This episode features Interact’s CEO, Josh Haynam, Growth Manager, Jackie Aguglia, and Digital Marketing Manager and Host, Jessmyn Solana.
In this episode we will cover:
- What is Generative AI?
- Why use it in your business?
- What is the benefit to humans?
- How do you make the most of it?
- How is AI changing the game in marketing and online business?
Ready to create a lead generation quiz in one click using the power of AI? Join the waitlist for early access to Interact’s AI-powered quiz generator.
Jessmyn:
Hi, guys and welcome back to Interact’s Creator Stories Podcast. So great to be with you all. As always, I’m your host, Jessmyn Solana. And today, we actually have a very special episode with CEO Josh Haynam. And to clarify that is Interact CEO. And for the first time on our podcast, we have Jackie Aguglia, our Growth manager, if I could speak properly. Jackie, it’s actually your first time on our podcast. So, can you really quickly… I mean, I’m sure people have probably seen you and all of our videos have talked to you, you’re everywhere.
But just in case, could you give us a little bit of a spiel of what you do at Interact and your role?
Jackie:
Yes, ma’am. Excited to be here, everybody. Hello, hello. You may have seen me on the YouTube channel. So, most recently, I have been creating YouTube content for us on how to use quizzes, how to use AI. Most recently, all different things, different questions that I’ve been asked by customers, I’ll turn that into video content to help you all out. Also, manage the affiliate program and our partnerships. We’re dabbling in sponsorships a little bit as well.
So, hit me up if you might be interested in that. And still, in the community as well, so helping you all build your quizzes in our community, The Quiz Collective.
Jessmyn:
Love it. And today, I love that you mentioned AI already because that’s what we’re going to be talking about. I know it’s been around for quite some time, obviously, but I feel like it really started to boom with the release of ChatGPT somewhere around the holidays, or at least that’s when I saw it. So, I guess really quickly guys, and we have some questions that we’re going to get through, but before we jump into everything, Josh, could you just describe what is generative AI and what’s the difference between ChatGPT and OpenAI and how do they actually relate to each other?
Josh:
Yeah, for sure. So, people on the ML side of the spectrum might make fun of me for my overly simplified explanation, and then folks on the marketing side might be like, “What the hell are you talking about?” But I’ll try to land somewhere in the middle from my understanding. Basically, there’s AI models, which you can think of as like AWS, which is where you store everything, and AWS is what runs the internet right now. You watch a Netflix video, Netflix is not storing that video, AWS is. And then, it serves through Netflix so you can watch it in a seamless experience. Similarly, AI is… and the reason why Netflix does that is it’s extremely difficult to efficiently store and serve a movie.
If you were to do it yourself, it would be super slow and lag all the time. I’m sure everyone remembers when my YouTube videos didn’t load and you’d have the stupid spinning thing. So, basically, you need a really strong engine that you can tap into. Similarly, with AI, there’s models like OpenAI, which is the one behind ChatGPT, and the one that a lot of tools are built on top of that allow you to tap into OpenAI’s engine so that you don’t have to build that model from scratch.
And the way that model in particular is trained, which most people are using that model is trained by reading the internet. So, mostly Wikipedia and academic articles. And then, it pretty much… it’s like the world’s most advanced autocomplete is really what it is. So, you give it a prompt and then you can tell it like, “Complete this.” And it does. So, then the tools that are built on top of it.
So, that’s like one layer. You have the model layer. The next layer is the tool layer, which is where names that you may have heard of. Like Jasper is a popular one out there. And then, Interact is building our own AI on top of the OpenAI platform as well. Those take that engine and they fine-tune it, is what it’s called.
So, again, the engine analogy. You fine-tune it with your own data so that you can make an Interact style quiz using Interact AI or a Jasper style blog post using Jasper AI, or you can have different tones and things like that that you can train the model on. So, you can adapt that model to your use case. And that’s where most of these tools come from and how most of them are powered. There are other models besides OpenAI. It’s not even necessarily the most advanced one, it’s just the one that’s accessible, I guess. And so, that’s what I understand so far.
Jessmyn:
But even though it’s not the most advanced, you’re saying that you’re feeding information into it, so wouldn’t it learn as it goes and probably become better?
Josh:
No, that’s the thing that is funny about this, at least to me. I’m like, “None of these models are learning.” They are not sentient. They cannot think outside of this specific prompt that you give it. They’re not teaching themselves. AI is not taking over the world anytime soon. It’s going to be a very long time before we get anywhere near that. Basically, if you think about levels of thinking, you have the basic level of thinking of like, “What can I do with these blocks?”
That’s one level of thinking. The next level of thinking is like, “Oh, what would I do if I combine these blocks with those rocks over there?” It can’t do that. So, as soon as you get to level two of like, “Oh, what happens if you combine two things,” it will fail. And so, it’s not teaching itself. You can only tell it like, “Oh, these are good blocks.” So, use more of these blocks. You can’t tell it like, “Oh, now go combine these blocks with rocks and make something better.”
It failed at that point. And that’s not to say that it won’t eventually get there. But in this version and from what I see, it’s a long-ways off before we reach those higher levels of thinking.
Jessmyn:
So, it’s not going to be iRobot where they become humans basically?
Josh:
No, I don’t think so. Basically, if it has a finite set of things to figure out, it can figure those things out. But as soon as you’re like, “Okay, now associate this with all these other things,” it doesn’t know how to make that transfer.
Jessmyn:
Right.
Jackie:
I can give a blog post marketing example to that. I was playing around in one of the AI tools the other day. Actually, I think it was Jasper AI. They have a content improver. And what I was trying to get it to do was… I keep calling it translate, but rather give me a blog post from a YouTube transcript that I put into it. But you have to put in little bits of the transcript at a time, so it only would read 800 characters of the transcript and then it would transcribe it, right?
So, it was copying and pasting and putting it all together. But when you read the finished version, it was very much so these were all great paragraphs. But when you read them all together, it was like, “Whoa, this obviously did not connect the point that I started with up here and then set again down here.” So, it felt very disconnected as a whole. So, I think that analogy, Josh, reminded me of that experience because it really couldn’t consider it as a whole thing. It had to go individually, chunk by chunk.
Jessmyn:
Yeah, I love that you brought that up because I think one of the questions for us internally was, is this actually saving us time? What is the benefit of AI or the generative AI to humans, especially in their businesses?
Josh:
Yeah. I mean, the thought that’s popping through my head is you’re saying that is, I get very salty about this, but the tech bros are already branding this term called Prompt Engineer, which is literally what you just said, Jackie, of just learning how to prompt it in a way that’s actually effective. You tell it, write a blog post about X, it’s going to be super generic and terrible.
But you tell it, write a paragraph about X in the style of my writing, and then you put an example of your writing, then it does really well. And that’s all Prompt Engineering is. So, if you listen to other podcasts and stuff, and they’re like Prompt Engineer, that’s all it is. So ridiculous. But that’s what I think. I mean, and at least from my perspective, I don’t know if either of you are video game people, but there’s a video game called Age of Empires that is built by Microsoft, and Microsoft has a deep partnership with OpenAI, so similar AIs.
And playing that game for years and years, you learn what the AI is good at. You give it a set of parameters and a pathway to set of directions, it will execute on those directions better than a human ever could. So, you can’t beat it if you just try to go head-to-head, but you can beat in by tricking it, doing something that it doesn’t understand. And so, that’s what I think is the way to use AI effectively, is learn what it’s good at, have it do those things, and then learn what it’s not good at. And then, you do those things. That’s my opinion on it.
Jackie:
Yeah, I agree. I’ve been dabbling in quite a few of the different platforms and just even seeing what’s out there, and I think right now anyways, the best way to use AI is to help you through the ideation process. Create an outline, create a shell of your content strategy, what you’re wanting to talk about, and then fill it in with your own expertise, your own stories, your own brand and offerings and all of that. Because it does… I don’t know if it’ll ever know, but for sure right now, it does not know your customers and your company as well as you do.
So, I think no matter what you’re putting into AI, you’re still going to have to embellish what it gives you back. But I do think it helps you get there faster.
Josh:
Yeah. And I would add one more thing too that we just learned in building the Interact AI the last couple of days we’ve been working on it, is you can teach it, right? So, it’s the best… I don’t know the analogy here, but it’s like, it’s the best student ever because it doesn’t have any problem with… teach it one thing and then ask it to do it 100 times. Whereas a human would be like, “Oh, that’s boring. Why do I have to do this 100 times?”
But for example, you write a perfect paragraph or your style of writing, or you write a perfect quiz once, and then you say, “I want five other quizzes that are written in this style for these other topics,” it does very, very well at that. Like almost spot on to how you would write it because it is writing how you would write it. But I think actually my opinion is starting to shift to the best way to use it is actually to do one thing really, really well once and then have the AI multiply your effort.
Jessmyn:
Right. Basically. I mean, killing two birds with one stone is not the right term, but you’re pumping out more in a quicker time period than you would if you were to handwrite it yourself. And by handwriting, I mean typing.
Jackie:
Yeah. I think of… when you give that example, I think of social media posts. Everybody’s always like, “Oh my God, it’s going to take so much time. What am I going to write about again? How do I say it differently?” If you have a great post, literally just go into your analytics and find your most engaged post or write a new one. If you put it into AI and ask it to rewrite it in that same tone, you can literally, with a click of a button, get hundreds.
I mean, okay, it’s 100 clicks, right? Because each time you’re clicking, you’re getting a new response, but it completely changes what it says, but it’s still saying the same thing that you wanted to say. So, I totally agree with that. It multiplies what you can do once you know what you want it to say, if that makes sense.
Josh:
Yeah. And the thing that people have been doing like influencers on Twitter and LinkedIn and whatnot, is they post the same thing, just slightly varied. And there’s Nathan Berry on Twitter. I like him because he just says that’s what he’s doing. He’s like, “I post this thread on Twitter once every three months, and it always does well.” So, it’s like, that’s how social media works anyways, so you might as well just… yeah, like you said, Jackie. And you can even prompt it.
Like, write me five versions of this. And then, you just put in your best post and you have it write five versions, or you could have it say, “Write me a post about X, about… I don’t know, working from home in the style of this other post.” And then, you paste in your other post that did really well, and then I’ll write you a post about working from home in the same style as your other post. And it just multiplies your output.
Jessmyn:
Would you need to already know what your brand voice is and exactly what your business is? Or let’s say you just got started three months ago, you’re still figuring out, you’re in the early stages of your business, you think you have an offer, you might have an offer, but you’re just launching in order to figure out what do people need. In order to use AI, do you need to be in a place that you know exactly what you’re doing, you know what your offer is, you know what your marketing funnel looks like in order for it to be effective for you?
Josh:
My opinion, 1000% yes. And it’s actually more important now. Because if you just go straight to AI, it will give you a brand voice that is incredibly generic. And everyone can tell it’s just AI. If you on the other hand, get to know your customers extremely well, learn what they resonate with in terms of writing style and social style, all that stuff, and you actually spend more time talking to people, then you can spend less time doing the boring stuff of like, “Oh, I need to write a template for this, or a template for that, or a post about how to accomplish this task.” That’s like, 90% of that was going to be taken from a template anyways, and now you can just spend zero time on it.
I think it’s actually more important now to actually spend the time understanding your customer base than it was before AI.
Jackie:
Yeah, I agree. And it’s just like starting a business without AI, you’d still have to know who your customers are. You’d still have to know what they need, what they want, why they would use your product, right? Because otherwise, you’re just putting something out there. Maybe it’ll stick. But if nobody knows what it’s for and how to use it and why would benefit them, they’re not going to buy it. So, AI is not going to change that.
Jessmyn:
Right. And I know, Jackie, you’ve been testing out a bunch of different tools specifically for AI, and we have a bunch of videos of them on our YouTube. We’re coming out with the blog post versions of them. Either they’re released or about to be released. In terms of, I guess you could say time wise, I know you’re saying we could do this faster. Exactly, what does that look like? How much time are we actually talking?
Jackie:
Yeah. So, for example, the last test that I just did was there is a tool called… hold on, gosh, it’s not… oh God, what is it called? ToWords, T-O Words, all one word. And I was able to take a YouTube transcript… I didn’t even take the YouTube transcript. I put in the link to the YouTube video into ToWords, and within three minutes of me doing nothing other than put that link in, it took three minutes for it to give me back the blog post. That is incredibly fast.
You really wouldn’t be able to on your own create a blog post from a transcript in three minutes. But the quality of that blog post was not something that a human would write either. And so, you just have to consider in the editing time.
But in terms of getting it into a place where it’s much easier to edit, I think that that’s where the time saver was. But then, that’s where you go in and you can add your own creativity because that’s a piece too of content creators usually like to create because they’re creators. And so, you’re not sparing your creativity by using AI either as long as you’re going in and embellishing it with what it is that you want to be creative about, your tone, your brand, your voice, all of that.
So, that’s one side. The other side is, we were talking about posting to social media before, right? You could have it prompt you with 10 versions of this same post. And usually with those, you don’t have as much edit work because the style of what you’re wanting to say and the information that you’re wanting to give is already there. It’s just rephrasing it. So, in terms of taking… how long would it take somebody to write, say, 50 social media posts?
You could do that in a matter of, I want to say seconds, maybe minutes. But I think literally seconds to just have it spit out 50 different versions of the same thing that you’re saying. I’ve also played around with using, you talked about new business owners and who are you serving and who do you want to serve and what do they need? I’ve even asked ChatGPT, “What are the fears of…” so I specifically was thinking I was an emotional eating coach, like a binge eating coach. “What are the fears of a binge eater?”
And it gave me a whole list of the fears that those types of people would have. Obviously, you can change binge eater out for anything. And so, that also saved a whole bunch of time in trying to figure out, “What angle do I want to give this?” Now, that’s not the final post that I would’ve posted. It would still take a little finessing, but it removes the… when you just stare at a blank screen and you’re like, “Oh my God, what do I want to write?” It takes all of that away.
So, as much time as you’re wasting on just staring at a blank screen and trying to figure out what to write, you can have AI help you get there, and then all of your time is spent creating, which is probably what you want to be doing anyways.
Jessmyn:
Right. And in terms of, Josh, I know we talked about this internally with our own blog posts and elaborate more on this for me, that what will make your content still stand out differently is the links that you link out to, the images that you include, and then of course your own flare of personality, which we keep talking about in terms of brand voice. But there is still some level of work that will be done or needs to be done in order for you to get to that final product. But this just speeds up the process.
Josh:
Yeah. Yeah, yeah, totally. And I think also stories is… this is a view of how this… stories are the only thing that’s unique. Everything else is already known. There’s nothing new under the sun. There really is nothing new. Even this AI stuff is just putting together old stuff that was already figured out. You put it together in a new way. It’s the same thing as when they invented tools. It’s like, “Okay, cool.
Now, we can save time.” But stories, you cannot replicate a story. Your story is different than everybody else’s story, experiences you’ve had. And so, that’s already what I think makes blog content unique or any content is the story. And then, your style of writing, your unique way of saying things. Yeah, like what Jackie said, it’s like, it gives you the starting place. You don’t have that, “Oh, I’m staring at a blank screen. I don’t know what to write.”
And then, I think depending on what content it is, sometimes I also think it’s better to start with one really good one and then have it replicate. So, we’ve talked about those two uses. But yeah. I mean, for us it’s like, infuse our stories that we have because again, going back to talking to customers, you’re not going to have stories unless you talk to customers. And I think that’s going to become super obvious. Companies that are just using AI to create content, there’s going to be no stories in it because they don’t have any stories.
They don’t have any customer experiences. They don’t talk to their audience. They’re just using tools and computers to write stuff. And that might work for a little bit. It might rank in Google or YouTube or whatever, but it’s not going to get engagement. And all those platforms want is engagement because the longer you stay on Google or YouTube, the more money they make. So, they’re just going to take your content out if it’s just made by AI.
So, I think, yeah, it helps you get past the blocks, but then allows you to focus in on the creativity way more. And I think that’s the general vibe that even machine learning engineers are doing that. In interviews for machine learning engineers, they’re asking you creativity questions. They’re not asking you necessarily just hardcore coding stuff. They already realize, and I think everybody’s realizing the next frontier is like, “How can you get creative with this stuff?”
Jackie:
And here’s a good use case for storytelling. So, I’ve used Munch, Lumen5 and Pictory are three options. If you, let’s say you already have a story. This is something… you’re establishing your business, you’ve already put your story out there, whether it be in a blog or maybe even a video. I mean, I guess with these three platforms, it would be video. Well, actually, no, I lied.
You could do either/or. But anyways, you could put your whole story in there and then have AI turn it into bite size pieces that you could put as a YouTube Shorts or on your Reel or as a post. So, the element of your story is still there because you’ve already created it, but you’re repurposing the same content that you’ve already done. Again, saving yourself tons of time.
And you don’t have to worry so much about the messaging there because you’re not asking it to change what you say. You’re just asking it to take it and put it into smaller versions so that maybe it’s more digestible pieces, or there’s a really good thing that you said at one point in there that would resonate with this certain side of your audience. Use that to get them to read the full blog post.
Or these people need to hear this message via video, so make a video on this little piece and then have that link back to whatever. So, that’s another great use case and way to save your time is you’ve already created the content, but the content you’ve already created into these AI platforms and repurpose it for different reasons in your business.
Josh:
And that’s the number one thing. Our waitlist is open for the Interact AI and everyone who’s signed up so far is like, “I want to turn my blog post in quizzes.” And so, that’s one of the main use cases we’re building it for. Put in your URL or put in a blog post, have it created a quiz.
Because you’ve already done that creative piece of you’ve created something valuable, now you want it as a quiz, you want it as a bunch of social posts, you want it as Shorts, you want it as… and how much time do we spend doing those things, which is just wasted time because you’ve already done the creative thinking, now you’re just trying to change the format. So, the AI is just really good at changing the format. And I think that’s the best use case for it.
Just do something once and then click a button, and now it’s a quiz. Now, it’s a social post. Now, it’s a blog post. Now, it’s a whatever.
Jackie:
Totally.
Jessmyn:
Something that I want to jump back to that you mentioned earlier, Josh, was being able to use AI to not only create content, but you could create, let’s say, templates off of it. Is it ethical to use AI to create a template and sell it as is?
Josh:
I mean, so Google’s official approach to AI generated content is like, “If it’s helpful, we don’t care who made this content.” If it actually solves the problem that the person’s trying to solve, then I think the same thing. In order to make something with AI that is actually worth buying, you’re going to have to learn a lot, either about prompting the AI, but I think more importantly about how to create things that can be replicated by AI. Because you think about what a template is, it’s a personalized version of a generic thing.
So, you’re going to have to get extremely good at writing that generic thing, which is super hard. We’re doing it with the quizzes right now. And it’s like for a result of a personality quiz, what does the first sentence contain semantically and how can I write that in a way that’s replicable a hundred thousand times?
That’s a really hard skill to learn. You don’t normally think like that because you’re not talking in a way where you’re like, “I need this to be replicable a thousand times.” And the most valuable skills that exist today, programmers, that’s why it’s so valuable because they write a piece of code and it can be used a million times or a billion times. So, now we’re basically just all becoming that. We all need to learn how to write things and create things that can be replicated.
And so, to me, I’m like, “Well, yeah, if somebody took the time to learn that skill, that’s an extremely valuable skill. And if they want to sell what’s created from that skill, that’s just how everything works already.”
Jessmyn:
Right, right. Question for both of you, because I’m curious to hear what you guys both think. From the consumer side, if I were purchasing, let’s say, swipe copy or downloadable freebies from somebody who used AI, and I’m going to use this in my own business, can I effectively just go to AI and do the same thing and just use it for myself?
Josh:
Yes and no. That’s where I’m like, the more I spend time learning to prompt the AI, the more I’m like… there’s so many layers to it. So, there’s the base layer of you go to ChatGPT, you play around with the prompts. So, that’s layer one. And layer two is like, “Okay, well now I’m going to prompt ChatGPT and give it an example to emulate.” And then, layer three is, “I’m going to go to the OpenAI API and create a fine-tune model, which means I need to give it 500 examples.”
And so, as you go deeper down those layers, it’s going to get better and better. And so, yes, someone could go and access layer one, but you’d still have a learning curve of what do you actually prompt the AI with to get a good response. And so, I think that’s where the value in these tools is from.
It’s like, they’re on layer three, layer four, layer five, now they’re cross-checking against what’s working and feeding that back in and getting people to rate content and feed that back in, which is how ours works. People can rate the outputs, that gets feedback in. And so, it just keeps… in that way, it does teach itself. And so, you get a very, very different outcome than just if you go to ChatGPT and say, “Write a blog post about X.”
Jessmyn:
So, there is still room for online business owners to sell their services even with the existence of AI.
Josh:
Oh, yeah. I think there’s a ton of room for every industry to just take their institutional knowledge and teach an AI and then sell that AI. I think that’s the next wave of innovation. It’s like when tools were invented, you’re really good at sewing, so you’re like, “Cool, I can now use a machine to make sewing.” And whoever was the best at sewing was best suited to make the sewing machine.
Whoever was the best at tilling fields was best suited to make the tilling machine. Same thing. Online business owners, whatever your expertise is, you can go teach an AI and then sell that. And that’s where I think value creation comes from too, because it doesn’t take as much time and it’s still valuable to the end customer. So, that’s where you get profits from.
Jackie:
And I just add to that. Every time you ask AI anything, even if I put in a prompt right now, if we all three put the same prompts in and hit enter at the same time, we would all get different responses. So, even if you create an AI quiz platform, we have one coming out, if our customers are making quizzes, every single quiz is going to be different even if they’re in the same industry.
And then, there’s another layer of how much of my own human creativity do I want to add into that is going to add another factor that’s going to be different than other people’s quizzes out there. So, sell away.
Josh:
Literally, that’s what Matt and I were working on this morning. That’s our CTO at Interact. And we were comparing just prompting the AI, which is the current private version of our AI that we have where you just prompt the AI and it gives you a response versus layer two which is what we’re at now, where we’re feeding in a proprietary example quiz. And it’s five times better.
And then, when we get to layer three, which we’ll get to really soon, which is now we’re feeding it 500 example quizzes, it’s going to be five times better than that. And then, we’ll get to layer four. And so, yeah, we’ll probably release our tool at layer three. So, it’s got 500 examples teaching it exactly how to write each type of quiz, and then it’s broken down by personality quiz, product quiz, assessment, trivia. So, those are all semantically entirely different. And if you just tell ChatGPT, “Write me a quiz,” it doesn’t know. Is this for a product?
Is this for a service? Is this for trivia? Is this an assessment? Is this supposed to be professional? Is this supposed to be fun? So, those are all things that you can teach the AI, which we are teaching our AI, and it’s incredibly intensive but super fun. And every time you teach it another layer, it gets so much better.
Jessmyn:
But I also want to highlight about that too is we’re not getting rid of our templates. So, I think one of the struggles, at least currently without the Interact AI, is like you could take a template, but it’s not necessarily going to be in your brand voice. It is conversion-focused, but there is still that work of going in there and making sure that you’re including your own offers or really specific jargon that you use depending on the industry that you’re in.
Whereas, when you pop it into an AI, you can say, “Do it based off of how I speak. Do it based off of what my business looks like and what I’m already saying.” And so, it’s writing it for you in a way that you would probably write it yourself, but in half the time.
Josh:
Yeah.
Jessmyn:
Not even half?
Josh:
Thirty-five seconds, instead of like-
Jessmyn:
It’s not dial up guys.
Josh:
Yeah. I mean, we’ve been doing this forever. I’ve been writing quizzes 10 years. It still takes me an hour and a half to sit down and write a really solid quiz. Maybe an hour if I’m really feeling creative. And I’m not joking when I say the outputs that it’s giving are very, very similar. It might take me 10 minutes to edit, once it outputs from the AI. But it’s basically what I would write anyways in 30 seconds.
So, it saves you a lot of time and you can put in your own URL, your own blog post and it will take in your site. And as we get more advanced on that, the first version will be pretty basic. It’s not going to learn your voice that well. But over time, it’ll learn your voice better and better as we advance our tools that allow us to understand your site just by you putting in your URL.
Jessmyn:
In terms of learning how to use an AI tool, what is that learning curve?
Jackie:
I would say just go in and do it. Find a website. I mean, go to ChatGPT, go to OpenAI, go to Jasper. There’s even a website I found yesterday called, There’s An AI For That. So, there’s like all… it’s one central place where you can find all of the different platforms that are available and just go in. Most of the platforms have a free trial or a free version that allows you up to so many character limits or outputs or inputs, whatever. Just go in there, start playing with it, see what you get. Maybe post a couple of things, finesse them, whatever you need to do.
But I think you literally just need to go into a platform, play around with it, see what comes out, and then you’ll learn very quickly on how to keep improving the prompts and what you’re putting in so that you get better outputs back.
Jessmyn:
With the different platforms that you tested, Jackie, was there a significant difference in terms of like, if I’m writing a blog post, I should use one AI tool over another?
Jackie:
I’m still figuring that out because something that I didn’t know or I wasn’t expecting was that each… I guess was that AI can do so much with video, with images, with words. So, I think a lot of the platforms that I’ve used, which you guys can check our YouTube channel for the different reviews, and you literally see me in action doing them for the first time. So, some of the videos might be a little funny. But I haven’t found one that’s ideal for blogs and one that’s ideal for social posts. They feel very similar. The difference is in the interface.
So, it becomes a personal preference on what you like. For example, ChatGPT was the first platform that I used, but Josh gave me almost all of the prompts that I would need to put in to start playing with it. So, I became comfortable with ChatGPT pretty quickly. Whereas, the other platforms when I started using them were forms that I had to fill out of like, “What’s the tone that you want to say? And what’s the topic and what are keywords that you want to highlight?”
And I wasn’t expecting that when I switch platforms. But for a beginner, if you don’t have the prompts, you’d literally be looking at a blank screen in ChatGPT or OpenAI. So, you might want to start with something like Jasper that tells you, “Hey, you might want to add a tone in here so that this comes out sounding like you.” You wouldn’t necessarily know that if you’re using other platforms.
So, that I think is the biggest differentiator. But in terms of what’s coming out, again, what you’re putting in is a little bit different because of those interfaces, but what’s coming out is the same things. It feels very similar in terms of the results that I’m getting from these different platforms. So, I’m definitely still playing around with them. And actually, some of the video ones where we cut, we took a longer YouTube video and cut little sections of it to turn into shorts and stuff. We actually posted and are testing them out.
Now, you’re putting me on the spot, Jessmyn, I can’t remember the one that did the best. Was it Munch? Yeah, you have to check out our YouTube channel, everybody. But yeah, there was one of them that got 100 something views within the first couple of minutes of posting it. Maybe it was Lumen5. Whereas, the other one on Pictory got next to nothing. They were very similar.
So, was it the time that I posted it on YouTube, right? There’s lots of different factors that go into the algorithm and all of that. So, I really think it just… a long answer short, I think I would just say it’s a personal preference on what the actual tool looks like. Like your email marketing system. Most email marketing systems, I mean, yes, there’s the advanced versions, but they do all the same thing. But they do them in a different way.
You’re clicking different buttons. They’re called different things, and people are attracted to different platforms for different reasons. And so, I think it’s very similar to the AI platforms that are out there. Play around with a few, get on some free trials, start some free plans, see what you like, what you don’t, and you’ll find one very quickly that resonates with you or that you like using. And you’ll figure out how to use it even better as you keep playing with it.
Jessmyn:
I think it’s so funny that… oh, go ahead, Josh.
Josh:
Oh, I was just going to say, as Jackie was talking, it made me think. I think a good way to vet whether a tool is worth paying for or not is to go check it against ChatGPT, because you can still use ChatGPT for free. So, if you put in a prompt and then you see it spit something out, go put in the same prompt to ChatGPT.
Because there are some tools, I don’t know which ones, but there are some tools that still literally just don’t do anything. They’re just prompting ChatGPT, there’s no fine-tuning on top of it, and you don’t want to be paying for something that’s not doing anything. So, that’s one of the big ways to test whether a tool is worth it. Can I easily replicate this using a free tool? And if so, this tool’s not worth paying for.
Jackie:
Yeah, and I guess… real quick, Jess. Just to that point, there was one plan that I saw, you had to upgrade to get this plan that included a plagiarism checker, which sounds really important. But correct me if I’m wrong, AI is not plagiarizing because it’s creating words on its own from what it’s read online every time. So, you don’t need to pay for an extra plan that includes a plagiarism checker because it’s not plagiarizing.
Josh:
Yeah. I mean, it’s just going to be like every time there’s another innovation where people are going to sell snake oil. So, that’s a good point.
Jessmyn:
I was going to say that I think it’s funny you mentioned ChatGPT was easy because Josh gave you the prompts already. Because when I went in there without any prompts, as someone totally knew, I was like, “What do I with this?” I had to call Jesy and be like, “What do I put in here?” And she was like, “Oh, I’ll show you.” So, I will say that for someone who just jumped in there, we had different experiences. Josh gave you prompts. But for me, I just got the login.
So, I was like, “All right, let me play around.” And I didn’t feel like I got I guess great things from it. It was really helpful in giving me ideas, but it wasn’t something that I could actually use. And I was using it I think for email subject lines. But after talking to Jesy about it and learning more about what prompts should look like, I will say that was a lot more helpful. But because you have to type in the things like, “Say it like this,” or the actual words say it.
Whereas, I think when we were on the call and you were using… was it Jasper that prompts you with a form that you said?
Jackie:
Yeah.
Jessmyn:
Yeah. When I saw that it had that on your screen, I was like, “Oh, that’s actually super helpful because I wouldn’t have known to prompt it in that way unless I had already had prior knowledge of how to prompt ChatGPT.”
Jackie:
And I guess I’ll just add a little shameless plug here for Interact blogs. Josh didn’t give me the prompts. Josh wrote blogs with the prompts that I then took to put to… so, you can get all the prompts that I started with right from the blogs on how to write a blog post. Because I think when you’re using ChatGPT or OpenAI, the number one thing to know when it comes to writing blogs is you want to instruct it to write you an outline first and then fill in the outline with specific information. I think in most cases you just went into the tool.
I mean, I would’ve for sure. I would’ve just asked it to write the blog about this thing, and maybe in this tone eventually I would’ve figured that out. But really, if you ask it to write an outline first and then have it fill it in, it gives you such a better output in terms of writing a blog than if you were just to say, “Hey, write this blog.”
Josh:
Yeah. I think with tools, another thing I was going to say to look for is the ease of use. The way we’re going to set ours up is put in an URL or start with a topic, and then it’s just done for you, the entire thing, and it’s already optimized. I think the more you have to tell it, “Do this, do that, include these keywords, do this other thing,” the more it’s just going to be like, “Okay, so why don’t I just go use ChatGPT then?” I think it’s going to be about… and that’s where I’m like, again, it’s going to go back to talking to your customers because how do you know which fields people should fill out and which they shouldn’t by talking to your customers? So, I think it’ll again be super obvious.
Businesses are started by humans. They’re run for humans. And even if it’s an AI tool, you’ll be able to tell if it’s made by a person who actually understands their customer or not based on how clunky it is to use or how intuitive it is to use where you’re just like, “Oh, this makes sense.”
Jessmyn:
I love that. Love it. Well guys, thank you so much for hopping on an episode with me and letting me drag you on here to talk about AI. Jackie, can you really quickly tell people how they could find the videos of the AI tools that you reviewed on our YouTube?
Jackie:
Yeah, just head over to our YouTube channel, Interact, and they will all be there for you.
Jessmyn:
Awesome. And we’ll link that-
Jackie:
I guess if you… maybe you could type in AI review, I think they’re all called something like that. So, if you typed in that keyword, all of the whole list would pop up. But they’re all listed right on there, right on our channel. They should be there.
Jessmyn:
They’re already on there. You guys could also check out our blog. We have some coming out in the coming weeks, but some are already being posted. And we will link all of that for you in the show notes. Josh and Jackie, thank you so much and we’ll see you guys next time. Bye.
Jackie:
Thanks for having us.
Josh:
This was fun. Enjoyed it.
Jackie:
Bye.