AI News: They're Putting AI into EVERYTHING!?

I really do need to trim my beard. Oh, hey, sorry, was just noticing how long and gray my beard was. Kidding. Anyway, let's talk about the AI news this week. There is definitely a theme forming here, and that's that they're putting AI in everything. And I know that itself isn't actually news, but in today's video I think you're going to see like, okay, maybe they're starting to put AI in too much. Let's like chill out. But I'll get to that in just a minute. Before I do, let's talk about some of the biggest AI news of the week that made headlines among the AI circles, starting with the fact that Adobe just released Firefly 3, their new AI image generation model, which is greatly improved when it comes to realism, which can be seen in images like this one and this one and this one. However, in my opinion, the images that Firefly 3 is generating are are still kind of not up to par with what we're getting out of something like midjourney. Now, it probably does realism better than Dalle 3, but not as good as Midjourney or stable diffusion 3 yet. So while I do think it's a cool big announcement from Adobe, it's essentially an announcement that Adobe is now almost as good as the other tools that are out there. However, the cooler thing about Adobe's announcements this past week is that they've added new AI features and improved upon the existing AI features that are currently in Photoshop and and you can use them right now. Now you do have to use the beta version of Photoshop. You don't know how to access the beta Photoshop and you're on Adobe's Creative Cloud. Simply go to apps on your left sidebar, come up to beta and make sure you install Photoshop Beta. A standard version of Photoshop does not have the new features yet. Once you're in the Photoshop beta, you'll get a nice little overview of what's new about it, including the fact that they've integrated Adobe Firefly 3, what we just talked about. The the fact that you can generate entire images now instead of just kind of doing inpainting and out painting, you can actually upload a reference image, which this is a really cool feature. You can see they highlighted the backpack here and then gave a reference image to replace the backpack. They've added an adjustment brush which I'm not sure is an AI feature or not, but it's in the beta right now. A generate similar option where you can highlight a certain thing in your image and have it generate another similar image. You can generate backgrounds. This is great for e commerce shots. Like the example here is they put a perfume bottle over some water and oranges and flowers and there's a new enhanced detail feature. In my opinion, this is a pretty big upgrade to Photoshop. Now, despite the fact that Firefly 3 still feels like it's trying to catch up with the other platforms and isn't quite there yet, the fact that we get all of this inside of Photoshop is what really excites me. If I create a new image here in Photoshop, a new button on the bottom left shows up or. Or a generate image I can click here. Let's do a wolf on a snow covered hill. We want it to be art or photo. Let's go photo for style. We can actually give it a reference image. I can upload my own reference image or pick from their existing reference images here. I can also pick some effects here like Bokeh and digital art and Synthwave, but I'm going to go ahead and leave that blank and let's click Generate. And while you can kind of see what I meant by it's still trying to catch up, this is what it thinks a wolf looks like. It did give me a few options though. This one's much better. That one's even better. But let's say I want a different wolf. Just select the wolf, click generative fill, let's do a gray wolf and click Generate. And now I have a few more options of wolves in our scene. I think I like that one the best. So some pretty cool new features. I don't want to spend this entire video playing with Photoshop, but there are a lot of new AI features in the beta version of Photoshop, so check that out if you haven't already. And since we're on the topic of Adobe, they also announced some new research called Video Giga gan and and this is video upscaling. So I can take a peek by moving this all the way to the right, seeing we have like a blurry low res photo and then you can see how it can upscale the video to something that looks pretty good. And there's a whole bunch of examples here, like this stack of pancakes getting slathered in, I don't know, strawberry sauce or something. It's a low quality blurry video. Now it's a high quality video. A blurry waterfall turns into a clear waterfall. Blurry man with beard. Clear man with beard. And again, plenty of examples here where you can see that it really, really up upscales blurry videos and makes them higher quality again. Now this isn't publicly available yet. I'm guessing it's going to be in something like Adobe Premiere at some point. But I will link up to this research paper below so you can check it out and learn more about the intricacies and details of what this does. If that's something that interests you. We also got a small update out of Mid Journey this week. They added the feature to give the images a random style. So so Ali Jules, my pretty much number one resource for keeping up to date with everything. Mid Journey showed off her example. So here's an original image that she generated with the prompt. A young woman with freckles wearing a yellow hat close up photo. The original is without adding the extra style reference random to it. This was the original photo and then when she added the style reference random, she got all sorts of random styles for the same concept. You can see here Alli shares the syntax that you need to use to get this. You basically type your prompt and at the end of the prompt you add dash, dash, sref random and you'll get the same image concept with a whole bunch of different styles. That's available for anybody inside of Midjourney right now. Blockade Labs, a really cool 360 degree image creation app, recently got an upgrade and it actually now generates images in 8K quality. We can actually check out some examples that are already sort of pre created created for us. If I go full screen here, you can see the kind of quality that we're getting out of Blockade Labs. Now I can actually move all around and look around in this image. Zoom in, zoom out. Really, really, really cool. Really impressive. Go ahead and bring our UI back here and we can see the prompt that generated this right here on the screen. Feel free to pause it and read the prompt if you'd like, but this new Model 3 generates these really, really cool looking 8K images. Now the other big news this week, and I did mention it in last week's news video, but I think it's worth mentioning again, is that the rabbit R1 is now available. They did a big launch party this week and let the first people pick up and get their hands on it. So the reviews are starting to come out on this. So far I've really only seen reviews from AI enthusiasts, so I kind of have to take those with a grain of salt because us AI nerds are going to think all of this stuff is cool. I'm still waiting for the, you know, MKBHD and Mr. Who's the Boss. Sort of reviews from people that kind of operate outside of the AI sphere. I do have one of these units coming to me, but I think I'm in like the fourth release of it or something like that, so I probably won't have my hands on one until like June or something. But when I get mine, I'll definitely share my thoughts on it. But if you aren't familiar, it's basically a little handheld device that uses voice for all of the commands. You press a little button on the side. It has Internet access, it has large language model access. You can ask it anything, it will answer. It's fairly quick, a lot quicker than the Humane Pin. It can also generate images directly within mid journey. It can order Uber Eats for you, it can order a ride for you, and you can take audio notes with it that get saved to their cloud and then those notes are summarized so that you can go and read them later. And it also controls Spotify for you. And for the most part those are like the only features it's really rolled out with so far. There's a few more that I'm missing, but again, I will follow up about this after I've seen a few more reviews from people that aren't giant AI nerds like me now. I personally think everybody should have a website. Whether it be to blog on or to share your portfolio or to sell products, having a website is really beneficial to pretty much everybody. And in this day and age with AI it couldn't be easier to create a website. And because I'm an AI nerd and I love building websites, I partnered with Hostinger for this video. With their AI website builder, anyone can have a website online within minutes. If you head over to hostinger.com Matt Wolf, click on Claim Deal and select the Business Website Builder. You're going to want to use this one because you'll notice all of the suite AI tools are on the business plan. So let's go ahead and add this to Cart and the longer the plan you go with, the more money you save. For this demo, I'm going to go ahead and use the 48 month plan so I don't have to think about my hosting again for four years. Hostinger already gives you a huge discount on hosting, but if you click this little button for a coupon code, enter the coupon Matt Wolf, click Apply, you save even more money. Once we're logged in, we get the option to start creating a website with AI. So let's go ahead and start Creating Just for this example, let's call the brand name Matt Wolf's Blog. Website type is blog for the description. It actually lets me use the microphone and just describe what I want the site to be. The personal blog of Matt Wolf, where he shares tips, tricks and tutorials on how to effectively use AI in your business. We have some personalization settings. We've got Surprise me, minimal, elegant or bold. I'm a minimal type of guy. So let's go ahead and do minimal. And then for colors, let's just let it surprise us. I'll click Create Website and then I'll just kind of sit back and wait. About 30 seconds later, I've got a blog online and ready to go. It's already got the design for me. Let's click into one of these blog posts. We can see the design of the blog post. Obviously it starts off with some filler text and some filler images, but that's easy to swap out once you start getting into creating your blog. So if you need to get a website online and you want it to be inexpensive and you just want AI to do it for you, Hostinger is the place to go. Again, head on over to hostinger.com Matt Wolf. Use the coupon code Matt Wolf for an additional discount. And thank you so much to Hostinger for sponsoring this video. Pretty much all of the big AI companies this week came together to essentially commit to prevent the use of AI in sexual harm against children. And this is really one of the first times I've seen all of these companies kind of all come together and agree on something, right? You've got OpenAI, Amazon, Anthropic, Civit AI, Google, Meta, Metaphysic, Microsoft, Mistral AI and Stability AI all coming together. And I imagine there's not many AI companies out there that are going to sort of disagree with these policies, but basically the idea is they're all working to develop and build these AI models and proactively address child safety risks. While we're on the topic of OpenAI, it looks like Jensen Huang, the CEO of Nvidia, and delivered the first DGX H200 to OpenAI this week. Which means, to the best of my understanding, that outside of Nvidia themselves, OpenAI now has the most powerful GPU on the planet as part of their system. Now. Now, at the beginning of this video, I mentioned that they are starting to put AI into everything. And for the rest of this video, you're going to start to see what I mean. We got news this week that the international Olympic Committee launched in Olympic AI Agenda I think it's sort of best summarized right here. AI can help to identify athletes and talent in every corner of the world. AI can provide more athletes with access to personalized training methods, superior sports equipment, and more individualized programs to stay fit and healthy. Beyond sporting performance, AI can revolutionize judging and refereeing, thereby strengthening fairness in sport. AI can improve safeguarding in sport and AI will make organizing sporting events extremely efficient, transform sports broadcasting, and make the spectator experience much more individualized and immersive. So it does sound like in the upcoming Olympics, AI is going to be leveraged to help in all of those areas. I just mentioned the people over at TED put out this video generated by SORA this week showing what TED will look like in 40 years. And I'm going to go ahead and mute the audio on this one because I don't know if it's copyright or not. Also not going to play the whole thing because it is a minute and a half and I don't really want this to take up that long of the video. But again, I will link it up below so you can watch this entire video. If I fast forward all the way to the end, you can see that it eventually zooms in on the TED logo. And the TED logo is literally the only thing in that video that was not generated with AI. But again, if you want to watch that whole thing, it will be linked up below. The Ray Ban Meta Smart glasses got an upgrade this week as well. They now have the ability to see what you're looking at and connect to the large language model to talk about what you're looking at. There's also a new feature to broadcast what you're looking at over WhatsApp or Messenger. Now again, I'm not going to play the audio on this video because I don't know the copyright status of the background music. But we can see she says, hey Meta, play some music. It starts playing music. We can see he is using the video calling feature here and the person he's talking to can see exactly what he's looking at. That's a pretty cool feature. I haven't tested that out yet, but definitely plan to soon. Again, those video calls, I believe only work over WhatsApp and Messenger and not just directly through like FaceTime on your phone or something. But still pretty cool. In fact, let's try this real quick. I'm gonna go backwards hat for this. I don't go backwards hat. It catches the bill of my hat in the sunglasses. Hey Meta, what am I looking at right now, it said you are looking at a desk with various electronic devices, including monitors and speakers. I don't know if you're able to hear that through the glasses or not, but I mean, that's what I'm looking at. This week, during the Tesla earning calls, Elon Musk claims that Tesla is not a car manufacturer, but that we should see Tesla as an AI company. He said we should be thought of as an AI robotics company. If you value Tesla as just an auto company, it's just the wrong framework. If you ask the wrong question, then the right answer is impossible. If somebody doesn't believe that Tesla is going to solve autonomy, I think they should not be an investor in the company. And we will, and we are. And this image has been circulating as well of the sort of Tesla ecosystem. And at the very top of the image they have humanoid robots and AI compute. Now I don't know if this actually represents the hierarchy of how Tesla sees the various elements of their business, but it does feel kind of telling that humanoid robots and AI computers are sort of the very top thing of the graphic and vehicles and charging and batteries and all that kind of stuff are further down. Maybe I'm reading too much into this, but the fact that Elon wants to be considered an AI robotics company and humanoid robots are sort of the top of the image. I don't know. Just putting two and two together here. And since we are talking about Elon and AI, this article came out literally as I was pressing the record button. Remember I do record these on Thursdays. And it said Musk's Xai is close to raising $6 billion from Sequoia and others. This values the company at $18 billion. Now we're talking about Xai. I'm not talking about Tesla or X or any of these other companies that Elon owns. I'm talking about the one year old Xai is now valued at $18 billion. I guess that's assuming he actually can raise this capital. Now the assumption is that this money is going to go to more computer so that he can train bigger, more powerful models even faster. But the article here itself doesn't actually say what the money is going to be used for, but I think that's probably a fairly safe assumption. Speaking of AI being everywhere, in Los Angeles, the Metro is installing AI powered cameras on buses to issue tickets to illegally parked vehicles. Now I do understand why they're doing this. They're trying to give parking tickets to vehicle owners parked in bus lanes. And if the vehicles park in bus lanes. It makes it harder for people with disabilities to get in, on and off the buses. However, I really have a worry that people aren't going to quite see it that way. They're going to see putting AI cameras on buses as a massive overstep of government and one step closer to big brother from 1984. It's just not something I see going over super well among the general public. This week, Microsoft launched Phi 3, which is a small language model. And this is something that we're seeing more and more of these companies are releasing smaller models because they want them to be able to process the large language models directly on mobile phones. I think the move is to try to get a lot of the large language model inference out of the cloud and onto your mobile device itself. We saw what happened with the humane AI pin where somebody would ask it a question and it takes 10 seconds to respond. That's because you're asking the question. Is sending the question to the cloud, running the inference on the large language model, sending the response back, and then speaking it to you? Well, if we can get smaller and smaller models, well, they can operate on devices like a phone and really, really, really cut down that delay. So when you do ask a question, you're getting your response within two or three seconds instead of 10 seconds, which does feel like a long time. This model is a 3.8 billion parameter model. It is available over on Microsoft Azure, Hugging Face and Olama. If you do want to play around with the model itself, we see here that compared to their larger counterparts, small AI models are often cheaper to run and perform better on personal devices like phones and laptops. And for simply looking up, like, quick bits of information, these smaller models work just fine. In some random AI news here, the rapper Drake created an AI diss track. least everybody's saying that Drake himself created it. I don't know if he's actually confirmed it. I don't know. There's some weirdness here, but he supposedly created this AI diss track and he used AI voices of Tupac and Snoop Dogg in it. And now the estate of Tupac is threatening to sue Drake if he doesn't take down the song and remove all traces of it, essentially. Now, the really interesting thing about this story is Drake has been one of those people who a lot of others have made AI songs, and he's kind of talked about how he's not a fan of this AI music generation yet. The rumors are now that Drake is actually making his own AI versions of songs. To create these diss tracks. Maybe just a touch of hypocrisy here, but I don't know, let's move on. This week the company Synthesia showed off their latest text to AI avatar, or whatever you want to call it. They call this new model Express one. What's different about this from previous versions of Synthesia or what had is doing is that it shows a lot more emotion in the face and the body language. It uses the context of what's being said in the text to drive the motion and the facial expressions and things like that. This little 10 second clip is probably the best example. They created this AI avatar and then they give it some text and then it speaks out the text. So check it out.

I am very happy. I am so upset. I am frustrated.

And you can see in the facial expressions that when she says I'm happy, she's got a little bit more of a smile. You can see in her face that she looks happy when she's upset. You can see it in her eyebrows and her facial expressions that she looks upset. Same with frustrated. So the facial expressions kind of change based on the context of what she's saying. Now they have a handful of avatars to choose from that you can do this with. And similar to hey Gen, you can actually create your own custom avatar as well. The process is a little bit more involved. They want you to have a green screen behind you and they even give you a list of certified studios where you can go record it professionally to get the best look and you can actually play around with it for free. But if you do want to use something like a custom avatar, you do have to be on one of their paid plans. Earlier this week, I made a video all about AI robots. I highly recommend checking it out. It's a really fun video at least. I really enjoyed making it. If you want to check that video out, the thumbnail looks like this. But one of the companies I talked about in that video was Sanctuary and they just released a new humanoid robot that learns faster and cost less. Here's their video once again, without music. But it's got improved commissioning and build speed, improved material cost, improved range of motion, improved hand durability, improved tactile sensors. I mean, everything is pretty much improved on it. And I just love robots. Robots are so much fun to me. That's why I made a whole video about robots. So again, check that one out if you haven't. But Sanctuary AI made this announcement like the day after my robot video came out. Otherwise I would have talked about it in that video instead of this one. Now, this isn't specifically AI news, but I still love virtual reality, augmented reality, extended reality, whatever you want to call it. I love that whole world as well as AI. But. And Meta did something interesting this week. They actually opened their mixed reality ecosystem so that companies other than the Meta Quest can actually use their operating system. So if you've ever used a Meta Quest, you've probably seen the Meta Horizon operating system. And we can see a few other companies that create headsets are already adopting this. The Asus Republic of gamers, Lenovo and Xbox all appear to be planning on using the new Meta Horizon os. Now, I think this is actually a brilliant play from Meta because you're obviously familiar with the Apple App Store and all of the apps they have available for iPhone and iPad and all of their devices, right? Well, Meta has the same thing for virtual reality. They have what they're calling now the Meta Horizons Store with all sorts of virtual reality apps. And right now, if you want to use one of those virtual reality apps inside of the Meta Horizon Store, you've got to have one of the Meta Quest headsets. Well, now, if Meta Horizon is in a whole bunch of different headsets, that also means that the Meta Horizon Store is in a whole bunch of headsets, meaning that Meta's playing this game of if we make our operating system and our store available in all of these different headsets, we're going to make more money off of the apps being sold, just like what the Apple App Store is trying to do. So it's like an Apple App Store kind of play, but for VR headsets, I think it's really smart of Meta. And on the topic of AI and everything, Rowan Chung shared this, and I have no idea who thought this was a good idea, but this is a robot dog that they put a flamethrower on. They said that this can be used in the prevention of wildfires as well as clearing snow when there's heavy snow buildup. But I don't know, man. I don't know if it's a good idea to put flamethrowers on robots. I mean, this looks like somebody's creating the future that we all fear. Yeah. Anyway, I'm gonna go ahead and move on. That's all I got for you today. If you like videos like this and you want to stay in the loop with the latest AI news and the latest AI tools, make sure you like this video and subscribe to this channel, and I will make sure more videos like this show up in your YouTube feed. If you haven't already, check out FutureTools IO. This is the website where I curate and share all of the cool AI tools that I come across and I make them filterable and easily organized so you can find exactly the tool for your needs. I keep the AI News page up to date on a daily basis and I have a free newsletter where I will share just the coolest AI tools and the most important AI news directly to your email inbox. And if you sign up for that newsletter, I'll also share the AI Income database with you. A cool database of fun ways to make money using these various AI tools. You can find all that over@futuretools.com I owe so thank you so much for tuning in today. I really, really appreciate you. Remember, I shoot these videos on Thursdays, so more news may have come out on Friday that I missed, but if it did, it'll be in next week's news video. Make sure you're subscribed to Check that out. Thanks again to Hostinger for sponsoring this video. I appreciate you guys. Thank you again for watching this video, nerding out with me and staying in the loop with AI. I'm having a blast doing this. Hopefully you're having a blast watching it. I will see you in the next video. Thanks again. Bye bye.

There's so much AI news happening all the time. They seem to be putting it into everything. So how does someone keep up with it all? They watch Matt Wolf's YouTube tube channel subscribe to Matt Wolf subscribe to Matt Wolf subscribe to Matt Wolf and hit that like button on this.