PROMPT ENGINEERING & GENERATIVE AI: Master Video, Images & Text + ChatGPT

1. Introduction: Hey, guys. Welcome. In this course, we're diving straight into how generative AI is transforming the way visual and video content gets made. If you've ever spent hours editing footage, hunting for the perfect image, or managing endless campaign version, this is your game changer, because now genitive AI can help you to do all of that way faster, way smarter. At a scale that was unthinkable just a few years ago, from generating staning visuals from a single prompt to turning scripts into dynamic animations to localizing campaigns for audiences around the world. Generative AI is reshaping the creative process from concept to the final cut. And in this course, I'll show you exactly how to use it. Why this course matters? The creative world is evolving fast, faster than ever. Brams want more content in more formats for more platforms, and that's where generative AI comes in. And that's not to replace creativity but to amplify it. You see, in this training, you'll learn what generative AI actually is, how it works, and how it's already changing industries around the globe, from advertising to entertainment to education, we'll break down how to work with generative AI to speed up production, enhance storytelling, and free up your time to focus on what really matters to creative ideas that you and only you can bring, what you'll learn. Here's what's ahead. I'll start with the fundamentals, what generative AI is, how it fits into modern content workflows, and how to write powerful prompts that deliver real results fast. Then we'll move into hands on creation, generating visuals, converting scripts into videos and animations, and mastering generative AI assisted editing. So you'll be able to move way, way faster. Also learn how to scale your content globally with smart localization tools and best practices for maintaining brand consistency across every market. By the end, you won't just understand generative AI. You'll know how to harness it confidently to supercharge your creative output and make your production process unstoppable. So buckle up, this is where creativity meets cutting edge technology with generative AI. So let's start building the future of content together, and let's get started. 2. The Birth of Generative AI: So to understand why we even need prompt engineering, we need to understand where it came from and how it relates to AI, the birth of chat, G, PT, and LLMs. From the simple rule based programs of the 1980s to today's smart creative chatbots, one thing has stayed the same. Our goal to make computers understand and respond similar to humans. The story of chat, GPT, and large language models, LLMs is really the story of how that dream became real. How we went from basic machines to powerful tools that can think and write using everyday language. Although it feels new, artificial intelligence AI has actually been around for decades. To understand where Chat GPT came from, let's take a quick look at how AI evolved over time. It all began in the 1950s when computer scientist Alan Turing asked a famous question. Can machines think? That one question started the entire field of AI. In the 1960s, a simple program called Eliza was created at MIT. Would hold short conversations by matching patterns in text. It wasn't truly intelligent, but it was the first step toward computers that could use language to communicate. In the 1980s, AI was used mostly for what were called expert systems, programs that followed rules written by humans. These systems could give medical advice, approved loans, or help design products. They were useful, but they had one big problem. Couldn't learn or adapt. If something changed, you had to rewrite the rules yourself. The 1990s and 2000 brought a big shift, the rise of machine learning. Instead of being told exactly what to do, computers started to learn from examples and data. This approach quietly powered many products we all use today. Spam filters learned which emails were junk. Google search got smarter at finding the right pages. Amazon began recommending products you might like while Google Maps learned to predict the fastest routes based on live traffic. Came alexa and other voice assistants, which could recognize speech and answer questions out loud, something that felt almost magical at the time. Computers were now learning from experience, not just following rules, but even then AI couldn't really create things. It could predict and categorize, but not write, imagine, or explain. That changed in 2017 when researchers at Google developed a new system called the transformer. It helped computers understand how words relate to each other in a sentence, not just one word at a time, but in full context. This was a huge breakthrough and laid the foundation for the next big step in AI, large language models. A large language model, LLM is an AI trained to understand and generate human language. It learns from massive amounts of text, books, articles, and online content by spotting patterns in how words and ideas connect. After the transformer architecture was introduced in 2017, models could finally understand context and meaning. This led to powerful systems like GPT, capable of writing, summarizing, reasoning, and chatting in natural human like ways. That next step came from Open AI, the company behind Chat GPT. They built on Google's work and created something called the GPT series, generative pre trained transformers. The first version GPT one showed that a computer could learn to write readable text by studying huge amounts of online data. Then came GPT two in 2019, which could write essays, stories, and even news articles that sounded human. A few years later, GPT three made an even bigger leap. With 175 billion parameters or neurons, it could write, translate, answer questions, and even code. There was just one thing missing, natural conversation. GPT three could give answers, but it couldn't chat as fluently. So Open AI improved it using feedback from real people, teaching it how to respond more naturally and politely. The result was chat GPT, a version that could hold a conversation. Remember what you said and reply in a way that felt personal. That was when AI truly became something everyone could use, and this is where prompting came in. The word prompt originally came from early computers. It was the line on the screen where you typed a command. The computer was waiting for your input. Over time, the meaning changed. Now, a prompt means the message or question. You give an AI a system like chat GPT to tell it what you want. At first, prompts were very simple. Things like write an email, summarize this text or explain this to me like I'm five, but people soon notice something interesting. The way you wrote your prompt completely changed the answer. A detailed prompt gave a detailed result, a clear question, got a clearer answer. The better your prompt, the better the AI's performance. This turned the act of prompting into both an art and a science. Today we call this art and science prompt engineer. 3. What is Prompt Engineering (and why do we need it?): Okay, guys, after learning how large language ubbles came to life, the next step is understanding how to tarp to them. And that's where prompt engineering comes in. If LLMs are the engines of AI, prompts are the steering wheel. They allow you to control the direction, quality, and creativity of what AI produces. At its core, prompt engineering means writing clear structured instructions but guide an AI model, chat GPT, Claude Gemini, or others. To produce useful, high quality results. A prompt could be as simple as write an email or as detailed as act as a marketing expert and design a campaign for my new product. The better you describe what you want, the better the AI's output will be. But how did this idea begin? Prompt engineering emerged as people started using AI tools like Chachi PT, Mid Journey, and Bendali more creatively. Users quickly discovered that two people could ask the same question and it completely different answers. And the difference came down to how they phrased the prompt. This led researchers, creators, and educators to study the patterns behind effective prompting. Early thought these was like Ethan Malik, Andre Carpathi and Seth Dobrin along with OpenAI's research teams began to share techniques that worked consistently. These evolved into the core prompt patterns used today, frameworks that help users think and write more strategically. Among them are the instruction pattern, giving clear direct demands, chain of thought pattern, guiding the model to reason step by step, persona pattern, assigning the AIA specific role or perspective, template pattern, creating reusable prompt structures, and the iterative refinement pattern, collaboratively improving the output. Patterns became the backbone of prompt engineering, a way to get predictable. Powerful results from any AI system. Prompt engineering has since become the bridge between human creativity and machine intelligence. It's powerful because it gives anyone, not just programmers, the ability to direct AI systems to perform complex or creative tasks. In short, it turns you into a kind of AI conductor, guiding the output you want. And it's not just for chat GP two, same skill applies to a wide range of AI tools used across different industries, Canva, create marketing designs and social posts from simple text prompt, Photoshop firefly, generate or edit high quality images using natural language, motion AI, draft reports, sunrise notes, and automate documentation. Runway ML, turn written scene descriptions into professional videos. Microsoft 365 copilot, write emails, analyze spreadsheets, and prepare presentations from prompt. Prompting is now a universal AI skill. Whether you're a marketer, teacher, designer, developer, entrepreneur or project manager. Learning to prompt effectively will help you work faster, automate tasks, and unlock new opportunities is also a practical way to earn more. At work, prompt engineering can save hours, automating emails, reports, and customer interactions. Entrepreneurs use it to scale content, analyze data, and generate business ideas without hiring large teens. Freelancers now sell AI based services from writing and design to strategy and automation. People who know how to communicate with AI are already leading the way in productivity and creativity. Here's what I want you to do first. Imagine you've just been asked to use AI to make your current role faster, smarter or more effective, or maybe even to help you land a new role entirely. Think about your day to day work. Are you managing projects, designing presentations, writing reports, planning lessons, selling products, or building marketing campaigns? We'll be using chat EPT throughout this training to show you exactly how it's done. But remember, what you'll learn applies to all large language models and many of the other AI tools we mentioned earlier. Like Canva, Notion and Microsoft copilot. The goal is for you to think in terms of how AI it can assist you no matter your role or industry. If you're a project manager, imagine AI helping you summarize project updates, create meeting agendas, and identify risks in seconds. If you're a marketer, think about how it could draft, add, copy, analyze competitors, and plime social media posts. A teacher could use it to generate quizzes and lesson plans. A business analyst could interpret data and find patterns. A freelancer could create proposals, automate admin work, and deliver projects faster. Make it practical, we'll use an example of someone running the day to day for an ecommerce business because it applies to so many roles. So take a moment to think about how an AI, I can help you in your role because by the end of this course, you'll know exactly how to make that happen. So let's get started. 4. How to Generate Images from Prompts with Adobe Firefly: So let's get into a demo. Here we're going to start creating visuals for our passion sports tracksuit ad campaign. And the goal is to create a set of branded visuals using Adobe Firefly. So the first thing we're gonna do is we're going to open up firefly, and then we're going to get to work. So the first thing to notice is this is the area that we're focused on. We want to make sure that this drop down is on image because you can also generate videos, but we haven't got to that stage yet. This is the particular model that's going to use to generate the image. And Adobe has a number of different ones that we can use, and they all, you know, were created at different times, and they have different levels of features that they can apply. So in this case, we're going to pick FireFlerimage three. Because I've used it before, and I know it will produce exactly what I want. In this case, for the hero image, we're going to go for a square, something that we could show on Instagram. We'll leave these settings at the moment. Haven't decided yet if we want to be three D animated cartoon. We're just going to see what the AI is going to come up with. So once that's all set up in here, this is where we're going to type our prompt. And so here's one I prepared earlier. So we've said we want a sleek black and gold athletic track suit displayed on a mannequin in a modern studio setting, soft shadows, premium brand aesthetic, realistic photography style. And that's about all we need to say to the AI. So, click Generate. So at this point, it will ask you to log in if you don't already have a login. So now that I've logged in, it's gonna set about actually generating our images. So we'll give that some time, and then we'll come back to it. And there we go. So, as you can see, we've got four generated images, exactly what we said. So we said we want it to be sleek. You can tell from the fit that it is a sleek black and gold. It's definitely that athletic track suit. You can see that, and it's displayed on a mannequin. In a modern studio setting, soft shadows, and it's got a premium brand aesthetic, realistic photography style. I think that looks very realistic and very premium. So I'm definitely happy with that. So now the next thing to do is actually refine the prompt for our brand and for our brand style. So the way we're going to do that is, well, we know our brands called passion sports. We know it's black and gold branding. So we want to edit the brand slightly. We also want to put it in a different setting. Like so so now it's going to generate us four more, which are a black and gold premium sports track suit displayed in a clean modern gym this time with metallic gold accents and dynamic lighting, professional product photography style. So there you go. Got the same items in a completely different setting now, and that might bring it to life a little bit more for our potential customer. Now what we can do is show it in some different scenarios. So we've said the track suit displayed neatly folded on a reflective black surface with high end retail lighting. Let's generate that. So there you go. It's not the colors exactly right there, but you can see that it's got them stacked up, well folded. So if we now go back and we choose any one of these, what we can do is we can go to the next level for this specific image type. If I paste in the exact same prompt and click Generate, colors came out a lot better. It's a lot more realistic and it matches with the original design. So that's the benefit of actually going in and editing one particular image. Now, what if we actually want to see a lifestyle shop? Now we're saying we want to see an athlete wearing the track suit mid run on an urban track, sunrise light, energetic and empowering mood. Let's click Generate. And the beauty is now we know that we're fully focused on this version of the track suit. And here we go. So now we've got an image that matches exactly what we asked for in our prompt. Obviously, there are various different things we can do to give the person the kind of actions and the character that we want. Next, let's create a storyboard frame for later video use. So here's the prompt. We want a dynamic composition of a sprinter in motion, wearing a gold trimmed black track suit, cinematic lighting, and motion blur. So let's generate that. So there you go. That's what was created from this prompt. So now we have an image of the track suit in a gym setting, futuristic. We have it folded. We have someone in mid jog, someone in sprinter motion. What we can do now is we can download all of these different versions of the image. So click Download. So there you go. We've now downloaded the images. Now, let's go back to the main page. Another thing that we can do is we can always use a reference image to influence one of our other images. So if we've got a few images together and we see that there's something we like about one of those images, we can now go back to our original concepts and influence those based on this new image that we've got. So let's do that for a second. Okay, so what we're going to do now is we're going to use one of our previous images as a reference. So what you can see here is that um, for every image, you can choose either the composition or the style to use a reference image. And so what we're going to use is we're going to create a new image and base it on one of our existing images. So first of all, I'll choose as reference from our device. We'll go to my desktop and find it. Let's use this Mannequin one, upload that. So what that says is that we're going to use as a reference, we're going to type in a prompt, and that prompt will use this image as a reference. So the prompt we're gonna put in is we want an image of the same style, different pose in an outdoor sports court setting. So what's done here is it hasn't exactly created them with exact same colors. If we add it into style, There you go. So now that we've added it into style, we can see there's the exact same colors, but just in different settings. And what we want to do now is we want to make it region specific. First of all, we can edit one. So let's say we go into here. We want to pick this one, paste in our prompt, which says the same black and gold tracksuit photographed in a minimalist Tokyo gym with clean white walls and soft daylight. We get a message here saying that Adobe's using something called partner models. So in other words, they're using other models with companies that they partner with, so I'm going to say, Okay, on that one. Fine by me. And now we're just waiting for it to generate. So, again, we said we were focusing the minimalist Tokyo gym with clean white walls and soft daylight. Then we can say the same track suit on an athlete jogging behind the sein at sunset. The seine is a place in France, a river in France. And we are saying that the mood should be elegant and modern. So let's go. Go. The next one we want to do is for Brazil. So let's say an athlete wearing the same track suit, running on a Rio beach at sunrise, vibrant and warm tone. Generate that. So there you go. Now we've got someone running on the beach, wearing the exact same track suit. We've now got a person also in the track suit, which is helpful because we started from a mannequin. We then went to the mannequin in more of a minimalist setting and then a real person in France to a real person on a beach in Brazil. So I could change this to I could change just to say during the day, see what it comes up with. So let's go over what you've learned. So to wrap up, you've now created your first AI generated hero product shot. You've produced multiple visual styles for one campaign. You've used prompt engineering for brand consistency. You can generate images for a specific locale, a specific country. You've built the visual foundation of your upcoming video ad, and you've got some shots that you can use on social media if you want to or just literally to show your concepts to people in your organization. And that's that. You see you in the next lesson. 5. How to Generate Videos from scripts & images with Runway ML: Welcome to this module where now that you've created strong visuals for the Passion sports tracksuit launch campaign, it's time to bring them to life through motion, through video. So in this lesson, we'll use Runway MLs video generation tools to convert your static image into a short branded video clip. In the old world, producing a video campaign meant filming, lighting, editing, a long expensive process. But with generative AI, you can write a script, feed in a reference image, and generate your video in minutes. The first thing we're gonna do is we're going to log in, and then I'll take you through. Okay, here we go. Now the thing to remember is that I know that many of you won't have paid for the subscription, so I'm going to do this all using the free version. But when you do upgrade, you'll get many more models and many more ways to use the software. So we can just drag this to one side. So first of all, let's start off on the left hand side. You can see this menu here. So let's just quickly go through this. Chat essentially allows you to type in prompts and to generate images and do anything you need to do simply by using prompts. And it gives you a little bit of a clue here as to what to do. Hello, I'm Runway's creative assistant. Describe what you want, and I'll write a prompt to generate it. What would you like to create? So you can actually tell it what you want to do and it will create a prompt for you. So that's in chat mode. In tools mode, you actually get two tabs here, image and video. If you want to do anything to manipulate images, you can do it with prompts and more features here, which we'll go through. And if you want to do anything to manipulate video, you can do it here by uploading an image. Typing in your prompt here and you've got some other options here. These are all of the different models that you can use. So they all have their various pros and cons. In the free mode, you get access to Gen four Turbo and previous model Gen three Alpha Turbo. Now, just from my experience, the turbo versions are usually quicker but slightly less powerful. And if you were to go for the equivalent, which wasn't turbo, it's slightly more powerful but not as quick. So that's the trade off. There. And so I'll be demonstrating that. And that's it, really. So, for the most part, I mean, there is so much that you can do with Runway ML. So if you go to apps, you'll see all the things you can do create a mockup, a vary ad. You can create an ad, expand an image into different dimensions. This is examples of different types of video that you can have at weather related one time of the day, lighting, backdrop. You can stylize images. I mean, there's so much that you can do. And obviously, we want to be very targeted in this lesson. So we'll come back to that at a later date. Then there's workflows. Workflows allow you to string together all of the different types of ways that you can manipulate images and videos and put them all together so that instead of you having to go through the same set of steps every single time with images and videos, as it says here, you can actually chain together multiple modules and intermediary steps. And what that does is it essentially makes you more efficient if you're doing the same thing all the time in your production. That's what workflows do. Open assets will show you actually opens a new down. I'll show you all the assets you've uploaded, but we haven't uploaded any yet, so any images, videos all in one place. And learn essentially is where you can go to get help to see examples of how to use workflows, how to build your own workflow, how to change voices. So there are various videos here that will allow you to get some help if you're not quite sure how to use RunwayML because it is very powerful and there's really so much you could do and some quite cool examples here as well. And that's pretty much it. The session is like your project, so it's already created a session for us called Untitled Session and that's what we're going to be using to do our work today. So let's get started. Let's get into. So today, because we want slightly more control, rather than using chat mode, we're going to go into tools mode. So let's click on Tools here. And as I say, there's two tabs, image and video. So our goal today is to produce a ten second motion clip, video for the tracksuit campaign. What we want to do is we want to upload a reference visual. If we click here, that allows us to upload a visual. So let's choose one that's already got our guy wearing our black and gold. I think that one looks great, but let's have a quick look through. That's a good starting point, I think, yes. Let's open that one up. So this will anchor the look and branding of our video. So then the next thing to do is to enter the scene prompt, and that's what we want runway to actually do with the image. So here's one I made earlier. And this one says, an athlete wearing a black and gold track suit jogs through a modern city at sunrise with the camera tracking alongside in slow motion, gold trim catching the light, final frame shows, brand logo, passion sports. So that's the prompt. That's what we want it to do. And here you can see we've got prompt selected, meaning it's going to read from here. Act two is another tool which we'll go through at a later date that allows us to model our face movements or various movements. In our video, but we don't need that just yet. We're going to keep it looking widescreen at 16 by nine. And if we go here, there's various settings if you click on that, and you can go and see the settings that we've got. We're not using any of this at the moment, so I won't go through just yet, but it's good for you to know you have settings in here. It's going to be in 720 P resolution. Down here, as I said, you can set the model. If we want to go for speed, this is the latest and greatest Gen four turbo. And if you go to previous models, you can go to Gen three Alpha turbo, which is the quickest using the previous model, Gen three, and they all have their pros and cons. But essentially, we'll stay with Gen four for now. So now that we've set the parameters to our liking, what we simply need to do is click Generate button and watch it do its thing. So that will go into a queue and we'll wait for that to complete and see what it gives us. There it is. So let's see what we got. So it's a start. It's not exactly what we want. It got the logo. We're not actually running through a street. But let's try some other things now. Let's try it step by step. One thing I found that works really well is when you guide the AI step by step through what you want. So first thing we're going to do is we're going to say to remove the background or replace the background with city street. There you go. So look at that. We've gone from the gym to outside. Now, for what we want, we actually need to do that in the image. And then what we can do is bring that image in and then manipulate the video based on that. If we ask it to do too much, sometimes, that's why we got the previous action where it was actually running through a gym instead of running through these streets. And we need to be very careful that we tell it every single step along the way. So if we go back to image mode, and then we add that image reference, we can pick the exact same one. Double click. And that's our reference. And we can be really specific. So let's replace the background with a New York city street and say exactly where are our guys standing. So we're now going to make this image. The Man stands on a modern New York City sidewalk walk in Manhattan. Let's generate that. There you go. That's a lot better. So what we're now going to do is we're going to use this as the reference for our video or the input for our video. So if we switch to the video tab and say input for video, that appears here. And now we can go on and we can be a bit more detailed with what we want to do with that. We're actually splitting the prompt up slightly just so we can be really accurate. So now we're going to say the man jogs through a modern city at sunrise, camera tracking alongside in slow motion. So let's see what we got. The beginning of a man jogging through. We haven't been specific about how he looks like he's jumping slightly rather than jogging. And we are using the turbo mode as well. So what you'll realize is if you start to use some of the other models like the exact same generation, but not the turbo mode, you'll start to get different results, as well. So now what we can do is we can further enhance it and say use as reference. And so now it's using this video as a reference for what we say next, and we can just continue to build upon it. So let's take this to the next level now. Let's say he's got an audience as he walks through the streets. So So we're going to say a group of 25 to 30-year-old ladies gaze as he jogs past them. He looks like he's somewhere in the age bracket, so I might as well make them the same age. Let's see what that says. And I've just noticed I was looking at the image tab there when we should have been in video. And so that's why we got a still image. But what we can do now is we can import that image into here and use that as reference. So now making sure that we're on the video tab, we can go again and say input for video. So that's what we should have done before. So now we can take this still that we've created and we can give some character to these ladies. So we've said the ladies smile, laugh, and point at the man as he jogs past them. They are dressed casually in jeans and casualty shirts. So let's see what we got. H So we didn't get so much going on with the different change of attire for the ladies, but they are clapping. I look like they're cheering him on just about, but he looks like he's cheering his self on. But you can start to see how he's taking shape, and it takes work to get the exact correct prompts. But as you can see, with a little bit of perseverance, you're going to start to get something that actually looks very much like the kind of ad that you would want. What we can do is we can always download things when we're ready is over here on the right hand side, click download. And there it is. Not bad. Excellent quality with the video. A lot more we can do with this. We're just getting started, but this is the beginnings of our video. 6. How to Lip Sync a Voice to a Face in Runway ML: And now we're going to look into how we can use Runway ML to do our lip sync. So if we go to the homepage, and then we can search for Generate Audio, and what we can do go to Lipsync video. We can select from assets. We can find our image, or we can find even this video, which has been uploaded. What we do there is we click on a microphone and we choose an appropriate voice. Here's an example of one. I'm not gonna lie. I am super excited to be here. I don't think that's right. Let's go. I don't think that's right, either. So let's just go through until we find one. Let's try Clint, see how that goes. We're not seeing you. The other thing we need to do is select the face that's going to go with. So we're going to go with face four. So there it goes. So now it's generating takes a little while to generate. So that didn't quite work. It said, The lighting changes too much in this content. Your credits have been refunded. Please try different input. Lip sync works best with a single human face that is medium close and front on avoiding extreme angles. So I can see why that didn't work. We had about four faces in here. So what we really want to do is we want to create a version that is a lot more close up and we want to really cut to that frame, and then we can go from there. So let's think of a different way to do this. So I've just uploaded a different image, the studio image, and it's detecting the face in the image, and there we go. So now we've got one face that should meet the requirement perfectly, and it's going to still say passion sports, move with power, train with precision in Clint's voice. So let's generate that and see how we go. So there you go. Let's see how we got on with that. Passion sports move with power, train with precision. Not bad at all, so it's definitely given us some lip sync, which is exactly what we want, and it's perfect for an ad. So we can now stitch together those scenes. We could animate this so that he's walking into the gym or walking out onto the street, walking back into the gym. So that's the next stage of our ad campaign. 7. How to Add Motion to Images in Runway ML: So let's see if we can get him to walk into the gym and walk to the front exactly how we want him to be to make an entrance in this ad. Okay, so we've got the gym starts empty. The man opens a door at the back of the gym, the sun beams in. The man The man walks to the front of the gym where he's now standing. Okay, so let's see what it came up with. Let's see if we can get it to start from the point of an empty gym. 8. How to Brand and Edit Video on the Timeline in Runway ML: Okay, so now we've created our individual video clips. What we want to do is we want to string this together into an ad. Now, what I want to do is I want to have an ad where at the end, however we put it together, we actually have the brand logo here. So what I did was I went over to chat because chat allows us to do things that sometimes we're not sure exactly how to do it. I know that we can generate images, but I want to do a few things at the same time. So what I did here was I went to chat and then I if I show you what I've put in, then it would show you exactly how you could generate the same image I generated. So first of all, I'll go here. So here's the session where I generated the image. Generate an image from my sports brand called Passion Sports in black and gold that I can overlay over this image and matches this brand. And so if I now add media, I can add media by getting stuff from my desktop or just select the asset from within runway. I like to do Add Media because I know exactly where the file is, and there's less looking around. If you look inside here, there's various folders, and sometimes it takes me a while. But assuming you've got it here, wherever you've got, so I'll go back and I'll do it my way. So assuming you find the image, go to Add Media, find it on your desktop. That's the one I want. That's going to be in our final scene. So now runway ML is going to use this to generate me an image. And so there you go. So here's the image. You can see it's nice, black and gold passion sports. Let's close that out. And what you want to do is you want to create a timeline, a video timeline. So if you go to home, and then you go down here to more and video editor projects, I've actually started creating some already, and I'm going to show you exactly how I did that. So let's create a new project. And in here, we've got the timeline. And what that allows you to do is down here will be the timeline, and we'll be able to drag various assets on. So the first scene I want is the scene outside with our runner, just with three ladies watching him run to our assets. Let's find the exact video that we want. If this isn't arranged and sorted properly, I often like to upload them, so I'm going to do it that way. So I want this video. Outside. Just check it's the right one. I also want And now, what we're going to do is we're going to add them to our timeline. So the first one we want is this. We'll just double click. The next one we want is him walking into the gym. Click the arrow to go to the end, double click. And that's that. Add it in here now. And then the final one that we want is the lip syncing one. And another way to add it in after we've clicked to the end is to just add it to the timeline. So now let's try out. Let's see the way this plays out. Ahi on sports, move with power, train with precision. That's our mini ad. We can see that he starts off on the street. A few ladies clapping for him. He then walks into the gym, walks to the front of the camera, and then he lip syncs and says, Passion sports. So that's pretty good, but I think he's made an entrance. We want him to make an exit. So what I actually did was created a reverse clip of that video. And so now what I'm going to do is I'm going to add that reverse clip in so we can actually see him walking away. So let's find that. And there you go. That's the one we want with him walking away. So let's add that to the end of the timeline. Here's what we've got. Y Now we're going to introduce the brand. We're going to put the brand logo in this final scene as he walks off. I think that will look very cool. There it is. So we're going to now add that to the right part of the timeline by double clicking again. So now we've got a nice transparent image as he walks off. However, it's not very visible. So what we're going to do is add background. Let's go here where it says solid. We're going to start at the beginning of the scene. Add a solid background. And these are actually layers so you can change the order. So if you want something above, something else, you just drag that to the top, drag this underneath. And that's sort of what we want. It is nice to add a little bit of opacity. So if I go to animate and inhere opacity, and I'm going to reduce that down. I think 50% should be about right. And also, if we now go back to the timeline, I think when he's turned his back would be a great point to bring us in. So let's start from here. Play that through. So now let's play that through from the beginning. Basham sports move with power, trained with precision. Excellent. That's our ad. Obviously, we've got some cleanup to do around the way he walks, because of the way that was generated when he walks in, isn't exactly where we want it to be. But as I say, there are some tricks up our sleeve to be able to do that. So now that we're happy with that, we simply go up to the top, click export in 720 P in free mode, and you can go up to higher resolutions in pay plans, MP four, export the video. And it says that you can view them in your assets page. So the way to do that is to go back and go back to home. And down here in assets, it says, In private assets. You can see that it's loading up here. Press play to check it's all there. As sports move with power, train with precision. Awesome. Obviously, a lot more we can do. It's simply a concept, but I'm happy to download that, so let's download and save that to my desktop. And there it is. That's our first all done in runway. Now, obviously there's a lot more that we can do to perfect this. We can perfect the way he walks. We can put more people on here. We can change the way that they're dressed. Obviously, the way he walks is a priority. As I said, there's some tricks up our sleeve for doing that. We could also add some fades to the way this pops up, add some text in relevant places. But at the moment, I think that's excellent. If you look at the video quality, it's very realistic. We could use different images for him. He looks very slight slightly unrealistic in his face, nonetheless a very good image. And obviously, we need to get rid of that, get rid of any branding. But this is a first look at exactly how you can achieve your result in runway and it's really excellent. 9. How to Remove Objects and Correct Issues in Runway ML: Now, we're going to further edit our ad, so we're going to clean up the fact that as this guy is walking in, he actually looks like he's walking backwards. And the way we're going to do that is I'm going to use it as an excuse to show you two things. First of all, how we can totally remove him out of the frame, and then second of all, how can we have him coming back into the room. So in our chat, essentially, we tell the AI to do whatever it is we want to do. So what I want to do is I want to upload a still, and then I want to actually remove the person from the still, and then we'll use that still to put that person back in but walking into the room. So let's add media. I like to do it this way, add media. And then let's find our still. Perfect. Remove Really simple instruction there, remove the man. So let's let that play out. And there we go. Completely gone, completely removed. No blemishes. It's really clear that the person's been removed and we've got a nice background. Now, we're going to say, show the man walking in from the back and ends up in the same position. So, show the man walking in from the back and ending up in the same position. And to be clear. As before, so let's go for it. Here we go. Excited. Excited to see. The way he's walking looks really fluid. There's no more strange backward motion. Okay, so let's download that. Awesome. So now the next step, we're going to go back to the timeline and then make sure that we put this in the right place, and now we'll have a really fluid looking ad. 10. How to Replace Video Segments in the Timeline in Runway ML: Okay, so, here we are back in the video editor. We're going to replace this piece of video. Let's play it where you can see he's obviously got some kind of problem with his legs. So we're gonna replace that piece of video with our new piece of video. So let's first of all, bring in again, I'm doing it this way I find it really easy to find what I'm looking for. So there's our video. What we want to do is remove this piece of video, and we want to go right to the end of the first piece or the second piece, I should say. And we want to slot in this piece of video. There you go. So hopefully now we've got someone who walks in so with power. Train with precision. Well, obviously, there's more sound that we could put in at different places. Pashun sports move with power, train with precision. I think that's a really awesome start. Obviously, some things we can do there. We can add music to every scenario with the footsteps, all that good stuff, more people on the streets. We can have it blend a lot more neatly from this scene into this scene. We can then when he's walking off, we could have music and footsteps in this part as well. And we could blend in the passion sports a lot better there. But ultimately, I think it's a really high quality production there. And I think it's good for an ad campaign. So there you 11. How to do Color Grading & Post Production in Runway ML: Okay, so once you've got your ad to a level that you're happy with, it's time to do some post production and some color grading and things like that. Now, the way you do that is here right here in the very same timeline. So let's take, for example, this first clip, if we select that, and then we go over to the right here, and then we go over to effects and filters. And if you just go into this dropdown, and pick color. You'll see a whole bunch of color options. Now, brightness and contrast allows us to change exactly that. We've got exposure and black level, hue saturation and lightness. Usually it's brightness, but hue saturation and lightness and invert and a whole bunch of different options here. So, for example, I can go into brightness and contrast. And down here underneath effects and filters, you can see that we've got brightness and contrast here, the effect settings. There you go. So we can then set the levels I really do like it almost how it was, to be honest, so I'm gonna leave it there. And then contrast, they really are very drastic, so you only need a very small amount of change. So you can slightly change your contrast there if you want to. Again, I like it pretty much how it is for now. And then if you want to add another effect, you simply go to the top, click the plus button and choose what you want to go for. One thing I often like doing is changing the, the black level and the exposure. So you can see it down here. Click on settings. And I like changing the Black level more a lot of the time. Today, I don't think we need to do so much. I think it's pretty good, but, you know, the professionals will tell you exactly how it needs to be set. But it's good to know that we've essentially got all of these options. So feel free to play around with these settings all on the right hand side, to change as many things as you think is necessary to get it to look exactly as you want. But it's good to know that we have the ability to do all our post production all in the same place and get it looking exactly spot on. 12. How to Localize Video with Synthesia: Now you've already created strong visuals and a motion clip, a video clip for Passion sports tracksuit launch. So we've got our ad or at least the beginnings of our ad. Now we're going to take those assets and make them globally ready. So another way of saying that is localized and personalized for different markets, languages, and audiences. We're going to use Synthesia for multilingual video, so creating video in different languages. So first of all, our goal here is to produce a Spanish localized version of the ad campaign, and the video should be dubbed into Spanish. The visual should also be adapted for a Spanish speaking market. So first of all, let's log into Synthesia And the first thing we need to do when we're in synthesia is we need to go over here into dubbing. Now, you can see here it says, Translate any video into 32 languages with the original voice. The next thing we're going to do is we're going to grab our video that has the speech in it. So now that's dragged in fine. I'm going to leave the project name the same. I like the fact that it's got 720 B three, because if we create different versions, at least we know the resolution and that this was the third version. Original languages English autodtected, that's fine. And then we're going to choose Spanish because we want to localize this for our Spanish market. We could choose more languages. Dubbing options, we're going to have lip sync on. And then we see that you need to upgrade for lip sync movements. So for now, we'll leave it off. In case you're using a free version, you'll see what it will look like. Remove Watermark. We can also see a premium option, so we'll leave that off for now. And video duration, you can choose either adaptive, which means that the dub video speed will be adjusted to fit translation, and that's best for instructional content, or you can keep it original, which means the dub video will keep the same duration as the original, and that's best for fast paced dynamic videos. I'll leave it adaptive. So now click Generate. In case you were wondering, there are two versions of the video. It actually has uploaded the English version, so that's what we see on the right, and it's now generating the Spanish version as we speak. You can see that there's 28 minutes remaining, so we'll come back when that's done. 13. How Video is Translated with Synthesia: So we can see now that both our videos are finish. On the right, we have English, the original version. On the left, we have our Spanish version. So let's see what it came up with. Phone sports. Ready for our Spanish market. Notice how it left the passion sports alone, didn't do anything to that. It left it in English. We would probably want sports to remain English because it's part of our brand, but it translated everything else. The other thing of note is that you can change you can switch between this version and the English version if you want to compare simply by changing up here. So that's nice and convenient for you. This is our first version, so there's nothing more to do there, do that. There's an area on the right, we can add a description. So I can say So just some text to say it's the passion sports tracks you add concept, Spanish localized, localized Spanish. And then if there's any other comments, they can go here in case I've shared this with someone and maybe an editor or someone that needs to leave some comments, they can leave them down here, and then they end up on the right here. That's pretty handy. So at this point, we can download our Spanish video, and we can include that in the runway ML timeline in place of the English one. And that would be job done, nice and easy. 14. How to Localize the Visuals in Adobe Firefly: Okay, so now our next step is to actually localize the video, the visuals. So we've localized the video in terms of the audio part, so that is in Spanish. But what if we want the background and the surroundings to actually look like somewhere in Spain? Well, that's why we're in Chat CPT. The reason we're in Chat ChiPT is because soon we're going to go over to Firefly, and we're gonna tell Firefly to localize the background, so make it look like we're in Spain in Madrid. Before we do that, we're going to knock up a prompt in Spanish. And the reason we do that is because my research showed me that when you actually type in a prompt in English, there's a chance that some things get misinterpreted. For example, if you're talking about a rooftop, it may put in a certain type of rooftop that doesn't look particularly Spanish, or it may use some kind of backdrop or interpret some kind of imagery that don't match with that locale, that don't match with that country. Pedo, if you actually ask the question in Spanish, that way, it assumes your Spanish, and it models everything after Spanish the Spanish characterization. The first thing I'm going to do is I'm going to put in a prompt for what I want, and then I'm going to get Chachi Pit translate it for me. So the first one is So I've asked Chachi PT to translate to Spanish, an athlete wearing a black and gold passion sports track suit running through a modern gym in Madrid with warm sunset lighting, premium commercial style. So let's run that Okay, so that's the first translation, and we're going to use these plus our imagery to translate some new images that are localized to Spain. So let's go. So now we're at Firefly. We're going to paste in our prompt. And this is the one saying we want to generate an image of An athlete wearing a black and gold passion sports tracksuit running through a modern gym in Madrid in a warm sunset with warm sunset lighting in a premium commercial star. So that's what that means. And what we're going to do is we're going to upload an image. We'll leave it Firefly five. There are a number of models here that you can choose from. We'll leave it at the latest and greatest for now. We want it to be 16 by nine. And so now we're going to upload the reference image. So now it's got a reference image. It knows the kind of scenario that we want, but it should just localize it to Spain. So let's go. Okay. Awesome. So it looks pretty much the same as before, which we like. It has shown the nice warm sunset light, but there's nothing about it that looks particularly like Madrid. So at least what we should do is it does look as if maybe there's some buildings or something outside. Let's tell it that the building should clearly be showing through. So let's say we can clearly see some Spanish buildings through the windows or some Spanish buildings are showing through the windows. Let's translate that. So let's say this says, So Spanish buildings and trees are clearly visible through the windows of the gym. And let's translate that. Awesome. So now we're going to take this. And we're going to add it into firefly at the end. Same reference image. Go. But There you go. So now, if we look behind, we can clearly see there's some imagery there. We can see that there's some trees, not so clear what they are. So let's change this again slightly, and let's say some Spanish palm trees are clearly visible. So I'm going to get rid of what I put in before. I'm going to go back to hachPT and I'm going to ask it to translate. Some Spanish buildings and some Spanish palm trees are clearly visible through the windows of the gym. Let's add that on the end. And generate. There you go. So now we've got the buildings, but we've also got some palm trees in the background. Excellent. So now we can use this as a starting point for our video. And so now when our guy runs out or runs in, we could use some of the same scenery coming into the gym, and when he goes out, if he does go out, use some of the same scenery going out, and it could all start from this one image. This would be enough context for the AI to know to preserve some of the outdoors or we could tell it to preserve some of the outdoors. Those look a little bit more like coconut trees, to me, to be fair, but you know, seems to have understood what we were trying to do, and we can always refine it further. This also kept the sunlight coming through just as we asked it to. 15. Introducing Automated Workflows with Make: Hey, guys. Welcome to this module on AI powered Workflows. This module really is going to bring everything we've done together and tell you how we can speed things up. So, to understand workflows, let's recap our workflow. First of all, we create our prompt. We might come into Chat EPT and create our prompt. Then we go into Firefly. We drop our prompt into Firefly and create an image. Then we'd go into Runway EML, drop in our image as a reference image and create our video, bring it into Synthesia where we can localize. Like so. And then we would end up with assets, as we see here, and our final video may be residing in Dropbox in our asset folder. So now the thing to realize is that seems pretty quick, and it can be quite fun. But this is the manual workflow. So in a true industry situation, imagine doing all of that, but manually for every single asset. Imagine creating 20 images, 15 videos, ten localized versions for eight markets, plus all the variants for social platforms, and imagine having to reimagine all of that and obviously have to put things on the timeline. That's hundreds and sometimes thousands of small creative assets. Nobody really wants to generate 1,000 images one by one, download them, rename them, and upload them, and then convert them and then localize them. And this is where automation enters the picture. See, AI doesn't just create content. It also allows you to manage the workflow around the content. And you can use generative AI to trigger the asset creation, pass the outputs, which are images or videos, et cetera, between different tools. Even quality check the results and deliver the final assets to wherever you want Dropbox or Google Docs or Air table, wherever you want them to be. And so the way we do that is using some kind of workflow automation tool or automation tools in general. This is where a tool like make comes in, make.com, and it becomes incredibly powerful. And that's what we're going to go over here is how to automate with make.com. So what is make.com? Let's explain it really simply. Make is a visual workflow automation platform. Think of it as a digital assistant that connects all your creative tools together, and it can connect many tools, in fact, but in this case, we're using it for creative tools. We can use it to move files from one place to another. Talks to generative AI tools like runway, and it runs task automatically. You don't have to do anything except set it up and then it can be set and forget until you need to change something. And the best part is you don't need to be technical to use it. Everything's dragon drop, and it's like building a storyboard, for example, of your workflow. How does it make use of AI? Well, M can trigger AI models and things like runway and even Chat GPT and lots of other apps through simple modules. For example, you can create an automation that says, Whenever an image drops into Dropbox or Google Sheets or Google Doc, send it to Runway. When Runway finishes generating whatever you want it to generate, a video or an image, you can get it to upload it automatically to Dropbox or upload it to Synthesia for translation and then drop it into Dropbox. So that's a fully automated pipeline, meaning that AI does all the work and with the automation does all the work of transferring it from one app to the other. So what we're going to do is we're going to do a demo, even though we could actually automate everything all the way up to Synthesia. Let's start with something really simple. So in this lesson, what we're going to do is we're going to automate creating an image using runway based on a reference image from Dropbox. So let's get started. 16. How to create a Scenario with Make: Okay, so we're going to get started with make so that we can automate our whole workflow. So let's get ready. Let's get started. First of all, let's log in. So here we are in M. This is a brand new scenario. I'm going to go quickly over the interface and how it works, and then I'm going to show you an existing scenario that I created before in my existing account. So the way make works is essentially you put together a bunch of components that all work together that allow you to reach your end goal. So you click this plus button, and then you can choose from various apps that you can string together that will work together to create your end product. So in here, for example, We were using runway. You can click on Runway. Then within that, you've got generate image, generate a video from images, generate a video from a video, make an API call. There's a number of things you can do with that. And the same goes for all of the other apps as well. So for example, there's an email app built in. We've got What's up Business Cloud built in. We have Dropbox built in, and so we can essentially pick a number of different apps and send information between those apps and get various outputs we want. And we can also access our own file system. So you can already see how this is really powerful for being able to create workflows and be able to send information from one app to another. So the way it works is you would string together your components and then tell them where you want the output to go. And then when you're ready to test something, down here, there's the run once button that you can hit. And so it would start sending your data from one place to another, from one component to another, which I'll show you soon. You've also got anytime you run a scenario, this is a scenario. You can re run and replay scenarios that you've run before. They were not in there at the moment. You can schedule it. So every 15 minutes, for example, at the moment, it says, every 15 minutes, you can turn that on and schedule so that you can keep running the same workflow. And then you can save your session in then there are basically various things, various bits and pieces down here that allow you to show all your inputs and outputs when you've arranged them on here. You've got some notes and you've got the ability to undo. You've got some various settings in here for your scenario as well. Flow control, aggregators and all kinds of different bits and pieces and components you can string together in here. You've got tools like the base trigger. And these are things you can access from the plus button here for convenience. Base triggers allow you to essentially just start off the whole process, a basic trigger. Text parser in here. These are things that get content from various elements. And this is the equivalent of this ad button here. When you hit it, it allows you to find all the different apps that you can string together. There really is a lot that you can do with M. It's really powerful. You can send things to your mobile phone or other mobile phones, and it's really as powerful as you want it to be. So just to demonstrate, I'm going to show you the kind of scenario we would build for our campaign or for generating any kind of multimedia that I've built before. And then I'm going to show you how to build it. So in here, I've got two scenarios. One is runway image from a prompt. So we can basically type in an AI prompt, generate an image through runway and upload it to Dropbox. So let's have a look at that. 17. How to Connect Runway ML with Make: Have a look at our documentation. So this documentation is telling us how to connect runway and make. Requirements. To use the runway app in M, you must have a runway account, which we do have. Connect runway and make. To get started, you must first create a connection between runway and make, allowing the two services to communicate. To create a connection, you need to obtain your API key in runway. So let's first so we're actually looking in. We're looking into dev.runwaml.com, and that's where we usually access things like API keys so that we can make contact with the API rather than through the front user interface that most people use that we've been using till now. So we're here. So let's Log in. Now, it says, create your organization, organize I'm a single user. So I'm going to name my organization and create So to get started, you must first create a connection between runway and make, allowing the two services to communicate. To create the connection, you need to obtain your API key in Runway. So let's do that first. Click CR. So it says, API key created, API keys are secret and should not be shared with others. Right. So this is the key. It's just a very long number that identifies our account. Securely copy the key above and store it in a safe place. Once you close this model, the key will not be displayed again. So I'm going to copy that. I'm going to store it somewhere safe. Name filled, so we've done that into the name and click Create. Copy the API key value as shown and stored in a safe place. We've done that. So you'll see this value in the APIkey field in Make. Create the connection in M. So once you have your runway API key, you're ready to create the connection in M. So to create the connection, first of all, log into your M account, add a runway module to your scenario and click Create a connection. And we're going to build from scratch. So we've got a runway module. Generation. We generate an image. So I've pasted in the key there. And I'm going to click Save. 18. How to Create a Text to Image workflow with Runway ML and Make: So here we are at my Assets folder. This is the assets folder with all of the assets that we've already created. So what we want to do is we want to look at a reference image as we've done previously in Runway. So the reference image has been this one. And so we want to take that reference image and we want to generate videos. Before we get to that point, we obviously need to experiment with how would we even get to the point of generating this reference image or any other image. And so what we're going to do is we're going to use prompts using automation to generate images, and we can use this reference image as a basis for our new images and then use our new images as a basis for video. So the first thing is to generate new images based on this as a reference. So let's go into M and we're going to go over exactly how to do that. This is the end game, and I'm going to show you exactly how we get here. What we're going to do is going to go from Dropbox, where we're going to download a file from Dropbox. We're going to send it into runway. And then from runway, we're going to process the image using text to image. In other words, using a prompt. It's going to reference the image and Dropbox that I just showed you. And then it's going to send it via HTTP, the same protocol we use for the Internet. It's going to send it to Dropbox and store it in our images received folder. So that's this folder here, and here are some images where I've done this already. So let's build this step by step. Now, the key thing here is these are apps or modules. And the key thing is every time we create one of these modules, we need to test it before we go into the next one. The reason is that the way make works is it understands what it's able to receive based on what you create here, and only after you run it does it know what it could receive. So let's get started setting up this workflow and then we'll see very quickly how we can automate our lives really and make our lives a lot simpler. So, first of all, what we're going to do, as you can see here, there are many different options. The one we're concentrating on is the scenarios menu item. And scenarios are essentially going to be the set of steps in our automation, and you can see some already that we've got here. So let's create a new scenario, and we'll do it all from scratch. So this is our scenario. And so what we want to do is we want to create the exact set of steps to get us from our Dropbox asset folder to creating our new image and then dropping that back into our Dropbox receive folder. So the first thing is to we're going to click Create there. Open up Dropbox. And what we want to do is we want to do it whenever we actually run it. As you can see down here, there's a run button. So that's the only time we want it to start the process of transferring the files. So let's look for we're going to download a file whenever we click the Run button. So let's click Download a file. And then in here, you need a Dropbox connection, so this is essentially logging into Dropbox so that make can use that login. So if you click Add here, and the way of selecting files, now, there are two ways to do this. One is to actually manually select a file, and then you can go into the file system on Dropbox and say which file it is that you want to select. And the other way to do it is a file path, so you can tell it where the file sits and get it to pick out from there. I'm going to choose a file because I know the exact reference image that I want it to pick. So we'll go to select a file. Then if we click into here, that'll allow us to select a file from Dropbox. There it is. So we're going to take this as a reference image and click Save. So now Dropbox will download this file, and then we can send that now. If we move that aside, we can send that to Runway. So let's just run this module only because as I said, the way make works is the information that comes out of this module will go into the next one, and we just get a lot less problems when we run the module first. There you go. So what you can see here when you run it, if you click here, you can always see the output. You see the input, you can also see the output. So what you can see here is that it's taking passion consulting, generative AI assets, and then passion track suit mannequin. And then the output, if we look here in data, this is basically a hex representation of the actual image. But the important thing is, what comes out is the actual image. So it's downloading the image. From Dropbox. So what do we want to do next in our workflow? The next thing we want to do is send this reference image to runway dot ml, so we can generate a new image. So if we click plus and then search for runway. There it is. So under here, we can see all the different things we can do with our API calls. And actually, what we want to do is we want to generate an image, and you can see it down here, it says, generate an image based on a text prompt that we'll say. And that's exactly what we want. So let's click on that. And in here again, you'll need to if you haven't connected to runway yet, you'll need to do that, but I've already connected. So we'll keep that my runway connection. Model, we're going to use Jen four Image turbo just like we did in the demonstration. Now, reference Images input method. There are two ways to do that. Again, you can have a URL or a link directly to it. Or you can input a file. We're going to do a file in this case because we're getting the file directly from Dropbox. So that's why we've picked this one. So now the way to grab the file is to click Add Image, and in here, it's already selected Dropbox download file, so it knows that we're actually using this module. We can give the image a tag, and that allows us to refer to it later as you'll see, so let's give it a tag. So that's going to be our way to refer to the file that we get from here. And that's it. So that's our reference image. And then in our promptex, we can actually put in a prompt to create our new image. So let's do something cool. Let's do an image of style. So an image of a man on Mars with two guys and three girls behind him all are wearing passion track suit. The man in front suit is branded passion sports in gold across the front of the chest. The image is in a cinematic style. And just so it doesn't get confused about which image All in a cinematic style. So let's leave it at that. So that's our prompt. The thing to remember here is we can't have more than 1,000 characters, but we're well under that. The ratio, so this is the aspet ratio. Let's go for what we usually go for 1280 by 720. So 16 by nine. And that's that. There are some advanced settings down here, but these aren't things that we're interested in right now. So let's click Save. And we've got two warnings. Two things must not be empty. So let's have a look at what they are. I believe we've got everything filled out, so let's just save Awesome. So that's worked. So as we can see here, everything's filled in Passion track suit is referenced in here. It can now generate an image based on a reference image. So the next step is to send it on to Dropbox in our image receive folder. Now, ordinarily you would think that we could just attach another Dropbox module here. But when we're uploading to Dropbox, the way it works is it needs the raw image file. However, this is going to send the link the URL, and that's not appropriate. So what we need to do is convert that URL into the raw image file by actually downloading it. The way we do that is to use the HTTP module. So it's very similar to how a browser works using HTTP. Click on here. Going to choose HTTP, and within there, what we need is a MAT request. And don't worry if this seems foreign to you. It is a little bit code related, so developers will know how to do this. And anyone configuring this who's done this before we understand what we're doing here, but I'll talk through it just for the purposes of this lesson. So we don't need to worry about evaluating all stages. That's fine as it is. For the URL, this is the link. And where we're going to grab that from is we're going to grab that from runway. Now, in here, you can see when I go over this, you can see runway pulsating. That means that all this information is what's going to come from runway, and that's why we had to actually run these modules first. And so we'll go in and we'll go into the generated image object. So we'll click on that. You can see when we click on that it populates the URL. What we need to do is put a one in there. And that means we're going to grab the first image in this list, which is this image, the only one generated. So that's fine. And the method we're using for this request is G. So we're going to get that image, link using the G method. We don't need to worry about the headers and query string. For body type, we do need to consider that. So what we want to do is turn that link into a raw binary image, so we will click raw as the body type. And content type, we actually need to set again, the main thing is we want this to be the raw binary image. So I happen to know that what we need is application octet string as the content type. And the way we'll do that is if we go to custom because none of these fit the bill, click on Custom and the value will B There you go. So that tells Make that this is going to be a binary file, and it's going to be the raw binary that we can send onto Dropbox. Request content can be blank, and pass response will leave it no because we don't need this to be Jason or XML. So there you go. This tells Make that we're going to take the link to an image and turn it into a raw binary for Dropbox because that's what Dropbox needs is the raw file itself. So click Save. There you go. Now, let's run it end to end and just check that everything works. Dropboxes worked. Runways running. That's worked. And this has worked as well. So that means it's successfully grabbed the file from Runway. Remember every time you run you are using credits, but this is what we need to do to make sure it runs every step of the way. Okay, now now on to our final step. So what we want to do now is we want to drop our file, which we've got from Dropbox. We've actually got from Runway, so we've got a reference file from Dropbox. We've created an image based on that reference file. We've grabbed that image based on HTTP, and we want to drop it in our folder. So, if you remember, this is our image receive folder where we're going to actually receive the new file. So what we want to do is click here. We're going to pick Dropbox again. But this time, so we're going to click Upload File. And this is the important part. So if we move this in a little bit, so we can see everything, we can actually shrink it down a little bit. So what we're going to do is we're going to use the same connection, and we're going to choose the folder that we're going to put that image in. Click folder. Let's search for it. In our image receive folder. There you go. So that's where we're putting it. In terms of the file, we're not going to download it directly from Dropbox. That's what this would do. We're going to click Map. And what that does is it allows us to map the file name and the data to grabbing the information from one of these objects. We want to, of course, grab it from this HTTP module. So we're going to double check it. I think it is correct, but we're going to double check it by deleting it out. And if you click in file name, you can see here if we move everything along a little bit. If I hover over this, you can see the HTTP module pulsating, so we know we're looking at the right module. And what we're going to do here is we're going to go down to the data, and then you can see that's the data we're going to grab. But before we do that, we need to name the file name. So what I'm going to do is I'm going to name the file name based on the ID of the file in runway because I know that will change every single time. So let's grab the ID, put that in there. We can give it an extension like dot PNG. And that'll be the funding. So now what we're gonna do is we're gonna grab the raw data from our HD HTTP module. And that's that. There are some advanced settings, but there's nothing that we need to worry about. We can say overwrite existing file. I'm going to say no for now because we should have a different name every time anyway. So I'm going to click off on advanced settings. And that's it. We've named the file an ID that we get from Runway, and we're grabbing the raw data that we've got from Runway via HTTP. So the raw data is the actual image. Click Save. And let's run that end to end. So this is now full end to end. It's gone from Dropbox, the reference file. It's using our prompt to generate an image. Then it'll go to HTTP to actually grab that image and send it onto Dropbox. Here we go. It's sending it to Dropbox, and it's done it. So let's go and look in Dropbox, but before we do, let's remind ourselves of our prompt. So we said we wanted to use Passion tracksuit, which is our reference image. Let's go and have a look at that reference image. So this is a reference image, so we're using that. And we're also using the prompt. An image of a man on Mars with two guys and three girls behind him all wearing passion tracksuit. The man in the front suit is branded Passion Sports in gold across the front of the chest, all in cinematic style. So that's what we said we wanted. And now this is it, exciting. So it should have dropped it in our images Receive. There you go. I can see the time is correct. Let's click on it and see what happens. There you go. So that's the first stab at it. Obviously, we would have liked a different font. But two guys, yeah, I think it's got the wrong wrong number of people on there. It's got two guys and two girls. Maybe it's an equality thing. But we have got something that looks very much like Mars. We've got a group of people standing to the front, and we've got passion sports on there. And obviously, we can work on this to make the font better. In fact, why don't we do that now? And we go to our prompt. We can further say, the brand name should be in a futuristic yet readable font with the logo. Let's see how that changes things. Run it again. I made the HTTP request, and it should now be in Dropbox. Let's flick back. See what happens. Go back to our folder. One has just landed in. There you go. So it's actually given all of them the brand, which I don't mind. It says Passion sports. You've got people of different races, which is always nice. I think that looks really good. That's nice for an ad campaign. I can't wait to see this in video. Hint, hint. That's what's coming next. So as you can see, that is a really nice, easy way to get stuff done. Our workflow now is really simple. All we need to do is go in and change a prompt and we get a new image. Can't wait to do the next step. 19. How to Create a Text to Video workflow with Runway ML and Make: Okay, so now we've got this workflow working for images. What we're going to do is we're going to actually make a copy of this workflow, and we're going to extend it to generate a video from our new image. So let's rename this Go back to our scenarios, all our scenarios. Safe changes first. Prompt image runway. And let's duplicate this one. To do that, we'll look at the three dots. Click Clone and we're going to call this So this is prompt to image to video for runway. Click Save. There you go. So we'll start off by remembering what we did. So what we did was we got a reference file from Dropbox. We sent it to runway to generate an image, and then we sent that via HTTP to Dropbox. Now what we want to do is extend this. So we're actually going to do two things at once, but we want to do them in the right order. So what this will do is it will drop our newly generated image in our image receive folder. With this HTTP file, we're actually going to grab the data, and we're going to use that to generate our video file. And we're going to do that using a new runway module. So let's create it now. Add module. Runway. And this time, what we're going to do is we're going to pick video from image, and this is going to complain that it needs to be connected. So let's connect it to this HTTP. Now, you'll see a router pops up because what we're going to do is we're going to transfer the raw data to Dropbox, but we're also going to transfer it to this runway module. And because it's going to two different places, we need a router, so it's put that there for us automatically. So now we can get rid of this actually because it's connected. There you go. That error should go away. It says the modules not set up, but we're about to do that. Okay, cool, so we're using the same runway connection, and we're going to pick Gen four turbo because we've used that so many times before. The method we're going to use is file because we're getting the raw data from HTDP. So we're going to map it to HTTPs output. So you click Map. The image file name, we're going to use the same ID from runway because it's a unique ID, and it's going to be the same thing. Dot PNG. The image data is going to come from our HTTP module, same as before. It's gonna be the raw data. And now the promptex. So what we want to do is we want to turn that image into a video. So we've said the scene starts from space. The people in the image are wearing jet packs and land on Mars from space with the jet pack firing, they end up in the same position as the image. So there's quite a lot going on there, so let's see how well it does. Let's make sure that runs to run it all at once. So it will still deliver an image file to Dropbox, and it will get as far as runway. So sends it across. There you go. So if we look here, we can always see the output and just check. Generated video. There you go. There's the URL. We could even put that in the browser, but I don't want to spoil the surprise. So the next part is we want to now do this exact same thing, create an HTTP. We won't need a router, but create an HTTP request to grab the raw image and then put it in Dropbox but in a different folder, so we'll still use the upload. So now create HTTP Me request. And it's the same as before URL. We're going to grab the URRL from. We're going to grab it from runway, and it will be generated video one, we're using method G. The body type will be raw again. The content type, custom value again. Application OctexStream. We don't need request content and a pass response is still no. So let's run that, make sure that works. That's worked excellently. And again, we can see our output here if we go to output and then click on this data and you can see the raw file here. So we're a step closer. Again, I won't spoil the surprise. And now the final step is to add in our Dropbox module, and we're going to upload the file we got from generating a video from image. We're going to upload this file to Dropbox. So we'll click Plus Dropbox. Upload File. It's complaining because we're not complete, but we're going to pick a folder. File we're downloading, we're actually going to map it. The file name actually is going to come again, from the ID. I want to actually call it. I want to call it the same ID as the image allows us to know which image relates to which video. So let's do that. And this time, we're going to call it dot mp four because it's a video. And the data is going to come actually from this HTTP. So we're just going to click on data here. And there's nothing to do in the advanced settings, and that's that. So click Save. So now let's run it. So what we've got now is we use a reference image. We pass it to runway to generate a new image from a prompt and this reference image. It uses HDDP to send it to grab the raw image, and it uploads it to Dropbox. And at the same time, that raw image gets routed to a different runway module that will generate a video from that image and a prompt. And it will transfer it via HTTP to Dropbox. And the prompt it's using this time is the scene starts from space. The people in the image are wearing jet packs and land on Mars from space with the JetPack firing. They end up in the same position as the image. Let's see how we get on. And there you go. A green. So that means we should now have an image sitting on Dropbox in our video received folder. New video not quite. They are definitely landing in Mars, but they look like they're starting from a standing position. Now, we did try and do a lot to be fair in that video. So what would be easier is if we get them to take off from a standing position. AI has a problem trying to do too much. So if there are people there, and I'm telling it to start from essentially space, which is a completely different scene, and then land on Mars, that's quite difficult. But if I tell them to start off in this position and then fly out from Mars, that's a lot easier. So let's change our prompt and do that. The position they're already in. So the people in the image are wearing jet packs on their backs and take off from Mars and propel up into space with the jet pack firing. We see the stars behind them. And in fact, I'm going to say we see the stars moving behind them. So let's save that. So what we should get is a very similar image, but the video will be different because they'll be starting from a standing position. Let's go. Okay, that's done. Excited to see what's happened. So let's first of all, go to our images receive and see what image we got. I'm kind of sad to see the old image, become a thing of the past, but let's see our new image. So a new image looks like that. Very similar. Passion sports now says fashion sports and it's got two Ts. Never mind. That's AI for you. Got three ladies and a guy in the background and one guy. I like it. Looks good. Looks space aged. Looks like it's Mars, so that's good. Now let's go and look at our video. So I see there's a new video that's popped in here. Let's have a look? Wow. So this is the new video, let's just hit Play. There you go. So it's one or two stars, but this is only a five second video. But again, these can all be extended, put together on timelines, amalgamated together using the API. What this demonstrates is that we can just generate all of this at the touch of a button just from a prompt. Great. So there you go. We have created a workflow that takes a reference image from Dropbox, passes it to runway, generates an image, and by HTTP, sends that into Dropbox, image receive folder, and at the same time, it roots that same generated image to another instance of runway running the video from image and a prompt. Sends that via HDDP to our video receipt folder. Now, hopefully you can see how powerful this is. Imagine if you had 1,000 images, we could have different scenarios, each one generating a different variation of the image, all from the touch of a button. You can link together all these different scenarios. And so therefore, you don't have to go into Dropbox, get the file, go into runway, generate it, play around, and then send it to all these different places to another instance, go and play with it in the image to video tab. None of that needs to be done. On top of that, this is just the beginning. From here, we can send this across to Synthesia, translate it into different languages. So you can get 1,000 more translations or I'm using 1,000 as an example, but as many as you want. So hopefully, you can see how powerful this is, and it brings it all together for you. There's so much more you can do. There are other apps that you can add on here, apart from synthesia and runway and those kind of apps. So really have fun with this, and I hope you've seen how easy it is to use AI, not only to generate content using generative AI, but also to automate your workflow. Make life so much easier. So go forth and have fun with generative AI for content and multimedia. 20. How to Create a Text to Image workflow with Runway ML and Make: Okay, so we're going to get started with make so that we can automate our whole workflow. So let's get ready. Let's get started. First of all, let's log in. So here we are in M. This is a brand new scenario. I'm going to go quickly over the interface and how it works, and then I'm going to show you an existing scenario that I created before in my existing account. So the way make works is essentially you put together a bunch of components that all work together that allow you to reach your end goal. So you click this plus button, and then you can choose from various apps that you can string together that will work together to create your end product. So in here, for example, We were using runway. You can click on Runway. Then within that, you've got generate image, generate a video from images, generate a video from a video, make an API call. There's a number of things you can do with that. And the same goes for all of the other apps as well. So for example, there's an email app built in. We've got What's up Business Cloud built in. We have Dropbox built in, and so we can essentially pick a number of different apps and send information between those apps and get various outputs we want. And we can also access our own file system. So you can already see how this is really powerful for being able to create workflows and be able to send information from one app to another. So the way it works is you would string together your components and then tell them where you want the output to go. And then when you're ready to test something, down here, there's the run once button that you can hit. And so it would start sending your data from one place to another, from one component to another, which I'll show you soon. You've also got anytime you run a scenario, this is a scenario. You can re run and replay scenarios that you've run before. They were not in there at the moment. You can schedule it. So every 15 minutes, for example, at the moment, it says, every 15 minutes, you can turn that on and schedule so that you can keep running the same workflow. And then you can save your session in then there are basically various things, various bits and pieces down here that allow you to show all your inputs and outputs when you've arranged them on here. You've got some notes and you've got the ability to undo. You've got some various settings in here for your scenario as well. Flow control, aggregators and all kinds of different bits and pieces and components you can string together in here. You've got tools like the base trigger. And these are things you can access from the plus button here for convenience. Base triggers allow you to essentially just start off the whole process, a basic trigger. Text parser in here. These are things that get content from various elements. And this is the equivalent of this ad button here. When you hit it, it allows you to find all the different apps that you can string together. There really is a lot that you can do with M. It's really powerful. You can send things to your mobile phone or other mobile phones, and it's really as powerful as you want it to be. So just to demonstrate, I'm going to show you the kind of scenario we would build for our campaign or for generating any kind of multimedia that I've built before. And then I'm going to show you how to build it. So in here, I've got two scenarios. One is runway image from a prompt. So we can basically type in an AI prompt, generate an image through runway and upload it to Dropbox. So let's have a look at that. 21. How to create a Scenario with Make: Have a look at our documentation. So this documentation is telling us how to connect runway and make. Requirements. To use the runway app in M, you must have a runway account, which we do have. Connect runway and make. To get started, you must first create a connection between runway and make, allowing the two services to communicate. To create a connection, you need to obtain your API key in runway. So let's first so we're actually looking in. We're looking into dev.runwaml.com, and that's where we usually access things like API keys so that we can make contact with the API rather than through the front user interface that most people use that we've been using till now. So we're here. So let's Log in. Now, it says, create your organization, organize I'm a single user. So I'm going to name my organization and create So to get started, you must first create a connection between runway and make, allowing the two services to communicate. To create the connection, you need to obtain your API key in Runway. So let's do that first. Click CR. So it says, API key created, API keys are secret and should not be shared with others. Right. So this is the key. It's just a very long number that identifies our account. Securely copy the key above and store it in a safe place. Once you close this model, the key will not be displayed again. So I'm going to copy that. I'm going to store it somewhere safe. Name filled, so we've done that into the name and click Create. Copy the API key value as shown and stored in a safe place. We've done that. So you'll see this value in the APIkey field in Make. Create the connection in M. So once you have your runway API key, you're ready to create the connection in M. So to create the connection, first of all, log into your M account, add a runway module to your scenario and click Create a connection. And we're going to build from scratch. So we've got a runway module. Generation. We generate an image. So I've pasted in the key there. And I'm going to click Save.