AI "Tells" & What They're Costing Us

Speaker A: Before AI, I never really used EM dashes in my writing. However, I recently started weaving them in because in many cases they really are the perfect tool, the perfect punctuation for the job, or in my case, sidequest. That said, I use them primarily in my personal writing, uh, creative writing, fiction, that kind of thing. I rarely. I never use them on my website or on my clients websites because what I find is they can be perceived as, and often are, a sign of AI generated content. In fact, someone in a networking call recently shared that she avoids EM dashes entirely if she's added them in, uh, due to habit or, you know, good writing practices, she actually has her assistant take them out. Now, she doesn't have anything against EM M dashes, but she shared that she does not want people even having the inkling that she used AI to write. So here's what we're going to be discussing in today's episode. Where do we draw the line with the AI tells? What are the tells, and are they costing us something? Because from my vantage point, it's come to this. You can't em um, dash anymore. You apparently don't want to delve or dive. That was gone long ago. We can't genuinely, quietly do anything. And we most certainly don't want to. Truly or honestly. Definitely don't honestly. In fact, it's best to avoid adverbs altogether. And what about writing lists or using emojis or the rule of threes or a semicolon? At what point does avoiding these tells stop being good advice and start being a slow surrender of your voice to a tool that you might not even be using? Surely there have to be other ways to ensure that you're not coming off as a flat, soulless, generic person. Surely we are not meant to eliminate punctuation altogether. Come for my Oxford commas and there will be hell to pay. Just to assure that we're not robots in this episode, we're going to get into it all. And side note, this is my first time doing video, so awkward. Uh, that's why. But either way, it's going to be a really fun episode. Welcome to Small Business Casual, where we take the dream we're building seriously with a, uh, hefty dose of laughing at ourselves along the way. I'm your host, small business copywriter Emily Aborn, and I'm here to help you answer questions like, does my business need a lead magnet? And how the heck do I talk to people at a networking event? All while making it fun, and, you guessed it, keeping it casual. Okay, I will not tell you any lies. This was quite an episode for me to put together because, number one, things change very, very fast when it comes to AI. And number two, because I like to do, uh, I started with one thought and went down this crazy rabbit hole. Pros and cons. And then this side of the coin and that side of the coin and how to condense and oh, me, oh, my. And number three, I genuinely, genuinely, honestly, truly, uh, I go back and forth on this stuff on a moment by moment basis. Do I surrender or do I resist? That said, I really want to try to keep this succinct. So as succinct as possible I shall be. And if we need a part two or part three, we shall have part two or part three. Real quick, before I get into the depths of AI, I want to absolutely, exuberantly and with gratitude, share a review from a recent podcast guest and a bit of a celebrity, in my, my opinion, maybe to. Um, These incredibly kind words come to me from Alex sanfilippo, founder of PodMatch. And this is what Alex has to say. Emily is a gift. This is such an incredible podcast. As an entrepreneur with a very small team, I resonate with so much of what Emily and her guests share. Thank you for this podcast. Emily is awesome. Thank you so much, Alex. And if you find it in your heart, dear listener or viewer, to write me a review, leave me a comment, leave me a review, give me, uh, a five stars, that kind of thing, I would just be over the moon. Um, fun fact, Alex is going to be a guest on my podcast for a longer episode than what you heard in the collaborative episode in August. Stay tuned for that. And P.S. i love POD match and PPS. I also love reviews, as aforementioned. So if you haven't written one, today could be your day. Today could be your day. Um, but that's not why you're here. You are here because you want to know about the AI tells. And that is what we're going to do. Here's where I will take you today. Number one, we're going to talk about how people are kind of feeling right now about the use of AI in content. I take in a lot of opinions on this stuff, so I'm going to share some of those. Um, two, we're going to talk about what are some of the tells, how can you spot them, and does it even matter? And number three, the ever faulty AI checkers, they're not even sure what AI wrote and what AI didn't write. So that's food for thought. Um, and then number four, we're going to talk about a serious issue for creatives and entrepreneurs and what we can actually do about it. So, like I said in the intro, I brought up the fact I was going to be doing an episode on AI Tells in that networking call. And, uh, the woman in the groups shared like, she definitely does not use EM dashes. She takes them all out, which I will define in EM dash in a minute. Don't worry. Um, for those of you who don't know what they are, and there were a few other things that she shared, she avoids as well, because it can signal that it was written by AI, even if it wasn't. Now, the fact of the matter is, is that while there are a of AI giveaways, um, those giveaways came originally from humans. So hear me out. They came from human writing. I, I never want us to forget that. What happens now? What happens is that AI kind of like scoped out the Internet and read it and this, that, and every other place, and they took from us. And now we as the humans, we are avoiding things that we have done for years and years and years and years so as to not be mistaken for the AI that took from us. So at this very moment, that's kind of where we stand. It's a little weird. And it's gonna get weirder. Um, I will say it kind of appears from the general tone of things and some articles I read, uh, that the split on like, what people think about it is basically like, still kind of. Some people are adopting it and some people adamantly opposed. Like, I thought it was more like 50 50, but I read a stat that 50% of Americans express more concern than excitement about AI. 16 to 34% are excited, and then 1/3, 33%, still feel mixed. So I don't think it's really that 5050 after all. I think it's more like 83. You know, like some of us are still like, I'm concerned or I'm really mixed. And I would safely put myself in that category somewhere between mixed and concerned. Mostly concerned. Another story for another day. But I do believe these things will take over the world if they are not reined, um, in a little bit. So surveys show also that a lot of people are distrusting of AI generated content. However, that has not stopped more people from using AI generated content in their to even in their text messages. I have been suspicious of somebody using it in a text reply. Um, also I'm 100% certain that one time I Got a birthday card written by AI and it just was like, not at all how this person talked. And it was so AI generated, and I hated it. Felt like, like, don't even get me a birthday card is how I kind of felt after that. So what's happening is this. While you are out there, um, perusing for content and reading something and listening to something, your brain and customers and your listeners brains, where we're on the alert for, like, fakeness, okay? We don't want things that are eroding our trust. We talk about that a lot on this podcast. But when we see or when your clients see or suspect AI, depending on where they kind of fall on the spectrum, their spidey senses go up and they think, okay, what are they hiding? Why couldn't they be bothered to do this themselves? Why are they too cheap to pay help to do this for them? Why would I be bothered to read or listen to something that they couldn't even put the effort into making? So I'm going to tell you what happened to me recently. While I was preparing for this episode, I read an Atlantic, uh, article called why AI is. I cannot pronounce this word. Incorrigibly didactic. It was a very long article, and I knew that I was going to be reading it, like, a couple times over again because I wanted to have the information to prep for you, but I first wanted to listen to it in my ears on my walk. So I hit listen on the little player thing. I knew it was going to be AI but it's like, fine, whatever. Some of them are fine. This one, oh, my God, it was boring as all get out. And I zoned out completely. I could not follow the text or the article at all. So, like, after five, six, seven minutes had gone by, I was like, oh, my gosh, I just wasted all that time. Like, I have to read it all again anyway, because I was not, not even paying attention because it was so monotone and just, like, boring. And this is what I think is happening when we're reading things too. Um, to the point of that article, actually, AI is really great effects and information and making everything into, like, a point and a moral and have a goal or a strategy. But it's not really great at creativity and being able to come with something with, like, a soul or generate something just for the sake of generating something. Like, sometimes we as humans, I don't know if they know this yet, the robots, we like to just have fun. We like to just do stuff because we can do stuff. You know. So, anyway, um, I think that this is why it's important to be aware of the AI tells at all. Because whenever you can, without overthinking every single last thing that you say or share or write, you would probably want to avoid them if you can, because you don't want people thinking that you are using AI. Now, if you use AI freely and vocally, this obviously does not apply to you. And if you all. If you don't care, it obviously doesn't apply to you either. Um, I also want to say I'm not trying to be the A.I. police. Okay. Um, Andy Lyons actually weighed in on a post that I put up on LinkedIn asking people about their AI tells and things like that. And she raised the importance of easing up on some of the judgment around this stuff. Because for sure, there are people who are too anxious or too unsure or too uncomfortable to write publicly, to speak publicly, to share publicly. And now AI is allowing them the ability to be able to do that. Maybe it's helping them organize their thoughts. Maybe it's helping them write when they couldn't write before. So folks with learning differences and neurodivergent folks and people who maybe freeze up at a blank page or face, major imposter syndrome. This could be their kind of gateway into sharing something. So, again, I am not. I thought that was a great point. Super valid. And I'm not trying to, like, recommend that I or anybody else walk around the Internet or monitor every single person in our network for, like, AI tells. Okay. I, um. At this point, I. Because I see them so often, I can just tell both with the tells, but also just. Just. I can just tell. I mean, there are just certain things that I see as a writer or that I know about the people that I'm interacting with that I notice and I catch. And I know it was done by AI, but in addition, sometimes it's just like an overall sense or feeling. And we'll get into that a little bit, too. There's a spider on my wall over there. Okay, so let's get into the actual tales so that you can be aware of them, because they may be costing you, and they may be costing you trust, uh, which means that they may be costing you business. And I do not want that for either of us. So I have ten of them. Um, all right, so first we're going to start with the star of the show, the EM dash. If you don't know what they are, it's like a long dash that sets phrases and asides apart from the rest of the sentence. So it's kind of like another way to parentheses or, um, ellipses, which are the. As my friend Lindsay Dreyer calls them. It's so cute. She calls them dot, dot, do ellipses, or parentheses would be your other options there. Or you can even offset it with commas sometimes, depending on the situation. Anyway, ChatGPT kind of started it with the em m dashes. They just aggressively used em m dashes. And now people who use em m dashes get accused of being AI, which is so stupid, because they're actually a very valid punctuation tool. And they show that you are adding an aside or that you're going on a little, like, side quest with the reader. And they can be very fun and helpful and. And, uh, I took them from us. Okay, second tell. This is the one I see most often is it's not X, it's Y. So, like, it's not failure. It's a learning opportunity. It's not you. It's your strategy. You're not broken, you're blah. You know, so it's not X, it's Y. That's the second one. Next up, the rhetorical question. The worst part, the honest truth. And then it goes on to, like, continue the phrase the honest truth. You didn't belong in that room anyway. You know, like that kind of thing. Also, we'll get into the sass in a moment because that's another thing. Third one. Third three. Uh, fourth one. Sorry. The rule of three. Uh, this breaks my heart. This is a highly effective rule in writing. The rule of threes is so wonderful. Creativity, marketing and entrepreneurship. Spoons, chopsticks, sporks enter in for fun games and adventure. Yeah, people say it. They say it all the time. And then now it's starting to signal AI because AI has copied this from us. So I've been doing this since I was, like, knee high to a grasshopper. Um, but now you kind of want to watch yourself sometimes. I'll use four now because I don't want people to think it was done by AI number 5. This is a big one. Uh, no X, no Y. Just see? So, like, for example, it would be like, no hype, no fluff, no noise, just peace. That staccato negative space where, like, good concepts become obviously, AI number six adverbs. This one gets my goat. Now, realistically, we shouldn't be using a ton of adverbs anyway, when. When we write, according to all the greats. Okay, we're. But obviously I like them. I've just used two in a row. I use them a lot. But the ones that raise red flags currently, and obviously this is just going to change. But the ones that raise red flags currently are genuinely, honestly, truly, actually. And quietly. Quietly. I'm saying that one a little bit louder or quietly because it's loud. It's a big one. It always says quietly. Meanwhile, you're quietly doing this or quietly doing that. Anyway, I can't stand it about the adverbs. It's like a whole part of speech that has just been taken from us. Thank you very much. AI, you may return to where you came. Number seven, there's some specific vocab delve. We know that one dive. Straightforward stuff like that. There's some specific vocabulary you can look for. Curated is another one. Um, aligned. It loves the word aligned, which is so sad because I like the word aligned. You know, um, genuine, as I already mentioned, authentic. Loves authentic. Uh, number eight, apparently even emojis and lists. I'm sure that this is incriminating me somehow right now with this list. Uh, and then the sass. Number nine, don't even get me started on the sass. Now, I personally think that the sass is impacting the entire Internet, including myself, and making us more sassy and more opinion riddled than we perhaps were before. Um, but specific to AI, Forbes calls AI's tone Ted Talk style. So like punchlines built for drama. And I've got to include the link to that article because it was really good and it is full of even more tells. So do please, um, read that article that will be in the show notes. Uh, and then number 10, lastly, but not leastly. See, I do love my adverbs. Um, this one is kind of just like the structure. So they're not doing. They're not just doing X, they're doing Y. So the examples would be like, Amazon isn't just buying content, they're buying credibility. This isn't just about revenue diversification, it's about survival. And it's always just so dramatic. It's so dramatic. Okay, so what are we supposed to do about this stuff? Do we avoid these things entirely? Do we sound completely, ridiculously over the top? Like I do. Do we rebel and just use adverbs and EM dashes like they're going out of style and. And use the rule of three? How rule of three? Hell, oh my God. You know what I mean? Okay, uh, before. Or do we use typos to kind of like throw it off the scent right before you race to these things? As your solution. Um, I would love for you to listen to this, this little excerpt from the Atlantic's article. I, uh, will include a link in the show notes. It is behind a paywall. My husband gifted it to me for the purposes of this episode. Um, and I'll link it so that you have a link too. But I'm going to read you an excerpt from it. Short. Don't worry. Pretty short. Okay. In the late 19th century, it was commonly believed that a criminal or lunatic could be recognized at a glance based on certain physiognomic tells. Enormous jaws, high cheekbones, and other animal like features. These were signs of an irresistible craving for evil for its own sake. Today, savvy readers use a similar approach to identify AI writing by hunting for supposed telltale signs. The EM dash and the it's not X, it's Y construction are the prognathous jaw of the large language model, betraying it hidden in humanity. The problem in both cases is that you can't always deduce what's inside from what's outside. A person might have enough rough features and a kind heart, just as a writer might use EM M dashes despite being human. It's not a giveaway. It's a style choice. And he did that tongue in cheek, because he used an EM dash and did the, uh, the AI styling there. So he goes on to say, as AI models evolve, their ability to mimic human writing is sure to improve. People have deliberately begun to make spelling errors to show that they are not chatbots. It is only a matter of time before the chatbots, uh, learn to follow suit. And that is exactly it. This a, that Atlantic, uh, excerpt in regards to AI it's exactly what I'm talking about. It also goes. There's another one too, I read, which talks, um, about these patterns being the sign of a problem, but not the actual problem. So I think that's important to remember too. Like, these aren't the problem, they're the sign of a problem. Um, that brings me to last point. So number one, paying attention, not to last point, but two points. Uh, number one, paying attention to all this stuff can have you overthinking it can have you washing out, watering down what you're trying to say just so that you can make sure that it does not sound like AI or you might over complicate it or just try on some like, wild out there personality like be like cuckoo for Cocoa puff so that no one thinks that it's accidentally AI. And so question number one is kind of like where do you draw the line? Like, do you want to overthink it all? Do you want to avoid everything? Uh, do you want to be really, really out there? Where do you kind of draw the line? Um, cut the EM M dash, then delve, then adverbs and the rule of three, then the emojis and the list. Obey every rule. And. And soon, before you know it, you've sanded off everything that sounds like you, you know, and then you've gone ahead and surrendered your voice to a tool that you might not even be using. So I believe that the more that everyone scrubs the same exact tells from their language, the more uniform all of human writing is going to be. And not for nothing, if we keep on scrubbing, the AI is just going to learn to navigate the way that we're navigating, and then we're all just. It's all just going to be identical blandness. And we're gonna have to really work hard to prove that we're not being bland. So think about that for a little bit. Uh, number two, I recently ran some of my own work through an AI checker, and it came back on various. So I put a couple pieces through. None of them were written by AI, mind you. And it came back on various pieces, anywhere from 7% written by A.I. to 46% written by A.I. now, in all of those instances, 0% was written by A.I. so it turns out, turns out A.I. doesn't even know what is written by A.I. and what is not written by A.I. which makes me wonder, are, uh, the tells even worth paying attention to, or is it worth that much time and energy? So, uh, I want to direct your attention to three things that I think are really important for entrepreneurs and for creatives right now. Three things I think you really need to think about for yourself as you move forward. Number one, if you decide to use AI, fine, but don't let it become a tool that you use so much that as such a crutch that you now have brain rot. Uh, and that you cannot come up with ideas for yourself or think for yourself. This has become widespread and problematic. Uh, a fellow podcaster who I'm connected to on LinkedIn, Joe Casabona. He talks about how he is using computers for computer things, but not letting AI be part of, like, making the final decision. Joe also has an article called 3 Signs yous've Got AI Brain Rot and How to Prevent It. And I will make sure that is linked in the episode Show Notes. It's super, super good. Um, but really getting Clear with yourself on where you draw the line with coming up, um, on ideas for your own and where you bring in AI Number two. I kind of alluded to this, but AI is going to become and already is, a snake eating its own tail. I called this for the record, with more and more AI generated content out there, AI is now starting to regurgitate its own AI generated content. And, well, that just as you can imagine, populates more and more and more of the the same. So it really is becoming like this circular thing, a snake eating its own tail. Number three, what I really want you to think about and look at is the heart and soul, the personality that you are bringing. And why in God's name would you want to outsource that? Why would you want to outsource who you are, your heart and soul, your personality? Um, are you just looking to save time? Because it's actually probably costing clients and money and time. Uh, so then does that become really worth it? Is it because you're stuck creatively? And if so, what did you do to get unstuck before AI? Do you need to bring somebody else in? Do you need another opinion? So the tells, for me, um, at the end of the day, I think that the AI tells are interesting to look at and consider and work on lessening as much as we can. But for me, the AI tells are actually surface level and semantics. The real difference between human writing and AI writing is not a punctuation mark. It's whether or not there is an actual person behind the words, an actual person bringing them to real life, uh, in written or spoken form. So you don't prove that you are human by deleting EM M dashes and getting rid of Oxford commas. You prove it by being the version of you that cannot be faked. So stop asking yourself how you look, how you avoid looking like AI and start asking yourself, how do I sound? Unmistakably like me. So, on, um, that lengthy note, I tried to keep it concise, but I did go a little bit long today. Um, I'm sure this wasn't and won't be the last that you hear from me on AI, but for today, it's enough. Next week we're going to be talking to the one, the only, Karen Kenny. We're delving into the topic of worth as we tackle the concept of charging your worth. There's a lot of fun stuff in there for you this month and hopefully always, as always, um, I will see you next week. Thanks for joining me on, on this week's episode of Small Business Casual. If you want more business advice that's a little quirky, not too buttoned up. One less accessory. And, uh, just right. Visit my website, emilyaborn.com and connect with me over on Instagram. Uh, Elyaborne. I'll be here, cozy slippers in hand, to help you design a business you love.