Welcome to episode number nine of AI after 50, the podcast for folks over 50 who want to understand AI without the tech speak and without feeling like it's already too late to figure it out. It's not too late, you're fine, and you are in the right place. Everything we discuss on the podcast is available in our show notes and whatever platform you're listening on, they should be available right there for you. But we also offer premium show notes and what those are those are items that didn't make the podcast for time or whatever reason, but things that are still relevant to the conversation and the discussion. So, the link is below in the episode description. Please subscribe. Not only will you get the premium show notes, you'll also know when a new episode drops. So, please subscribe to AI after 50 so you don't miss an episode and again, you can do that inside of your platform that you're listening to right now. Well, welcome back. If you've been here before, thank you so much for still being here. I took a break. A longer one that I really had planned and I want to explain that because I think the reason I stopped and the reason I'm back are actually the most interesting part of the episode. So, an honest explanation. Why did I stop? Well, the short answer is the show wasn't working the way I had set it up. The format wasn't right. I wasn't being myself the way I needed to be and I had a choice, keep pushing through something that felt forced or stop and fix it. So, I stopped and fixed it. Hopefully. Here's what I decided. The best version of this show isn't me or anyone else talking about AI in the abstract. It's me actually using AI to build something real and bringing you along while I do it. Not theory, not frameworks, not here are three things leaders should know. Real work, real results, real mistakes, and I had something to build. So, the idea was a number of years ago, at least 15 years ago, I think, I had an idea for a local newsletter. I live in the extreme northeastern portion of the northeast Ohio and in the middle of one of the most underrated weekend destinations in the state, actually in like a three or four state area. It's called the Grand River Valley Wine Corridor. We just call it Ohio Wine Country. And if you've never heard of it before, that is kind of the point. There are over 30 wineries within a driving distance of each other, live music most weekends, a beautiful Lake Erie shoreline, there's local events, and nobody was putting something like this together for people who wanted to plan a weekend in Ohio Wine Country. So, I thought somebody should be doing that. But, you know, it's a lot of work to publish something weekly. There's curation, there's formatting, and just putting it all together, the time I would need, and I had no idea how to do that. Now, here's where all this becomes relevant. I'm not a publisher, I'm not a journalist, I've never run a newsletter, and I don't have a design background or a marketing team or a content calendar, anything like that. All I am is a guy who knows our area, Ohio Wine Country, pretty well. And now now I have AI and now I have Claude. Just so everybody is on the same page, Claude is an AI created by a company called Anthropic, and it's very similar to ChatGPT created by OpenAI. And they look a lot alike, and I also work with ChatGPT daily. But as for Claude, I've been using it in my work for a long time now, over 3 years for sure. But I had never tried to build something from scratch with it. So, back earlier this year, in the spring of 2026, I heard of a new feature in Claude called co-work. And many models now offer a similar work feature. There's co-work in Claude, there's work in ChatGPT, there's co-work in Copilot, and this work feature, it's just that. It's like sitting down with a very smart and intelligent coworker who knows much more about anything than I ever will. You can sit, you can talk, and you can have conversations about practically anything. And the AI actually works with you, asks you questions, it builds a framework around your idea, and it helps you build whatever you want. So, that's just what I did. I downloaded the Claude desktop app, I opened a chat, I selected co-work, and off we went. The first thing I did was sit down with Claude and just talk through the idea. Not write me a newsletter, not give me a business plan to write a newsletter, just here's what I'm thinking, what do you think? And that conversation alone was worth more than I expected. We talked about the audience. Who exactly is this for? And we thought, well, people who like wine, but maybe not just for that. That's kind of broad. We needed to get a little more nichey and more specific. So, our primary reader is a socially active woman 25 to 65 who enjoys wine and often is the planner of their family or their group. You know, she's busy, she's on her phone, she scans everything before she reads, and she doesn't need more information. She just needs someone to decide for her. And that one sentence became the whole editorial mission of the newsletter. And that is, we're not here to tell people about every wine or the history behind every winery. We are here to make their weekend. I didn't write that, we wrote it together, Claude and I, in a conversation. And then we named the newsletter. We went back and forth for a little bit, and we decided on this, The Lake and Grape Weekly. Over the next several weeks, I built an entire publishing operation with AI, with Claude co-work. And let me tell you what that actually means, because I don't want to sound like it's something futuristic or complicated, because it really isn't. So, the first thing I did was I built a master database with Google Sheets of every winery in Ohio wine country. The name, the location, what makes each one interesting, their website, their Facebook page, 33 wineries organized into tiers based on popularity, how reliable and how feature-worthy that they are. And I did the same thing with breweries. There's a few of those in the area. Distilleries, same thing. And I created a sheet of artists that frequently play in the area. Clyde Co-work learned everything and created skills or markdown files that the AI reads before and every time we start to build a new issue. With Co-work's help, I built a newsletter structure, five sections, specific word counts for each issue, rules about voice and tone, and what words we never use. We have a banned words list, which by the way, you should have one, too. So, picturesque, that's on it. Nestled, that's on it. So is don't miss, that's on it. I built skills that I just mentioned, which essentially they're a set of saved instructions that Clyde follows every time it curates, creates, and it writes the newsletter. That's so I don't have to re-explain everything from scratch for every issue. The skills, it knows the audience, it knows the mission, it knows the rules, and I build a weekly workflow, and the newsletter is delivered at 6:00 a.m. to our subscribers every Friday morning. As far as social media is concerned, right now we're only on Facebook and soon Instagram. And we're only on one platform for many reasons. And Facebook and Instagram, of course, are owned by Meta, one platform. And mainly this is where our audience is, and it's easier to manage one platform than it is many. Plus, if we ever have to run ads, we're all set with Facebook and Instagram. So, it was just easier to pick Meta, Facebook. So, in addition, Co-work, it creates Facebook posts from content that didn't make the newsletter, which is really cool. And he and I, it and I, the AI and I came up with a great system for posts, and we'll share that with you in upcoming episodes. And I tag all the wineries, all the distilleries, and all the performers in the newsletter and on social media, so they know that the Lake and Grape Weekly is out there. And I've gained a good bit of traction doing that. So, as of today, I've published 11 issues. Number 12 is this weekend. Every Friday morning to approximately, about 50 subscribers, but we're growing and we'll get there. And I will tell you how I did that and do that as the series moves forward. Here's the thing I want to spend a couple of minutes on because I think it's the most important thing I can tell you. I made mistakes, lots of them, real ones, and ones that went out to real subscribers. So, for the first five issues, the emails that went out were in the wrong colors. That was supposed to be a nice navy blue frame around a warm reading surface. Beautiful design, spent a lot of time on it, clean, branded. But for five issues, it went out just white, plain white. The design wasn't applying. Well, that was because every new post in Beehiiv, that's the publishing email platform that I use, was silently overriding the template I had created. And there was one specific setting panels that I wasn't touching, but the AI, with its help and Claude co-work, we we decided to tackle that problem. We fixed it, and it was just one setting I didn't know to check, and five issues went out before I caught it. So, there were lots of mistakes and other missteps, and there are rules around those now. The AI knows, so mostly built into skills, and we talked about those a little bit ago, and we'll talk about more about skills and markdown files in future episodes because they are important. And I'm telling you this because this is exactly what building looks like. It's not a clean before and after process, it's a process. If you make a mistake, you figure it out, you figure out why it happened, and you write a rule so it never happens again. Then the next mistake teaches you the next rule. So, 11, soon to be 12 issues in, the system is significantly better than it was at issue number one. So, here's what AI after 50 is going to be starting with this episode. It's going to be a documentary. I'm going to take you through the full story of building the Lake and Grape Weekly, what worked, what didn't, how AI collaboration actually functions day-to-day, how that works, and what I'm building next. And all the while, I'm going to show you how I'm doing it so you can do this yourself. If you've ever had a notion or a desire to create a newsletter, maybe not a wine newsletter necessarily, but whatever your version of this is for your community, your expertise, your niche, your quarter, whatever that might be. Because here's what I've learned. The people with the most to offer AI collaboration are experienced people. Us folks over 50 come to mind. People who know something, people who've built things or led things or solved problems. You bring the judgment, the knowledge, and real-world credibility. AI just brings the speed, the structure, and all the mechanical work. You don't need to be technical, you just need to be curious and willing to try, and not afraid to make a mistake, lots of them, in front of yourself. I've made plenty, I'll keep making them, and I'm going to keep telling you about every single one of them. In the next episode, I'm going to walk you through the very first conversation I had with Claude about the newsletter, the naming conversation, the audience definition, the mission statement. And I'll show you exactly how that kind of conversation worked, what I asked, what came back, how I responded, and how we got better together. After that, we're going to go into infrastructure. We're going to talk about the master database, which is just spreadsheets, uh the skills, the reference files, the folder structure, the stuff that makes this repeatable instead of one time. And eventually, we're going to probably do something else with this. But for right now, if you want to build a newsletter, a publication, a local media brand using AI as your co-publisher, I want to teach you exactly how I did it. And together, we'll get there. If you've been listening to AI after 50 before and you stuck around, thank you so much. It just means the world to me. Believe me, this is a better version of the show, and you deserve that. And if you're new, thanks so much for dropping by, and please subscribe wherever you're listening. Leave a review if you feel like it. I'd appreciate that. And if you have any questions at all, I am at mark@marktennant.net. I read and respond to all my emails, and come find me at aiafter50.com, where I'll have show notes, resources, and eventually some other surprises as well. So, until next time and next episode, keep building, keep learning, and don't let anybody tell you it's too late to figure this stuff out, because it's not. The best is yet to come, and we're going to be there together. Till next time, take care.

AI After 50 - Episode #9 - The Relaunch - Creating a Publishing Company Using AI

Show Full Transcript

Repurpose this content with AI