Leveraging AI

26 | The Roadmap to Efficient Creativity: How to Generate Mind-Blowing Visual Assets for Your Business with AI in Minutes with Drew Brucker, Midjourney and growth-marketing expert

August 22, 2023 Isar Meitis, Drew Brucker Season 2 Episode 26
Leveraging AI
26 | The Roadmap to Efficient Creativity: How to Generate Mind-Blowing Visual Assets for Your Business with AI in Minutes with Drew Brucker, Midjourney and growth-marketing expert
Show Notes Transcript

Should you use AI to replace your graphic designer? 

Artificial intelligence is transforming how businesses create visual content. In this episode of Leveraging AI, hosts Isar and Drew explore whether AI image generation tools like Midjourney can boost efficiency and lower costs compared to hiring human designers.

Topics we discussed: 

  • Why every business needs custom visual content
  • How AI creation tools work and main benefits
  • Detailed walkthrough of Midjourney and prompt setup 
  • Creative ways brands can use AI visuals in marketing
  • Easy tips to get better image results from the AI
  • Parameters you can tweak to customize the image output
  • Finding the right balance between AI and human creation

Drew Brucker is a visual content expert who helps brands improve efficiency with new technologies like Midjourney. Connect with him on LinkedIn to continue the conversation!

Check out Drew's Midjourney Mastery Guide and get a 20% discount.

What do you think - is AI a creativity machine or creative apocalypse? Let us know in the comments!

About Leveraging AI

If you’ve enjoyed or benefited from some of the insights of this episode, leave us a five-star review on your favorite podcast platform, and let us know what you learned, found helpful, or liked most about this show!

Isar Meitis:

hello and welcome to Leveraging ai. This is Isar Metis, your host, and I've got a different show for you today. In most shows, when we talk about generative ai, we talk about. Text related applications of generative AI like ChatGPT, and Claude and Bard and so on. In today's show, we are going to dive into how to use text to image capabilities in a business context. We're going to dive into how to create amazing visual assets. Any image you need, exactly the image you need that is unique for the scenario that you needed, and also unique for your business, whether it's for a blog or for your website, or for a marketing campaign or for print, whatever the use case may be, what are the exact steps? Everything you need to know in order to create these amazing images in minutes without a camera, without a crew, without editing, and have amazing results that only you and your company have. At the end of the episode, I'm gonna share some really exciting and interesting news that happened this week in the AI world that you should be aware of as a business owner, and now let's dive into this fascinating masterclass on how to create visual assets with Mid journey. Hello and welcome to Leveraging ai, the podcast that shares practical, ethical ways to leverage AI to improve efficiency, grow your business, and advance your career. This is Isar Metis, your host, and we are going to have a blast today. We're going to have a blast because our guest today is good or bad, is very much like me in several different things. we both really like photography. We both are marketers in our core, and we both really like using AI right now to expedite a lot of things that we're doing. And he's an absolute expert when it comes to Midjourney and how to use it in a business context, how to use text to image in a way that can help businesses be more efficient and get better results and faster, which is really winning across all the different aspects of how you run a business. And so I'm really humbled and excited to welcome Drew Brucker to the show. Drew, welcome to Leveraging AI.

Drew Brucker:

Thank you so much for having me. I'm really excited to jump into this topic. Me too.

Isar Meitis:

So really, let's start dive right in, in a business. Why should I even care that there's this capability out there to generate images by writing text? And then later on we're gonna go into more details, including use cases and how to actually do this. But to get started, why should I even care?

Drew Brucker:

Yeah. it's a great first question and a way to kick off this conversation. the default right now, I think for a lot of people, if they've heard of AI or they've heard of ChatGPT, that's the default, right? Just like in marketing, when you think, let's create content, it's a blog. No, it doesn't have to be a blog, right? let's get outta that mindset. But it's that idea where I. ChatGPT made all this noise, and just by that scale and that volume of how much noise it made, that's what most people hang their hat on in terms of, hey, this is what we think should be important. And of course ChatGPT is super important and there are a lot of other tools out there, with AI components that are very important. Sure. I think with Midjourney specifically it come down, it comes down to a few important things, right? Which is that text to image while is not, I think there probably is a little bit alert different of a learning curve than something like a ChatGPT. It does have some similarities and it's got practical use cases in various ways for virtually any kind of company there is. So that's why it's really exciting and why people should be paying attention to it just in general. Yeah.

Isar Meitis:

so let's dive a layer deeper, right? So what can I use this for? in, in a business context, okay, I can generate images, but what kind of images, what do I do with them? Why should a business create images using ai? Or should they consider doing that?

Drew Brucker:

Yeah. first we have to address this elephant in the room, which a lot of people are bringing up right now, which is in relation to copywriting, infringement, these regulations on AI images. Because the long story that is now short is that somebody like Midjourney has scraped, the web for various images, copywritten or otherwise from various artists, various styles. So it could learn, right? And so it could take these things and adapt and have things that relate to each other and create that connective tissue for the software in general. Yep. But, with that said, and we don't really know where those regulations are going to land. There are things that you could do both theoretically, right? Or even right now. And so I think if we're talking about right now, let's address that first. it's creative campaigns. So I could create something on the fly as a marketer. it doesn't matter how practical, real life oriented it is or if it's just colloquial or, funny, so I work for a live event production software company. We provide software to that vertical. And speaking in general, it's very hard to create certain scenarios, both with photography and with graphic design. Staying on the design side, we could say, Hey look, we want to do a campaign that. Talks about our technology all the way back to the Stone Age or some major events that occurred in the past where they could have used our technology. Yeah. But it just wasn't around. So maybe that is Shakespeare doing one of his plays? Maybe it's Abraham Lincoln doing his four score, speech. Or some event like that where we can make a little bit of a comedic play, but also a very visual understanding play of what's happening. And we could stand that up immediately. there's no obstacles that exist in terms of the typical way that you would go about doing that. Which would be, Hey, I gotta go looking for a graphic designer, assuming I don't know anybody that could create exactly what I want. I have to vet them. I have to have conversations. I have to explain what I want to do. I have to create a creative brief, maybe a mood board, assuming, they come back with something. provide iterations, who knows how many iterations that takes, and then I gotta pay'em. Graphic designers aren't cheap. So depending on what the project is or how that's being billed, whether that's by the hour or by project, it's expensive. It's very expensive. And just looking at the environment and the ecosystem that we're in right now with the economy, everybody's being asked to do more with less. Budgets are shrinking. Typically how that plays out in a business context is sales and marketing headcount or budgets are shrinking. Yeah. so this just provides a, an easy way to provide several iterations, up your quality of the things that you're producing in Midjourney which again, is the text to image generation. And like you and I were talking about, create endless cycles to refine exactly what I'm looking for. I can do that in a matter of minutes. Versus days and weeks. And I think ultimately what it comes down to is clearing out the obstacles, clearing out the worry of budget and clearing out the worry of time. so it's, it's that combination of all those things. So creative is where I land right now in terms of an absolute use case you could start to execute on, because you could layer, you could take that into Photoshop, you could make some changes here or there, whatever. As long as it's not commercially sold or monetized, it's good. It's good to go. So I think that's where the line in the sand exists right now in terms of, is this something you're monetizing off of? how do we feel about a particular image coming out of the package and being used just as is, right? Because that's where the murky water exists right now in terms of regulations and understanding like what's fair game and what isn't.

Isar Meitis:

I agree. I think you raised a few very important points. One is that the creative side totally avoids the whole, is this somebody else's image or very close to somebody else's image? And then, am I infringing the rights? Or from an ethical perspective, do I want to use a tool that has infringed somebody else's rights? And both you and I are photographers, so we have even a softer spot in our heart for the people who actually create the images. Yep. That being said, I can tell you just again, a small use case. Where I use this literally daily, is to create images for stuff that I need, whether it's blog posts, whether it's social media posts, whether it's stuff like that. And my images are always so specific that I can guarantee you they're not just out of the pack because they're very specific. Now, does that avoid the ethical aspect of me journey stall other people's images in order to teach their model? Probably not. so I, I still think there's an ethical gap. I assume that's gonna be resolved by them paying a fine that's gonna be distributed somehow. I don't even know. But the capability to have something, and I'm, when I'm talking about my stuff, is not setting up what you do, which is, you're talking about setting up a location in a cameraman and filming at an event and edit. That's a huge work. I'm just talking about going through stock photo websites. yeah. For literally hours to look for something that I feel that kind of describes the thing that I'm trying to describe in my blog post or in my social media post, which takes hours. And you usually end up with something that's a, it's okay, it's not brilliant, and B, it appeared in 25,364 other blog posts because it's on a stock photo

Drew Brucker:

website. That's right.

Isar Meitis:

Versus. In minutes, whether three or sometimes 10 minutes, I end up with exactly what I wanted with the right feel, with the right setup, with the right mood, with the right statement that I want to create that fits the exact format that I wanted. So square landscape, portrait wide, narrow, like whatever I want. And I don't have to crop it to that, it generates it the way I want it. So I feel that from a alignment with the goal of the content, which is what we do as marketers, right? you create, you, you need an asset that drives people to take action, which means you wanna push them in a specific mood. You need to grab their attention, you need to do all these things. The ability to do that today with these tools, text to image, and from my perspective, Midjourney is currently number one, is uncomparable to anything. That existed before, in addition to everything you said before, I don't need 20 cycles with other people on the team. Like I can literally just do it myself as a marketer. Yeah. or as any other person.

Drew Brucker:

Yeah. I'll just add one more point to that.'cause I think that was terrific. I think social media, you think of any sort of content asset in general. if I want to turn a long form piece of content into a P D F or design that again, am I gonna go and rely on stock photography that people have seen over and over again if I'm making a, if I'm redoing a website, which is, this is a real use case right now. we've, we have, AI generated images on our website right now. Me too. And I'll tell you what's great about that is the points that you just made is you can create very specific images. They're unique and they have a tailored or, specific feel, to, to them, right? Perfect example. our brand colors are black and yellow, our primary colors. So everything that I was able to create has this undertone of this dark black contrast in this yellow light. And so now everything feels very cohesive as a brand. Which, you just told me that you would spend hours going through a stock photography website. I am. I've been there and done that too. And it sucks. Yeah. like you're trying to manipulate the filters, the system, the keywords, just to get like something that feels close and you're probably making compromises on ultimately what you land on because that perfect item doesn't exist. And to your point, to your last point about the dimension of it, you only got square. Landscape and portrait, and sometimes you'd see a good portrait photo that you wish you could use for landscape or a good landscape photo. You wish. Yeah. so that, that was a pretty real and sucky scenario too. That, was definitely something we've experienced as marketers.

Isar Meitis:

quick summary to what we touched on so far, and then we're gonna go to the how. So the why is very clear. Regardless of the needs, every business has needs to create visual content. This could be for product PDFs, it could be for marketing, it could be for sales, it could be for any kind of visual content. That visual content historically took a very long time to produce was really expensive. And in many cases you would make compromises because time, money, or just availability of the resources you needed. And now you can, without a large team, with one person who knows what they're doing, create the perfect. Assets that speak in the language of your company to the feel you want to create in that content, to get your customers to take action. And it's never been accessible in history before. So if that's attractive to anybody who's listening, let's not move to the how. So how Drew do you, what are the different components that goes into a prompt that needs to create the outcome we just talked about?

Drew Brucker:

Yeah. And we got into a pre-chat about this that I think is really interesting too. I told you, and I believe this, that there is, this is different for everybody, this learning curve and what that looks like. And here's what I mean by that. I think for people that have been creative directors or on this graphical side, visionary side, really high velocity and seeing and feeling and touching these things and creating campaigns, they have an advantage. and probably so too do photographers, in fact, I, they do, right? Like people like you and I do have an advantage there because what comes into play with these images is goes far beyond just what you're seeing. it's the composition. and it's really the complexity. Of the end result and the output quality that come into play. So with a photographer in mind, I'll just give you some very specific examples, things that you and I are thinking about. if we're out in the field and we're shooting photography, we're thinking about, first of all, light. Light is the most important factor with photography, right? And, what time should I take a photo can play a large part in the quality of the photo, right? if you and I are traveling, which you were just telling me you were on a trip, sometimes you sacrifice getting up early or your sunset to get the right light naturally, especially if you're not using any other equipment besides that. So that's one example. different angles, d so different times a day. Light, shutter speed. One other component of photography is how fast the shutter is opening and closing, right? So a long exposure, me taking a photo of something in downtown the city, I could go for a long exposure and with a long exposure that's gonna capture all those light trails from cars. Or if I'm looking at the sky and it's a, nice overcast day and the clouds are moving and it's gonna, create this nice soft feel of the clouds that are passing by. And so that's just another manipulation of a camera that you can use that I don't think the average person is really familiar with, right? And so in large part, it goes back to the depth of knowledge and experience that you can bring into this system. So while you and I are thinking about camera lenses, angles, shutter speed, focal length, which is how much depth of field I want to capture. So a landscape, and maybe I want everything in focus. I'm going for a higher f stop. If I've taken a photo of you up close, I want. Some BCA in the background or just a nice drop off, right? I'm shooting an F two or something like that. And so all of those components can affect the quality. So let's just start there. I think for the average person that's interested in Midjourney marketers, graphic des like maybe people that have these like parallels or they've got interest in aesthetics, or just the average person in general. There are, you don't know what you don't know. And so what could be a prompt for them is they're thinking about, okay, I wanna create this scene. I want to create a dog sitting on a couch, with a cat and they're maybe eating popcorn, right? And so while there are some very short prompts that could work, it's less reliable and less consistent in terms of the quality of the output that you're gonna get. Versus if someone like you or I were doing that same thing, right? We would be thinking about all these other factors that could come into play. What color is the dog? what kind of light am I looking for? am I looking for a composition that's going to, include the whole scene? Am I looking for a wide angle or am I looking for something up close? Or where's the shot coming from? Is it a, from above? Is it from the floor? Because all those things affect the mood and, how you interpret the image. And so if you're just starting out, it can be difficult. But even if, it's somebody like you and I, a learning curve exists and you have to be willing to spend the time experimenting, playing around just like you do with ChatGPT, right? There are gonna be some people that know how to Google better than you, but the reason they know how to Google better than you is just because they've done it more. They know, they've been through so many scenarios and practice learning and iterations of what that looks like, that's how they've gotten better, right? And yes, some people have handicaps, but that gap can close.

Isar Meitis:

Okay. I love what you said. I think let's break it down. Let's give people a list, a checklist of things they need to consider. Yeah. Some people understand what these things are better, like you're saying, okay, what's an f-stop? an F-stop is how much the actual aperture is open within the camera. Okay. What does it do? It defines how much is actually gonna be in focus within the picture. Does that sound like Chinese to you? maybe then maybe research that later. But there's 15 other things you can pay attention to. And you mentioned a few of them, but let's just break it down as a list to people so they can really go step by step and understand what are the Legos, right? So what building blocks can I use to build my next image on mid, oh, by the way, any other text to speech, tool out there that people need to be aware of, and then they can start playing, as you said, and learning as they go.

Drew Brucker:

Yeah. I it's a long list, but you don't have to use all of these things, right? Yeah. I think the idea is that, You ultimately want to get specific, okay. if there's an underpinned component to this, it's the specificity of what you're trying to prompt. So what I mean by that is, look, we mentioned camera angles and this and that, but maybe that's not super important to the image that you're trying to get, right? so maybe you don't have, you don't have to include that. There's no, there's nothing that says you have to include that. But what I think will help, is thinking about things like, just like a template of what you're going to use. okay. I think you could think of like things like adjectives, right? So things like if it's food, like delicious, if it's, a portrait, maybe something like breathtaking. if you're looking for a dirty scene, maybe it's filthy or dirty, maybe, so describing that, second Midjourney is not just for realistic images, right? You can create. The entire spectrum of what art imagery looks like. are you looking for something that's surreal? Are you looking for something that's realistic, cartoonish, isometric, right? what are you trying to get? So that's like another option. There would be like an artistic style. one of my favorites is inspiration, is there a particular photographer, artist, a filmmaker that I enjoy that kind of look, one that I've seen quite a bit on Midjourney and it's a fun one, is Wess Anderson. he's a filmmaker. He is a very distinct look to his movies. It's very whimsical, very uniquely colored. The composition is a certain way, it's very symmetrical. And if you were to prompt with that inspiration of saying something in the style of, then that creates a very specific image too, right? Because it's gonna pull, what it knows about that particular artist or that inspiration. I wanna pause you.

Isar Meitis:

I wanna pause you just for one second, for the thing you said before as far as the artistic style. Yeah. in some of the courses that I teach on purpose, every single slide has a completely different style. And the reason I do this is after about 40 slides in, which is like day three of the course, I'm like, okay, now we get to the point of talking about text to image. And I'm like, okay, I want to take you back. Here are all the images you've seen so far. Some are this, some are that, and I do this on purpose to show people that you can generate a cartoon and it could be a specific cartoon, right? I want, is this a Simpsons scene? Or, I want this to be a Pixar scene. And they're both cartoonish, but they're very different kind of cartoons.

Drew Brucker:

Very

Isar Meitis:

different. Yeah. And it will create it exactly that way for you. I want it to be a highly realistic photo. I want an imaginary world that has this, in that thing, in it. Whatever you want as far as the style can be generated. And again, it has to be there to achieve the goal that you're trying to be. and I want to go back one thing and I, and then I'll let you continue with the list. yeah. None of this takes anything away from good marketing, right? The fact that you can now generate 10,000 photos in a day exactly the way you want them, doesn't mean you should. There still has to be a strategic design process of what you're trying to achieve, what's the atmosphere on what level you're trying to connect, who you're trying to attract, what, like, all these things are still in place. It's just the production became extremely simple. So now I'll let you continue. You said great point adjectives, you said artistic style, you said inspiration. What other components do you have in your

Drew Brucker:

list? Yeah, I can, I hope if you're listening to this, I'm gonna just give you a chance. You got a note. note app or on your phone or you got a pen handy, just like this would be a great place to pause and just grab that. Yeah. Yeah. But because the options are really unlimited, but like to give you, more of that entire sense. So we talked about looks, but you could also go for like sensory details, right? edges like materials, like things like that are super interesting too, like smells, sounds, so like that becomes super interesting and materials. So are the reflections, are there optics are there, are you looking for a soft material versus a hard material? physical properties, right? So like we mentioned, dirty already, but Something that's charred or something that's perforated or something that's mesh or like anything like that. And so you've got that materials, is it stone mineral wood paper, like materials is another great one. I think one that gets used a lot in almost all my prompts is color. I color matters a lot to me, and I think that's probably maybe like the photographer side, but I'm constantly thinking about that color grading, if you will, of what I'm trying to achieve. that could be just naming a color or it could be like, Hey, I'm looking for this kind of hue or tone or a gradient or something that's vibrant, something that's bright, something that's dark. how many colors do I want? What are the colors? so like that factors in, And then we mentioned kind of the inspiration side, but that's so deep because you can get into, let's just use photography as an example. Are we talking about. If you're trying to create a portrait, maybe it's a good opportunity to use a portrait photographer style that you really like in the style of, I dunno, Peter Hurley comes to mind as a portrait photographer, like in the style of Peter Hurley. And, that way I know what to expect and it's something to play with. But that could be very different from street photography and an inspiration I wanna do there. So if I'm matching up that kind of image with somebody that is an inspiration in that space, that's a nice little shortcut to get very specific. we've talked a lot about lighting already, but, natural lighting, ambient lighting, directional lighting, time of day, we've talked about angles and shots. So I'm really thinking about it from a more, more or less a photographic perspective because I do think that knowing those things will get you some extremely better outputs than. One or two of these things, or maybe even three or four that don't include the lens, don't include the light, don't include the depth. I think those things really matter.

Isar Meitis:

I agree. I, so I want to talk, I wanna mention two things. One you said in the beginning, and I think that's the summary of all of this is be as strict descriptive as possible. The more details you are gonna give to the prompt, the closer the outcome is going to be to what you imagine. And every now and then, it'll be better than what you imagine. It happens to me multiple times where it comes up. Yeah.'cause and you're like, holy crap, this is good. yeah. And so it happens, but it happens more often, like Drew said, when you give it more details. So the more details you're gonna give, the better. The second thing, which is the way I always think about it, and I must admit I'm probably more of a marketer than a photographer. I always start with really what am I trying to convey? Which means what I really start with, what I invest the most amount of time in is the composition, what is actually in the image. And it's not just, oh, it's two people shaking hands. I'm like, okay, where are they standing? Is it an, are they in an office? Are they outside? Are they in the park? Yeah. what kind of people are there business people? Is it a female and a male? Are there two males? What are they wearing? So I'm trying to understand, okay, who am I trying to connect with? These are that kind of people. What will attract them to the image? So somebody like them, somebody exactly the opposite. what is it that I'm trying to, so I spend a lot of time in describing the composition in a lot of detail, including here's what's in the background, in the middle of the image, I want this and that on the sides. This is on the right side of the image. There is this. What you get is what you like dreaming in your head, described in a bunch of sentences that then tells it how to draw that image. And only then I go to all the additional layers and sometimes I don't. Sometimes I'm lazy and I just start with that and see what's happening. And sometimes it's good enough and then I don't have to invest all the rest. But then I go into all the other layers of what kind of lighting do I want, what kind of materials do I want? And sometimes I don't really care. But sometimes you see the image like, oh, I wish it was this and that. This is a one piece. I wish there was a reflection on the lake. I wish. His suit was actually blue instead of black, because it would've looked better contrast to what is created. So then I go and manipulate the prompt and try again. but you can start with less details and just add as you go on a, as a learning process. And b, you can save time. Maybe just describing a clear composition that you want will get

Drew Brucker:

you to where you're trying to go. Totally with that, and I was just gonna say there are a few, I think, getting back to the goal of this too, and trying to be as practical and, hands on and actionable as possible here, a few things, maybe the beauty right now, the way that it's set up in Discord, Midjourney is, especially if you're new, you can see other prompts coming into the, to the channel, so you're not in this space by yourself, which I think can be overwhelming, And hey, look like I'm in the same channel. And you just created an awesome prompt, right? So I could say, Hey, you know what, I like that style, but I don't like, I don't care about this person, but I would rather do this as like a young child, or I wanna do it in this scene. Or, and so that can help your learning curve quite a bit, right? Like maybe take note of that and run that prompt for you, change a few elements of it, run it, see what it, see what you get back, right? So a lot of it is trial and error. And so there's that. I think that's a great sort of, I don't wanna call it a hack, it's obvious, but take advantage of that, and same with Midjourney is going to store all of the images or the, like these prompts that you've run, the outputs that you get, and you get those in sort of your own space or hub if you will. I don't know the proper name for it, even with that, you can click in an image that you've created. And it'll bring up other images that it associates in that same category or synonymous. And again, like you can see other people's prompts. So it's hey, what did I miss by this? Re this one looks better than mine. Like, why is that? did I miss something? And so a lot of it's just like having this curious mindset and of trial and error and to your point, right? Like just iterating, because why not? this is the fastest way possible to do that. So like those are two really good things that I try to point out because you're not in this alone. it can feel overwhelming. Sure. And especially if you're not in used to discord, which I mean, I've been in there a few times, but I don't love it. And it's a little confusing to me. Like I, I know a lot of people struggle with that. And so those two things are super important. I also just wanted to say really quickly that. Again with the practical use cases for these, I think there are different like strategies you can have, but if I really one of the results that I get, right? You get this box of four, one, 1, 2, 3, 4, and you'll start to understand that as you're in there. But like you can say, okay, I really like this image, I wanna upscale it. So yeah, you one, you two, U three, U four, basically trying to identify which one I want to upscale. Okay. I wanna upscale number one. And the reason I'd wanna do that is'cause I really like that image. I don't really care about the other three, like this one. I could then take that image, right?'cause basically what's gonna happen is it's gonna generate that, it's gonna enlarge it. I could say, okay, like I like this and now let me like vary this slightly and let's see if it generates something I like even better than this, because maybe I like her looking into the camera. Or maybe I like, this face that's in there if it's a portrait, but what they'll do is they'll slightly alter the face, right? So some people, just some people are more photogenic than others, right? some people just work better or match that vibe that you're going for. so that's a use case. Or another, another point is I like parts of this and I wanna vary it, which is the other option you have, right? Which is v and I like parts of this, but I, it could be better, right? Let me just see if I hit vary, what it'll bring me back, right? So it's gonna again, generate a new, result for you and you just you can really, that's a strategy within itself because depending on what you like to do, you can go in a few different ways with it, and Midjourney's rolled out a few really cool features, especially lately with this whole zoom out thing, right? Which is the out painting, which is, if you're not familiar with Midjourney which is may what you've seen with Photoshop, The beta access where you can almost say, Hey look, I wanna stretch this image out and it's gonna basically, create something that it feels like should fit this scene. it's essentially that. Let's just leave it there. But, again, that's another strategy that you could take or right. Like you can always re-roll it, which is basically just refreshing it and saying, Hey look, I don't really like any of these four images, or, I like this image, but I want more. I want to, I wanna have a pool of them to choose from. And I think that's like a photographer thing too. Like we, we wanna have a bunch of good photos to choose from, to get the, yeah. To get the best of the best.'cause we're vain like that. But that's also a nice little use case. And so these are all super tactical and should help you feel a little bit more comfortable, like just poking around in there.

Isar Meitis:

awesome points. I'll summarize them quickly and then I have a follow-up question. Yeah. Yeah. That is also very tactical, One is look at what other people are doing. And it's a huge benefit. It's incredible. Like you can literally see what everybody else is doing. I don't know of any other, literally any other artistic environment where you can literally see exactly what your competition is doing. Horrific, horrific point doesn't exist. nobody shows you what they're doing on Photoshop. Nobody's showing you how they're shooting the pictures. Nobody's showing. It's a community. Yeah, it's a community aspect of it. You can see exactly what everybody is doing. There is no veil. Like you can actually see how things are done. So when you see things you like, you can see exactly how they're done and you can copy them. So that's an exceptional point. the other thing is we, like you said, there's a lot of trial and error built into the platform itself that is allowing you to manipulate and change and vary different things. I wanna deep dive into one thing that I think very few people know. Even the people who actually use me journey regularly. There's all these parameters you can put at the end of the prompt. You can put a dash and you can put something. Yeah. what are some of those and how they're actually affecting the picture that you're

Drew Brucker:

creating? Yeah. Great. Great question. Sometimes I use them, sometimes I don't. this is, so parameters, let's just break down what those are. it's a terrific point. So these go at the end of your prompt, and what's gonna happen is this is basically, like this is affecting things like style. and what I mean by style is like how much detail is within the style,

Isar Meitis:

aspect ratio

Drew Brucker:

as Yeah, aspect ratio. If I want. Like a super diverse set of images that feeds back. I may, attach something like chaos to this, which is just like a, so all these parameters are used, dash, and they have a shortcut. And then they all have a spectrum, a number spectrum, with the, I guess exclusion of dimensions. But they all have a number spectrum of which basically you're saying, okay, like I either want to do it a little bit or I wanna do it a lot. And then, so all of these things can drastically impact your image as well. So one thing I like to use, because I do a lot of photorealistic stuff, is throw in like a style raw on there and it doesn't work all the time, because of some really complex things in the background. But I would say if you're going for a more realistic look in general, eight or nine times outta 10, it's gonna give you a, just a slightly more. Photorealistic take to it. you and I

Isar Meitis:

are probably, so yeah. What's the actual thing you put in the problem so people understand how it's actually built? Yes. So if you pick a raw thing, you put dash.

Drew Brucker:

So if you're going for this specifically, you're gonna push dash style. And then just like all these parameters, you're gonna connect the shortcut to the dash. Dash, yeah. So dash, dash style space raw. if I'm doing chaos, it's dash c. I think you can also do chaos. And then it's a spectrum of, I think it's still one to a hundred. Yeah. And all these also have default levels attached to'em. And so it's probably worth digging into those to see. Okay. if I don't enter this, How will this affect anything? Because that's why I use, I said I use it sometimes and sometimes I don't. In your settings, in Discord, like in this server, you can set your defaults and based on what those things are, you may not have to do these things. Correct. Is my point. all

Isar Meitis:

the reason I'm asking this is the first time I saw them is exactly what you said. I was looking at other people and I'm seeing these dashes like, what is that? yeah. so if you're curious about it, that's what they do. They allow you to, if you want, manipulate the engine behind the scenes by literally tweaking the parameters on which there's default parameters. It's the same defaults for everyone. If you want to move away from those defaults, those dashes in the end allow you to do that. The other thing that is very common is just the aspect ratio, which is dash, dash ar, and then whatever aspect ratio you want. So one to 1 69, 3 2, 9, 16 if you want, portrait versus lance, so you can define what. Aspect ratio, you want the outcome to be, so you don't have to take the default image and then crop it to whatever you need, which I've seen people do. That's right. So you can actually tell it what the aspect ratio, needs to do. anything else you wanna add? I think this was, you know what, I'll add one thing and then I'll let you summarize, yeah. Or add anything you wanna add. But I wanna go back to how we started this. There are huge business benefits to this, and they go from, you get a better outcome, you get it faster, and you get it cheaper than any other way. You can create visual assets. And we talked in this particular episode a lot about stuff that is photorealistic, but it's not. You can create logos, you can create graphics, you can create backgrounds for slides. You can create literally any visual assets you can think of. You can create with this tool. And it's limited only by your understanding of how it works and your imagination. That's it. And so the ability to generate high quality assets very fast with practically no money because the access costs, I don't know, 10 bucks a month or something ridiculous like that, is mind blowing and is life changing if that's what you do in your role or it's life changing if you're dependent on the team to provide you these assets and how you don't have to depend on them anymore. So that's the premise. Understanding how it works. Drew gave. Amazing examples and ideas and concepts of how you can use it and the different details that you can add in order to get better and better results. And he also said, at the end of the day, he's looking at other people experimenting and learning how it's being done to get better and better business results faster and cheaper. Anything else you wanna add? So hard

Drew Brucker:

because this is so complex, right? there, there are just so many ways we could take the conversation, but I think we did a really good job and I think you did a really good job of recapping these certain segments that we were talking in. to reiterate the point that you made, the real restraint, the, the real restraints that we have right now, time, resources, expenses, and the compromises that we make, which I think is just a great takeaway. And then you also brought up a very terrific point, which is, Just, in other tools, you don't have this in front of your face. Crowdsource, community, educational learning alongside component, right? if I wanted to learn Photoshop, I would have to go into YouTube or find a course directly and really like focus on understanding what it is I'm looking for and do all this and that, right? Where this is much more of a tactical or a tactile feel of trial and error. What am I seeing? like you said, I thought, which it brought me back to, like, when I first started, you see all these prompts and all these parameters or things that you have no idea what they mean, like dashes and colons, and you're just like, what the hell is going on here? Like, all I know is that this image looks good. let me play with this, right? don't be afraid we all start in the same place. in some sense with this, there's gonna be a learning curve no matter who you are. But, the use cases are infinite too, right? you just think about even stuff like interior designers, right? you may not be necessarily selling or monetizing that, but you could use that right now to drop a concept for a client or real estate for somebody building a home or, like things like that exist, right? And so you just need to think about it in terms of what you're trying to get out of personally. And if you are bringing this to the business side, how does this apply to your business right now? And of course we talked about more of the use cases you can immediately implement versus theoretical, because we, the truth is we don't want to necessarily lead anybody in the wrong path either because we don't know. We like, we don't know what some of these things are gonna result in. But the thing that you can unquestionably say is there is a need that you have right now that could be solved by this, and it's being slept on so far. The reason I like Midjourney specifically is just like with ChatGPT, you're putting your money or your bet on a certain horse in the race, and I like what Midjourney's doing, right? There are other players and other tools outside of Midjourney that you could use for sure. But what if Midjourney disappears, right? Then I, and then I gotta go learn a different program and upskill and take the time to do that, right? So yeah, there are probably certain things that are very similar, but I gotta start over. So pick your tools wisely too, right? And just assess these things at scale and really think creatively about how they can become practically use cases for you or for your business.

Isar Meitis:

Phenomenal summary. Great advice. Drew. Thank you so much. This was absolutely awesome.

Drew Brucker:

Enjoyed it man. I, I felt like we could've talked for hours, so this was fantastic, man. Thank you for, really I think putting the guardrails around the conversation, because like I said, I think this. This is such a big topic, but if you are not familiar with this topic, this is your sign to really dig in and just try to, try to keep up and learn what's going on here, because it's gonna be extremely impactful regardless of industry. It's gonna be industry agnostic.

Isar Meitis:

Agreed. Thank you again. Thank you. Amazing. Right? Absolutely amazing. The outcome that you can achieve today, following the guidance that Drew just shared with us are unparalleled and never existed in history and can save hours and thousands of dollars every week probably, and definitely every month for every business out there, for any marketing need. And now let's dive into some really important news that are gonna affect your business, that come from the AI world in this past week. The two most important aspects of news this week that have a significant impact on AI and as it relates to businesses. Actually come from the legal world. The first one is a judge called Burial. H Howell have denied copyright registration to an AI created artwork. Why is that important? It's important because it aligns with the US copyright office decisions from earlier this year that basically said that AI generated content has no copyright protection. What does that mean? It means that anything you generate with AI that you put on your website, that you want to use in a book that you create for courses that you generate for your clients. I. Is not copyrighted, meaning anybody including your competition, can literally copy and paste it and use it to compete against you, and there is nothing you can do about it. Now, whether you or me agree with that or not, and I do not agree with that because I think what needs to be copyrighted is your ability to connect the dots together and to write the right prompt that generated the outcome that you were seeking that potentially nobody else could create. So I think that should be able to be copyrighted and hence the outcome needs to be able to copyrighted. But again, the fact that I don't agree with that and the fact that I think that in the future this will have to change. Right now, this is the case and this is. Extremely important to anybody in business, and even more important to anybody who has like a marketing agency who provides these kind of services at scale to clients the clients of the marketing agency assume that whatever the marketing agency created for them is copyrighted, and that is not the case. And hence, everybody needs to aware that that's the situation. And while before this was a copyright's office decision, now there's a federal court that has put its stamp on the same decision. Next piece of really important and really critical legal aspect comes from the New York Times and OpenAI. So for a while now, OpenAI has been in negotiations with New York Times to put in place some kind of a licensing deal that will allow OpenAI to pay the New York Times for using its stories to train its AI models. apparently these negotiations has failed and now the New York Times are planning or maybe already suing OpenAI for using New York Times copyrighted data to train its previous models like GPT 3.5 and GPT-4. Now, if they win One of the potential outcomes of this is that. OpenAI will need to take out this information, this copyrighted information out of its model, which as far as I understand, the way it works is impossible to do. Meaning they will need to retrain their GPT model basically from scratch without having access to that data and some other piece of data from other people who are suing them for the same thing. Now, if this actually happens, this. May put open AI's livelihood at risk because it requires a huge amount of money to train such a model from scratch. Not to mention the fact that during that time they may not be able to use the existing models, meaning they will have no revenue. Now, do I think that's where it's gonna go? I seriously doubt it. I assume there's gonna be some check written to compensate the New York time for that, and potentially other people are gonna sue them and these are gonna pile up as soon as the New York Times wins this case. But I don't think it will actually make open ai, take any technical actions other than that, but again, this may evolve in multiple different ways, and this will impact A, the future of how these models will be trained, meaning most likely under licensing mechanisms and agreements, and B, what may happen to ChatGPT and OpenAI as a whole. Another big piece of news. Comes from Google. Google tested a very interesting feature this past week on Chrome. And that feature summarizes webpages in Chrome using generative ai. What does that mean? That means that in addition to what they've been testing so far, which is providing an AI driven answer for the search query that you write, even for specific websites, it will tell the users the general knowledge that is in that website without the need to go to the site. It means that non-branded SEO as we know it, meaning website traffic that generated from general queries that does not include your company naming them is gonna be dramatically hindered. Which puts the whole concept of SEO as we know today at risk. Whether or not that test is gonna continue and it's gonna spread out to all users. I don't know. we talked about this before. Google are playing a very delicate game here between its current revenue to its future abilities to win the next variation of what AI search looks like. And so they will extend and try to milk the existing revenue they're generating from search, but in parallel, they will test all these things moving forward so they can win the next variation of search. How will this evolve? I'm not exactly sure, but definitely worth paying attention. The last piece of news I wanna share with you relates to how generative AI will potentially impact the global workforce. IBM one of the largest companies in the world, and definitely one of the biggest service providers to multiple levels of companies in the world, has done a survey across 3000 C levels, executives across 28 countries, and another 21,000 workers in 22 countries. The executives surveyed said that they believe that 40% of their workforce will have to be re-skilled in the next few years due to the impact of AI and its implication on the way they run their business. That means 1.4 billion people out of 3.4 billion people in the global workforce. So that's a lot of people that need to be re-skilled. Diving a layer deeper into the research. They've asked these people, what do they believe are the most important skills that they will hire for in 2023? And if you look back, On what were these skills in 2016 and compare them to 2023. The interesting thing is, in 2016, the most important skill was proficiency in STEM, meaning science capabilities, math, and so on. And that has dropped from number one on the list to number 12. And the flip side is time management skills and the ability to prioritize went up from number seven to number one. Meaning the ability to understand science and use advanced tools became significantly less important because of the ability to use AI to perform a lot of these tasks that previously required a much deeper knowledge in those specific fields. The flip side is that if you know how to time manage, prioritize, and then use AI tools to generate what you need to generate, that's a very needed scale right now, and hence, it jumped to the top of the list. The last interesting piece of information from that survey and now I'm quoting technology driven job changes, report a revenue growth rate premium of 15% on average, meaning companies who will adapt to technology in general see a 15% premium, But for those who focus on ai, see a 36% higher revenue growth rate than their peers. Meaning investing in AI skilling training and technology gives companies a 36% higher revenue compared to the competition, which is incredible hence the really strong need and urgency of companies and leaders in businesses to understand the gaps they have right now. Develop a strategy and a plan on how to do that in the most efficient way without losing people and without hindering the business in the process. This is one of the things that we do at multiplAI.ai. If you visit our website, multiplAI.ai multipai, spelled M U L T I P L A I ai. You will find multiple ways of how we can help your leadership team and your employees understand the gaps that you have right now and adapt to AI in the most efficient way. We do it across multiple levels of companies and different levels of involvement from us, and we can definitely help you with that. So if it's something you're looking for, either for yourself or for your business, go and check it out. And until next time, have an amazing week.