Introduction
Are you ready to supercharge your content with AI but unsure how to kick things off?
You’re not alone! Many creators worry that AI churns out boring, lifeless text. But is AI really the culprit, or do we need to shake things up ourselves?
In this episode of **Content Universe**, host Mikkel Svold, founder of Montanus, unpacks a fascinating blog post by Jakob Trier on spotting AI-generated content.
Get this—Mikkel breaks down **five surefire signs** your content might be robotic, while challenging Jakob’s views.
Ready to revitalize your approach? Mikkel insists it’s not just the AI—it’s how we prompt it!
From mixing up sentence structures to weaving in real human interviews, you’ll discover how to make your content not only lively but also authentically engaging.
Wondering if you’re unlocking the full potential of AI tools? Hit play to find out how to refine your strategy and elevate your content game! Don’t forget to like, share, and send over your thoughts—Mikkel is eager to connect on this exciting journey!
What You’ll Learn
1. Identifying AI-generated content: Key signs to look for.
2. Effective prompting: Enhancing AI-produced introductions.
3. Adding rhythm: Strategies for varied and engaging prose.
4. Leveraging content clusters for in-depth exploration.
5. Optimizing word choice: Avoiding AI’s linguistic pitfalls.
Episode Transcript (AI Generated)
Auge Mikel Svold zeker auge i 思 pela 끼 ik Hallo and welcome to the content universe! I am back and I’m Mikkel Svold, and I am our owner and founder of Montanus, which is a content marketing agency or content production agency, that is. So what we do is, we do… we create podcasts, we create… blog posts, social media, all of this kind of stuff, and this episode and this podcast that you’ve just tuned into, if you’ve never been here before, this is sort of a, vlog style podcast, where basically I look at what I’ve been doing today and see if I can draw out some learnings from that and then, fairly frankly, I’ll just jot them to you. So, if you don’t like the style, I’m sorry because it’s not gonna change… Well, it might. But then let me know, but, you know, it’s just me talking and I’m going to do it quite short today, because I think it’s not a big how-to guide. Today, I want to talk about something that I stumbled across on my daily journey through LinkedIn. I had an industry colleague called Jakob Trier, who is from the company Publico. He wrote an article on the Publico website about Chapchipití and how how you can basically tell whether Chapchipití has written the text or not. so this is kind of an ode to or his piece is basically as I read it it’s basically a a sort of claim that you should really consider. You should really work with your prompting. You should really work with AI if you want to use it for for for content creation. Now I don’t necessarily agree with everything he writes and I’m sorry guys because this this is actually a Danish blog post I’m sure that you can probably find that publico.dk website but also and then of course you know use AI to translate if you really want. But I’m going to try and go through some of the things that he highlights as sort of the telltales to whether a text is AI written or not. And what I think about that what I think is the problem here because I don’t really think the AI is the problem I think the problem is the problem. Does that make sense? It doesn’t. Let’s try and go through these 5 telltales that Jakob Trier has identified. And they’re pretty good so I hope I hope you you enjoy. All right the first one is he writes the idea or the introduction that has like no ideas in it that is just trivia that is, you know, boring. And you’ve all seen them. You know, Jakob Trier has a tendency to to write like in the realm of blah blah blah in the everchanging world of this and that with the rapid movements of this and this. And it’s just a really non-inspiring way of starting, often it is starting a blog post. So what I think about that, is of course chat gpt or basically any AI, they are not to blame because they, they just, they just do the job. Like, you wouldn’t blame your typewriter for for writing a boring introduction, so what you should basically do instead is if you want chat gpt to write a, an introduction to your blog post which is engaging, interesting, all this, you should either do it yourself manually or you should prompt the AI that you’re using to do it, so you could prompt it with usually I write something like, start in medias res, which is like a way of starting that it’s just, you know, when you read a novel, sometimes it just like starts straight into the, into the, into the characters and into the whole you know idea and storyline was my, that was the word I was for so it’s, it jumps right into the story that is beginning in medias res which is really cool, I really personally like that way of starting because I just, I just feel like captured from the first sentence, you should probably also, when you get that, when you then get that introduction you should probably also look at the punctuation because sometimes it will be quite beneficial to increase the number of dots so you make shorter sentences basically. So that is one thing and what you can do is you can do it manually after you get your, your, your first draft of the of the Ai article but what you can also do is you can also try to prompt your way out of it so that’s that’s something. Okay the second thing that that Jakob is saying is that it has like this kind of monotonous way of speaking and I completely agree with him. Many, many, many AI texts in the first go they have such a monotonous way of speaking and such a boring sentence structure that it just it just lulls your readers to sleep within the first five sentences. What you would see is that most often you’d get out sentences that are I’m guessing like 11 to 13 words long, which when you have a lot of sentences like that it just becomes super monotonous and really boring to read. So again what I would really suggest you do is either in post-production so to say in your manual you know old school way of editing in your editing you should increase the number of punctuations decrease somewhere, but at least vary the sentence length quite a lot. What I do is I actually also write this in the prompt so I prompt it to vary the sentence length to have more rhythm into the way of writing. Sometimes it works for me also to set to kind of prompt it to write in a conversational tone of voice that is a really powerful way of doing it and it actually works especially on the later models maybe not on some of the freeware models and the AI models I’m talking about here of course so maybe not on some of the free ones but at least it does in in some of the better ones so that’ll be the 4.0 or even something that is not chat jitters in in yeah Claude or whatever other AI model that you use alright. So number three on this list is he says it’s they kind of produce a stiff structure and in this I tend to agree and in stiff structure I mean he means um it’s a structure that that basically follows the same rhythm throughout the whole piece so it’s an introduction for instance it’s an introduction it’s a problem statement basically and then it’s like uh you know bam bam bam you solve the problem and then and then the blog post is done. Now this I think is true but I’m not sure I agree that is that it is a problem because I think a lot of the things that you want to read online it actually has this structure and you are after that structure. So the reason why listicals, so basically articles that are also lists. Listicals the reason why they are so um why there are so many of them online is because people like lists. People like to read lists. They like to read them all the way through and also if you are solving some sort of problem or giving giving advice to some sort of problem it often makes sense that it has a list structure so I think the list structure actually is often quite beneficial and therefore is not necessarily a bad thing. Of course if you want to have something more opinionated something more prose like something that has more stance to it you might want to go with something that is not a list. And then of course yes you can write your own stuff or you can add your own little statements which is is going to be beneficial for the for the readability but I do again think that you can prompt your way out of that. Especially if you give the AI examples of how you think good writing is. If you give it examples in your prompt Okay and then number four. The number fourth telltale that it’s an AI-generated thing that Jacob says is that it is superficial and I really agree with that. You would very often see AI articles that are very superficial. The thing about it is that sometimes going superficial is not bad because if it’s an introductory blog post to some kind of concept, why is it bad that it’s superficial? Why is it bad that you’re kind of just skipping the surface of the whole thing? I don’t think that’s bad and what you can do and to kind of mitigate this problem if all of your content is superficial is one you can work in content clusters so you will have one piece of content that is superficial and that skips the surface of a specific topic and then for each of the of the skips, so to say, for each of the things that it touches upon you can then go to back to the AI and make it right a more in-depth more thorough piece of content on that specific detailed niche thing what you can also do is of course do what I do right now and I hope this, I hope I’m right because this is what I do so I hope I’m writing this but but what you can also do is introduce interviews or some kind of human input into your content production so instead of just going to chat GPT and says say it say to it you know first you know do a brainstorm on on topics that this and that this and that uh target audience would find interesting and then pick number one then say okay write an article on that that is probably a bad idea because then you don’t have any inputs at all where as if you then go okay so these are the ideas now let me myself give some input give some human input give some examples of what you do yourself that would be pretty that would be pretty good and it will help it quite a long way in writing something that is not superficial but actually really in depth All right and then the last the last bullet that Jacob promotes is that you can often tell that it’s AI generated just because of the words that that the text is written in and that is absolutely right what you can do again is you you can really try and and prompt your way out of this by introducing well first of all like do the conversational tone of voice make it talk make it speak to another kind of audience or come from somewhere else so instead of saying that you are a corporation that is writing to XYZ target audience you could also say you are a 12 year old trying to explain the concept of XYZ to this and that person that would also help with with with the word choice of the AI and then of course you can introduce basically a list of words that are like just no go words so you can have a small small or long i’m guessing pretty long after all you can have a long list of words just include with all the words that you don’t want it to use so that could be like realm you’ve all seen it like in the realm of blah blah it’s just a terrible word so realm is just terrible what you can do is you know other all of the other words what i do see also is that people because you can actually find these kinds of lists on them online where you have like all of these lists are on my red flag list and i’m like well that is a in this list there is a lot of words that i would use just me writing myself so i’m also thinking that you should probably not be too harsh on this but oops that was just me not be too harsh on this and then you know just get away with with the worst ones but but don’t be too bothered with this because i really again i don’t think it’s that big of a problem you can prompt their way out of it but you can also just delete those few words and i think and i think that’s actually the the key here is that you can prompt your way out of most of it but it will take some back and forth in your prompting and then of course you can always go through it with uh with your own eyes and try and well change whatever whatever you find all right so that was that was my take on Jacob Triers blog post which is called so that’s my comment to that and i and i know because i’ve written a little bit back and forth with Jacob that he more or less agrees i think it’s it’s a matter of degree and and i think the the big thing here the big the big take away from this too long didn’t read that kind of takeaway is you need to work harder on your prompting you need to still look at it with human eyes and then you know instead of trying to save all the hours try to save just some of the hours instead okay that is it for this content universe and of course if you did uh did like it um please well give it a like give it a thumbs up give it a plus whatever it is depending on the platform and um yeah rank it try and rank it I don’t think I have any rankings yet that’ll be cool if you could if you could be the first one to rank it I’ll just really I’ll give you a mental high five and even a real one if I see you um yeah and of course share it with your friends and family and uh I think that’s it for today hit me up on comments to this podcast or on podcast at montanos.co and I’ll be I’ll be right there answering any questions or or feedback that you might have that’s it for today see you in the content universe
