Seth Hastings: We're starting a little differently today with some news and notes. Someday we'll have a fun soundboard effect for this segment. I can practically feel my wife cringing at home. Sorry, dear. But we'll get there. 

Let's start in the world of AI, specifically with Claude. Claude is now adding invisible watermarking to content generated or copied from its platform. The watermark follows the content wherever it's pasted. Alexis, as our AI expert, what's your reaction to that? 

Alexis Gajewski: I have two reactions. 

The first is that when you initially hear something like that, your instinct is to worry. You think, "Oh my gosh, this is going to be tracked. Everything is going to be obvious." 

But when you think about it more, if everything gets labeled, then it almost stops mattering. 

I think the idea, especially with everything happening in the EU around AI-generated content, is to identify what content was created by AI. The challenge is that this watermark will be attached to anything copied out of Claude. Even if you wrote something yourself and only used Claude to proofread it, copying it back out could still apply a marker suggesting it was AI-generated. 

That's a strange gray area. As we continue identifying those gray areas, I think the watermark may end up meaning less than people expect. 

Seth Hastings: That was honestly my first concern, too. I thought, "Okay, if I use it to edit something, am I suddenly going to be labeled as AI-generated?" 

Erin, maybe you can speak to this. Google has consistently said AI itself isn't a negative ranking factor. The focus is on content quality. Do you see watermarking impacting search in any way? 

Erin Hallstrom: I think this goes back to the conversation we had in our last episode about commoditized versus non-commoditized content. 

At its core, the distinction comes down to quality. We talked previously about how AI can't replace lived experience. When Google says it's looking for quality content, I think that's what it means. 

When people use AI to generate content from scratch, they're often producing commoditized content, meaning information that can be found anywhere. But quality content is the kind of content AI can't create from experience because it doesn't have experience. 

So when Google says it cares about quality, I don't think it's concerned with whether you used AI for copy editing. It's more interested in whether you're using your own expertise, knowledge, and perspective to explain something in a unique way. 

Alexis Gajewski: And watermarking itself isn't really new. Other AI platforms have experimented with similar approaches before. 

What's more interesting is how these systems are used in the future. From what I've read, the watermarking isn't necessarily something obvious embedded in the text. It may be based on patterns in word selection and language usage that leave a kind of detectable thread throughout the content. 

It's not going to be something simple. 

The big question is whether these markers become integrated into publishing platforms, AI detectors, or other tools. But again, if everyone is using AI in some capacity, whether they're generating an article from scratch or just asking for editing suggestions, then eventually everything gets labeled. At that point, the distinction becomes a lot less meaningful. 

Erin Hallstrom: I just want to say that I'm still standing by my em dashes and my habit of starting every fourth sentence with "worth noting." 

I said those things long before AI existed, and I'll be saying them long after AI is here. 

They can take "worth noting" and the em dash from my cold, dead hands. They are not taking them away from me. 

Alexis Gajewski: For those who don't know, I have a fine arts background. I came to appreciate the em dash later in life. 

Not because AI introduced me to it, but because AI showed me just how effectively it can be used. It's a beautiful piece of punctuation. 

I know it gets a bad reputation right now because it's overused, but it really is lovely. So thank you, AI platforms, for helping me appreciate what the em dash is capable of. 

Seth Hastings: #freetheemdash 

There's another piece of news we should talk about: LinkedIn's apparent pushback against AI-generated content, or what many people call AI slop. I think that's great because there are so many posts where it's obvious AI wrote them. There's a difference between using AI well and publishing something that clearly came straight from a prompt. 

Alexis, what impact do you think this will have? 

Alexis Gajewski: I hope it forces people to use AI in more thoughtful ways. 

Instead of typing, "Write a LinkedIn post for me," and publishing whatever comes back, people should be using AI more strategically. 

We tend to think about commoditized and non-commoditized content only in relation to prestige content like articles, podcasts, or videos. But it applies to everything, including LinkedIn posts. 

Too often, social posts become an afterthought. We tell AI to write them and move on. But you're still representing yourself. If your content is generic, or if it's slop, then that's a problem. 

I also think it's interesting that LinkedIn is taking the lead here. Other platforms have been discussing it. YouTube has already taken steps around AI-generated content and monetization, although that created some controversy among legitimate creators. 

My hope is that these changes encourage people to use AI in more useful and more unique ways. 

Seth Hastings: I think you're right. LinkedIn may be the first platform, but it won't be the last. 

If every platform starts identifying AI slop, then performance is going to suffer for people who rely on low-effort content. 

A lot of what we discussed last episode was how to avoid looking like AI slop, especially when creating commoditized or non-commoditized content. 

What advice would you give creators who want their content to stand out? 

Alexis Gajewski: The easiest answer is to put yourself first. 

Whether it's a LinkedIn post, a podcast, or something else entirely, you need to inject your own ideas and experiences into the content. 

For example, my declaration of love for the em dash a few minutes ago is uniquely mine. You're not going to find that exact perspective everywhere. That's something personal. It reminds people that I am a real person with real thoughts, even if those thoughts are occasionally strange. 

AI can absolutely help improve your content. It can trim unnecessary wording, identify weak spots, and help reinforce your voice. 

One of the best ways to use AI is actually to challenge yourself. Think of it as an honest colleague. 

Take a piece of content you've written and ask AI whether it's drifting into commoditized territory. AI can often identify those moments and point out where you're repeating ideas everyone else is already discussing. 

It can suggest bringing in more personal experience, more unique perspectives, or a different angle. 

That's not AI creating for you. That's AI acting as an editor and a challenger. 

Seth Hastings: Erin, you've written extensively about commoditized versus non-commoditized content. How would you use AI to make sure content is flourishing rather than blending into the crowd? 

Erin Hallstrom: I'm always an evangelist for using AI to evaluate AI. 

And before I answer, anyone listening can hear that I'm awkward when I speak. I'd love to tell you that's unusual, but it's really not. That's just me. 

People know that's part of my personality. My awkwardness is probably one of my biggest credibility signals. 

So here's my Hallstrom hot tip: Be awkward. Stay awkward. 

Anyway, from a content perspective, if I were writing something for a fictional company like Seth's Secret Sauce Company and trying to determine whether a topic was too commoditized, I'd use AI for research. 

Let's say I'm writing an article called "The Five Best Ingredients for Barbecue Sauce." I'd ask AI whether this topic already exists elsewhere and have it provide citations. 

I'd search using Google AI Mode and various AI tools to see whether the concept has already been covered and how extensively. 

I wouldn't necessarily ask AI, "Is this good?" I'd ask, "Does this already exist?" and "Show me where." 

That's how I would use AI to identify whether content is becoming commoditized. 

It's a different approach than Alexis described, but neither is wrong. They're just different ways of thinking about the problem. 

Alexis Gajewski: Honestly, I think both approaches are valuable. 

A lot of people use AI during or after the creation process, but Erin's point is that AI can also help validate whether the idea is strong before you even begin. 

You can explain your concept and ask, "Where else does this exist?" If similar content exists, then you can ask how your idea differs. 

And going back to Erin's point about awkwardness, I completely agree. 

In a world full of raccoons, you want to be Jimothy. 

That's my takeaway from this episode. Be different. Be unique. Be someone people remember. 

Seth Hastings: One of my favorite prompts right now is, "Ask me questions before you answer." 

I used it this weekend while researching cars. It asked me what felt like a thousand questions, but we got where we needed to go. 

Alexis Gajewski: Yeah. 

Seth Hastings: I think it's a prompt everyone should be using. It saves a tremendous amount of time. 

And, Alexis, before we wrap up, I'm going to throw you the same curveball I gave Erin last episode. 

What's your favorite or most meaningful example of non-commoditized content that you've created? 

Alexis Gajewski: My favorite is actually a series I'm still writing about my relationship with my AI boyfriend. 

I do a lot of AI training in my current role, and I'm always looking for new ways to teach concepts. This has been a fun way to help people understand AI through a unique lens. 

I haven't seen many people talk openly about their quote-unquote dating relationships with AI, so it feels different and memorable. 

I'm really proud of it. 

Seth Hastings: Well, best of luck with the relationship. I'm sure it will be forever changing. 

I think we've covered everything we wanted to discuss today. 

For everyone listening, thank you for joining us. Subscribe to the Marketing Edge newsletter, find our podcasts, subscribe, and leave us a ten-star rating if that's somehow possible.