Surprising AI Use Cases For Everyone

It's another week full of novel AI use cases. We'll be looking at everything that came out over the last seven days that you can actually use today. For example, there's a new tool that's free and with just one image of your face you can generate any scenario with you in it. There's a curious new way to brainstorm inside of ChatGPT and Pika has a brand new mode that turns vertical video into horizontal video and more. So without further ado, let's get into the first one and let's show you what's possible with AI today.

What makes this new tool special is that it uses stable diffusion and this thing allows for you to create custom images of yourself in any scenario and the kicker is this thing is free right now. Plus you don't even need the 10 images that were needed with Dreambooth when you were training a custom stable diffusion up until now. That was the workflow that you needed if you wanted to get results like this. So there's two sites where you can use this right now. First of all, it's hosted on Hugging Face or this Replicate page. I'm gonna be using both. So for Replicate, we could just click this and take an image right here. I took a bit of a weird expression because we're gonna be putting this into scenes. So neutral works, but it also transfers your expression. It's gonna alter the prompt, a photo of an impressed performer in a circus. And yeah, let's just keep the photographic style, let's generate. We're gonna do the same thing in the second model. In this case, I'm gonna take the screenshot of a YouTube thumbnail where I look surprised and we're gonna be transferring this face to Indiana Jones collecting a golden statue and I'm just gonna add this EMG at the end because it needs that generated. In this case, we're gonna do a cinematic style, submit this. And all right, here we have an impressed performer in a circus. I'm just gonna say juggling knives. Let's see if it can do this. Yeah, so that's something. Now, you could regenerate this many times, but you can see this really resembles it. And as opposed to many of these face swapping apps where it just cuts out the face and blends it in, this is seamless. It's integrated in the model from just one picture. So if you upload multiple, it's obviously gonna get better. But this is actually really good. Look at that, Indiana Jones stealing a statue. So have fun playing around with this.

Moving on to Microsoft's Copilot Pro, and this is gonna be rather fast because we talked about this a few weeks ago. It allows you to use GPT-4 for free in different apps, but now they released their Pro version. And the main difference is the Pro version is natively integrated into their Office 365 suite, which includes Word, Excel, PowerPoint, Teams, Outlook. So that's really great for people who are using those apps already and wanna pay $20 a month to have easier access to GPT-4. That also considers the context of the app they're using. Especially for beginners, this is a real win because the prompt suggestions and the integration right there don't have to do a lot of copy-pasting. It just works. And if you're asking yourself, what's the difference to GPT-4? Well, there's not many, except that it's less capable in some ways, I guess. So it's Paul AI here on xSharedThisChart that compares the two. As you can see, GPT-4 can do everything that Copilot can, except that it doesn't live inside of the app and it doesn't give you those prompt suggestions. When I looked through all of this, what really caught my attention was this mention of the Copilot Studio that they're coming out with because they're introducing Copilot GPTs. And the naming of this is all over the place. In some places, it's called Copilots. Here, it's called Copilot GPTs. But the point is that this direction of creating custom versions that are more specialized of GPT-4 is the one not just OpenAI is gonna follow. Also, Microsoft is gonna follow that. And with a rather pricey subscription at $200 a month, you're gonna be able to create your custom GPTs. So here in the announcement post from two days ago, it says that this will be slowly rolling out. So I think they have some of them out now. I will be definitely diving deeper into this. But yeah, this seems quite promising. You're gonna have custom AIs for your company and you don't need to be enterprise with over 300 users to use this now. Powerful stuff. And if you're already using the Office Suite, you now have the option of a subscription that adds GPT-4 right into it.

Alright, moving on to an interesting one. So this is just a GPT that was created by Ethan Mollick here on IGGS. But I think this one is particularly interesting because whatever industry, everybody agrees on one thing and that chat GPT is insane for idea generation. It's a fantastic companion when it comes to generating new ideas. At the end of the day, you do want to go a little deeper than just give me five ideas for XYZ. And this GPT makes that possible. He actually uses various techniques of first questioning you and then using the various tools like internet search and the answers to the question to give you ideas that are a little more out of the box. So yes, this is for chat GPT Plus users. But as you can see, when I prompt this, it's gonna go ahead and it's gonna ask me multiple questions as to which line of reasoning do I want to follow here. And then it goes ahead and generates ideas in various forms. And in the end, it saves all that to a work document. A really simple little use case. But if you're using chat GPT for ideation, this is worth trying out.

Okay, moving on to the next one, which is a text to 3D model by Luma Labs here. And I'm not saying this one is great, but just my personal feeling is that it's better than the other ones I've seen. Meta Research had some, but Luma Labs now came out with this and you can simply try it in the web interface. It's so simple. Look, just go over to this link and we're gonna test the use case from the video right away. Okay, so the ancient Greek face. Let's see how well this actually works. Because you know how it is with previews. They're always polished and perfect and then you try it and it's not everything you hoped for. But let's have a look at this. This doesn't look bad. So you just click on it, creating interactive 3D model. And yeah, I mean, that is decent. Look, we got this from a simple text prompt. God, the future is gonna be wild when you combine this tech with XR and your personal context. You're going to be able to speak worlds into existence. Anyway, let's do one more here before we move on. Let's do something a little harder. The good old classic, a cat with a hat. I cannot deny myself this pleasure. All right, so that's actually not bad. Look at this. Again, I'm not saying it's great. I'm just saying it's better than what I've seen so far. Okay, so nowhere near photorealistic, but for a stylized world like this, it kind of works, I guess. Here you could make it high-res in a second step. And there you go. LumaLabs Genie, freely available and kind of fun to play around with.

You should try this. Next up, let's talk about this video's sponsor, which is Neira Studio. It gives you access to multiple AI tools that are powerful and they work in tandem. You can transform any text into engaging video with custom AI avatars that speak over 140 different languages. You pick one of their 20 pre-made AI presenters, customize the voice, set the motions, and watch the magic happen. So it does both video and engaging voices. And they focused on adding a layer of personality to the voices. They can even create a fully customized AI avatar just for you. All you need to do is send them a clip of you talking in front of a green screen, and then you have your own AI avatar. Once the video is done, you can export it in both horizontal and vertical formats, making it optimal for all social media platforms. Just for a second, imagine the possibilities that a multilingual AI avatar can provide. You can create marketing videos, product demos, social media posts, and more in multiple languages with just a few clicks. But Neira Studio does more than just AI avatars. They have a text-to-speech feature that's great for turning a script into a voiceover. You get quite a bit of control over the language, voice, and motions in your outputs. Here's how it sounds. This is another thing that AI can do. One of the things I personally appreciate the most about Neira Studio is how intuitive the user interface is. It's so user-friendly that you don't need any prior video editing experience to create videos in here. To try it out today, check the link in the video's description. And now let's get back to brand new AI use cases that appeared over the course of the

last week. Okay, so here's another one. And I won't be extensively testing this today because I feel like this is one of the things where it really depends on what you give it. But let's have a look at the demo. And the demo shows a new peak collapse feature, which, by the way, is paid now. At the beginning, they claimed that it's going to be free because they want to democratize video creation, but it's paid now, as to be expected, I guess. But the point is, this is a first, okay? We had this for images since a while, and Photoshop could do that quite well already. But for video, there's nothing like it. Inside of After Effects, they do have a smart function that is not AI that could fill in certain parts of the screen, and that was deemed revolutionary. But it only worked really well in, like, skies or regular surfaces. Here, it literally takes a vertical video and turns it into horizontal video by generating three times the content that the vertical video provides. This is really impressive, and what I tasked my team to do is to test this extensively. What we'll find, we'll be sharing on our Instagram. So, follow that if you care about an update on what works here and what doesn't. But my guess would be this really depends on the use case. And yeah, it did the woods well, whereas people, it could obviously not do. And I don't know about this restyling. In practice, this never works as well as in these demo videos. But yeah, this is a thing now. We have video expansion with AI. I'll have to spend some serious time on this before I give you an opinion, though.

Okay, and as our last use case of the week, we have Cassette AI, which is a brand-new music generator. And this one doesn't generate lyrics like Suno does. That's really the best in class for lyrics plus music. This just creates AI music. And what I really like is these suggestions down here, because as somebody who's not very technical with music, it's really helpful to have some of these down here. As you know, generally speaking, I'm a big fan of prompt suggestions. I think it makes these interfaces so much easier. It's a quick Google sign-in later. I can click one of the presets up here and generate a chill lo-fi hip-hop track with a relaxed beat, subtle melodic elements, and ambient soundscapes for studying and relaxation. Yeah, this is legit. So, lo-fi hip-hop is definitely one of the easier genres to do, but in other tools, this doesn't work as well. So, this is one of the good AI audio tools now, which you can play with. And like so, you can create copyright-free kind of original music, right? All right. So, if you want more AI use cases that are worth your time and actually usable today, then subscribe to this channel because I'm doing this every single Friday. Here's a playlist of all the videos if you want to explore even more tools. And now it's your turn. Let's get started. And now it's your turn to go ahead and put some of these tools to work.