Podcast: Play in new window | Download
Are you ready to take your content creation to the next level? Jeff Sieh his insights on crafting eye-catching, original visuals with AI tools like Midjourney and Ideogram. Whether you’re a marketer, solopreneur, or content creator, Jeff’s tips will help you create standout images that amplify your brand while saving time and effort. Generate Eye-Catching AI Graphics Summary
Generate Eye-Catching AI Graphics Episode Transcript
Rich: My next guest is a hirsute content creator, live streamer, and host of Social Media News Live. He specializes in content repurposing, helping creators maximize the value of their content by distributing it across various platforms. He’s known for his engaging and humorous style, often mixing practical advice with wit, and has built a community of solopreneurs, marketers, and content creators.
He’s also passionate about using AI to repurpose content and craft images, and he teaches others how to leverage AI in these areas. Additionally, he’s involved in video production, podcasting, and speaking at events. His work is all about making content creation smarter and more efficient while keeping it fun.
And today we’re going to be talking about how AI is going to impact the images and even the video you create with Jeff Sieh. Jeff, how’s it going?
Jeff: Man, thank you so much for having me on. I’m excited to talk about this. It’s been a while. We shared the stage, I guess it was Lima a couple of years ago, but we’ve always run in the same circles. We’re old. We’re we’ve been in this for a while.
Rich: Yes. We are OG creators for sure. So it’s really funny, we just actually got off a call that we were both on. They asked you to be there, I just showed up. But it was it was for Mike Stelzner’s AI paid membership group, and you were basically the star attraction for how to create images using AI.
And I’m just curious, there were a lot of questions about consistency, about which image creators to use, about free software. As you’re out there, are those the kinds of questions you’re hearing the most, or were there any questions that you are used to hearing that maybe you didn’t hear in today’s session?
Jeff: They’re always about how to prompt and get the best-looking images out of there. And then there’s always the questions like, hey, is this ethical and all those kinds of things, too. And you have to look and see what the models are training on.
And we talked about it a little bit as well, but you can’t create Spiderman to hawk your real estate business and expect Disney not to come after you. You just, you can’t do that kind of stuff. You got to use some common sense.
But the cool thing for me that I get passionate about is that now, and you’ve probably done the same thing, it’s you’ve struggled with for years and you have this idea in your head, and you can’t get it out on paper, or into Photoshop, or Illustrator. And this is the first time I’ve ever been able to actually do that. Like have something in my head and then make it happen and have an original image.
It was a couple years ago, I spoke at Lou Mangiello’s event, he’s an OG podcaster, and he does a thing in Florida, and it was the first time I was ever able to do a presentation using all original slides. Always before, we’re always trying to find stock photo and shove it in to make a point, and it works, and you’re like, I wish this was all original. And it was all original. It was amazing. And it was so exciting, as a creator and as a speaker, to have it be totally originally and be exactly what I wanted to communicate to the audience.
Rich: That’s awesome. And so for me, from my standpoint, is I’ve been using a lot of these tools for a while. I love using them, but if I’m being honest, I love using them for the creation standpoint. I love seeing what the AI is going to come up with. I love playing around with it. It’s a real creative outlet.
I’ve been less than impressed sometimes with either the final product, or how long it takes me to get to the final product. So I’m hoping through today’s conversation we can turn this from, at least for me and perhaps the people listening – because I care about them too – about how to turn this from something that’s just fun, into something that’s really a good tool that content creators, marketers, and business owners should be using.
So starting at the beginning, obviously you’ve always been involved with visuals. You were the manly Pinterest guy. I’m curious about what drew you into using AI for image creation. Was there a tipping point or an ‘ah-a’ moment where you’re like, oh my God, this is the way to move forward?
Jeff: So I’ve always liked retro images, like The Hardy Boys, that kind of style, that kind of stuff, the cool Norman Rockwell kind of images. And it would be too expensive to have a graphic designer do that. And so I was playing around with some of this, and it was two years ago when I did that conference. Midjourney was based in Discord then, you had to do all these weird things inside of Discord. But I was able to create an image that I wanted, and it looked like it could come from the artist.
And I was like, okay, this is real. I’m finally able to take what’s in my head and get it out there. I’ve been able to do Photoshop and stuff like that. And just being in the business, you’ve learned how to manipulate things to try to get stuff to work. But this was like, oh my gosh, I can actually get what I want.
The thing you mentioned before, the struggle with trying to get, it’s great for ideation but you’re never really happy with the final result. For me, as of now, I can’t create a prompt and ever have the first thing I generate be the one that I pick. It’s always like moving down. The thing that I’m able to, when I create something, I can hone in on that.
And that’s why my favorite program is Midjourney, because it gives me four images and I can vary those images, wide variety of images, but then I can drill into it and go, okay, I want this to be that, I want this to move this over to this side, I want this guy to have dark hair, make him angry or whatever. And I can do that inside of Midjourney and get that final image that so many people struggle with. And it does take some work, but it’s a lot faster than me trying to make it in Photoshop or put up on fiverr and have somebody try to recreate it or that kind of thing.
So the key is to find the tool that works for you and giving you the images that you want. I go with Midjourney because of the variety and the control that I have. Dall-e is one that a lot of people use, and they get disappointed in. It reminds me a lot of Canva when that first came out. Like you would go and you’re like, “Oh, there’s a Canva image.” “Oh, that was made with a Canva template.” And now I feel that same with Dall-e.
I can usually see somebody who’s made an AI generated image is going, “Yeah, they use the ChatGPT and that’s from Dall-e.” You can just tell Mid journey I can do whatever I want, photorealistic art, abstract, and it’s just the jack of all trades for me right now.
Rich: Yeah. And I agree with you that you’ve got to find the right tool. And if you have the budget to be able to use multiple tools is definitely the key.
I, like you, love Midjourney. I prefer it. That’s my go to tool. But there was a time when I was creating rebuses, business themed rebuses, word puzzles, and I was able to create them best in Dall-e. So sometimes Dall-e seemed to be really good at creating crisp, black and white clip art based on my descriptions, and usually within one or two iterations.
If you do have the budget or if you’re willing to try some free trials and keep on coming up with a new Gmail address every week. I’m just kidding. I’m just kidding.
Jeff: So a perfect example is Midjourney doesn’t really do a great job with text. It’s getting better, but it’s still not the best one. So when I need to make text I’ll go to Ideogram. That’s the one I found that does the best job with creating text-based graphics, and they have some really good quality art and photorealistic images. So those are the two I play with because you know if you’re doing social media stuff or you’re doing marketing, you’re going to need both. You’re going to need stock kind of photo-ish kind of things, and you’re also need those text-based graphics And so I have paid for both of those because I use them interchangeably.
Sometimes I’ll create a background in Midjourney, and then I’ll use a Ideogram to create texts, and then I’ll remove the background and put them on top of each other. And so I think that’s also what helps you stand out in the feed when you’re not doing the same thing everybody else is. If you can combine tools and come up with something kind of new, then I just think your images are elevated.
Rich: I would agree. I think it’s almost a year ago when you would do a simple prompt to create an entire blog post, and you’re like, “Oh my God, it’s brilliant!” And now if you saw a similar blog, you’d be like, oh, that is one lazy prompt. And you wouldn’t think anything about it. It’s using these tools as an artist might, where it’s once you start to use them and then combining them in new and fresh ways, I think is where you’re going.
But at the same time, some people are just looking for the fast, easy way out. And one of the things I’ve done in the past, especially for slide decks, is I will create a background image in Midjourney, and then I put text layered on top of it in another program, either Keynote or maybe I’ll pull it to Photoshop and layer on that way. So that’s another option to about using multiple tools.
You referenced some of the AI or some of the ethical concerns people have when it comes to AI generated art, especially regarding originality and copyright. What do you say to people when they ask you about their concerns when it comes to this?
Jeff: I say that if they’re really concerned about copyright, I would go with Adobe Firefly, because they have trained on stuff that they’ve purchased or they own, and it’s probably the safest and best bet out there if you’re concerned about that. My thing is these artists have put stuff out there and they’re usually being trained on stuff they’ve already put out on the web, but even artists learn by copying. And I’m not saying that’s right. We’re just, it’s copying it at superhuman speeds.
Like even Leonardo da Vinci, when he was learning, he would go to the masters and copy their style and chisel out and try to copy exactly what they did. AI for me is like that, that it just does it a lot faster. Now, if you are taking somebody’s art and passing it off as your own, then I have a little bit of problem with that. If you’re selling it, Like there’s some really great comic book artists and people are using them as models. I have a problem if you take that and print it on a t-shirt because you changed it up a little bit. That’s not cool to me.
What I like to do is take different styles, mix them together to something new, and make my own. Because that’s what I think artists do anyway. There’s really nothing new under the sun. It’s just so easy now and it’s so fast. And we’re not talking about music, but even music is a whole other thing. The stuff that you can create now with AI generated music. Where is that line? We still have to have conversations about it, but for me, for what I’m doing, I’m doing it in a way that I feel like I’m not really stealing from them. I just, I don’t have a problem with it.
Rich: All right. Now for the marketers who are just getting started with this, and we heard some of these questions come up on the segment we were just on, is there a first tool or a first step that you would recommend where they should get started?
Because with so many tools that are out there, and seemingly more AI tools coming on the market every day, each with its own little twist on the original formula, where do you suggest people start?
Jeff: So I’ve been pointing people to Magai. I think you’ve used it as well, too. But because it gives you access to everything, you can play with things and find out for one monthly fee, instead of trying Midjourney for a month and then canceling, and then go to an Ideogram and canceling. It lets you use most all of the big image tools.
Now Midjourney is not part of that because of their API, but Dustin, the founder and creator of Magai, he’s always adding new ones. And so it’s really cool to go in and be able to play and see how those different models do different things with your prompts.
So that’s where I’ve been pointing people. If they’re getting started and they want to kind of experiment with AI image generation, I would go to Magai, and go in there and start playing and finding out which one you like the best.
Rich: I would agree. I definitely think that if you’re just starting, it’s the smorgasbord of AI tools because you’re able to do so many else. And also you can tap into a bunch of text and even video editors all within that same license. So that’s definitely an interesting place to start.
So there was a conversation about how much time or money this will save using AI generated tools. And I’m going to be honest with you, this is just my own experience, but I love using Midjourney and other AI generative tools to create images. But I often find that I am spending more time Trying to create the right image, when at the end of the day, I end up going over to Adobe Stock, finding something that is damn close to what I had in my head in the first place, and just using that. And, it’s a monthly fee, but it’s virtually free because I get 700 images. How can I break that habit? How can I create consistently high-quality images in a time effective manner?
Jeff: We talked about Flux is a free one, and there’s some other ones. The problem with those is they usually generate one image. The one thing I love about Midjourney is you get four. One prompt gives you four, it gets me there faster. And I can also turn on what they have “fast hours”, which you can actually render more than four at once if you want to. It’s going to burn through them. But a lot of times I’ll do six or eight when I’m trying to come up with something. Especially if I’m in a deadline for a presentation that I’m doing and I need some slides, I’ll use those to go in and then drill down.
My thing is, even the stock photo libraries, I could go to a presentation and somebody else could be using that same image. They could. I’ve seen it happen with presenters. There used to be a stock photo of a guy holding a banana, and it seemed like every person used that for their presentation.
These are all mine. They’re all original. Nobody has them like what I have put together. That is really cool to me. And even we do it for our blogs, we do it for Instagram, we do it for all the pieces that go out. They’re mine. There’s no stock photos anymore. And for me, that’s really important and really cool.
Rich: Prompt engineering is something that some people swear it’s the future of AI. Other people say it doesn’t really matter even anymore, that the AI is smart enough. How important do you feel prompt crafting and generating the right types of images is, and what tips do you have for creating more effective prompts?
Jeff: So it’s going to be based on platform. But a lot of times you can actually go inside of ChatGPT and have it craft a prompt for Midjourney. There’s a plugin called Gibletree, I think, is in ChatGPT that you can use. It will actually give you four different images and four different styles, and also give you Midjourney prompts that you can copy and paste over. Which I find really helpful when I’m trying to explore styles.
I think that especially for Midjourney, every word matters. Sometimes you want to get super specific, and you want exactly what you thought in your head, or you’ve used ChatGPT to give you the F stops and all the things, the camera and all that kind of stuff. But sometimes you want to have a kind of a more general because you’re exploring, you’re not really sure how.
I go back to this all the time. We had a lady on who is a professional voiceover artist, and it was about the power of sound. How do I portray sound other than just a waveform that everybody uses? So I had ChatGPT go in and craft me a prompt. And I was able to go in and use that and get an abstract image that I could use for a thumbnail background and paste that into Midjourney.
So you want to have, there’s a balance. If you want super specific, you got to use super specific words and maybe a longer prompt that you’re putting into Midjourney. If you want something more general, you need to lighten up a little bit and let m Midjourney play. And then you can go down some rabbit holes and find the perfect image.
Rich: Yeah, I found that sometimes I like going in there and literally just seeing what happens with a one-word prompt, like “darkness” or “entropy”, and it just starts a really interesting creative journey. It’s not always something that ends up in flyte new media or Agents of Change social media feed necessarily, but it’s a lot of fun to play and experiment with these tools.
Jeff: You can use emojis as well. It’ll actually go off of emojis and you can see what that generates.
Rich: Oh, that is going to be what I do right after this interview is over! So I’m definitely going to have to check it out.
What are some of your go to commands in Midjourney or some of these other tools like aspect ratio, 16×9, or dash raw? What are some of the ones that kind of give you the most control or the most interesting outputs?
Jeff: So I go between raw and the other toggling it on and off a lot. I really like going in and going dash random when I’m bored, because I can unlock those S ref codes, which are the cheat codes that you can plug in later for Midjourney, which is fun to do. And a lot of times I’ll make a mood board to see how Midjourney will actually treat faces, colors, all that different things with using those S ref codes.
So proflow.ai is a great resource to go in and find those codes and some prompts that people have unlocked and uploaded there’s whole YouTube and Twitter accounts where people are just sharing these S ref codes. And there’s some really cool ones out there. So that’s probably my.
My biggest hack, I use that dash R in Midjourney, and then put a number to have more than just those four come up if I’m really cranking stuff out for a presentation aspect the thing that’s been amazing is the new web editor compared to the Discord thing. It was just a nightmare. And now dragging images that I have that I’ve created before, or that I actually are of me that I can just drag into the editor and say, “base it off of this” is really cool for coming up with really images that fit your brand and your style.
So those are the big ones that I use. The web-based editor and being able to upload and combined images to come up with a new style is really pretty powerful in Midjourney.
Rich: So there was a lot to unpack there in terms of it got a little geeky, which is awesome, but no, it’s great. So let’s talk, define SREFs for me.
Jeff: So SREF stands for ‘style reference’. It used to be in Discord, you’d go dash SREF, and that’s where you’d upload your image URL. Now you don’t have to do that. You can just upload an image and click on an icon, and it will be that style reference.
Style reference codes are almost like hex codes in Canva. There’s actually a S like a style code that you can put in to have your styles look the same. A cheat code. That’s what these SREFs are. And there’s like a billion of them and nobody’s unlocked them. But you unlock them by going ‘dash Random’, and it’s just going to give you a random thing to your prompt. And people are posting those and cataloging them and all sorts of cool things. But that’s what SREF stands for, a style reference.
Rich: So if I were to upload, or if my prompt was something like, “a bouquet of flowers”, and then I was to do one of these SREF or the random, what kind of things happen once I put in random or any of these things.
Jeff: It’s totally random. What it’s going to do is take that prompt and whatever that code is that it’s going to give you randomly. It’s going to apply that to your image. If you already have that code or you have done some random ones and you like some of them, you can actually use that code on all your images whenever you want to apply that style to whatever prompt you want.
So it’s a great way, if you find them on the web or you find a certain style that you like, that you can just copy that and paste it and it’ll apply that style to any prompt that you put up there.
I’m hoping, and it’s been rumored, that eventually you’ll be able to upload your own image and it will unlock a code that you can use, which would be very handy. But so far, it’s all based on random, and you just have to find them on the internet or do them yourself.
Rich: All right, so here’s an issue that I’ve run into quite a few times. And this comes down to I’m often frustrated by the lack of control that I have over these AI images. For example, what I would love is we’ve got these six great Agents of Change characters illustrated by Joshua Fisher, he created them for us, and occasionally I’d like to have a different perspective on them, or have them doing something else, or putting them into a situation that they weren’t drawn for. Josh is very busy, he can’t always get to me. So is there a tool that I could take these illustrations or perhaps it’s a photo or whatever it is and put them into different circumstances and get an image that looks like the original image, in terms of people would recognize, oh, that’s the AI agent, or that’s the SEO agent, and so on and so forth?
Jeff: Yeah. So you could actually upload that image and then use that image as a style reference for anything else. Or you could even experiment with a character reference. You see a lot of those Midjourney images where it looks like the same person is in different scenes, they’re using character reference on that.
I did a thing, and this is what got me the gig for speaking at Social Media Marketing World next year, is I downloaded Stelzner’s banner image. You know, he has that kind of signature look of a kind of an illustration. I was able to download that, upload it to Midjourney and say, “I want Godzilla coming down from a mountain”, and it did that in that style. And so he was like, oh, okay.
And then I was able to go in with, there’s what’s called “in painting”, where you almost like in Photoshop with a Photoshop brush, brush it out and change it. And I said, “I don’t want Godzilla, I want King Kong”. And it kept everything the same and it brought King Kong in.
So you’re able to do that now with your own images or even styles that you like. So for your Agents of Change kind of graphic, I would upload that as a style reference. Upload that to Midjourney, use it as a style reference, and create something new based off of that.
Rich: How specific… like, as specific as you can be, I upload the image to Midjourney. And now there’s midjourney.com. we don’t have to bother with Discord anymore. And then is there a button there where it says “style” or “prompt”?
Jeff: So you have three… when you upload an image, you do the image drop down and you’ll see that image there. And there’s three different things you can click on. It looks like a profile photo, it looks like a paper clip, and another one looks like a scene or a photo. The middle one with the paper clip is the style reference. The other one’s an image reference. And the other one is a character reference. But you want that style reference to go and create that based on the image that you’ve uploaded.
Rich: All right. So I know that many AI tools struggle with letters and words. And then, like we were talking about, I’ve often brought it into Photoshop and added text there. What are your favorite tools? You mentioned Ideogram. Is that the only tool, or are there other tools that we should be experimenting with to create words that are actually spelled correctly?
Jeff: Ideogram is the best one that I have found. There’s some other ones. And once again, it’s an arms race, they’re going to get better. This is the worst everything’s going to be right now.
But a hack that you can do that I’ve used a lot of times, especially for slides, is I’ll actually inside of Ideogram say, “I want you to write this phrase”, like Agents of Change, right? I want to have a really cool Agents of Change lettering. I’ll say, do it on a green screen or do it on a solid color. Then you can upload that to Canva and remove background, and then you can put that wherever you want and it’s transparent and you don’t have any of the issues. Or you can do it in Photoshop. But it’s a great way to move something around and split it up so you don’t just have a background that you can’t take out later.
Rich: So is Ideogram creating new fonts for us?
Jeff: I think so. I actually took it in, there’s a program with Adobe where it will recognize fonts, and it didn’t recognize them. So I’m thinking it’s braiding it generatively and combining different styles and stuff. So it’s a lot of fun to play with if you’re trying to use some quote graphics or something like that.
Rich: All right. Now branding is really important to me, to my company. We actually, flyte new media, offers branding as a service. It seems to me that a lot of these platforms out there are not fantastic when it comes to staying on brand.
One of the exceptions would be Canva, which comes with its own brand kit, or you can create your own brand kit and upload it. But then again, I don’t find the Generative AI imagery, all that fantastic on Canva, so it’s a little bit of a trade-off.
How do you recommend we find a balance between staying on brand and using some of the better tools that are out there right now?
Jeff: So this is what I’ve done for Midjourney. I use Canva to go create a swatch palette with the logo and then the main colors of the logo or whatever. You take that image, and you make that a style reference or a character reference inside of Midjourney. And then whatever you create is going to use your brand colors. And it works really well.
I did an example for Stelzner, where it was his logo and we created a bunch of robots and they were all in his colors and the look and feel. So that’s what I would do for branding.
The other cool thing you can do, especially if you have really like Red Bull or a racing really strong brand kind of presence, is you can actually upload an image. Let’s say I had a friend who did the dirt cart racing things. And so he was a PR guy. So we took one of his images, upload it to Midjourney, you can also upload it to ChatGPT. And you can say, “describe this image”, and it will describe that image.
And then you can take those descriptions for both Midjourney and also a ChatGPT combined together and say, “Combine this, make it a really precise prompt”, and then you can upload that and create all these different assets from that image that matches. So we were able to create all these race cars in his logos, in his colors, all that stuff with the same motion blur and everything in Midjourney.
By doing it that way, having it describe it and then taking that, creating a prompt and then putting it back to mid journey and having it render out all these cool race cars.
Rich: In ChatGPT, I can create custom GPTs. I can basically train it in my own voice or my company voice, and then I can actually specifically ask questions of the virtual Rich Brooks, or whatever the case may be. What you’re describing sounds similar, but is there a way in Midjourney where I can always say, “use the flyte brand”, “use the Agents of Change brand”, or whatever it would be? Or is it each time I’ve just got a prompt library offline, and I’m bringing that prompt in each time that I want to create it?
Jeff: So they do now have a really cool folder system inside of Midjourney, which is really helpful, and you can search inside of it. We have thousands and thousands of images, and so if I want to go find a pink Cadillac that I created for a client, I can go and type that and it’ll pull it up, and I can get the prompt and everything’s still there that I use for that, which is really helpful. The images that you upload stay in the image dropdown. I don’t know how long they last or how many you can have in there.
There’s quite a few that I have, but what you could do if you’ve got a style that you’re using for a brand, that would stay there, and you could always use that as a reference and you wouldn’t have to go find it every time. There are not any great tools like Magai has where it’s got a whole built in prompt library, something like that, but they are getting better with organizations.
Rich: Alright, maybe they’ll listen to this podcast, they hear you speaking, and we’ll have this tool in a couple weeks or so. This might be me, I always feel like I get tired of things sooner than the average person does. But I feel like most of the AI-generated images I’m seeing in my social feeds, at first it was interesting, but now they’re just not that good. And the novelty, which used to keep me hooked, has worn off. Maybe that’s not true with everybody.
What is your recommendation for social media managers, people who are responsible for creating images in AI for their social, to make these images really stand out and pop in a sea of AI-generated images?
Jeff: I think it’s going to come down to blending things together, like coming up with your own style. Don’t just hit the easy button and create a prompt and just say, “Hey, that’s great. I’ll post that in Instagram”. Actually go in and…
What I have done is I’ve really drilled into the look and feel of what I want. It’s that kind of fifties retro. We took a lot of time going into the way we want it to look and the fonts and the colors and all that stuff. Dive into that so you stand out and you can create amazing images. There’s a lot of great, amazing AI images, but you’re right, it’s the sea of great images.
And then where do you fit in? Because we have that nostalgic kind of feel to it, I think that’s what makes us stand out in the feed compared to others. And so you need to find out what that is for you and learn artists. Like for an example, I was trying to do like this Jetsons feel. Now I’m really dating myself. Jetsons was a cartoon way back in the day. But they have a certain kind of architecture. I didn’t know what that was. Turns out, it’s called Googie, like not Google, but Googie. And it’s the kind of this retro style. Once I was able to learn that, then I could go in and tell these AI models to create images based on that.
So you need to learn artists, you need to learn artists, and keep a little notepad. And then start blending those together to come up with your own look and feel that will help you stand out compared to everybody else.
Rich: Side note: we are closer to when the Jetsons is supposed to take place than when it actually aired. So just thought I’d throw that out there.
Jeff: That’s good. Yeah.
Rich: Yeah. And you have that consistent look and feel it. And as you’re talking, one idea that I was just thinking of is I could take flyte or Agents of Change brand guide that we’ve created, which has our fonts and our colors and examples of the photography we choose and all this sort of stuff, and upload it to a tool like ChatGPT and ask it perhaps for some feedback and some descriptions that then maybe would go into a prompt, that we’d start to get some consistent look and feel for some of these images, too.
So if you are lucky enough to have one of those brand guides, that might be something worth trying out at home, too.
Jeff: And also, your logos and all of the images that you’ve already created, you can have multiple style references on Midjourney. So I would put three or four of those and then do a prompt and see what it pulls in from your existing catalog, and it may really surprise you.
Rich: Yeah. So I mentioned that we use Adobe Stock here at flyte, and I’m very used to going in there and finding the right images and things like that. Some of them are great, some of them are terrible. Do you see AI generated images as being a replacement for traditional stock photos, or do they serve a different role, in your opinion?
Jeff: I think the big stock photo libraries are probably on their way out. I don’t know the last time I used them. For me, I just don’t anymore. Because I want to have those original images and have them exactly the way I want, even for my blog post.
So I think there’s always going to be photographers. There’s always going to be great photographers. There’s always going to be for art photography. There’s always going to be a need for that. But I just don’t think there’ll be… Adobe will probably stick around, just because they’re Adobe. But I think some of the other ones will just slowly go away.
And I think the same thing for stock music libraries, I think. Now I can generate any of the sound effects that I need with AI. So I’m like, will I use those libraries anymore? Why not create my own music for my YouTube videos that’s original and I don’t have to get hit by a copyright thing or a license thing because the license thing didn’t go through.
Anyway. I think there’s going to be a lot less of them. You still might use them if you’re going to have a picture of the constitution of the United States, you’re probably going to have to find that from a stock photo library, because as of now it’ll get close maybe, but it won’t be the original Constitution that everybody knows. So in the same thing, if you have a picture of Disney that you need for a blog post, it will create a Disney-like thing, but it’s not going to create that Epcot and the things that people associate with the actual place.
Rich: And again, that goes to the whole copyright issue anyways. And nobody’s going to sue you faster than Disney.
Jeff: That’s right. They’ll get you. You can’t even say, we can’t even say, “Mickey Mouse”. Oops.
Rich: Exactly. We’ll bleep that out. Are there any AI tools that can take a flat image and turn it into a 3D rendering? Again, thinking about my own characters for the Agents of Change and just filling them out as you might.
Jeff: One experiment with Midjourney, because they have some really cool making things look Pixar-ish and 3D, that’s a great one. Ideogram has rolled out in the last month, actually, where you can actually go and tell it what kind of image you want and there is a 3D button that you can play with. I have not played with it too much, but that’s where the, one of the first places I would go if you’re wanting to make a from your 2D art to 3D.
Rich: Alright. And then the next logical step would be to animate them. I’ve got one of my characters called “Mobile”, and she’s got a jetpack, and she flies around. I’d love to see her actually zipping around. What tool might I get started with or play around with to get her flying?
Jeff: I think you could probably do it today with Kling. And if not Kling, then I would use RunwayML.
Rich: Talk to me a little bit about character references. What exactly are they, and how are they used to create more consistency in some of our image creations?
Jeff: Yeah. So that was a big thing for AI generative tools, it was hard to keep a character going through scenes. If you’re trying to storyboard something or make your own comic book or something like that, character reference actually does a pretty decent job, and it keeps getting better in Midjourney.
So you can upload an image and say, “This is a character reference. Now create that same…” let’s say it’s a dark-haired lady and you say, “I want that same dark-haired lady on a park bench. I want that same dark-haired lady walking a dog.” And you can do that now with character references, and you can actually have different weights. So you can tell it how much of that character you want to pull. And so you can actually adjust that inside of Midjourney.
Rich: All right. Can’t wait to play around with that. Are there any AI tools or updates on the horizon that you are specifically excited about?
Jeff: I think the more we’re going to be able to do with video generated from our still images. It keeps getting longer and you keep extending some things. That’s where I think I’m really excited being able to upload your own images and then adjust the canvas and actually go in paint your own images. Now that came out last week in Midjourney and Ideogram as well.
So those things are cool. The thing that’ll excite me is when some of the reins are loosened a little bit. They’ve got a pretty tight right now which what you can generate just to be safe so you don’t… especially at the election today, they’ve I think held back on some of that. But being able to go in and adjust an image.
So I do Amazon videos, and let’s say I had my wife modeling a new blouse and I’m like, “Okay, I want the same blouse, I want to change the color.” You can do that now. That stuff’s really cool being able to go in and do that kind of thing. So that, and then if I could actually go in, upload my own image and create a style reference code, like we talked about earlier, have it be my own for my brand that I could apply every time. That would be just amazing to be able to constantly create images just by using those style reference codes.
Rich: That would be great. The other thing that I wish that we had is I would like to take myself and be able to put it into one of these tools and get me in different situations. And I know you say that we can do this now, I have yet to be able to do it.
Now you would work better with because you have real character with a long beard and there’s things about it. And I think back to when we spoke at Social Media Week Lima and we all got bobbleheads, and yours was amazing. And I got mine and I’m like, who is this guy? It’s just the most generic, white guy I’ve ever seen in my entire life.
Jeff: So you’re going to be on my show later on, I think this month. And so I will create a thumbnail and I’ll use you, and I’ll show you how to do it.
Rich: I cannot wait.
Jeff: So you get it as close as you can and then use a face swap program that swaps your face to that image. And so that’s the best way to do it as of right now.
Rich: Awesome. Jeff always amazing talking to you, and you get my creative juices flowing. Where can people find you online?
Jeff: Best place is at jeffsieh.com. That’s I before E, especially in Sieh. So that’s Jeff S as in Sam I E H. That’s how my mom told me to do it.
If they want, if you want to find out the best tools that I have, jeffsieh.com/toolbox, that has all the AI stuff we talked about today, and I’m constantly trying to update it. And so you guys can go play there. That’s all the tools that I use in my business.
Rich: Fantastic. We’ll have all of those links in the show notes. I’m going to get there before you do. Jeff, thank you so much for being on the show. It’s always great to get in contact with you again.
Jeff: Thanks for having me. This has been awesome.
Show Notes:
Known for his engaging style and expertise in content repurposing, Jeff Sieh helps marketers and business owners maximize their content’s potential. Explore his resources and favorite tools here.
Rich Brooks is the President of flyte new media, a web design & digital marketing agency in Portland, Maine, and founder of the Agents of Change. He’s passionate about helping small businesses grow online and has put his 25+ years of experience into the book, The Lead Machine: The Small Business Guide to Digital Marketing.