what if I told you that now anyone can recreate the most iconic TV commercials ever using nothing but AI tools ads and commercials complete with lovable consistent characters in every single scene by the end of this video you'll know exactly how to implement this six-step process that took me many many hours to develop this is exactly the kind of transformation I saw coming when I made the difficult decision to shut down my marketing agency i've seen firsthand how much time money and complexity goes into producing commercials like these we're now witnessing the same pattern that's happened in every single industry that AI has touched first it seems impossible then suddenly it's inevitable this isn't just about cool tech tricks this is about understanding how AI is fundamentally changing creative industries and positioning yourself to capitalize on this so buckle up it's time to make the donuts so the first step the first thing that we got to do is create a prompt that will create our main character and there are two tools that I like for this there's midjourney as well as Gemini's imagine feature so here we are at imagine i'm going to click into Gemini here and I have a picture of this Fred the Baker he is the notorious baker from the famous Dunkin Donuts commercials i'm going to load him in here once I've got that picture of Fred the Baker this is my reference image i'm going to go into the cheat sheet i create a cheat sheet like this for every single video that I make these are all instantly available to anybody who joins my Patreon so there's a link in the description check it out if that's interesting to you i'm going to grab this prompt here and copy this right into Gemini here this just says "Please describe this character in a way that I can use in an AI image generation prompt." And while that's running I want to show you another way and maybe a better way to do this is inside of Midjourney inside of Midjourney you can use this describe feature so just backslash describe and then you can upload the image right here into MidJourney and it will create a description of this so this is instead of going from text to image this is going from image to text is what we're doing here to create this prompt for our main character back in Gemini it's given us a few options which is pretty cool not only does it give us one but it gives us three full options which are really nice including some call out key elements man Gemini just keeps getting better every day similarly here in Midjourney it's given us a bunch of different um prompt ideas for here and it can um now what's really cool is you can just click this button and it will imagine all of these so it will create images automatically based on all of these and we can decide which one is the closest so I'm just going to click that it is generating that and this is right inside of Discord if you've never used MidJourney before it's very easy to get set up in uh Discord with that or you can now I think access Midjourney directly via the web this is always really interesting to me uh Midjourney creates a bunch of images these are pretty good that one is pretty close obviously it has trouble with uh the letters but we can fix all that later here's what Gemini created pretty interesting and at this point you want to go back and forth a little bit and fine-tune your character the way that you want it i've found that really less is more so looking through those prompts for those key words and creating a very simple image prompt can be the best way to go i've got a few examples of those right here in the cheat sheet as well after a little back and forth this was the image that I decided to settle on all right so we've made it through step one which is creating the prompt and then step two creating the actual image now we're moving on to step three which is to get some different angles of this image and this is really the key to character consistency this was the big thing that I figured out recently is how do we get this person the same character doing different things in different settings different camera angles etc it's basically impossible to do that just directly in Gemini or in um you know midjourney because every time you prompt it it creates a brand new image and you know you have to really tighten the screws on the prompt to try to get a similar image but they're always a little bit different and that's where this killer tool comes in this is called enhancer e N H and it is naturally an upscaler so taking a lowfi image and making it into a hi-fi image but it has this character consistency feature which is just incredible so we're going to load in our picture of Fred here and now we want to create the scene if you remember it starts when he's in bed sleeping and the alarm clock goes off the prompt I'm using is just man in bed with an alarm clock in the foreground let's generate this image and look at how cool that is it generated this image it's really amazing that we can now take that same character and do whatever we want with him you can tell it's identical it's the same guy same hat i like the fact that the hat's on there it adds the humorous effect you can take the hat off i experimented with that in the cheat sheet I've included this really cool section all about these visual language so how to describe these different camera angles uh understanding this different cinematic vocabulary can really help when you're trying to create you know this character closeup far away whatever use the right vernacular the right vocab in your prompts can really really help here especially when you're looking at lens choices we'll get to camera movements in a second when we get to the animated parts but lighting etc that can be really really helpful uh when you're making these prompts and you're positioning your character in different ways using that enhancor tool all right so now that we've not only got our character but we have our character in all the different scenes and camera angles that we're going to need um we're going to now go and figure out the voice of this character we're going to use 11 Labs to do that to pick out a good voice and to create the dialogue 11 Labs has been leading the charge when it comes to text to voice right along they've been really good at keeping up with some of the major players this is just 11labs.io and from here we're just going to go to voices click voices and I'm going to search for an old guy we got Lucifer you can audition them by clicking on them and hearing what they uh sound like and then this little plus button allows you to pick the ones you're interested in these will then appear in the my voices section so these were a handful that I thought might be good for our Baker voice i settled on this voice Joe Davis for my character and let's just generate this the classic saying Time to Make the Donuts time to make the donuts love it it gives us two different performances here that we can choose from you download those by clicking this button and that's without any instructions about how you want the dialogue to be performed this is one of the really cool advancements in 11 Labs is you can now guide them how you want them to perform that dialogue and this is what they're calling an alpha feature because it really doesn't work great all the time but it works enough to get some really cool stuff done this is the button for that advanced alpha feature i found the best way is just in uh square brackets to say we want him to do this perform this really tired you can see when you put those square brackets in it turns into an instruction and again oftentimes 11 Labs will read that out as well it's probably a bug they're working on but you can edit that out later i'll show you how to edit that out and just get the um performance that we need so let's generate this time to make the donuts listen to that that was a really good one so you can go and not only create the dialogue here but create it with the emotion that you need using that same voice over and over that's something you can't do inside of VO3 you know anytime you create a new video it's going to be a different voice once you have your voice performances complete it's time to create these videos for this we're going to use Hey Jen they have a really cool new feature called Avatar 4 that does this here inside of Hey Jen we're going to click on this avatar 4 we're going to upload our photo this is Fred Embed and then we're going to upload our audio that we created in 11 Labs i've labeled that Fred inbed as well just so I know that they go together add audio and generate and here we have Fred inbed time to make the donuts and here we have him at the shop time to make the donuts so once you've taken all of your photos and animated them with the dialogue that you've generated inside of Hey Jin you may want to create some scenes that do not include dialogue and include you know fancier actions or camera movements and that's where quite honestly Google's Flow came in really handy i was able to upload that same image of Fred and give this prompt that says "The camera moves behind this man to reveal a crowd of cheering people waiting for their donuts." And I thought this was really cool so that's definitely going to be part of the video here and so for any you know fancier scenes without dialogue Flow seems to be the place to go unfortunately inside of Flow you can't create dialogue with VO3 quite yet so I know some people were disgruntled to find that out as I was while I was experimenting but just wanted to let you know uh that doesn't work currently i'm sure they will fix that soon before we move on I just want to reiterate something that I mentioned at the beginning of the video this is not just about these cool tech tricks but this is really about fundamentally understanding how AI is changing creative work and I want you to think deeply about that as you pivot into your new AI reality i had to make my pivot from my marketing agency now into creating these videos teaching as many people as I can about how to use AI and enrich their work and their lives with it and I think everybody's going to have to go through a similar transition my whole goal with this channel is to help you position yourself to capitalize from this major transition if that's something you're interested in make sure to subscribe and consider checking out my Patreon if you want to support this channel so now that we've created all of our video snippets all of our shots it's time to edit this all together the editing tool that I like the most is Dcript so that's what I'm going to show you how to use although Cap Cut I think can be a pretty good one or if you're really professional and you know how to use Final Cut Pro you can obviously do that here in Dcript I'm just going to click create a new video project i'm going to pull in all the clips that we've made right here into our project files and then I'm gonna pull them into our workspace here in the right order so we start with him in bed then here he is at the doughnut shop and now turning to the crowd and down here in the timeline is how we start to position these so we're going to take out all that blank space here take a quick look and we'll just add a little audio here there we go and here's the final version time to make the donuts think running zebra donuts is easy time to make the donuts we're slaving over hot grease making sure every doughnut is perfect time to make the donuts our reward more sugar craze maniacs [Music] all right I got a real kick out of making that i hope you enjoyed it and I hope you can think about how you can start to get the word out about whatever you're doing with this process and with these consistent characters creating a you know Geico lizard all on your own for your own brand for your own startup whatever it may be i've got a ton in the cheat sheet all about this again getting into all those different camera movements and camera angles and cinematic vocabulary you can use in all of your prompts as you begin to animate your characters animate the camera movements and describe the lighting all these things that you need to master to really create some compelling ads or stories online so make sure to check out my Patreon you get instant access to not only that cheat sheet but over 130 others i've got some other options in there as well next I want to point you to a video all about using V3 in combination with Notebook LM that expands on some of this stuff there's a link to that i'll see you over there make your dreams come true