How Writers Can Harness the Power of AI

Lauren: Hey everyone, welcome back to another episode of Publish & Prosper. Today we are gonna be talking about a topic that absolutely nobody in the marketing or content entrepreneur space has ever discussed in the history of time and space. And that is AI and generative AI and how to use tools like ChatGPT as an author or content creator. 

Matt: You might wanna re-record that intro. Unless you meant to be extremely sarcastic. 

Lauren: I am always extremely sarcastic. 

Matt: Okay, I can never tell with you, but also to be fair, I wasn't looking at you. I was reading something, so I wasn't quite sure what I was walking into. 

Lauren: I meant this is a hot topic. It was really what I meant to imply. 

Matt: Alright. 

Lauren: Yeah. I guess we should just go ahead and get started with that. I don't know, I don't have anything interesting. 

Matt: I mean, that's quite the intro, so maybe we should just get started and make sure everybody understands it. 

Lauren: So let's dive right in. I am definitely gonna be advocating for this tool in this episode. 

Matt: Yeah.

Lauren: I'm definitely not trying to argue against it. I actually made a note as I was doing research for this episode, there was an article that I came across that… The article, you know, didn't wind up being particularly useful, but the title of it was How To Use AI to Become a Better Writer. And I thought that was such a great title and kind of explanation for what they were talking about and then also like what we're talking about here, because I think that is the point. And what we're trying to accomplish here and we're trying to explain and help people do here is not the, like,  fear mongering subject of how to use AI to replace human writers, but how to use AI to help yourself become a better writer and help improve your writing workflow. 

Matt: Yeah, definitely. 

Lauren: Yep. 

Matt: I do think - you alluded to it a little bit - but I will say that as marketers and also as people that work in the publishing industry. I do think that authors and writers and people in the publishing industry in general had a much more visceral response to AI and AI tools when they really became popular last year, like when ChatGPT was released to the public last year, that's when it really sort of dominated every headline. Marketers, there was still some mixed reactions - you know, a lot of marketers were like oh, there goes my job, but I think most marketers were like, okay, this, this could be really cool. And then most other industries, at least around digital content and other stuff, I think also were somewhat mixed, but authors really went to one end of the spectrum almost immediately. 

And it's been really interesting to watch that shift over the last year. And it's thankfully been spearheaded by some really good technophiles like Joanna Penn and some of the others who really early on took a lead on really trying to help authors understand, and writers, just how much these tools can help you versus replace you. It's been cool to watch, but I still believe that I haven't seen any other group of people, any other industry really have as strong of a reaction to the onset of generative AI tools as the writing and publishing community. 

Lauren: Are we including, like screenwriters - writing and publishing? Cause there was, you know, a pretty massive writer’s strike last year. 

Matt: Yeah, well that's what I'm saying. 

Lauren: Okay. 

Matt: They took it. 

Lauren: I was gonna say, are we including that in the - 

Matt: Of course, yeah, they're writers. I mean.

Lauren: Yeah.

Matt: Again, authors and writers, definitely they hit the wall the hardest at the beginning. But you know, the point is it's been cool to watch a lot of them sort of come back over to the middle of the spectrum or even convert to more pro generative AI tools to help, so. 

Lauren: Yeah.

Matt: And again, you know, spearheaded by people like Joanna Penn and Ann Handley and a bunch of others. So that's been really cool. 

Lauren: Yeah, I actually was kind of one of those people, honestly. 

Matt: You? 

Lauren: Yeah. 

Matt: Why? 

Lauren: I was really not in support of AI and generative AI when we first started talking about it here. And I would like, you know, kind of, kind of downplay that at work, because I knew you guys were all excited about it and were like very interested in it and I was trying to keep an open mind about it, but I, you know, from this side of me that is, has a background in writing and probably had some of my writing used in the like internet scrape to teach the AI how to, how to write and stuff like that. Like there were a lot of like initial concerns and fears and a lot of things. Also that it's one of those echo chambers where all it takes is a couple of loud voices talking about how bad this is for it to kind of like permeate the culture. And then everyone's like, wait, yeah, yeah, yeah, it is bad. And it really wasn't until I started doing more research and attending some of these sessions at conferences and actually using ChatGPT for myself and stuff like that, that I started seeing how that initial reaction to it was definitely not reality. That it was - not that it wasn't reality, but like that the fear of being replaced by AI is probably outlandish and instead it's a really potentially powerful tool to help people become better writers and also to help like, I don't know make writing more accessible? I think?

Matt: I think the reality of it is is you just needed some new drama in your life. 

Lauren: Oh, I always need I always need something to like fixate on hating at any given moment. 

Matt: I don't understand your reaction to to AI, your job wasn't even solely dependent on writing. 

Lauren: No, it wasn't and no and there was never a point where I was like, oh my god, I’m going to lose my job over this. 

Matt: Okay. 

Lauren: Like that I never, I definitely, it was actually, it was much more personal than professional. 

Matt: Okay.

Lauren: It was much more like fear of, like it's just gonna start using people's writing and regurgitating it verbatim and that it was gonna be just like plagiarizing outright. And then the more that I started actually seeing the results of generative AI and like how, it's like so obvious when it's something written by ChatGPT versus like written by a human being. 

Matt: Yeah. 

Lauren: Which was why like, you know, we've talked about it in the past. We've always said like, make sure that you proofread anything that it puts out for you before you publish it anywhere. Cause you know, you want to take a look at it for sure. Or like, talked to my sister and her friends who are teachers and how they'd be like, ‘yeah, it's so obvious when a student hands in, hands in something that was so clearly like a ChatGPT answer. 

Matt: But I think that's why people need to jump into these things and instead of being afraid and running away from them, because when you talk about some of the repercussions of a tool like this, like let's say ChatGPT, or if you want to talk about the others, that's fine, but we use ChatGPT, so we'll stick to that one. Just like a lot of other things in our history that came on the scene, and people were afraid of them so they ran from them, including governmental regulation things, you ended up having this wild wild west effect, which. It started to happen with AI, but thankfully a lot of people took that fear and they… it turned into action. And so you have right now, thankfully, a lot of regulation efforts that are happening. You have a lot of other tools that are now popping up to help safeguard people against bad AI generation. You have these other tools that are helping authors with the ability to watermark their content or exclude their content from being scraped for the large language learning models. So I think that if you don't have people jumping in and using that fear or that curiosity to drive them to do these things, that's when you're left with something like the crypto industry that went - and still does go - unregulated for so long that the bad actors just continue to permeate throughout and people continue to get hurt. But when you do have people leading the charge, a lot of the writers that either took up lawsuits, if for no other reason to at least make sure that the spotlight was put on the ability to, to plagiarize. Or again, other influential people in the world of publishing and, and, you know, Joanna Penn, Ann Handley, all these other people that are getting involved. You need to get involved. You need to check this stuff out. You need to get your hands on it. You need to understand what it is. I think the fear for a lot of people was rational, but I think the way that a lot of people chose to process and then act on that fear was not helpful. Mainly for them, but also in the overall cause.

Lauren: Yeah, I do actually agree with you on that. That was what changed my mind about it was actually getting to, like, getting into ChatGPT and like using it. And I now use it every day for work. I used it literally today. I will explain some of how I used it to help me outline this episode. We joke about how many different sessions that we've been to at different conferences where like, you can't go to a conference these days without there being at least one AI session. And you know, sometimes maybe there are a little bit too many, but still like if it weren't for that many people talking about the importance of AI and how people are using it in modern content creation, in marketing, in writing or whatever it is. If it wasn't, if I wasn't exposed to that, if I was just exposed to the, like, panic that I saw on my like specific niche corners of the internet without ever actually getting like the alternate perspective on it, I think I'd be in a very different mindset about it right now. But having gotten to really immerse myself in it and use it and learn how useful and helpful it could be totally changed my mind. 

Matt: Yeah. 

[9:54]

Lauren: So let's talk a little bit about how helpful and useful it can be. So like I just said, I was using ChatGPT. Let's, we probably should have mentioned this already, actually, backing up a second. We use ChatGPT as a tool here at Lulu. We all have ChatGPT accounts. We use them pretty regularly. I feel like we're going to default to saying ChatGPT just in the same way that you would say, like, I'm gonna Google something instead of saying I'm gonna search something. But this is definitely applicable to any generative AI tool that you're using. If you have one or you're interested in trying to use one that isn't ChatGPT, go for it. These are like general practices and tips and ways that you can use it and stuff like that. So don't - it's not ChatGPT specific. 

Matt: Yeah, no, that's a good point. We do tend to solely basically use ChatGPT. At the time of recording this it's probably the most stable generative AI tool that's out there right now. 

Lauren: Yeah. 

Matt: Some of the others are having issues and things like that. So we've just been using that and most of what we'll be talking about, like Lauren said, will be examples we pull from ChatGBT. But most of the stuff we're going to try to include will be helpful across any of these tools. They're all basically built kind of the same way and designed to return kind of the same results. 

Lauren: So anyway, I was using ChatGPT this morning because I procrastinated on writing the outline for this, for this episode and I was feeling some Monday morning writer's block for sure. And I was staring at a largely blank Word doc so I opened up my little ChatGPT window and I asked it for help with this outline. 

Matt: Glad I didn't compliment you on how well this outline was done like I was going to earlier.

Lauren: You know what, I did most of it. I'm just saying. 

Matt: I'm just kidding. It's a well done outline, so.

Lauren: No it’s fine. Thank you. 

Matt: Thanks to you and ChatGPT. 

Lauren: Thank you so much to my little helper over there. 

Matt: Your co-pilot. 

Lauren: Yep, exactly. So I asked it this question, I said: what are some creative ways that authors can use ChatGPT to help them with drafting and editing their book without specifically relying on ChatGPT to actually write the content of the book? 

Matt: Nice. 

Lauren: Which, yes, is a long-winded question, but you kind of have to be long-winded, and we'll get into that a little bit more later. 

Matt: Long-winded is actually just a replacement for parameters. 

Lauren: Yes. 

Matt: Yeah. 

Lauren: You want to be specific, but we'll talk more about that in a little bit. But it gave me, actually, a great answer. It was a long, it was a like 10 item list and it gave me like, you know, like a little bold heading and then one or two sentence explanation for each one of them. And I just wrote down the headings for this, but it was: brainstorming ideas, research assistance, character development, dialogue generation, writing prompts, structure and outline assistance, feedback on coherence and consistency, editing and proofreading, title and headline creation, blurb writing, perspective and tone adjustments, generating alternate endings on scenarios. 

And then I went over and looked at my outline and what I had so far written on the outline, which was: outlining and brainstorming, researching, summarizing and curating, smashing writer's block, providing name and title ideas, summaries and author bios, and editing, revising, and providing feedback on clarity, grammar and coherency. 

So, pretty much everything that I had already spent time thinking out as like, this is what ChatGPT could be helpful with, was included in the answer that ChatGPT gave me. It was right on the money in terms of like, the answer that it gave me was right on the money for like, where I was already going with this. It was very helpful in elaborating on those different ideas for me and giving me a couple of additional ideas that I hadn't thought of or hadn't already included in here.

Matt: So had you just used ChatGPT and not spent the time doing it yourself, you would have actually saved some time is one of the points, I think.

Lauren: Yes.

Matt: Is what you're making here. Yeah. 

Lauren: Yes, absolutely. And that did occur to me. 

Matt: Yeah. I think for me, I love the tool for a couple of things. Obviously we write a lot of marketing copy here focused on authors and publishing and all those other things. So it's a great help to get… for me, I'm somebody who is always stuck at the beginning, always. 

Lauren: Oh yeah. 

Matt: I've always been that way since college. Like, man, if I had this tool in college, holy cow. A, my grades would have been better. But B, there would have been a lot less long nights where I literally spent hours trying to get started on whatever the writing assignment was, as a creative writing major. So having something like this that helps me get started, even if it's to just start an email that is… potentially that I have to write to somebody where maybe the email is a little sensitive in nature or something to do with their book or something like that. Like, just having a tool that within seconds helps me get a start that I can then build from, that's really what sold me out of the gate. Like, the one of the first things I played around with it was like, help me create an outline for a book based on this topic for this audience, blah, blah, blah, blah. And when I saw what it spit out, even though I wouldn't use it verbatim, like when I saw what it spit out in the form of an outline and how close it was to what the outline should have looked like, I was just like, wow. Okay, it would have taken me hours, days. And again, not because, I don't think it's because I'm not smart enough to do it, obviously, as a creative writing major and just working in publishing. Like I can do an outline, but for whatever reason, there's part of my brain that just, it's getting started that I just struggle with so hard. 

Lauren: Yeah, I think it's super helpful for that and kind of like tackling that first step for you. I definitely used it the same way where I've been like, here's the topic I'm going for, like help me give an outline for this. I'm pretty sure I've mentioned this on the podcast before, so sorry if it's repeat, but even if you get a outline response that says something like, maybe 50% of it is actually relevant to what you're doing. It still gives you that sense of direction in the sense of like, I don't wanna talk about this. I'm gonna rule this out. I don't wanna talk about this, but that reminds me that I do wanna talk about this. And I do wanna include this in my outline. And this is a really good point. I didn't think about this. So it gives you a starting point.

For me, one of my favorite things with ChatGPT and one of the ways that I've been using it a lot more and more lately is one of the primary roadblocks for me in writing is when you get tripped up on like a little piece of minutiae that is not really important, but it somehow winds up being the brick that you trip over and then you're like face-planted on the ground. Whether it's something as simple as like I have to name a not even minor character, like a character who's going to be on this one page in this one scene like they're going to a party at somebody's house. And it's not relevant whose house it is. They just, you can't just say I'm going to a party at somebody's house, you have to have a name for the character whose house they're going to. And I will sit there for two hours trying to figure out a good name for a character that is never actually gonna appear on page, ever. And then that's it. Like I've completely stalled out, I'm stuck on that detail, I'm never gonna move past it. And I've always had the advice given to me by like, other authors, writers, creators, it'll be like, just put a placeholder in there and keep moving. Like that is such standard writing advice is literally just write NAME in all capital letters and bold it, so you remember to go back to it, and then just keep writing. And ChatGPT is another way to handle that. 

[17:22]

Matt: Yeah, I think before we get too much further into like some of the specific tips and tricks, which we're gonna offer some on how to use it, how we use it, how others use it as a way to help, I think we should spend just a minute talking about, I guess for lack of a better way to put it, how you talk to ChatGPT, right? We tend to treat tools as tools. It's just this extension of our laptop or whatever that might be. But given the nature of what ChatGPT is and some of these other tools, from what we've seen and even from how I use it now versus how I started using it and just, you know, every best practice you'll see out there, you really should talk to it like it's a human. Now that's probably going to generate some weird emails coming back to us, but hear me out. When I say talk to it like a human, you should be polite. And most importantly, when you want something from another human, we tend to at times stroke their ego. And so people have shown that they tend to get better results when they structure a prompt in the beginning with something along the lines of… if I need a couple of subject line choices for an email, that's how I'll start it. I'll say, you know: as an expert in email marketing and understanding what it takes to create a good subject line to generate high open rates, could you give me a couple of subject line examples based on this type of content? So talking to it like it's a human, say please and thank you by the way. I always say thank you to it. 

Lauren: I was, wait, that was literally, I was waiting for you to finish talking because I was gonna say that my friends and I had a conversation about this over the weekend about whether or not we say please and thank you to our various AI, whether that's like ChatGPT or like Alexa or Google Home or whatever or Siri whatever you're using. Like, I always say please and thank you. 

Matt: Listen if for no other reason, and I don't want to stoke the fires out there. But if for no other reason, let's just say that the Terminator prophecy comes true in five to ten years, and AI is running the world. Don't you want to be known as one of those people who was very polite to it? 

Lauren: That was literally exactly, exactly what I said. Cause this all came up because we had an Alex- we were using Alexa as like a kitchen timer, and when I told Alexa to stop, because the timer was going off, I said ‘stop.’ And then I said, ‘thank you.’ And one of my friends was like, did you just thank the Alexa? I was like, yes, I did. Because one day when the AI take over the world, I want them to remember me as somebody who was polite. 

Matt: Yeah. 

Lauren: Like. But yeah, no, I mean, jokes aside, yes, I completely - speaking to it conversationally, speaking to it like it is a two-way conversation. It's not a search engine where you're typing in like a search query in a blank bar, and you can't react to the responses that it gives you. 

Matt: That’s right. 

Lauren: This is a conversation that you're having between two entities. One of them happens to be generative AI, and one of them happens to be human. I also asked our Content Marketing Manager for his favorite tips for using AI, and actually he gave me an answer that was something that I hadn't heard before, I thought it was really interesting. And he said that lately he's been experimenting with shifting the tone of his prompts and instead of giving it like a directive prompt, like write me an outline, he will instead ask it to put itself in the point of view of whatever. So he'll say, like: you are a podcaster writing an outline about whatever the episode topic is, what would this outline look like? And he said, he's gotten some like different interesting results from that shift in how he phrases the prompt. 

Matt: I think yes, no, that's pretty cool. I could see where you would get to. I think the point here is that you really need to prime the tool each time you're going to ask it or request something from it. And again, by prime it, I mean either stroke its ego or just put it in the, the quote unquote mindset that this is what we're going to be talking about. And I'm positioning you as the expert, right? To really dig deep and give me what I need here on this particular topic. So whether he phrases it as, hey, as a podcasting expert, please help me create an outline for this particular topic that should, you know, support a 65 minute episode on whatever, whatever. Or as you said, by shifting that in a way where he says, instead of saying as an expert on podcasting, he actually puts the tool in the position of the podcaster. It's kind of very similar. Either way, what you're doing is priming that tool. So I think that's the real important thing that people should come away from.

[21:48]

Lauren: Yeah. For sure. It is something that, like we said with the, when you're coming up with the prompts, the more specific you can be, like the more parameters you can give it, the more information you can give it, the better your results will be. But it's also, and this is another one of those things where, like I said, it was conversational and we have to treat it that way. This was one of those things that when I first started using it, it was actually a hard habit to break where I would just take the first answer that it gave me and didn't really… like, even if it wasn't an answer that I was like really happy with, I would just be like, ‘okay, thank you.’ And like, go back to whatever I was doing with this not really super helpful input from it.

Matt: Yeah. I kinda did something a little worse actually. 

Lauren: Oh? 

Matt: So I didn't just take the first answer. What I did was I would say, can you expand on that? 

Lauren: Oh.

Matt: And then, well first I thought I was being clever, right? And - 

Lauren: Right, but it's very open-ended. 

Matt: Well, that's the problem. And so it would, but it would expand on the entire thing it just gave me and so I'd say yeah, but can you really expand on that in a more detailed way? So it would try to.

Lauren: Right.

Matt: And I would just keep doing that. And I thought I was being clever, but the problem was all it was doing was just repackaging what it had previously given me in some other way. And it wasn't until I learned or saw that, hey, I needed to focus on whatever it gave me that I thought was good until tell it to expand on that component, that theme, or that set of bullet points or whatever that might, instead of just, you know, cause really what I was wanting was more of a certain aspect of what it gave me. 

Lauren: Right. 

Matt: Not the whole thing just regurgitated in a different way, so. I'd say what I was doing was probably a little bit worse, because what you were doing was probably just taking what it gave you, using it as a jump off point, and then spending some time really building around it. I was trying to get ChatGPT to really do the whole thing for me by baiting it to keep expanding, but I wasn't giving it the parameters for how I wanted it to expand. 

Lauren: Cause it’s almost, what you actually want it to do is narrow down, not expand.

Matt: Yeah.

Lauren: What you're trying to get it to do is to narrow its focus in on the specific thing that you want it to. But by asking it to expand, you're doing the opposite. 

Matt: It's like when I tell my son ‘go clean your room.’ And I walk in his room later, and really the one thing I wanted done - which was his desk - is a mess, and there's papers and books everywhere, is not really done. Technically I didn't tell him to clean his desk or straighten his desk up. I just said clean your room. And so, you know, it's the same concept. If I wanted the desk cleaned or I wanted specifically that to be done, I should have said ‘please go straighten your desk and put away all those loose papers and pens and markers and stuff.’ It's the same concept, yeah. Parameters and specifics are going to get you much better content. 

Lauren: And ChatGPT itself does provide that advice. 

Matt: Yeah.

Lauren: Cause I did ask it. I did the same thing, and this is another long-winded question again, but you know, as we said, you need to be specific. I asked ChatGPT: I'm writing an outline for a podcast episode on how authors can use ChatGPT as a helpful assistant or tool. And one of the sections is about how to use prompts or questions intentionally to get the best answers from ChatGPT. Do you have any advice on how we as users can structure our prompts to get the best results possible from you? So right away in there, I'm providing content for it, I'm asking it as an expert, like I'm asking it to explain itself to me. Giving it as much information as I can. I would never type a full two sentence long query into Google. Like I would never put that in a search bar.

Matt: Yeah. 

Lauren: But I'm asking ChatGPT for its input. And I'm giving it as much information as I can ahead of time. And it did, much like that first question that I asked it, it gave me a pretty lengthy list of tips and strategies for how to use it. And the first three things on it were: be specific, provide context, and use clear and concise language. That's the advice that it'll give you right off the top is like, be specific about what you're asking for and provide as much context as possible.

Matt: Yeah, I think… there are times where you can be a little bit looser with the query when you're starting pretty high up in the funnel for what you're creating. So like I said, I'm notorious for getting stuck at the beginning and given the context of what I do and the environment I'm in, I constantly have ideas for books or content that should be written. But again, I have such a hard time getting started. So a good prompt to use that I've been using, and I've written some of these down is I'll start by saying: hey, you are very specialized in generating creative ideas for books, or you're an expert in generating creative ideas for books. Let's brainstorm some ideas for a new, um, book on, and then I'll give it a topic, right? So if I'm, I'm working on a topic that day, or I just happen to be dealing with some content that day and something will hit me like, hey, this is something that people could really benefit from. So I'll use this prompt that I have written down. And then I'll ask it, please suggest a few themes or chapters that you might actually create in a book like this. And it'll give me back some really cool stuff to get started. So again, talking to it, priming it, and then setting it up hey, you're the expert, you're very creative. I don't know, I've had good luck with that. 

You mentioned that ChatGPT also has some resources to help you with prompt generation, things like that. At this point in the game, there are so many resources out there for the top 10 best prompts to use when doing this, or as this type of person, or as a marketer. Like just go out there and Google search that. 

Lauren: Oh yeah.

Matt: Like don't even ChatGPT that, go Google search the best prompts to use as a… whatever, a real estate attorney, I guarantee you there's a ton of great ones you don't need to recreate the wheel. There are people out there that over the last year, couple of years, that's all they've been doing is like just practicing and experimenting with different prompts. So go out there and find some really great ones and then just tweak them to your needs. 

Lauren: Yeah, I found a whole bunch of articles like that when I was working on this outline and they ran the gamut of everything from like some of the biggest names in like digital marketing, resources that we're all very familiar with, to Reddit threads. There are also plenty, you know, like we have referenced different conferences and different people that are speaking about this regularly. I watched a webinar this morning that was from an author that was saying, this is what I've learned about AI in the last year, and had a very similar experience to me: When I was initially introduced to this idea, I was very opposed to it. And the more that I've learned about it, the more that I realized how useful and resourceful it could be… There's also, I - Marketing AI Institute has a virtual conference coming up that is a AI for writers summit, specifically. 

Matt: Yeah. They also have an in-person event every year that's actually a really good event and we go, we'll be there again this year. There's a lot of great speakers and sessions on all different facets of using AI. So yeah, shout out to Marketing AI Institute. 

Lauren: I actually - I don't want to have to edit this out. So I'm just going to correct myself retroactively and leave this in the episode because the virtual event is upcoming for right now, but it will have already happened by the time this episode airs. So. 

Matt: Oh, I guess to be clear, yeah. I guess to be clear, the in-person event that I'm alluding to that they do every year is in September, by the way. 

Lauren: Okay, so that there will be plenty of time for, but the AI for Writers Summit that I just referenced will have already happened. We are recording this episode ahead of time because we're getting ready to go to London Book Fair. but yeah, we're getting, we're getting ready to travel. So we're recording a little bit ahead of time, so. Sorry if you cannot attend that writer's summit because you didn't hear about it until after the fact. 

Matt: I doubt this is anybody's main source for event and other news, but nice of you to to put us in that category. 

[29:27]

Matt: One of the other things I was surprised to see people using AI or ChatGPT for was feedback on their writing. Like, and I know that there's some people out there that are like, not only would I not use an AI to help me write, I would never trust one to give me feedback. So when I saw that people were using it, I decided to try it as well. And with a well-written prompt, you can get back some very good feedback on your writing. Just as good, if not better than you might get from peers or professional editors and proofreaders and people like that. As with anything, especially AI related, you should obviously always get another opinion or do your own research. You can't just take it at face value, but if you've been writing long enough you'll recognize good feedback when you see it, because it's stuff that the minute you read that feedback you know it's right. You just weren't able to see it because you were in the moment and you were the one doing the creating. 

Lauren: Yep. 

Matt: But again, using prompts similar to something like: as an AI trained in providing writing feedback, read this dialogue from my book. How can I make this more… whatever, whatever you're looking for. You can use prompts like that. And again, by priming it to say, as an AI trained in giving really thoughtful writing feedback, and then tell it what you want it to do, you'll get good feedback. You can do the same thing for… I'm still not sold on using it for research yet, so.

Lauren: I'm not really either. 

Matt: If you're using it to help you summarize research - 

Lauren: That, yeah. 

Matt: That's amazing. Because if you're, you're writing anything that requires you to cite, or rely on research for some of that writing, taking a 20 page article written by some academic in some far reaching corner of the earth that you know is going to take you six days to read because it's the most boring thing on the planet. You just dump that thing into ChatGPT and be like, hey, as an expert on this, whatever the topic is, can you give me your expert summary of this entire article? And it'll give it to you. 

Lauren: And I - I think another way that you can use it like that is to summarize your own content. Sometimes it's a really interesting way to see… this happens a lot with our podcast episodes. You know, we'll come in with a topic intention and a clear outline of what we're planning on talking about. And then because Matt and I couldn't stay on topic without deviating on some kind of tangent if you paid us with an all-expensive paid trip to Disney.

Matt: Yeah, isn't it surprising they let us do a podcast as two of the most verbose and opinionated people in this building. 

Lauren: I know.

Matt: Like, they're like, yeah, sure, go ahead, do a podcast. That's a great idea.

Lauren: There's no way this could go wrong. 

Matt: Yeah.

Lauren: But no, we do. So we definitely, like, you know, we always come in here with good intentions and then whether or not we actually stick to the plan or wind up deviating remains to be seen. But sometimes I will be going after the fact and I'll try to like write up a little show notes summary, and I'll write it based on the outline that I created, which is not always actually the direction that the episode went in. And then recently I've started uploading the transcripts that we get for these episodes into ChatGPT or some of the like GPT custom tools that they have. So there's one called Cast Magic that you can use for podcasting, and I'll upload the transcript into that and say, can you give me a three sentence summary for my show notes and a list of the top 10 key points that are made in this episode? And I'll go back and read through that and be like, oh, the details that I thought were the main points of this episode are not necessarily what ChatGPT identified as the main points based on the transcript that I provided. 

So it's a good way… if you're writing, if you're trying to get a better feeling for your writing, or if you're trying to see if the blog post that you wrote gets the point across that you're trying to get across. See what happens when you ask ChatGPT to summarize it back to you. And if the summary is not accurate to what you were attempting to say, maybe you need to rework some things in order to make it work better.

Matt: Yeah and sometimes what I do, because it's me. If I'm still not really getting it or if I've dumped something into it and I wanted a summary, sometimes I'll have to go further and one of my favorite prompts is explain it to me like I'm a fifth grader. If you are sort of struggling to summarize a large piece of content into something smaller and more simplified for whatever audience you're trying to present it to. That's always a good prompt. Or in general, if you're really just trying to create a piece of content for an audience where you do need to simplify something or break it down, if you're trying to explain the concept of a short stock or something like that, that might be a little more technical in nature. Now, I would also again heavily caveat that by saying, you should always check the facts of whatever is spit out from ChatGPT, but nonetheless, I love that prompt and I use it often.

[34:16]

Lauren: Yeah, I've been using it since you first mentioned it. And it can be very useful. I also, I think kind of like hand in hand with using it for feedback for your content. If you're not necessarily interested in asking ChatGPT for like opinion feedback, you can also just ask it for the more factual kind of feedback. You can ask it for like clarity of sentence structure. Is this grammatically accurate? Are all the words here the correct words? Is this supposed to be used a different way?

Despite the fact that I literally have a master's in creative writing, I studied this for most of my life. No matter how many people have tried to explain it to me, I do not know the difference between past and passed. Like, P-A-S-T and P-A-S-S-E-D, I do not know the difference between them. I literally avoid using them in writing and sentences because I do not know the difference between like when is the correct time to use the correct one. I know. I know. So many people have tried to explain this to me, including my high school English teacher sister, but…

Matt: I can't tell if you're trying to be funny here or if you’re serious. 

Lauren: I’m not, I genuinely cannot like distinguish between.

Matt: Give me an example. 

Lauren: So like if I'm saying, like, if I drove past something on the highway. I don't know which form of past I'm supposed to use there. 

Matt: Really? 

Lauren: Yes. 

Matt: Oh. 

Lauren: Yeah. 

Matt: Okay. 

Lauren: It's one of those, it's just like at this point, it's - that ship sailed. 

Matt: Out of curiosity, which one do you think you use in that situation? 

Lauren: I'm always wrong. So my instinct is to say P-A-S-S-E-D, which makes me think it's actually P-A-S-T. 

Matt: It is P-A-S-T by the way. 

Lauren: Dammit.

Matt: Okay. Wow. Alright.

Lauren: But something like that, I can now ask ChatGPT to help me out with that. And I do.

Matt: Okay. 

Lauren: Yeah. 

Matt: Yeah, I mean. 

Lauren: You know, use it, use it as a glorified Grammarly or something. I know. Look.

Matt: God.

Lauren: It's, it's fine. It's fine. It's okay. There are plenty of other things that I can do. Some people can't tell the difference between they're, there, and their, okay?

Matt: Listen, you're 100% right. Let's not get into Mary, merry or marry. I just didn't know that about you. Thank you. 

Lauren: It's okay. 

Matt: Yeah. Thank you for sharing. 

Lauren: I feel like I get dumber with every episode that comes out. 

Matt: You might not, our listeners probably do. 

Lauren: No, it's okay. 

Matt: Yeah. 

Lauren: But I will actually, ChatGPT is my primary grammar and proofreading tool these days and this is another one of the things that I love about it, in what makes it conversational versus a search engine where it is more one sided. 

Matt: Yeah. 
Lauren: I'm a big fan of using a thesaurus. One of my writing pet peeves, that's one of those silly hills that I'll die on is repeating the same word too many times too close together within copy.

Matt: Yeah.

Lauren: And I'm always the one that's pointing that out when we're reviewing copy for here, where I'll be like, hey, we just used the phrase, join us three times in four sentences, let's find another way to put this. 

Matt: Yeah.

Lauren: As helpful - 

Matt: It's worse when it's descriptors or things -

Lauren: Yeah.

Matt: Like the word great. 

Lauren: Yes.

Matt: Like, sir, excuse me, my brother in Christ, do you realize you just used the word great six times in two sentences? Like. Can we not be a little more creative? Yeah. 

Lauren: Yeah, no, I'm right there with you on that. And as great as a thesaurus can be for that, I literally have thesaurus.com bookmark on my computer, I use it all the time. But sometimes you can't provide context for words that you - 

Matt: Right.

Lauren: Words or phrases that have more than one meaning for them. And thesaurus.com will only give you synonyms for one definition of that word. ChatGPT, I can provide context. ChatGPT, I can say, I'm writing an email about inviting subscribers to attend an event that we are hosting, and I'm looking for synonyms for the phrase ‘join us at this event.’ Can you give me a list of alternative phrases that we could use here instead? 

Matt: The time has passed for that event, which form of past did I just used. 

Lauren: I don't know. P-A-S-S-E-D? 

Matt: If I say the time has passed, yes. If I say the time is past, it's P-A-S-T. 

Lauren: Jesus Christ. 

Matt: Alright, let's move on. 

Lauren: This is why people have such a hard time learning English. 

Matt: It's also why people never let me know their weaknesses in the office, because I'll just keep poking at them. 

Lauren: It's okay. It's okay.

Matt: Don't think I forgot about your deep-seated fear of birds, by the way.

Lauren: Shit. 

Matt: Yep. I'm still trying to figure out how to get a dozen birds into this building, but I'll figure it out one way. 

Lauren: I won't be here that week. 

[39:00]

Matt: Alright. 

Lauren: Alright. Well, I don't know. Is there anything else? I'm sure there's many, many, many other things that we could do. 

Matt: I think the goal here was to, to help people maybe understand some ways to better use it in their journey and careers as writers and creators and authors. And I think that if you want a more exhausting list of particular types of prompts or things like that. There really are a ton of resources out there where you can find that stuff. But I would say, just like we've done, A. don't be afraid to try it out, like just jump in and do it. If you're afraid to try it out still, you're missing a boat on some pretty creative things that could help you free up a lot of important time in your day to be used towards more creative activities or other things you'd like to be doing. The amount of time that these tools can save you as a writer and creator is, is immense. So you really should look into these and again a couple of quick Google searches, you'll find lots of cool prompts to use and I would just keep playing around with them and finding ways to to make sure that this tool helps you get a better jumpstart on what it is you're working on. 

Lauren: Yeah. I think that if you're still hesitant, if we haven't really sold you on it yet, or you're at least at the point where you're like. Okay, I want to try it, but I'm not quite at the point where I want to commit to using it for my own writing efforts or for my business, experiment with it. Try playing around with it for something low stakes. My sister - this is the third time I've referenced my sister in this episode, and she doesn't even listen to this podcast - my sister uses it as a DM for Dungeons & Dragons. She uses it for like if the party is getting a letter from somebody that's moving along the plot, she'll tell ChatGPT like, this is what I need to get in this letter, and this is the style that I want it written in, and ChatGPT will write the letter for her. So that way it's like in somebody else's voice other than her own. So it doesn't sound like it's coming from her and it's low stakes enough - 

Matt: Yeah.

Lauren: That the only people that are gonna hear it are the five of us that are - 

Matt: Yeah.

Lauren: Playing D&D. 

Matt: And it probably saved her at least an hour. 

Lauren: Knowing her, yes, because she is even more of a perfectionist than I am. 

Matt: That's a great thing probably to end on is just knowing when to use the content that's generated by the tool and when not to. And then whatever the middle of that spectrum is, like there are times where you're probably not gonna use anything that spits out. 

Lauren: Right.

Matt: There are times where you're gonna be tempted to use all of what it spits out. Spoiler alert, don't. 

Lauren: Nope.

Matt: You want to fall somewhere in the middle and just use that content as  a jumping off point. But there is low stakes content that Lauren just alluded to where, yeah, I mean, you could take 95 to 100% of what it spit out and just use it. Because like she said, only three or four people are going to see it. It's not for anything that's going to have any significance on anybody's life, other than possibly the outcome of this game that a bunch of nerds are playing in the basement of their mom's house.

Lauren:  None of us have basements, but…

Matt: Just kidding. I don't think you're nerds either. I'm actually, you know, 

Lauren: I am a nerd, that’s okay.

Matt: I wish when I was younger, I had gotten into something like Dungeons & Dragons or something, but I didn't. As an adult - 

Lauren: It’s never too late.

Matt: I just - no, I feel like I feel like the time has passed, Lauren. The time has passed me by. 

Lauren: Oh, well, too bad. That one passed me by too. But no, I do think you're right. And I know we've talked about this in the past on episodes.

Matt: I think we should get past this. We should move past this. 

Lauren: It's in the past. It's fine. 

Matt: We just lost whatever few listeners we had left, by the way. 

Lauren: Really? It wasn't when I brought up Dungeons & Dragons? 

Matt: No. 

Lauren: Okay good. 

Matt: This was it. This was the final straw. They are going to move past our podcast every time they are scrolling through  Spotify and Apple now. 

Lauren: Well, if you are still listening, maybe we can just end it on the idea of, we think that you should give ChatGPT or whatever other generative AI you want a try. Test it out, play around with it. I think that there's a lot of talk about it out in the world right now. And if you haven't actually used it, that talk only means so much. 

Matt: Yeah, and honestly, if there's anything that we left out, you know, of this this high level overview of it, send us an email and tell us. 

Lauren: Yeah, I'm sure, I mean, this isn't even the first episode that we've done on AI in this podcast, and I'm sure it won't be the last one. So if you want to give us some more ideas for what you want to hear us talk about for it, or you don't want to hear us talk about AI ever again, and you have other ideas you want us to talk about. Questions, comments, concerns. Want to attempt to explain to me the difference between past and passed in a way that no one else ever has before? Give us a give us a shout. Send us an email at podcast @lulu.com