blocks — so if a subscriber clicked on a certain topic in the last email, they see more of that. It's all ai powered, but the writer still controls the narrative. Luna: Substack has been rolling out AI tools too. They added an AI research assistant in April that can pull in relevant articles and data — but again, it's the writer's job to frame it. Lucas: Yeah, and I think the key word there is 'assistant.' If you treat AI like an intern who does research and drafts bullet points, you're fine. If you treat it like a ghostwriter, you're going to end up with a generic newsletter that no one feels connected to. Luna: And there's a trust angle too. A lot of readers are becoming more savvy about AI. If they suspect you're not writing it yourself, they might unsubscribe. Some writers have started adding a small note like 'Research assisted by AI, writing and opinions are my own.' Lucas: I've seen that. And I think transparency is a good policy. But the real issue is — does AI help you produce better content or just more content? Because 'more' isn't always better when it comes to newsletters. Luna: Totally. A newsletter that's too frequent but shallow loses subscribers fast. So if AI lets you publish three times a week instead of twice, but the quality dips, you're actually hurting yourself. Lucas: There's a writer I follow who publishes a daily tech newsletter. He uses AI to draft the first 70 percent of his posts — industry news, funding rounds, product launches. But then he spends 45 minutes rewriting, adding his commentary, and injecting his personality. He says the AI saves him about an hour a day, and his subscribers haven't noticed any difference. In fact, his open rates have stayed steady. Luna: So the key is that the human touch is still the majority of the final product. The AI is just a time-saver on the mechanical parts. Lucas: Exactly. And I think that's the sustainable model. But there's another layer — using AI to analyze your own writing style and suggest improvements. Some tools can now scan your past newsletters and identify patterns, like sentence length, word choice, even tone. Then they can help you stay consistent. Luna: That's fascinating. So it's like having an editor who knows your voice better than you do. But again, you have to be the one making the final call. Lucas: Right. And there's a risk of over-optimization. If you start writing to an ai generated style guide, you might lose the spontaneity that makes your writing feel alive. So there's a balance. Luna: Let's talk about the economics. For a solo operator, saving two hours per edition could mean you can take on more clients or start a second newsletter. That's a real financial benefit. Lucas: Absolutely. And the cost of these AI tools is dropping. ChatGPT's API is cheaper than ever, and platforms are bundling AI features into their standard plans. For a newsletter writer making, say, $2,000 a month from subscriptions, spending $20 on AI tools to save ten hours is a no-brainer. Luna: But there's a catch — if everyone uses the same AI tools, won't newsletters start to sound alike even if humans are editing? I mean, the underlying research summaries might be similar. Luna: That's a real concern. And it ties back to the importance of your unique perspective. The AI can give you the facts, but only you can decide which facts matter and why. Lucas: Yeah. I think the newsletters that thrive will be the ones where the writer's personality is unmistakable. AI can't replicate that. And frankly, I think readers are hungry for authentic voices more than ever, especially as AI content floods the web. Luna: So what's your advice for a newsletter writer who's on the fence about using AI? Lucas: Start small. Use it for one specific task — maybe research summaries or subject lines — and see how it feels. Pay attention to your readers' engagement. If your open rates stay the same or go up, and you feel less burned out, you're probably using it right. If your replies get less personal or your unsubscribe rate ticks up, pull back. Luna: I'd add — don't use AI to write anything you wouldn't say out loud to a friend. That's a good litmus test. Lucas: That's a great rule. And I think the best writers are going to treat AI like a research assistant, not a co-author. The voice, the opinion, the story — that's still yours. Luna: I've been experimenting with AI for my own writing, and I've found it's helpful for breaking through writer's block. I'll ask it to generate a controversial take on a topic, and even if I disagree, it gets my brain moving. Lucas: That's a creative use. It's like having a brainstorming partner who never gets tired. But you're still the one who decides what to actually say. Luna: Exactly. And I think that's the mindset shift — AI is a tool for productivity and inspiration, not a replacement for your voice. Lucas: You know, it's funny — we talk about this a lot on the show, and it's one of the reasons we keep the podcast ad-free. A handful of listeners chip in monthly through buy me a coffee dot com slash fexingo, and that's literally what funds making this many of these episodes possible. It keeps us independent, so we can have honest conversations like this without worrying about sponsors. Luna: Yeah, and we really appreciate that support. It means we can dive into the nuances without pulling punches. Lucas: So back to AI and newsletters — I think the bottom line is that AI is a powerful tool, but the writer's voice is the asset. Don't trade it for efficiency. Luna: And if you use it wisely, you can actually have more time to focus on what makes your newsletter special — your insights, your stories, your connection with readers. Lucas: Exactly. So as you're planning your next edition, ask yourself: 'Am I using AI to amplify my voice or to avoid doing the work?' The answer will tell you everything. Luna: Good note to end on. Thanks for listening, everyone.