
Unleash Possible · 2022-11-01 · 23 min
Key moments - from our scoring
Substance score
33 / 100
Five dimensions, 20 points each
Frank Prendergast joins host Samantha Stone to explore practical applications of AI image generators for content creators and marketers. The conversation centers on Midjourney as the primary tool, chosen for its artistic aesthetic and intentional design philosophy that steers away from photorealistic outputs that could enable deepfakes. Frank explains that unlike traditional software, AI image generators work unpredictably - even their creators don't know which prompts will yield great results, making experimentation essential. He shares his workflow for creating multiple blog post images by combining complex and simple prompts, then layering in style descriptors like "editorial illustration" or "magazine illustration." For maintaining brand consistency, Frank demonstrates converting AI-generated images to black and white in Pixlr, then overlaying brand colors using layer blending modes. He references Google's unreleased Imagen tool (superior at text rendering but shelved due to deepfake concerns) and Meta's competing efforts. The episode emphasizes that these tools excel at filling resource gaps in organizations with limited design capacity, suitable for blog headers, social content, and presentations where polish expectations differ from commissioned design work.
Midjourney is specifically designed with an artistic bias rather than photorealism, a deliberate choice by its founders to avoid enabling deepfakes. This artistic bent comes built-in, making it well-suited for editorial-style blog and content illustrations.
Convert the generated image to black and white in a tool like Pixlr (free) or Photoshop, then add a new layer filled with your brand color and adjust the layer blend mode (overlay or screen) to control how the color overlays the image.
Combine a simple prompt (a single key word or phrase) with a complex prompt (a paragraph from your content), then add consistent style descriptors like "editorial illustration" or "magazine illustration" to both, and compare the results.
Most current generators like Midjourney struggle with typography, often mangling or dropping letters. Google's unreleased Imagen tool reportedly handles text brilliantly but hasn't been made public due to deepfake concerns.
They're excellent for filling volume content needs (blog images, social posts, presentation graphics) but shouldn't replace professional designers for high-stakes, complex design work requiring precise visual specifications.
Our reviewer’s read on each dimension, with quotes from the episode.
The episode surfaces a handful of genuinely practical tips - using both complex and stripped-down prompts in parallel, building a prompt library with terms like 'editorial style illustration', and the Pixlr black-and-white-plus-brand-color technique - but the majority of the runtime is casual enthusiasm, filler, and mutual commiseration about being beginners rather than dense, transferable advice.
what I'll often do is I'll take a really complex prompt...and I'll pop that in as a prompt. I'll also do a really simple one...and then I'll just add illustration to the end of each one
bring the image into Pixlr and turn the image into black and white...add your brand color on top of it. Change the layer settings so that the image is showing through the brand color
The episode offers no contrarian or first-principles arguments; the main frames (Midjourney is artistic by design, prompting is unpredictable, iterate like you would with a human designer) are widely circulated takes in the AI image community, and the schizophrenia AI anecdote is interesting but tangential rather than structurally generative.
with artificial intelligence, you know, it is very much like a person in terms of we don't know what a person is thinking
somebody managed to create an artificial intelligence that could predict with incredible accuracy if somebody was going to develop paranoid schizophrenia. But the tool turned out to be completely useless because they never were able to reverse engineer how the AI was actually figuring it out
Frank explicitly positions himself as a non-expert enthusiast who experiments on personal blog posts distributed via LinkedIn Newsletter and Medium; there is no evidence of deploying these tools at organisational scale, managing a team, or producing measurable B2B business outcomes from AI imagery.
I'm not an expert in AI and actually that's kind of why it's interesting that I'm here right now
I've just been experimenting, playing, having fun, and then just trying to figure out, like, how can we actually use these tools
The episode does name specific tools (Midjourney, DALL-E 2, Pixlr, Photoshop, Canva), a concrete post-processing workflow, and a specific business use-case (one long-form blog post per month with six images), but there are no business metrics, time-savings data, cost comparisons, or before/after evidence of impact to anchor the claims.
I can do six images for my blog post...in way less than the time it was taking me to do one image
Meta Facebook are working on one at the moment. Google are working on one called...Imagen
The host asks a few genuinely useful tactical follow-ups (notably 'what do you use today to make the tweaks?') and draws out a concrete workflow, but she frequently injects self-deprecating commentary and frames questions as open invitations ('what other advice would you give people?') rather than pressing on claims or introducing productive friction.
what do you use today to make the tweaks? Are you using a design tool or filters or how are you doing it?
What other advice would you give people?
Computed from the transcript - who did the talking, and the words that came up most.
Looking to add AI art generated images to your content strategy? Join Samantha and her guest, Frank Prendergast , Digital Strategist at Frank & Marci, as they discuss the latest AI technologies that can expand your image library.
Transcribed and scored by The B2B Podcast Index.
Speaker A: I spent a huge amount of time just playing with it because it was fun.
Speaker B: You're listening to Unleash Possible, a podcast dedicated to sharing strategies and tactics that drive revenue for complex products and services. Get ready to learn from the best sales and marketing leaders in the world. Let's get into the show.
Speaker C: Welcome to Unleash Possible. I'm, um, your host, Samantha Stone, and today we're going to talk about a topic that I'm really excited about. I think there's so much potential, but we're just getting started on it. And my guest today, Frank Prendergast, is a digital strategist. I've had the opportunity of meeting him through some, um, Discord groups are both a part of and, um, have been really impressed with the sort of experimentation he's doing and welcome him on the show to talk about AI generated images and some of the tools that are out there that we can start using. Welcome, Frank.
Speaker A: Thank you so much. I'm delighted to be here. And yeah, as you say, I've. I've just been experimenting, playing, having fun, and then just trying to figure out, like, how can we. How can we actually use these tools, like, right now? What, what, you know, what use are they for us right now? So, yeah, I've been having a lot of fun in this area.
Speaker C: Well, I think it's fantastic. And I will just let the audience know that I have made some absolutely horrible images and using these tools. Um, and I've made a couple that are. Aren't terrible. So I'm definitely personally in the very early stages of figuring out how to do this. But I thought we could start our conversation today, Frank, talking about some of the tools. I know you've sort of been experiencing multiple tools. Tell us a little bit about the tools that you've been playing with and why you chose those.
Speaker A: What I will say is that really Mid Journey is the one that I've been using primarily. And the reason that I've been using Mid Journey is it just seems to have a more artistic bent than the other image generators that I've played with. Now, I will say that because I've played most with Mid Journey, there is an. There is a certain element as well of, like, I know how to get what I want out of Mid Journey. So there is that element to it. It could be that if I explored more deeply the other image generators that I could also get, uh, artistic, more artistic feeling output from them. But it just seems like Midjourney has that kind of automatically built into it. And I remember I do remember reading somewhere the, one of the founders of Midjourney saying that that was a very intentional play on their part because they wanted to stay away from the more, you know, creating a highly photographic image that could potentially, you know, they didn't want to get, they didn't want to allow people to get anywhere into the realm of uh, deep fakes or anything like that. I will also admit there's a tiny part to me, so I read that and I thought that's brilliant. I love that. I love that they're really consciously making that decision to be more artistic so that it can't be used for any nefarious purposes. And then I have to admit the cynic kind of piped up and there's a little voice at the back of my head that wondered are they saying this because they just haven't managed to crack the brilliant photographic results?
Speaker C: So you know, it's, it's so funny that you say that about it because I played a little bit with Mid Journey and I actually found it challenging for me in that I wasn't, I found it very um, fantasy driven and I, and I didn't do the homework to understand, you know, it's actually by design, which is quite smart and had I done a little bit more work up front I probably would have understood why they went that route. And I do respect that. That, that is, you uh, know whether it's real positioning for the sake of we haven't figured it out or whether it is truly by design and intention, it definitely has a specific feel to it and I think that's an interesting piece of this. So as we look at using technology to create images for us, each of the engines does do things slightly differently because of how they've been trained and written. And I uh, think in my head I, I don't know why but I sort of expected them all to work the same like some of the other technology that we've gotten very used to. If I look at marketing automation as an example, they all do slightly different things but they, you could almost log into any marketing automation tool and without any training, if you've used one, probably figure out how to create uh, a template and send an email without any real thought. Whereas these tools required a way of asking for a response that's, that's different than that. Tell us a little bit about your Mid Journey and you talked about, you honed in on how to get the image you want.
Speaker A: This is, this is a fascinating area and uh, like I will, I will just say up Front I'm not an expert in AI and actually that's kind of why it's interesting that I'm here right now is because I'm not an expert. Because what's beautiful about this is that you don't have to be an expert anymore. Like a couple of years ago you did, you had to be an expert or you at least had to be really technical because you had to install something on your machine and you had to figure out how to do commands in a particular way, whereas now the interfaces are getting to a point where anybody can actually do it online. But what's fascinating is, and the bit that I do not profess to understand is that ah, with artificial intelligence, you know, it is very, it is very much like a person in terms of uh, we don't know what a person is thinking, we know what they say, but we don't know what was going on in their head beforehand. And AI image generators, fascinatingly are the same. So the creators of the AI image generators don't know what prompts are going to yield brilliant results. And it's up to the user base to just experiment and throw things at it and try things out and come up with prompts that generate amazing results. So, ah, it's not like the image generators are pre programmed with certain things that are going to get great results. It's that uh, they're an artificial intelligence and they figure things out from prompts and they figure things out from the images that they've been trained with. And so nobody knows, nobody knows what, what, what might be that thing that's going to spark an incredible result. I read another, like I read a fascinating thing about artificial intelligence way back when, where it's just an example of kind of a similar thing where somebody managed to create an artificial intelligence that could predict with incredible accuracy if somebody was going to develop paranoid schizophrenia. But the tool turned out to be completely useless because they never were able to reverse engineer how the AI was actually figuring it out. So it's a little bit like that. It's like the image generator. There could be hidden prompts out there that we haven't even discovered yet that will yield the. You were saying you were struggling to get the results that you wanted and there could be this prompt out there that is just waiting for you to discover it that will yield the exact results that you want. And I just find that absolutely fantastic. Fascinating in terms of how it all works.
Speaker C: It is fascinating and it's I think also part of the art of image. Right. And so, um, we are, we're, we are creating through technology something that is a little abstract. And you know, even with really detailed prompting, I did find, over time I've gone in and I've probably requested maybe 50 images. So not a huge number of them, but 50, 50 images. Um, mainly because I just wanted to practice asking prompts and see could it get better and better and better over time. And I did find, you know, over time trying to sort of. I had something in my head and repeatedly asking and changing a little bit one thing and a little bit one thing, a little bit one thing. Um, I was getting closer and closer to what was in my head, which I imagine to your point is very much like working with a person. You. If I'm working with a graphic artist, I've got to ask the question in a way that they can understand what I might be envisioning because my stick figures isn't gonna do it right.
Speaker A: Yeah, exactly. It's kind of like you've gotta create a brief that gives the general kind of outline. Um, and then frequently when you're working with a real designer, the results of the brief will come back and you'll be like, oh, okay, yeah, this is close. But what I really meant was. And then you clarify. And it's very much like that. Absolutely. Um, like I'm, I tend to, I, uh, I spent a huge amount of time just, just playing with it because it was fun. And then, um, I used up my free credits like almost immediately. So I did a paid account and then I used up the, uh, I just went on the low tier. So I used up those images in no time. So then I ended up on the full price tier and I was like, okay, I'm, I'm paying for this now. Now I need to actually make this useful. And so I started looking at, well, how could. To create images for my long form content. So for that I now just have a really, really simple process where you, you, you referenced a minute ago about, you know, adding. You kind of start simple and add things on, which is a really, a really good way of, of doing it. And what I'll often do is I'll take a really complex prompt. So I'll take like, it could be a paragraph from the long form content I've written or it could be the headline for my long form content and I'll pop that in as a prompt. I'll also do a really simple one. So I'll just take like, what is the absolute nub or what's the most important word from that, and I'll pop that in and then I'll just add illustration to the end of each one because I just want, I want that kind of editorial style illustration. So that's another prompt I have is editorial style, illustration, magazine illustration. So I've built up this kind of library of prompts that I will just try alongside the words that I'm, that I'm uh, trying to create images for. So you just end up with this, this library of prompts that you tack onto the end of your um, of your concept. And yeah, that, that works really well. And I think the looking at the complex and the simple, you can then decide, you can look at each one. So it's kind of like getting um, a range of concepts back from the uh, from the generator. And then you can decide, okay, the complex one is close but not quite right, so maybe I'll simplify that. Or the simple one is close but it's not quite right. So I'll add some extra prompts to that. Uh, and I work my, basically work my way backwards and forwards at the same time and see what, see what works best.
Speaker C: I love it. The way I've been thinking about the image generators is that it doesn't replace, uh, this is an overly simple example. It doesn't replace needing a cover design for a, um, book I'm in the process of writing. But it may create an image, get used in the COVID design or it may be used, um, I've been thinking of it a lot for like blog posts and things, uh, content that has a different level of expectation of polish for me. Where are you, you mentioned long form content. How are you thinking about and using these images in your work?
Speaker A: So I'm using them for my blog posts which I also. So I write, I write one long form blog post a month. I push it to LinkedIn newsletters and I push it to medium. And you know, way back when I was just using like unsplash and things like that. But you know, they're unsplash and other free stock images. They're a fantastic resource but they don't tend to be particularly exciting first of all. And secondly, you know, there is that good chance that you're going to see that con, you're going to see that image pop up on somebody else's blog post as well. And so then I do have a design background. So I thought, okay, this is silly, I'm going to start designing my own for my blog posts. But they were just taking too long. Uh, it was just taking me ages to create just one image for the top of my blog post and then I still had this wall of text underneath. And so that's what I'm finding it absolutely fantastic for is that I can in, in the, in way less than the time it was taking me to do one image. I can do six images for my blog post. And it's very much what you said there in terms of the polish and the finish because you're not, you're never going to get exactly, exactly what you want. You're never going to get the precise, you know, you might, you might visualize an image in your head, put a prompt into the image generator to try to get that image back and that's not going to happen. But you'll get something vaguely close or vaguely related or it has some of the elements in it and it's, it really is kind of a case of going, yep, good enough. Um, because really for the blog posts I just want the header image to be in some way, you know, visually attention seeking or attention grabbing. And then the other images are really color, introducing color and space to the content. Uh, obviously I don't want them to be completely irrelevant, but I don't need them to match perfectly or absolutely. And if you flick through, you know, a business magazine and you look at the editorial images, they often, they're often the same. They don't, you can't necessarily see the absolute direct link, but you see the, the elements that they've drawn from uh, in to create the images. So it's those kind of editorial style images that I think this is perfect for.
Speaker C: I think this is really interesting. I'm glad you brought up stock photography. And look, I've used stock photography many times myself so and uh, on my website I've tried to limit it, but it is there certainly in some of the blog posts that I have and I use it sometimes in my presentations because you know, but, but you can almost look at a stock image and like I know what keyword you're searched for, right? And, and, and, and it, you know, it serves its purpose but it's not ideal. One of the things that I've been doing is encouraging clients and friends, you know, uh, who are peers in other organizations to start playing with these image generator technologies. But the question I get back quite a lot is well, how do I make it in my brand standard, right? How do I, you know, is it a color palette? Is it a, uh, kind of look? Do you have any guidelines for people? Is it by choosing the, you know, AI generator That best suits your style. Is it in. I've been playing with putting like a, like teal as one of my colors. So when I've been using my prompts, I've been trying to incorporate that color quite a lot in it. What other things might people do?
Speaker A: Yeah, so what I do, I was actually. I spent a long time trying to get the image generator to give me images in a particular color palette. And I asked on the, uh, mid Journey Discord what people were if people had explored this. And really nobody has come up with a great solution to it. So what I do is I tweak them afterwards. So I create the images and then I tweak them afterwards. And I do think this is surely something that will become available. I mean it has to be. It's quite a simple thing. So I would imagine that at some point you will be able to just feed in your color palette and have it work from those colors. And I can't wait for that.
Speaker C: Because right now, what do you use today to make the tweaks? Are you using a design, uh, tool or filters or how are you doing it?
Speaker A: Yeah, so I use Photoshop, but there's, there's a free online tool, uh, called Pixlr. I think it's Pixel or dot com. And so what, what I do is I will. You can, you can bring the image into Pixlr. You add a new layer and just fill that layer with your brand color and then you change the layer setting to be. I think it's overlay. I mean there's a whole bunch of, um, layer settings that you can play with and it kind of depends. And basically each one of them actually apologies, I left out a step. Bring the image into Pixlr and turn the image into black and white.
Speaker C: Ah, okay, that makes sense to turn
Speaker A: it into black and white and then add your brand color on top of it. Change the layer settings so that the image is showing through the brand color. Um, and then you can use overlay or you can use screen, depending on how light or dark you want the image. Ah. So it's just a question of playing with those layer settings. Um, and that's how I get this. These consistent and they're monochromatic, but I love monochromatic. So it's, it works perfectly for me
Speaker C: and it gives incredible are listening to Frank and doing what I'm doing, which is like a knot in my stomach. A little bit of like layers in Photoshop, like just not my skill set. You can even do simple things like put a frame around the image that has, you know, a branded frame that you have created or a logo that you create, uh, that you put in a corner. Things that we can all do. Even in PowerPoint, we can do that or Canvas or some of these other tools that don't require sophistication. So, you know, don't let, I guess my, my point was like, don't let that be a barrier to playing a little bit and experimenting and see what comes out.
Speaker A: Right, absolutely. And, um, you know, I haven't, um, I haven't explored, for example, canva, but canva has incredible tools in it. So like, for all I know, canva has like convert this to one color mark.
Speaker C: Ah.
Speaker A: Schaefer was telling me that you can do it in PowerPoint. Ah, I think it was.
Speaker C: Which I do it in PowerPoint all the time, which seems so silly, but I do most of my Image Manipulation in PowerPoint. I find it easy to crop and color change and add a, uh, text over and, you know, add image layers. Not that the photo tools are difficult. I just, I don't use them and I'm not familiar with them. So for me, I occasionally will manipulate an image and it makes it easy for me.
Speaker A: Sure. Yeah. No, exactly. There's, there's, there's bound to be a tool that people are using right now that they could use just to bring the images back into their kind of brand, their visual brand identity a little bit more. For sure.
Speaker C: I love it. What other advice would you give people? Did you play with any of the other tools before you found Mid Journey? I know. Maybe you looked at a couple other ones or was it just midjourney?
Speaker A: Mid Journey was the first one and then I had to look at the new, uh, the Dall E2. And I had a look at. Oh, no, I've actually forgotten the name of the third one that I had a look at. Completely forgotten the name of it.
Speaker C: That's fine. If you go in and people do AI Image Generator, they'll. A host of things will pop up that people can use.
Speaker A: Host of them. Yeah. And there seems to be more popping up. More and more popping up all the time. I mean, uh, Meta Facebook are working, uh, on one at the moment. Google are working on one called. I don't know how to pronounce it. It's either Imagen. Imagen maybe. Um, and that tool looks absolutely incredible. But they won't release it to the public because they're saying that it's so incredible that they're afraid that it would be used for nefarious. Purposes and that it would be used for deep fakes. Um, and I'm a little bit sad that they're not releasing that one for one simple reason, which is that it looks like it handles text absolutely brilliantly. Um, because I've tried to get Midjourney to do typographic treatments for me. Um, but nine times out of ten, it completely mangles typography. It completely mangles words. It leaves out letters, it adds letters. Um, yeah. So I would love. I would love one that could do typographic treatments. So if anyone from Google is listening, come on, guys, give us that, too.
Speaker C: Find us a way to stamp it right. So that the image can't be manipulated. But, yeah, and you know what? I. I do appreciate that the people who are creating this technology are thinking about those things, because it is important. Um, and, you know, it is. It. We don't want to make it too easy to do something that's manipulative in a. In a bad way. But what a great opportunity that we have to use these tools. And I will say I've used two or three, and honestly, I can't remember the names of them either. But what I do know is in all three that I played with, the images with the same prompt were really different stylistically. And so I would encourage people to go out and play with different things, and you could do them for free. There's, you know, I think all of the tools let you have a certain number of images at no cost. So you can kind of go in there and play with it and see, See what you think and, um, you know, give ourselves a little bit of a chance to experiment and pilot.
Speaker A: Absolutely. And I think, you know, the tools, they're relatively easy to use at the moment, and I think they're just going to get easier and easier. And I. I would hope, and I would. I would think that at some point you're probably. Whatever blogging tool you're in, you're probably going to be able to just, say, create some images for me. Um, and hopefully you'll be able to say, create some images for me and use my brand colors.
Speaker C: I love that. And I'm, you know, thank you so much for these tips, Frank. This is really helpful. Uh, you know, I think I've been working too hard at my image descriptions, and I haven't been doing some of the basics things. You just, like, take a piece of the content. I'm trying to associate with it and put it in as the prompt. I've been overthinking what I'm doing and trying to get too precise and I need to let the tools do the work a little bit for me and, and be a little bit more open to that. So thank you for that reminder. For those folks who might want to get in touch with you and say hello or, um, take a look at some of the work and the blogs that you've created images for to kind of see how these things actually work, where they might find those.
Speaker A: Yeah, so my, my blog is@frankandmarcy.com and uh, in terms of connecting with me, I would love if people would connect to me on LinkedIn. That's kind of my, that's my social platform of choice and I love hanging out there, connecting with people and having chats on there. So, yeah, please do connect with me.
Speaker C: That's fantastic. Thank you. And, um, for those of you listening, I hope this has inspired you. Just even go out and play. Whether you use it or not, I suspect you might find a purpose for it that you're not even thinking of it. It's an image for a presentation or an image for a blog post or maybe a social post. Something that is low risk and low weight. It gives you a chance to go out there and play because this really is going to allow us to scale creating images in a way that keeps us off. Dependence on things that aren't particularly effective today doesn't replace the need for all those talented graphic designers we have that do really complex things for us, um, and really important things. But it allows us to fill gaps which all of our organizations have because we have limited resources and limited numbers of graphic talent that we can tap into. So I encourage you to go out there and use it and keep us posted on how you're doing. At, uh, the Marketing Advisory Network, we specialize in B2B high growth environments that need practical advice to drive results. Our analytical expertise enables revenue growth, new opportunity identification, improvement, improved efficiencies and stronger collaboration between marketing and sales functions. To learn more, visit marketingadvisorynetwork.com youm've been
Speaker B: listening to Unleash Possible. To ensure that you never miss an episode, subscribe to the show in your favorite podcast player. Thank you so much for listening. Until next time, unleash new possibilities each day.
Other episodes covering the same guests and topics, from across The B2B Podcast Index.