Introduction
[music]Ryan Weber: Welcome to 10-Minute Tech Comm. This is Ryan Weber at the University of Alabama in Huntsville. Like almost everyone in the past few months, I have been fascinated and sometimes terrified by AI writing, best embodied by ChatGPT, which is making headlines for generating sophisticated text that sometimes comes across as human and can perform feats like passing essay questions on college and MBA exams. Of course, this new technology has raised a lot of alarm in my related fields of writing studies and technical communication. Professors wonder how to deal with AI writing in terms of teaching and plagiarism, and some instructors have found ways to incorporate AI writing into their classes. Meanwhile, writers wonder if they’ll be replaced or at least if their jobs will significantly change.
To get a different perspective on AI writing, I invited today’s guest, Justin McGill, the founder of Content @ Scale. Content @ Scale offers automated content creation that allows companies to upload keywords and receive blog posts and articles about those topics. According to the Content @ Scale website, the software is “The first ever content automation platform that was built to scale content marketing by replacing the need to hire a team of writers.” As a writer and writing teacher, I find the statement concerning, but McGill argues that AI content creation is nothing to fear, and that it will change rather than replace the role of writers. In our interview, we talk about how Content @ Scale works, some of the ethical issues involved with automated content, and the potential future of AI.
I found this conversation extremely interesting. I tried to ask tough questions, and I also crowdsourced Facebook to get some of my questions. Though I imagine some listeners will wish I had gone much further in pressing Mr. McGill. I’m interested in getting listener reactions to this interview.
Die-hard fans of 10-Minute Tech Comm will recall that I already did an episode on AI back in 2018 with Seth Earley. However, the situation feels different now, so I wanted to revisit the issue. You can find more information about 10-Minute Tech Comm, as well as transcripts for this interview, at tenminutetechcomm.com with everything spelled out. Please enjoy the interview, and I wanted to again thank Justin McGill for joining me to talk about AI writing.
Interview
Justin McGill: Justin McGill. I founded Content @ Scale. We started building this in the summer of 2021 and launched the beta in September of 2022. Essentially, my background is in content marketing and SEO. Initially, when I launched an agency and scaled that into a seven figure agency, and then in 2014, launched my first SaaS product, built that up actually through inbound traffic and SEO as well. So the challenge I always had was scaling it. You need either a lot of head count or a lot of freelancers, whatever. It’s an expensive foray into content marketing when you’re trying to scale it, and if you don’t have a really good team of writers or a good process, it can get muddied up pretty quickly. That was kind of the genesis. I felt like AI was advancing to a point where we could probably do something here pretty unique compared to some of the tools out there that just are kind of more general purpose and kind of a single layer on top of GPT3 or something. I wanted something that was more specific for long form, SEO driven content and so that’s what we built.
RW: Cool, so tell me a little bit about what you’ve got right now. I know you’re continuing to develop the product, but as it stands today, tell me a little bit about Content @ Scale. What does it do and, without giving away your secret sauce, how does it work?
JM: Yeah, so essentially what you do is you can just upload a list of keywords or you could have one at a time, but you could add 100 different keywords and then in, I don’t know, 15-ish minutes, if you upload that much, you’ll have 100 blog posts, they’re 2,600 plus words. What’s unique about us as well is it’s undetectable AI content. So obviously from an SEO perspective, that’s pretty important. If you’re using AI writing type tools that are out there, I’ll say, you’ll often see that those end up detected as AI, which is not gonna help you from a search ranking perspective. But on top of that, we’ve built in table of contents and click to tweets and key takeaways, and even we’ll pull in, ’cause we crawl the top ranking content in Google as well, and so we use natural language processing to identify different keyword phrases and whatnot that need to be included. On top of that, we’ll also pull in, like people also ask questions that are relevant to that primary keyword, and then we add those with answers, and build QA schema into that as well. So it’s just a lot of SEO-driven content there. But with that, we try to make it helpful. Table of contents might seem like, oh, that’s not a big deal, but when someone clicks on that, that’s sending a signal to Google that there’s engagement happening on this page, that scroll down helps.
The key take away is help for readers to digest what you’re talking about. So all of these things are built with the end in mind being helpful to the visitor. So that’s what we built. You can get granular if you wanted to customize what that brief looks like or even just add some contacts to help shape the post the way you want it, but that’s kind of it in a nutshell.
RW: You mentioned this a little bit, like it’s mining the top content on Google. What content does Content @ Scale mine to generate these posts?
JM: Yeah, so what I always recommend too is following this whole topical authority, meaning you write your primary piece or your money post for a given topic, and then what you do is you use something like Content @ Scale to supplement that with just 20, 40, 80 different keywords around that main topic. Content @ Scale is ideal for informational keywords, so we’re working on, AI right now, if you’ve ever tried to build a listicle type post, it’s just not that great. We’re working on a solution there as well. For informational content, it’s best in class. We hear it all the time. I just had a live chat request come in right as I was clicking through, and this guy was like, “Oh my gosh, we went from a post costing us $380 and 5-6 hours to $13 in 13 minutes.” So he’s super jazzed. But that’s really the whole use case there.
RW: So you do suggest that people are doing some writing and then supplementing it with this tool.
JM: Yeah, definitely. Where I see this is we’re kind of focusing a lot of training around what we’re calling AIO, artificial intelligence optimization and basically where this is all heading. The people that are fighting this, they sound like the people that were scared of e-commerce at the very beginning, “I’m not giving you my credit card information on a website.” It’s so obvious where this is heading, and so you can either ignore it and likely just get passed by or adopt stuff like this into your tool set and see it as an opportunity, because that’s what this is. And it allows you to scale so much faster, so you can use AI to get your base content, but you should be going through that. You wanna fact-check some things, you wanna add links, you wanna maybe add some images and add your personal touch to this content because I’m telling you we’re growing massively fast. When you have thousands of companies all producing content through systems, the next step is, well, how do you stand out from that? Because now our content, systems like ours, that becomes table stakes because so many companies are able to produce so much more content. Well, the way you stand out is through all of those additional reasons, through that optimization process that writers should start to look at how they can transition to using stuff like this and then just adding those extra layers on top.
RW: That is an interesting point that I hadn’t quite thought about of like, so if there are 20 travel companies and they’re all using Content @ Scale, you’re gonna get pretty similar stuff for all those companies. And now the next step is, well, how do I stand apart from the other 19 or 100 companies or whatever with my content?
JM: Yeah, you gotta think about that. Right now, that’s obviously not the case or something you need to really worry about. But fast forward in a year or two, that’s likely the scenario, and so you should be looking at this as not, oh well, it’s just all gonna be the same. Of course, so there’s gonna be similarities, but that’s the thing too with AI. Whether it’s AI or a person. When you gotta write a brand new blogpost, what do you do? You go Google and research and you map out what people are talking, or you at least you should be. And you’re putting your outline together based on what’s out there. So people are like, “Oh, it’s all things that have been said before,” everything you are going to say has been said before. You are not going to come up with brand new informational stuff that no one’s ever thought of. You might have your own little spin to it or add some level of expertise to it, but the general concept, believe me, has been there. And so there’s nothing different here, and it’s just that the AI is doing all the heavy lifting, and now you can go in and add those unique elements to it to make it your own. And that’s what you should do. Definitely.
RW: So if I generate these posts for a company, do I, does my company own the copyright? Who owns the copyright for the posts?
JM: Yeah. You do, yep. And then obviously, you have your contract or whatever with them to pass that off to them or something.
RW: How does that, I guess it seems like a new area in a way, because we’re taking, you’re taking copyrighted content, repurposing it and generating something new that’s also copyrightable. Is that right?
JM: No, we’re not taking copyrighted content. Our content comes out, it’s never been used, and we have built-in Copyscape and stuff, so you can actually just run that scan. You’ll see sometimes maybe it is like 1% or something, and there’s words in there that have shown up in other content, but out of the gate, it’s not using other people’s words. And so one of, we use multiple AI models, and one of them being GPT3, we share that one because that’s public and it’s a great generative AI. We had multiple layers on top of that, which is ultimately how we get undetectable AI, that’s trained on billions and billions and billions of words, like 175 billion. It’s through that, there’s formulas as far as, okay, well, the next most probable word in this sentence as it’s being generated would be this word. It’s not that it’s looking at this piece of content that’s on page two, let’s take that paragraph and just rewrite it. Tt uses it as context, so there’s a little more information, but it’s not just rewriting that.
RW: So it’s like ’cause I use Google Docs and it’ll be like the suggested thing will pop up and sometimes I’m like, “Yeah, Google Docs. That is what I was going to write. You are correct.” So it’s similar to that. It’s not like I took this post and this post and blended them together, it’s that I’ve read, the AI has read so many things, it knows, well, statistically, this would be the next word that would likely come through.
JM: Within this context of this category or whatever. Absolutely, yeah.
RW: Okay, that’s interesting. And it’s an interesting way to kinda help understand, ’cause I also think a lot of people, myself included, just still aren’t quite sure even how this technology works right now.
JM: It’s so new, right?
RW: Oh, yeah. It’s new. It’s interesting, it’s scary, I’ve never seen conversation on social media amongst writing professors about anything like this, so just with interest and concern, but also intrigue. I guess that’s a good place to kind of pivot into some of these questions that people might have. As someone who teaches writing, trains writers, I feel like I’m representing that group of people. My first exposure to this was I saw an Instagram ad that was for one of your competitors that you actually mention on your site, and they were like, write AI content without a human, and I was like, “But I train my students to do that, what am I… “
JM: If you look at the comments, I’m sure you did, you look at the comments on those ads, it is brutal. And we see it in our social too. People are just so scared and so it’s like I see people like, “Just write it yourself, stop being lazy!” and all of this and it’s about working smarter. It’s not about working harder or trying to spend your time doing it. If that’s what you wanna do spending your time, cool. But just understand, so many people are adopting technologies like this to do it faster, you’re not gonna be able to keep up. You’ve gotta understand where this is going, and see the bigger picture, like you’re talking about how I’ve never seen writers all talk about something and here it’s happening. It’s happening because it’s changing, it’s changing the game forever, and that is for sure happening. You’re either going to adapt to that or you’re gonna get left behind. And so that’s why I was saying earlier, transition into thinking more AIO so that you can optimize that AI content and still make it your own and add your own unique spin, but there’s a reason this is blowing up.
And it’s the fastest moving industry I’ve ever seen. It’s just, when we were working on this, it was not really all that talked about, and then shortly after launching, it’s just everywhere now. It’s all just kinda culminated into this perfect little storm here, but it’s nothing to be afraid of. It’s just something you’ve gotta adapt to and learn how to take advantage of the opportunity, really.
RW: So you see companies still having some sort of writer, editor, content generator that would work with the AI to generate content.
JM: We have a writing team. Absolutely. So what we do, obviously, we have access, direct access to the software, but then we have a done-for-you service layered on top. And basically our team runs that and then edits all the content, adds links, add images, they do the AIO work, and then they take that and publish it on the client site. That’s exactly what you should do. The cool thing is, a writer that would struggle to do 20 posts a month, now we could get 10-20 done in a day. It’s just a totally different ball game. And so think of the velocity you’re able to accomplish when you’re able to publish that much content in a given month. And the reason you would do that is because you’re able to just start ranking for so many more keywords which drive more traffic. I remember with my prior SaaS with LeadFuze, over six years, we had done 300 and something blog posts. It was one a week consistently for six years. You can do that in a couple of months now. It’s unbelievable what it allows you to do, and so yeah, absolutely.
Somebody should look at this as an opportunity to transition from creating that content from scratch. Instead, you’re optimizing, you’re able to do so much more. And so you’re able to potentially charge more for that with your clients, if that’s what you’re doing. Or if you’re working within a business that is trying to scale up your content marketing, this is how you could do it, is just by making that transition from writing it all from scratch to taking a 2500 word blogpost and just making it your own basically.
RW: So that’s an interesting point too. You’re saying if I’m a freelancer, I could almost be like, “Hey, I am… You gotta pay me more because I’m good with AI.”
JM: Absolutely.
RW: I’m not just the guy or woman who’s gonna sit down and write this stuff from scratch, I can handle AI, and so I’m actually worth more as a writer/editor.
JM: Right, and then what you do is you sell more. You can sell higher volume packages because chances are that your client has heard of this, they’ve heard of AI content and what it’s doing and how it’s changing the world. And so what you do is say, “Listen, I’ve adopted this, I’m able to scale this much faster, what this allows you to do is publish more content, rank for more keywords, drive more traffic.” But here’s the thing too, and this is the selling point, you gotta understand where this is going, and your competition is gonna be doing stuff like this soon too. So now it’s like, let’s get ahead of the game before they start doing it as well, and so there’s some urgency, right? And that’s the nice thing when they start to hear about this stuff on their own. I have a police officer friend who just messaged, he’s like, “Hey, what’s going on with this whole ChatGPT thing?” You should not know about this.
RW: Oh yeah, no, I have my cousin was mentioning it at Christmas, a friend texted me the other day and wrote ChatGPT messages to me. Everybody is talking about this stuff. I think something interesting that I don’t think I quite caught before I talked with you about your company is you’re writing kinda sort of for robots, you’re writing for Google to maximize the SEO of your site in a way, right?
JM: Almost. Yeah, that’s an interesting way of looking at it, but yeah. You’ve got essentially AI and algorithms that are reading your content, why not have that made by AI as well? And then here’s the thing. Where I see this going beyond the content is because when there’s so much more content out there, it becomes harder and harder for Google to discern which should get those rankings. I think links have always been important, but I think they’re gonna be even more important as this starts to really unfold and hit mass scale. ‘Cause that’s just a way for them to really tell which of those sites are actually people are trusting and linking to and I think that’s gonna become just a bigger factor overall. But like I was saying earlier, I think systems like ours just becomes table stakes. You’ve gotta be popping out the velocity. At the same time though, you really need to think about that link building aspect.
RW: So again, the other thing. I’m a writing teacher, I see these conversations all the time. Obviously the big thing that everybody is scared of is plagiarism. My students are gonna use ChatGPT or Content @ Scale or whatever. I’m going to give them an assignment, they’re gonna use this to write it and turn it in. How would you respond to that kind of concern?
JM: Yeah, I think, in fact, we’ve gotten some of those. So we have a free AI detector, and we had a professor use that to give a failing grade to a student. And that student reached out and wanted to know what they could do and she was like, “Well, I have full proof that I wrote this and it wasn’t AI,” and so here’s the thing. You can get it to where it reports back that it was human written even though it was from an AI. That transition can happen pretty easily. But it’s something we’ve solved. To have content that you’ve hand written, and I’m talking you’re a couple hundred words into it, and it comes back as 100% AI, the statistical improbability is through the roof because that’s basically on every single word, you’ve somehow managed to use the exact highest, most probable word choice for everything that you’re saying. It’s so impossible.
Listen, can it happen? People win the lottery too. That happens. Sure. But I said, “Listen, this should be used for entertainment purposes type of language,” but I was like, “If it’s showing up as 100% that you wrote this or that it’s obviously AI is how we show the grading metric there, it’s gonna be pretty hard to convince us that you didn’t,” but he or she was like, “But I have full proof there,” I was like, “I don’t know what your fool proof is,” I guess maybe he had a camera sit behind him as he was typing the assignment maybe, I don’t know what full proof meant. But I said, if you have that, then I would go ahead and present that to your professor, but yeah. If your stuff is getting flagged as obviously AI . It’s gonna be a problem for people and it’s gonna change how universities do grade things. Obviously, they used to just check for plagiarism, but this is a factor now too. We’ve had requests to utilize our detector in a way where they could just supply their grades or the assignments could run through that first, and you better believe universities are all over this and they’re not as slow moving on this as maybe you think that they would be. But yeah, it’s absolutely gonna be a factor.
RW: You’ve got… I saw it on your site, it’s cool. You’ve got a thing. You can plug text in there to see if it’s AI written and on one hand, you want your content that is produced for companies to pass this test so that Google doesn’t say, “hey, this is AI written” and lower it down in the SEO rankings. But at the same time, you’re saying it would be very hard for a student to pass that test if they legitimately just wrote an AI, plugged something into ChatGPT and turned it in. Am I understanding that correctly?
JM: Totally, yeah. That content oftentimes is gonna come out fully reading as AI. Listen, here’s the thing too, ’cause some people get confused like, “Oh well, I copied this right from ChatGPT and this one’s saying that it was human-written,” and it’s like, “Cool, that just means it reads as human written.” It’s not that this is wrong, this is just giving you this metric of like, “Okay, yeah, this comes off as human or this comes off as AI.” Just because it came out of AI and it’s coming up as human, that doesn’t mean that it’s wrong. It’s just letting you know that, okay, yeah, this passes the test. But yeah, generally, if you’re pulling content out of tools like that or just tools, any AI writer. That’s just a layer on top of GPT3 like that. Yeah, it’s gonna show up as AI written and you’re gonna have some problems.
RW: But it is possible, these tools are just testing. I saw there’s a new app that a college student created that’s kind of looking for, well, humans tend to do this, AI tends to do this. But you’re saying it is possible for AI-generated text to read as human and kind of pass the detectors.
JM: Yeah, so that’s a differentiator for us. We do that and you can take our content and run that through any AI detector you want, we’ve got a free one, there’s paid ones, but run it through and you can see that it comes out as human written. That’s something that was a focus for us out of the gate because that was obviously a big concern that people had. And so that was something we set out to solve very early on with this. But yeah, it’s definitely… It is possible.
RW: If I, say I’m the CEO of a company and I run, I use Content @ Scale to generate a post and I don’t really do the thing you’re talking about of going in and adding my own spin, just like, “Okay, here’s my post,” boom, I slap it up, I put my name on it. Should there be some kind of flag or some kind of indication that this was AI written? Is that an ethical use of the program, do you think?
JM: Yeah. Listen, ultimately that’s on you and how you wanna do that, but you see Bankrate.com and cnet.com, this has been in the news recently because they’ve got that in their author byline that this was straight out from AI and it’s ranking all over page one. And so Google was asked, “Well, what’s going on here? You said we can’t do this.” And so they made a very subtle, but very massive tweak to their guidelines, which was before any content created by automated means is against our guidelines.
RW: That’s Google or that’s these companies?
JM: That’s Google and where now it’s like, any content made through automated means without regard for the helpfulness for the end user, something along those lines, but they made a little tweak. Then that’s the problem, which is all what we set out to do is have helpful content. That’s a very subtle but pretty massive change in terms of their handling of that, and they had to because they know where this is going. So when Google is adapting their guidelines because they can’t keep up, right, and there’s just so many people doing this. As a writer yourself, that’s what I’m saying, you have to understand where this is going, you have to see that. So just stop fighting it and adopt it or just understand that you’re probably gonna need a different career path if you’re stuck on that angle because that’s just where it’s going.
RW: So to get back to these companies, just to be clear, it is good to say, to have some indication that you use AI to generate the content.
JM: Yeah, if you’re not gonna add to it, and you’re just gonna basically, which I don’t recommend doing. I mean, I don’t. So it’s gonna still come out as undetectable AI from our system and everything, but I still recommend, you’re gonna wanna do something with that to make that your own piece of content. And if not, you just wanna go straight from it, from an ethical standpoint, that would make the most sense, having some sort of an author byline that just says that this came from automated means or whatever. You don’t have to share where it came from, but if you wanted to. Honestly, I don’t think we have a single customer that’s doing that.
RW: It’s just taking, spitting the content out just straight to their website with no adaptation.
JM: Well, I think there is some of that, but I’m saying, no one is adding that this just is AI-generated.
RW: Oh, I see. So no one is indicating that it’s AI-generated.
JM: Yeah, I don’t think you’re gonna see that be the norm, I don’t. Because I don’t think they wanna raise any flags to Google. Well, I think they view it as, it’s okay if these big sites are doing it, but I don’t wanna raise suspicion. I think that’s the general kind of consensus there.
RW: And I guess that spins to another question that’s kind of related, and I imagine this is mostly just anecdotal or in your experience, but how do people respond to content when they find out that it’s AI-generated? Do you have a sense of how people approach it?
M: You know what’s funny? I used to measure this where, in advance, where we would tell somebody, “Okay, this article came from AI,” and they go into it a way more skeptical. So we’re actually running a survey, writers that we’re hiring from different services, and we’re going to compare that and ask people that are not our customers that don’t know how our content, all that, the formatting and everything, and we’re gonna actually survey them, just show, just pick which one you think is best. Because that’s the thing. Human doesn’t mean better. I don’t know what your experience is like in hiring writers, but let me tell you, it is not easy to find good writers. And so I’ve seen it time and again, this is why this stuff exists, because there are so many bad writers. And so AI out of the gate beats most of them, unfortunately. And so I think when people are aware that, oh, this is AI generated, they look at it through a little bit of a different lens than if you were to just give them two pieces of content and they don’t know which is which, and I’ve done both. And I’ve seen different ways that people handle it.
So I think there is a natural skepticism when they find out that, oh, this was AI-generated, but honestly, I don’t know that many could tell the difference between something that was handed to them. If you handed two articles, one from a person that you got off of Fiverr or something or even the Textbroker, whatever these different services or an AI-generated one, I think a lot of people would, that are kind of not in the know, just your average typical reader, I think what you’ll find is that they’re not gonna really care.
RW: Interesting. Yeah, and I also imagine those attitudes might change as AI becomes more and more common, like the first time you know you’re talking with a chatbot versus the thousandth time or whatever. Do you think that there is anything, you’ve talked about this AIO, is there anything that humans can do in terms of writing that at this point, AI is not as good at?
JM: I think sharing those personal stories, and those unique, maybe they’re screenshots, maybe they are photos or images, or tying that story arc into whatever content piece, that’s just something that the AI is not gonna be good with in comparison to a human that has that experience. And that’s the stuff that I think you should try to incorporate when you do that final layer of editing, is tie that stuff in. When it comes to everything else as far as your typical blog post, I think AI does great with that stuff. So you have to look at, well, what can I do that would make this unique? You really stand out, and it is those personal stories or a super deep expertise on a topic that you’ve gained through years and years of experience. If you can add those to it, that’s how you would set it apart, but those are the ways that you could differentiate from just your typical AI-generated.
RW: I was also curious: this runs on kind of estimating what the next words might be. It seems like this is less helpful, let’s say if I was a scientist and I ran an experiment that generates new knowledge, getting it to write up the results of that. Is that accurate?
JM: Definitely, yeah, and so now you can build custom models, and we’ve had some conversations with folks on that, where you supply data and then there’s expected outcomes and stuff from that, and there’s ways to incorporate AI there, but yeah, you’re much better off, or if you’ve got something like unique data like that, that’s a great point. That should be something that you add to a piece of content because no one else has that information. When it’s uniquely yours like that, absolutely, that’s something that you’re gonna be able to provide better than an AI will. But you could build some models to feed that would then generate maybe for context and create a better overall post, but yeah, if you’ve got data like that that’s uniquely yours, that’s a great opportunity to differentiate it.
RW: Can the AI write… Someone asked me this and I was curious about it. So if my audience is a regional audience say, or an audience that maybe is bilingual and has a particular dialect or whatever, can it customize to particular dialects, audiences, populations in any way?
JM: Yeah, so there’s ways for the AI to detect tone of voice and things of that nature, or you could assign that upfront if you wanted. As far as what you’re talking about, you almost need to fine… It’s called fine-tuning a model, but basically you would give it, maybe it’s all of your prior content, and this is something that OpenAI and these different platforms need to get a little better with actually enabling those custom modeling options, but you would supply all that content for it to then basically extract your tone of voice, your dialect, all of that would be encompassed, but you’d have to fine-tune the model.
RW: So the model would need to read off of either you or people who write in that group that you’re targeting or whatever?
JM: Yeah, and if it’s out there, if there’s content out there that you can steer it that way, it’s gonna know… I’m in Phoenix, so it’s gonna know the Grand Canyon’s in Arizona, and so it could piece those things together without an issue, but when you have certain slang or things like that…
RW: Like a turn of phrase or a colloquialism that it may, would need fine-tuning to be able to do. All right, I’ve enjoyed this. I just have a couple more questions for you. So, it sounds to me like efficiency-wise, SEO-wise, AI content generation is going to happen, and it is something that people need to grapple with or incorporate or work with. In terms of like, what do you see is beyond that profit for businesses, efficiency? Other benefits for society for this technology, society at large.
JM: Yeah. Man, are you talking just AI content specifically, or just AI as a whole?
RW: Let’s start with AI content specifically, ’cause AI as a whole, it’s a whole other issue, but yeah, just in terms of AI-generated content, what other kinds of benefits can that bring for us above efficiency and profit?
JM: Yeah, obviously, those are the big ones, the efficiencies and the profit, but you’re able to convey your message more frequently as a business. We’ve got a content library brewing of hundreds and hundreds of ideas, and we’re really kind of working on this whole AIO model and really try to train people on how to make that transition. And so what we can do is use our platform to pump out a lot more of that type of content to help people make that transition. And for us it’s education and all of that driven. For other businesses it might be just getting their products or different use cases or things like that out there that either would have been bottlenecked with and not done it, or it became too expensive for them to do or something like that.
So that’s where I think from a business perspective, there’s a lot more benefit, but ultimately at the end of the day, it’s gonna come down to dollars. That’s the bottom line, I think. And I think that’s why you’re seeing AI get adopted more and more frequently. Things like going beyond the content side, I think things like Siri and Alexa, and I don’t really use them because they’re pretty basic things, and I think that’s gonna change here very quickly, and I think there’ll be a lot more usable, a lot more resourceful. Take that ChatGPT experience and give that to Siri or Alexa or whatever, and it’s like, that’s a different ballgame now. You’re getting a lot more information, not just the things that they’ve been kind of programmed with, it’s a much bigger deal.
RW: It can pull from more and develop more sophisticated answers.
JM: Right.
RW: So I mentioned impacts on the writing profession, plagiarism, that kind of thing, copyright, were there any other ethical issues that your company has or continues to have to contend with as you’re developing this product?
JM: Honestly, those are the ones. That’s it. Those are the ones that people are maybe not as familiar with or they’re concerned about all of those. It’s interesting with the images, it’s like a new lawsuit going on with Dali and Stable, or actually, I think it’s Midjourney and Stable Diffusion from artists because…
RW: Yeah, I saw that. I saw the news about that the other day. Yeah.
JM: Yeah, that’s gonna be interesting. That’s gonna be a fascinating lawsuit there to see how that turns out. At the end of the day, I think there’s gonna be, it’s gonna be pretty hard for there to be proof that somehow their image was factored in, but who knows? Who knows where that’s going. I think there’s millions and millions of images that have free rein, free license control and all that, so I think it’s gonna be a tough one to win, but we’ll see what happens with that.
RW: In that case, in case listeners aren’t familiar, was that an artist was suing this AI over basically using their image to generate new images. Is that correct?
JM: Yeah, to use their collection of artwork or images or whatever, and to help train the AI model to generate more. Yeah, that’ll be interesting. I don’t see that going very far, but it’ll be interesting to see where it does end up.
RW: So you’ve mentioned, I’ve talked about people that I know who have concerns, you’ve mentioned the social media comments on your ads and other ads and the concerns that people have. If you had a chance to alleviate those people’s concerns, besides what we’ve talked about already, what would you tell them?
JM: Listen, it’s like with any change. People normally don’t like change. It’s just human nature. But you have to just understand what’s going on here. When Google is adapting their guidelines for their search quality, when you’ve got random cousins and friends in just totally random professions that are asking you what’s going on with this whole thing, with ChatGPT or AI content as a whole, it’s happening and it’s happening very quickly. Like I said earlier, I think it’s the fastest moving industry that I’ve ever seen. Just understand that this is happening whether you want it to or not. And so the choice is yours ultimately to adapt and see this as an opportunity, a brand new opportunity. Listen, they’re saying now, they think OpenAI could be the first trillion dollar company or whatever, the opportunity is massive. And everyone knows it. So this is not going away. It’s just, you can ignore it, but just understand that this is absolutely happening.
RW: And that’s interesting, ’cause as a writing teacher who teaches technical communication, I tend towards practicality in that way of like, well, this is happening. We need to deal with it. On the other hand, I imagine there are listeners who are like, well, just because something is happening, doesn’t mean that it should happen. Is there a sort of, can we make an even stronger case than “This is happening, deal with it,” for why this is a value add for companies and humanity?
JM: Listen, I think the reason this is so valuable is because there are so many bad writers. Honestly, if everyone was a great writer, this wouldn’t be a problem we’re solving. You know what I’m saying? This has happened because there’s just so much bad content, there’s bad writers or you’re off-shoring it and they can’t speak the language well, and so your content comes… It’s just… I’m telling you, this is why this is in existence and why it’s in such high demand. Or the other is that you’ve just charged so much for it that people are looking to cut some corners there. And so that’s not the sole reason obviously that this is taking off, but you have, it’s absolutely part of the reason. And so everyone’s contributed to it. I used to run an agency, I’m sure that that was a factor too. So if people can utilize technology in a way that helps them become more efficient, businesses are going to do that, whether you think they should or not. You’ve got robots flipping burgers at McDonald’s now. This stuff, it’s happening all over the place, it’s not just in this space, it’s AI as a whole.
And they talked about this for years in movies and everything else. And we’re starting to see it, we’re at the cusp of it right now. There’s not much you can do. You can ignore it. But again, there’s nothing you, I, or anyone else is gonna say that’s gonna stop everyone and just like, “Okay, you know what, that’s enough for AI, no more. Everybody just stop.” It’s just, that’s just not gonna happen. So if you have that big of a problem with it, I think you should just continue on, keep doing your thing and fighting the good fight, and I think there’s gonna be people to do that and that’s fine. And who knows, because maybe everyone else is going this route, you could stand out in a way, and be just so unique because you’ve not gone that way. So later on, maybe that becomes a differentiator for you.
RW: Sure, I was wondering about that ’cause it’s kind of like a restaurant that’s like, all our food is local. It was that backlash against kind of like factory farming, fast food, and you’ve seen a whole industry emerge around that, and it’ll be interesting to see if something like that happens with writing as well. It sounds like your take though is this change is happening, so let’s be smart and proactive about it.
JM: Totally, absolutely. You can see this as an opportunity, which is what you should see it as and understand that you can actually still sell more, you can charge a higher premium because chances are your business clients haven’t figured out this whole AI thing. But they’ve heard about it, and so you could go in and kinda piggy back off of that whole wave of things happening right now and let them know that you know what, we’ve adopted this into our offerings, and this is something that’s gonna help you because of X, Y, Z, and you can publish more or rank more or whatever. And so because of that, we’re gonna adopt this for you, here’s how we would do that and incorporate this into the strategy. I think there’s a way for you to upsell it quite honestly, and capitalize on this whole movement really, but it’s a personal preference too. Like I said, you could wait for this to all be mainstream and everyone’s doing it, and then you stand out by doing things completely custom. So I think that will be an opportunity for some people, at some point.
RW: As a writing teacher, would you, for me, would you recommend that I help students work with this technology, incorporate this into teaching them how to do this AIO type of stuff?
JM: I think you it owe to them, honestly. Because this is just so prevalent in this space, in particular with content writing, that if you’re avoiding that, I think you’ll almost be doing them a disservice. Even if it’s to just, listen, understand what they’re doing, and then, this is what I teach, right? Even if it’s that, but just giving them the insight into what’s happening, like I said, I think you’d be doing a disservice if you weren’t teaching them how to take that technology and incorporate it into their workflow. If you don’t wanna do that, here’s this option too. I think that’s how you would handle that.
RW: Well, Justin, this has been really interesting, I really enjoyed the conversation. I appreciate you coming on the show. It’s not always easy to talk to skeptics of your technology.
JM: No, that’s fine. I understand that that is a lot of people right now, and we’ve got a lot of work to do in front of us as far as just educating people on why it doesn’t need to be so scary, and I appreciate you inviting me on to discuss it.