AI News: Amazing New Tools You Can Use NOW!

This was absolutely one of the most fun weeks in AI. We got all sorts of creative AI tools from Dream Machine to sudo updates to stable diffusion 3. Apple revealed a whole bunch of announcements, a lot of fun stuff. And the coolest part is most of the stuff I'm going to talk about we can use today. So let's go ahead and jump right into it. Starting with Luma AI and their release of Dream Machine. This is their competitor to Sora and Bayo and Cling and Pika and Runway and all of the AI video tools. Now, a lot of people have claimed that this video generator is on par with what they've seen from Sora. Personally, I don't quite think it's there yet in most scenarios, but there are some scenarios where I think it did really, really well. So just scrolling through my account here, you can see I've played around with it quite a bit, generated a handful of videos already. The first day that I got into this platform it was taking forever. In fact, if you look, it says generations take 120 seconds. However, due to extremely high demand, your request will be queued. Well, the first request that I made all the way down here at the bottom took seven hours in queue. It only took 120 seconds to generate, but I had to wait seven hours for it to start on that 120 seconds. Now they seem to have scaled up enough to eliminate that huge wait time, but I can tell you it was a bit frustrating to enter a prompt like a monkey on roller skates, wait seven hours and get this as my output. Yeah, that was a bit frustrating. Even more frustrating than that was waiting seven hours to get this little error triangle on it. And when you click into the video, generation failed. Please try again. That's what happened when I entered one of my prompts, waited seven hours and then got an error message. Not fun. But thanks to some guidance from the AI community over on Twitter, I kind of figured out the proper way to use this. Now, truth being told, it's not very good at text to video. Here's my video of a wolf howling at the moon. The wolf looks like it's three legged, the moon's behind it, it doesn't appear to be howling. Here's another generation of the wolf howling at the moon. And the wolf and the moon aren't even in the same shot together. This was one of the auto generated suggested prompts of a teddy bear in Sung is playing electric guitar. For some reason he was like a butterfly or something that all of a sudden morphed into a bear in the water. Here's another monkey on roller skates, which isn't horrible, but I mean, those definitely don't really look like roller skates either. Then I wanted to test some of the prompts that we did see from Sora. So we have this video from Sora that really impressed people. A stylish woman walks down a Tokyo street, etc. Etc. I copied this identical prompt, plugged it into Luma to see what Luma would present. And well, here's version one, which looks kind of funky. She also looks like she's missing an arm. But then later on the arm appears. And here was my second attempt, which is even more scary than the first attempt. I also wanted to try this prompt, a movie trailer featuring the adventure of 30 year old spaceman. Etc. So I copied this exact prompt, plugged it into Luma, and it's actually not horrible. The big problem with this is Luma only generates five second videos. So there just wasn't enough time to generate anything extra than this. But out of all the text to video I did, this one was one of the more impressive ones. It's also really good at these like flyover shots. I did a time lapse of the northern lights. That one came out pretty good. An aerial shot of a lighthouse standing tall on a rocky cliff. That one's pretty decent. Another lighthouse shot, which is pretty decent. And I tried again with a wolf howling at the moon and really just got a wolf looking around. And that was it. That was before I learned that image to video is where this really shines. Now my first image to video got a little weird. I don't know what's going on with this video, but I put in a thumbnail of mine and it turned me into a news team or something. I kind of love it. Here's another one. An explosion of color happens behind a man's head. This was another YouTube thumbnail concept that actually looks pretty legit. Once again, this is now image to video. Here's a colorful futuristic city, which I thought came up pretty good. Another image to video. A pixelated video game. Wolf howls at the moon. I actually really like that one. These are all image to video. Now up here, there's another image to video. I generated another one of like a cabin in the woods that actually is super realistic. This one I thought was actually one of the more realistic videos I've seen. Now he is spraying the water in the wrong direction and his helmet looks like a giant plastic bucket over his head. But that's not Luma's fault. This was an image generated with midjourney. Midjourney generated the bucket head and the water sprang in the wrong direction. Luma actually did a really good job with the animation of it all. Here's a few other quick ones. Walking towards a window with a city skyline out in the distance. Looks pretty good inside of a mansion with large windows and an ocean view. Again, these are all image to video. At this point, I learned that image to video is what really works with Luma. Right now. We've got a mad scientist mixes up colorful chemicals in a lab. It looks pretty cool, but you'll notice the face sort of morphs as the camera moves. And it looks like a different person in the beginning than in the end of the video. A man wearing robes with a hood and looking in a crystal ball. Definitely got some funky fingers going on. And once again, you can see the face sort of morph as it moves. It doesn't look like the same character from beginning to end, but what's going on with those fingers? A man shaking hands with a robot. Pretty good. The face morphs. Bright, colorful fantasy world also looks pretty good. Pretty cartoony. And then James Bond looks into the camera and then quickly looks away. He doesn't ever actually look away. And it looks like Daniel Craig in the start, but if you watch as it moves further and further away, it looks less and less like him. It's got some of those little issues there. Now, one thing to keep in mind about Luma Labs here is that right now, while it's in Research Preview, you get 30 generations for free per month. That's probably not going to last. And then after that, the math works out to roughly 25 cents per video that you generate. So if you want to generate about 100 videos, it's going to cost you about 25 bucks. But as of right now, you can use it up to 30 times a month for free. So get in there and play around with it before the free version goes away, because I can't imagine that's going to last forever. Tika, one of the competitors to this Dream Machine tool, also made an updated image to video model and showed off what it's capable of. Now people are asking in the thread what's different about the model. They actually never came back and responded about what's been updated or different. I guess it's just a little bit better. I don't know. I think Dream Machine might be doing a little bit better job right now of doing the image to video. This looks like it's sort of morphing a little too much as it changes where the other one looks a little bit more consistent. But if you want another alternative to try, Pika Labs now apparently has an improved image to video model. We got a lot of news in the AI image generation world as well, including finally we have access to stable diffusion 3 medium stability. AI teased stable diffusion 3 all the way back in February, saying it would be coming soon and the weights would be made available. Well, this week they finally made them available. So you can actually use stable diffusion 3. We could see some examples of images that it generated. It seems to be pretty good at getting text into the image now and the weights are actually available to download over on Hugging Face. So if you want to actually download the weights and run this model locally on your computer or on a separate cloud computer, you can, you can. And it appears that a handful of people have also made them available inside of Hugging Face spaces. So if you want to generate images for free right now on Hugging Face, you probably got to wait a little bit. But you can do that. Let's go ahead and click into the stable diffusion 3 medium here on Hugging Face. I will make sure all of this is linked up below. It might be in a Google sheet because I'm starting to put so many links down below that I'm running out of space in the description. So I'll probably link you up to a Google sheet with all the links from this video. But here's where you can test it. Let's use my old standby. A wolf howling at the moon. Let's see what we've got under advanced settings. I'm pretty much just going to leave this all by default here and see what it comes up with. We'll click run. Not too bad. As of the recording, it's claiming it's going to take about 16 seconds. And sure enough, it was about 16 seconds. My wolf has two chins or two mouths. It looks like this is a wolf nose and this is a wolf nose and this is the wolf mouth. So it didn't do a great job. But Stable Diffusion is also kind of known for sort of requiring advanced prompts to get it to work. Well, here's what I get with the prompt. A monkey on roller skates, which isn't half bad. I mean, each skate only has two wheels and this wheel is kind of janky. But this time it generated in less than 10 seconds. So it's actually moving pretty fast. As of the time I'm recording this video, I'm just kind of underwhelmed by the images that it generates right now. If you give it much more detailed prompts, it does tend to work better. Here's an astronaut in a jungle cold color palette, muted colors, detailed 8K and a pretty solid looking image. A lot better than the images I was getting when I was just putting sort of short, brief prompts. Here's an astronaut riding a green horse. And here's a delicious ceviche cheesecake slice, which actually looks pretty good until you start looking at a fork or whatever. This is supposed to be on the side, so it's capable of making really, really good images. I just feel like you still need to be a prompt engineer to get real good images out of stable diffusion 3. Where a lot of the other AI image generation models that are out there, the Dall E3S midjourney, custom Leonardo models, things like that, you don't have to engineer your prompt so hard. You could put more normal text and actually get good images out of it. Here's one of my favorite generations I came across on X from Max Wolf here. Taking a look at people testing out stable diffusion 3. And to be honest, this goes hard. Yeah, just try not to think about this one too hard. Somebody else followed up with this image. It's impressively consistent. Yeah. Speaking of Leonardo, they rolled out their own new custom model. Now, full disclosure, I am an advisor for Leonardo, but they have zero control over what I say. And if something is funky about it, I'm going to point it out. Being an advisor to them does not impact what I actually say about them. But they did just release this new Leonardo Phoenix model, which is their own foundational model. A lot of the previous models that Leonardo was using were stable diffusion models. This one was trained specifically for them from the ground up. It is not stable diffusion. The major features of this new model is it's got enhanced prompt adherence. It can actually do coherent text and images, superior image quality. So it's actually gotten better than previous models, more creative control. But right now, some of the things like image guidance and elements and the photo real versions aren't available yet. They're still working on some of those extra features. So let's go ahead and click. Let's go. We'll do an image generation here. Up under preset in the top left, I'm going to select Leonardo Phoenix. We'll leave prompt enhance on auto preset. Let's go cinematic. For generation mode. I'm going to go quality over fast. I like to generate in 16, 9 and let's try a simple one. A wolf howling at the moon. And I can already see how it improved my prompt. A wolf in a cinematic digital photograph situated atop a moss covered granite boulder in a misty forest lit clearing, silhouetted again. So let's see what it comes up with. So here's the images that we got out of it. Here's the first one, here's the second one, the third one, and the fourth one. And I don't know about you, but these are quite a bit more impressive than what I just saw out of stable diffusion 3 medium. Let's take a closer look at the entire prompt that it gave us. I'm not going to read the whole thing, but if you want to go ahead and pause the video here, you can take a peek at the more detailed prompt that it automatically generated for us. Let's try adding text to an image. A penguin holding up a sign that says Mr. Eflow. We can see the advanced prompt here that it generated for us. And there we have it. Now, these first two generated a couple extra wings on the penguin, but the final two here look pretty good. The words are all right. Definitely a penguin holding a sign that says Mr. Eflow. Here's another one. Spelled it correctly. In fact, it spelled it correctly on every single image. So no issue with the actual text. These ones, it just made kind of a funky penguin image. But in the same week we got SDXL and we got Leonardo Phoenix. If you were to ask me, the Leonardo Phoenix feels like a better model than sdxl, but I highly recommend playing with both and seeing which one works best for your needs. We also got some news out of Midjourney this week. They just released a new feature called Model Personalization. Now, the way Model Personalization works is if you've used Mid Journey for a while and you've actually voted on images, well, then Mid Journey kind of knows what images you like because you've done a lot of the voting. So you get essentially your own personalization code. And if you use that personalization code, then it tries to make images as close to the images you've shown to like as possible. Now, if you want to train your own version of Midjourney on what you like, here's how you can do it. You can come to the midjourney website, midjourney.com, assuming you're a member, you can come over, click on Tasks and then click on Rank Images. Now, what it's going to do is it's going to show you two images and between the two you click on, the one that you like better, you can see I've ranked 305 images already. And they say that for the best results, you need to have ranked at least 200 images. But you'll just go through, click through the images that you like best between the two. And over time, it learns your preferences and knows what kind of images you like. Now, when you're ready to create an image, you can click on the Create tab here, enter a prompt, let's do, I don't know, a wolf howling at the moon. We'll turn on our little preferences here and turn Personalize on. Make sure we're on the latest version here. And we'll go ahead and hit enter to run our prompt. Now, just for a quick comparison, I'm going to go ahead and turn Personalize on off and run a version with that turned off. Now, if we look at all these images that it generated, this top row here are with personalization turned off. And you can see we got some, like, illustration styles. And I don't know, this kind of looks like an oil painting style. And then down here we can see the ones with personalized turned on. And these ones down here are much more tailored to what it knows that I like because of all of the ranking that I've done. Now here's another tool that I just came across. I actually don't know how new this is. This could be months, weeks, days old. I have no clue. I just came across it today before recording this video and thought it was pretty cool. This is from Google and it's available inside of Google Labs over at Labs Google, and it's called Gen Type, and it will actually do letters in whatever style that you ask for it to do those letters in. Now, Adobe Firefly has something similar, but this one's available for free to play around with right now over on Google Labs. So it asks, what is your Alphabet made of? I'll do colorful electronic circuitry. And click generate. And we can start to see it generate each letter one by one in this electronic circuitry style. And there we go. Now we have the entire Alphabet in electronic circuitry. Now I can come up here and type anything like Mr. Eflow and you can see it typed it out in that circuitry style. I can make it larger so we can get a better look at it. And then I can save it as a PNG file. And just like that, you could see the image that it generated with all of the letters I just created. Again, this could be older than I realize. I Just thought it was really cool and wanted to share it with you. Now, a couple weeks ago, Suno teased a feature where you can make songs by, you know, tapping on something or adding whatever sound effects you want. And then you'd be able to generate a whole song from that. Well, we finally got access to this feature. If you're on a Suno Premium plan, you can use it right now. So I head on over to Suno AI click on Create up here and then we've got this option to upload audio. So make sure you're on 3.5 here. Click on Upload Audio and you don't actually have to upload the audio. You can actually just click this audio button and it will allow you to record directly into your computer. One sec. Let me. Let me do something real quick. Let's go ahead and try to put like a real simple guitar piece in here. I'll go ahead and click Upload one file. Click continue. And now we can see that it put our recording in here and I can actually play it. But if I want Suno to do its thing with it, I need to click on Extend. So let's go ahead and click Extend here and you can choose where it extends from. Now, I basically played the same little arpeggio two times in a row. I don't really need them both. So I can have it extend from seven seconds after I played through the first time because the second time I just repeated myself. So let's go ahead and extend from 7. Let's put acoustic options, acoustic guitar. Let's throw a beat in there and see what it does. I'll let it just make random lyrics and we'll click Extend. Here's our first version. Not too bad. You could hear the chords that I was playing underneath. I was hoping it would add like a little bit more of a beat to it. It basically just took the little riff I was playing and put lyrics over it. But not too bad. Also made it sound a lot better than me actually playing it. So let's try this second one here.

Stars are singing love full of world beyond my closed eyes but you get the idea.

It's pretty cool, especially if you can start with your own music and then have it extend and improve upon the music you're trying to create. Here's another little guitar thing that I uploaded the other day, And here's what it turned it into.

Let's take flight in the kitchen you and me dance around and sing Weepy toast is popping Butters hot.

I mean, it's Something I'd probably listen to, like got some Death Cab for Cutie vibes or something. If I'm being totally honest, out of all of the AI tools that I play with, Sudo is probably the one that I spend the most time playing with lately. Like, I just have so much fun with this tool. Anyway, I'm going to move on. In other AI art news, Adobe's overhauling their Terms of Service to say that it's not going to train AI on its customers works. People freaked out last week about a terms of Service update that implied that they would train on anything you made in Adobe Cloud. So if you made anything in Photoshop, then Adobe has the right to train on that. I think it was probably just poor wording, but they're now backtracking and saying that's not what we're saying at all. We're not actually going to train on your work. And now they're rewriting those terms of service. Now the other massive news this week was the big Apple WWDC event that happened on Monday where they essentially announced they're putting AI and everything. They're not putting artificial intelligence and everything, they're putting Apple intelligence into everything. Now, I don't want to go into too much detail about the Apple Keynote, because I did make a fully dedicated video called Apple's Huge AI Unveiling that I put out on Monday, the same day as the keynote. And well, it's 22 minutes, so I can't really break down all the same stuff in this video because that would add 22 minutes ish to the length of this video. But here's the ultra fast rundown. Apple's building their own AI and it's going to be in everything. It's going to be on iOS, it's going to be on iPad, it's going to be on your Mac. No announcements about AI in Apple Vision Pro, but pretty much all of the normal consumer devices are getting AI. You're going to be able to do things like proofread and rewrite and summarize and get the key points from pretty much any text, including things like email. They're adding AI vision capabilities into both the Notes and the calculator apps so you can take notes inside of the Notes app. It'll summarize it. It will actually be able to read any sort of images that are on the screen. If you write any math on the screen, it will actually be able to solve those math problems for you. That was pretty cool. Your email inbox, if you're using like an iPhone is going to get completely overhauled. It's going to make the messages that it feels are priority messages rise to the top. And it's also going to give you a quick summary. Typically, this first line of an email is like the first piece of a sentence from the email. But now it's actually summarizing what the email is about into one sentence using AI. So invited to is a Kaya Soft opening by Florence tonight. Check in for flight to New York. Order lunch and coffee for Anita's 12pm meeting. It sort of summarized all this stuff instead of just showing you the first sentence, making a quick glance at your email much more useful. It's also going to summarize emails. You can see the emails down below and up at the top we get a quick summary. It will also generate smart replies. They selected the text of the message and then it asks questions. Will you be joining? Will you be driving or taking an Uber? You answer these questions and it writes a reply with the responses in it. They're also doing this with your notifications, which is going to be awesome for text messages. You can see the example here. This person has a text message and it says, invite for dinner with Sandy and her boyfriend. That's not the message of the text. It's summarizing the text message so you don't have to open it and read it right this moment. You can just quickly glance at your phone and see what the text is about. And that seems to work with all notifications. Here's one from Instacart. Here's one from Slack. Here's another example where it determined this one was possibly important. It says, Chloe needs to be picked up at 2:30pm this, to me is probably going to be one of the most handy features to just add a glance, get a summary of any of your notifications. They're rolling out a new image generation feature that they call Image Playground. This is Apple's model that generates images just like Meta's Emu, just like Adobe's Firefly, Google's Imagine. Well, Apple has Image Playground and it doesn't do realistic. It only allows you to do animations, illustrations and sketches. So they're obviously trying to avoid their tools being used for things like deep fakes. I also really like this UI where it's got these images, Cat Party Chef, and then it makes a Cat Party Chef, right? So it's kind of got this different UI where you can see the various contexts that it's going to try to blend into a single image. It's also going to be able to know what people look like in your contacts. So if you've got a contact and it actually can see their image, you'll be able to generate images of those people. But again, only in cartoony styles, no like deep fake stuff. Here's another example where you could see the image here and then a new image was generated. They have a new Gen Moji feature where you can create your own emojis with AI. And it appears that after you create a genmoji, it saves it inside of your normal emojis so that you can reuse that same one again in the future if you want. Those can also be used as tap backs. So instead of just thumbs up, thumbs down, haha, exclamation, etc. You can now use any emoji including the Gen Mojis that you created. There's some new photo editing features. You can remove people in the background, clean up your photos a little bit. New contextual awareness of photos so you can do more semantic search across all of your photos, find exactly what you're looking for. Will also create memories. You can give it a prompt like everything we ate in Japan Leo trick or treating with a spooky vibe last summer in our garden. And it will generate a custom photo storyline for you based on your prompt series. Getting a bunch of updates. You're going to be able to type directly to series. You don't actually have to talk to it anymore if you don't want. You can type what you want from it. When you do use Siri, it's also going to try to run the AI directly on device and it's going to know the context of your apps and your emails and your text messages and anything you want to give it access to it will know that stuff. So you can say, hey, what was I supposed to do today? And it'll look at your calendar, your email, your messages, and across all of the context of communications. Figure out if there was anything you were supposed to remember to do today that it's going to be able to do directly on device. There's also a Apple cloud which is supposedly completely secure, so they're not storing any identifiable information on you. In fact, from what I understand, all the information is encrypted, so they don't even know what your information is when it sends to the cloud. And then a third option is they're partnering with OpenAI. So if you ask Siri a question that OpenAI is better suited to answer, it will actually ask you, do you want us to send this question to ChatGPT for you. In fact, here's a quick screenshot of what that looks like. You ask Syria question, do you want me to use ChatGPT to do that? You could either cancel or use ChatGPT, but really there are so many things that they announced in the AI world from that keynote. So again, check out the video if you haven't already, about the Apple event, because that breaks down pretty much all of the AI stuff that they talked about. I kind of skipped over a bunch right now. Now, during this entire hour and 45 minute keynote from Apple, they spent roughly two minutes of it talking about how they partnered up with OpenAI to integrate ChatGPT. And a lot of people were curious, well, how did that deal look? Is ChatGPT paying Apple to put it in there? Is Apple paying ChatGPT to use the tech? Well, it turns out that nobody's paying anybody here. Apple is getting access to the tech. OpenAI is getting more exposure for the tech and that's essentially the exchange that they're doing. There's really no cash exchanging hands. And well, this Apple event that happened on Monday, it went over quite well among investors. In fact, Apple took over the top spot on the world's largest companies. They're now the number one company by market cap, once again passing Microsoft. I really think we're going to see these three companies here kind of constantly flip flop with each other. I think Nvidia is going to rise again, I think Microsoft's going to rise again and I think these three companies are all going to kind of bounce around being the top companies for a moment. It's going to be interesting to watch. After Apple's keynote, Elon Musk tweeted, if Apple integrates OpenAI at the OS level, then Apple devices will be banned at my companies. This is an unacceptable security violation. Just to clarify, OpenAI is not operating at the OS level. In fact, they made it very clear that when you ask Siri a question, it's going to try to use the on device or Apple cloud service first and then try to use ChatGPT if ChatGPT is going to give the better answer. But even when it uses ChatGPT, it's going to anonymize any sort of information. ChatGPT is not even going to get your IP address when you asked the question. Elon also went on to tweet Apple using the words protect your privacy while handing your data over to a third party AI that they don't understand and can't themselves Create is not protecting privacy at all. Obviously Elon Musk has got some beef with OpenAI lately. If I had to guess, he probably wishes it was GROK that was getting integrated and not OpenAI's tech. But this got an added context memo to it that says assumption made here misrepresents what was actually announced. Apple Intelligence is Apple's own creation and ChatGPT access is entirely separate and controlled by the user. You control when ChatGPT is used and will be asked before any of your information is shared. Remember this screenshot, do you want me to use ChatGPT? To do that you have to actually allow it to use ChatGPT. Marques responded. I asked Tim Cook, he said Apple at no point sends any of your data to OpenAI without explicit user permission. And since we're talking about Elon anyway and elon's beef with OpenAI and if you remember a couple months ago Elon sued OpenAI. He claimed that there was a breach of contract and fiduciary duty, claiming that OpenAI team had set out to develop artificial general intelligence for the benefit of humanity. But then the project was transformed into a for profit entity that's largely controlled by principal shareholder Microsoft. Well, this week Elon decided to drop the case. Now back when this was originally announced, Sam Altman and the team showed a whole bunch of emails back and forth between Elon and the OpenAI team showing that Elon was on board with all of these plans back when they happened. And I think Elon realized eventually he just doesn't have a case here. In other OpenAI news here, OpenAI brought on two new C level executives. They brought on Sarah Fryer as the cfo. She was previously the CEO of Nextdoor and previously the CFO of Square. And they also brought in Kevin Weil. Kevin Weil as the CP pio. He was recently the President of Product and Business at Planet Labs. He also worked on Libra Cryptocurrency, worked at Facebook, worked at Instagram and worked at Twitter. I don't really know too much about these two people, but they both seem to have fairly impressive resumes. Microsoft is making a change to their Copilot Pro. If you remember, OpenAI rolled out GPTs in the GPT Store and allowed people to make custom GPTs. Well then Microsoft went okay, we're going to add that same feature into Copilot. You can now create your own custom GPTs using Copilot outside of OpenAI's ChatGPT. But it's basically the same thing just in Copilot instead of chatgpt. Well, to be honest, they still haven't really totally caught on inside of ChatGPT, and they definitely never really caught on over at Copilot. So Microsoft is just getting rid of them. I'm guessing it just wasn't getting used enough We've got a new best open source model this week, Quin 2. According to all of the benchmark testing on this Quin 2, it outperforms Llama 370B in every single benchmark, as well as Mixtral 8x22B and the Quinn 1.510B that came before it. So it's actually a model with less parameters than the last Quinn model, roughly the same amount of parameters as the llama 3 model. And if you look at these scores in a lot of these benchmarks, it blows llama 3 out of the water. Like human eval 64 verse 48 in this C eval 91 vs 65.2 CMMLU 90.1 vs 67.2 Now I don't actually remember what all of these benchmark tests do. I turn to people like Matthew Berman for that kind of stuff and I'm pretty sure he did a video about Quinn 2, so check that out if you haven't already. If he did actually make one. If he didn't, I bet he will soon. But it seems to be getting really, really impressive scores, especially since Llama 3 was already really dang impressive. Now Quinn272B is beating it and everything. Photographer this week was disqualified from an AI image contest after winning with a real photo, which is interesting to me because we've seen the reverse of the news quite a bit where people sent in AI generated images to photo contests and got disqualified because they found out it was AI. This image actually won an AI contest. It was generated by Miles Astray, and it turns out it's a real image. It's not AI generated. And he was trying to prove a point that human creativity is still valuable and people still tend to prefer that over AI generated art. And that's what I got for you today. I hope you enjoyed this video. I hope you feel much more looped in on what's going on in the world of AI. I really enjoyed making this video because I always love it when I get to actually show off tools that we can use now and show you results from what those tools can create. Really, really fun week in AI. A lot of cool stuff for creators that love to make video, that love to make art, that love to make music. So much cool stuff. If you like staying in the loop with all this stuff, make sure you check out Future Tools IO I curate all of the cool AI tools that I come across. It's updated almost daily. I also keep the AI News page up to date, so any news that I miss in these news videos, they get shared on this news page and I record these on Thursdays, release them on Fridays. So there is probably even some AI news that I missed in this video, but it will be listed over on this AI News page so that you can be fully looped in. Also, make sure you join the free newsletter. I'll send you just the most important news and just the absolute coolest tools that I come across. It's totally free to join. When you do join, you get access to the AI Income Database, which is a database of cool ways to make money using these various AI tools. I'm also constantly doing giveaways and freebies and raffles and things like that to people that are subscribed to both the YouTube channel and the newsletter. So make sure you subscribe to the newsletter again totally free at FutureTools IO if you want to stay looped in on the AI news. You like videos like this. You like AI tutorials, you like learning about the latest AI tools. Make sure you like this video and subscribe to this channel and I'll make sure more stuff like this shows up in your YouTube feed. Thank you so much for tuning in. I really, really appreciate you nerding out with me. I have so much fun making these videos every week and recapping everything that I find fascinating that I think you'll find fascinating. So hopefully you come back and watch more of them. Thanks again. Really appreciate you. See you in the next one. Bye Bye.