Three Content Writers Interview Each Other About AI
Thoughts, Concerns, and How to Use It Responsibly
Artificial intelligence doesn’t feel things (yet!), but we’ve all got a lot of feelings about AI. Recently we asked Big Sea Social Media Strategist and Content Writer Aviva Campbell, Big Sea Content Manager Dan O’Keefe, and Big Sea Senior Content Strategist Kerry Haze to sit down for a conversation about AI-generated content. What follows is that transcript, lightly edited for length. We talk through what AI looks like in our day-to-day work, what parts make us uneasy about it, and where it’s genuinely useful in the copywriting world.
Dan: AI, and especially our collective reaction to it, has gone through so many phases since ChatGPT launched back in 2022. How do you feel about AI in general in the year of our Lord 2026? Like, from a civilizational level?
Kerry: I used to be afraid of AI and worried that it would consume a lot of our humanity and many of our human-focused tasks, like writing and content marketing, but now I’m confident in saying that it will never replace us.
It’s been a while now since the main AI platforms launched, and it has become increasingly clear that you need real people to do real work. I don’t like the narrative that AI is here, it’s happening, it’s everywhere, and it’s not going anywhere. We can choose how we use AI, and we can choose to use it in ethical, moral, and sustainable ways.
Aviva: Yeah, AI is impressive and terrifying at the same time. I’m not anti‑tech, obviously, but I don’t trust Big Tech to roll out world‑changing tools with any level of caution. We’ve already seen AI used for harm, surveillance, and warfare. But we’ve also seen it detect early cancer precursors and make social causes more impactful. So when people talk about AI “advancing civilization,” I’m like, sure, maybe. I’m hopeful about the good it could do, but I’m not naive to the bad it already has done.
Dan: Yeah, I can be a bit of a Luddite. My natural pessimism makes me susceptible to some of the more catastrophist talk about what AI could do to society. But I try to take that with a grain of salt, since a lot of the doomerism about AI is just as much a part of the hype machine as the utopianism.
Aviva: Outside of work, do y’all use AI in your everyday life?
Kerry: I don’t use AI in my everyday life, no.
Dan: I resisted for a while, but I’ve been impressed enough with Claude that I got a personal subscription. I don’t have a lot of occasions to use it, though. When I do, it’s usually as a more detailed search engine. I’ll ask it to recommend a handful of articles that approach some topic from a few different perspectives, or I’ll use it to track down some anecdote from 17 years ago that I half remember. It’s surprisingly good at that if you can give it enough context to work with.
Aviva: I use it when it’s genuinely helpful, like for quick summaries or reviews. Most recently, I used it to compare the price of cat trees. It’s revolutionized the way I search for things. Beyond that, I use it with extreme caution, and I’m very aware that my daughter’s generation is growing up with this stuff baked into everything. She’s 14, and she’s learning how to do things without AI first. I believe it should be a tool she uses later in life rather than a shortcut she relies on while her brain is still developing.
Kerry: How do you feel about different models of AI? About ChatGPT vs. Claude vs. Gemini, etc.?
Dan: Claude’s definitely my favorite. There’s something both about the quality of its responses and its user interface that just feels more convincingly thorough, “thoughtful” (I’ll put quotes around that), and trustworthy than ChatGPT does. Chalk it up to good UX design and a dash more transparency.
Kerry: Yeah, I appreciate that we aren’t using ChatGPT at Big Sea anymore. Claude has a bit more of a content-focused lens around it, compared to other AI platforms. For content writing specifically, I used to use something called Jasper, and I really enjoyed that platform as it was so specifically attuned to helping someone create outlines, website copy, social media posts, ad copy, print copy, and other written marketing copy.
Aviva: Claude is also the one I trust the most right now; as a corporation, Anthropic seems to be the most transparent and grounded. That said, Gemini is exciting for its Google Workspace integrations. I’ve been using its email synthesis feature more and more. And ChatGPT, while powerful, hallucinates the most and has the shadiest data ethics and privacy policies, so I rarely, if ever, use it.
Dan: Where in your workflow do you find AI most helpful? Strategy? Brainstorming? Editing? Do you use it a little in each step?
Aviva: I use it almost every day when brainstorming, researching, editing, and creating reports. If I need a list of angles or trending topics related to a client’s niche, it’s awesome. It’s also useful when I’ve got long video transcripts that I need to chop up to create short-form videos for social media; it can help me pinpoint story arcs and key timestamps, which is a huge timesaver.
Kerry: Yeah, it’s really helpful for brainstorming ideas, drafting first outlines or rough drafts for blogs or landing pages, and then reporting and gathering data on how pieces of content are performing. It’s also proving to be beneficial for keyword research if you connect your visibility platform to it, like SEMrush.
Where I draw the line is that I will not let it fully write something and call it done. A first draft is fine as a starting point, but human eyes need to be on every single part of it.
Kerry: Aviva, do you feel like AI is making it harder to cut through on social media?
Aviva: Yes and no. There’s a lot more social media content since the genesis of generative AI, but much of it is copy‑pasted from generic chatbots. You can absolutely tell when a bot wrote a post, and your target audiences don’t love that.
I have a question. Do you think we’re heading toward a world where people assume everything is AI until proven otherwise? How do we adapt to that?
Dan: I do worry about that. It’s definitely made people more suspicious of a certain sort of polish that used to take real time and effort to achieve. And that’s a shame, because those skills are still worth developing and that kind of writing has some real uses. Lately I find myself tempted to throw in little asides or structure things a little unevenly just so my writing doesn’t have that characteristic AI homogeneity.
Kerry: I think we’re headed toward a world where it’ll be increasingly clear when something is AI and when something isn’t. Sure, things have gotten more advanced and more intense, but I think we’ll start to recognize and value human-created content that’s wild, free, has an opinion, and maybe even includes a few errors over the shiny, glossy, AI slop that’s on the internet right now.
Dan: How is AEO/GEO changing how you write and organize content?
Aviva: It’s pushing me to structure content in a way that’s easier for models to parse. Content needs to be technically sound and factual, but in digital marketing, it also needs to be evocative. My outlines focus on how information gets chunked and retrieved by large language models (LLMs) now, but the meat of my writing relies on the same logic it always has: clarity and connection.
Dan: Yeah, a lot of what makes for good AEO/GEO builds on search engine optimization (SEO) fundamentals that have been in place for our writing process for a while—clear writing, easily scannable section headers, accurate and comprehensive answers, attentiveness to real people’s actual search habits online. All these virtues transfer over from one specialty to the other.
I do worry sometimes that writing for AEO can make our writing a little choppier, since the impulse is to make major statements as retrievable as possible, shorn of their context in the rest of the blog. That’s not a bad impulse, but it is a little in tension with making a longer piece an enjoyable and smoothly flowing read from beginning to end.
Kerry: AEO isn’t really changing how I write, but it is changing the structure of our content. Blogs should include FAQs if possible, and you should answer the main question in the first few sentences of the blog. You want citable sources in your content, and you want to have an author listed for each blog. We were doing a lot of the same things when SEO was the only acronym we were using, but AEO has just exacerbated it a bit and made it more important to have a solid structure for content going forward for all our clients.
Dan: At Big Sea, we’ve always prided ourselves on writing for people over algorithms. Does the need to get cited in LLMs and AI overviews make that harder? Are the algorithms winning?
Kerry: I think this is a big issue and probably my biggest worry when it comes to writing with AI overviews and answer engines in mind—homogenization. You brought this up earlier, Dan. AI recombines what already exists, so the more writing it handles, the more your AI content writing ends up sounding like the 10 other articles already ranking on search engines for that topic. We’re already seeing this everywhere. You read something, and you think, “I’ve read this before,” because, in a way, you have.
I don’t want the algorithms to win. I want real, honest, human, thoughtful content to win in the end. Yes, we might have to play by some of the rules to get cited by Google and AI search engines more and more often, but that doesn’t mean we have to regurgitate the same content and remove our humanity from our writing.
Aviva: Yes. Writing for AI citation requires adherence to formulas, but there is still plenty of room for color, charm, and originality.
Kerry: Aviva, you’ve worked on Big Sea’s official AI Policy and Task Force. Could you talk about what we’re doing to define how and when to use AI and what its limits are?
Aviva: We’re identifying where AI helps us move faster without compromising the quality or humanity of our work. And we’re creating clear norms around usage, data privacy, and transparency in our processes.
Dan: Any final thoughts?
Aviva: AI is powerful, and it’s changing everything, but we shouldn’t treat it like an inevitability we have to succumb to. We get to decide how we use it. If we stay skeptical and intentional, we’ll be better off.
Kerry: We shouldn’t be afraid of AI. We should learn how to use it and adapt it to what we need when we’re creating content, but it’s never the end-all-be-all. Humans will always be needed in the content creation and content strategy process. Full stop.
Want to join the conversation? Tell us where you land on AI in your marketing and content work. Follow us on LinkedIn and drop your take in the comments, or send us a message directly. We read every one!
