The Best AI Tools Just Got Even More Powerful

In the last couple of weeks, pretty much every major AI platform got a big upgrade. So, I wanted to make one video to share with you the top 10 major upgrades across some of my favorite AI apps. Okay, the first one is the big obvious one took place inside of chat GPT where we got a new image generation model that is going completely viral right now and it's a huge upgrade from Dolly 3 which was the way image were generated inside of chat GPT but now this new model is extremely good at generating very accurate text. It's good at character consistency. You could upload and restyle all kinds of different images. And it's a whole lot more photorealistic than anything we've had inside of Chat GPT before. And I made a dedicated video showing the top 10 use cases inside of this one. Let me just show you one quick use case here. I'll do it right on my phone here with the Chat GPT app. So I just took a picture of this Ray-B band here. I opened this Meta Rayband, okay? And I uploaded to ChatgPT. Now all I said was turn this into an Instagram ad here. And this is what it created for me. Then I said, "Make sure you include the meta text and put this in a blurry tech background. " This is what we got next. And then I followed up with one more prompt. By the way, it found the meta logo here on its own. And this is the image I got literally after three prompts. The prompts were not at all optimized. And then I said, "Turn it into a YouTube thumbnail for an unboxing video. " And then I got this thing. And since covering AI tools since 2022, this definitely ranks in the top five, if not the top three top AI upgrades of all time. And as part of that same upgrade, we also got the ability to generate images here inside of Sora. So now inside of Sora, if you click this type option, you have image and you have video. So if you create an image and then try to turn it into a video, it still pretty much doesn't work. Okay. So, would I recommend Sora? Absolutely not. If you're using an image and you want to turn it into a video, I mean, look, I don't know what is going on here inside of Sora, but the image generation inside of Sora is really actually good. It's kind of the same image generation you get out of chat GPT, but I wanted to show you this update specifically because you can now turn those images into a video. I just wouldn't recommend Sora. Sora does have this upgrade. So that brings me to this other upgrade. This is inside of a platform called Cling AI. This is a great platform for turning any image into a video. And I'll show you one example. So I was making a little short film inside of Chat GPT in this animated style. And I wanted to bring this to life. And I said this tree is waving gently with one of its branches. Okay, let me show you. This is Cling now and this is new cling 1. 6 that recently came out. Okay, how good is that? I mean, look at the expression. This little pigeon here is a little bit messed up here, but the tree is unbelievable. And I literally did this with one prompt inside of Chad GPT. I got this. I didn't correct any of it. I could have erased this part and have it regenerated again before I brought it to Cling. Making these type of films, I think these short films for YouTube, I think it's really close to being a reality. The next update is not a major update, but it's inside of one of my favorite AI apps of all time that I cover all the time called Notebook LM. This is from Google where you could give it all kinds of different sources including YouTube videos, PDFs, websites, and then interact with it using AI right in the center here and ask your own sources that you created here or added here any question you want. Create study guide. But they just recently added this mind map about a couple weeks ago. So if you click on this right here, it generates this mind map. I created one before. So let me click on it. This is what it looks like. So you could explore all the topics that I have inside of my knowledge base in this specific notebook and dive a little bit deeper into each one. So, if I wanted to learn more about I was studying retail industry in 2025 and I want to see how they're doing loyalty programs, I could click on that and you could see it expands it out a little bit more and then I could have a discussion here and ask AI any question related to loyalty programs. It always gives me a reference to where it got that actual information from. A great way here to have these mind maps and explore topics. Okay, the next one is Manis AI. This is a new platform out of China that is a Gentics. So you could type in a text prompt and it will literally go do all those things for you. And this is still invite only. So you have to go to the website and you have to ask for a code and they send you that code. But this is what I did. Let me just go back to the top of this. And I asked it to create a website that focuses on AI education and AI tools. So basically what we already have, but I wanted to see what it's going to do. And literally you just give it that one prompt and it does all these

different things for you. And then if I click this, it creates all these different pages. It writes code for all these pages. It creates all these documents here. These are all the different code files. And eventually it created this link for me. And by the way, it published it to this URL on its own. I could literally send this link to someone and they could view this, right? And it created a name. Look at the modern formatting here. It decided on a single sentence prompt, by the way, with no context, with no information to create something like this. I could go to the AI education side and look at this. It's missing some images, which I could do with a follow-up prompt. I could go to the AI tools section and see the AI tools that it chose, and you could see some of the popular AI tools, which obviously these are. So, it did the research to find out what the popular AI tools are. And that was just one example from one single text prompt. But I'll show you exactly what the new update is with Manis right here. They added this whole knowledgebased section. So I could add knowledge here, name the knowledge and say Manis will use this knowledge when and then here's the content of that knowledge. So if I want something specific, I could do that with copy and paste text. And this is probably the biggest upgrade here. They actually added a plan to it. So before I had a free plan, but I only got one or two prompts a day. So if the prompt didn't get me all the way there, sometimes it will run out of space, the context window. Again, it's in beta. It's in private invite. So I'm sure a lot of that is going to get fixed. But right now, we have a $40 a month plan, which gives you about 4,000 credits, and you can run two tasks at the same time. And it looks like they have a $200 a month plan. So if you are getting useful things out of this, this might be worth the upgrade. Right now, I'm going to start with this one, test it out, see where it could get me. I do have a Manis AAI video where I show all kinds of use cases there. But I'll do a new one once I subscribe to one of these plans, and then review this, too. The next one is inside of a music generation platform, one of the best ones out there called Refusion. And I actually got to partner with them for this video. They were already on my list, so I really like when the sponsorships kind of work out where I'm already going to talk about it. So, Refusion right here has two major upgrades they recently rolled out that I wanted to show you. And if you haven't used Refusion before, it's pretty straightforward here. You click on that box, you have a prompt. So, you type in your text here. You could choose if you want lyrics or if you just want instrumental, you could choose that there with that slider. And then you could go ahead and generate the song. It's going to add it here to your library. But what makes it unique is two new updates that I wanted to show you. One of them is the ability to understand what stems are. So, with the song we generated right now, I'm going to just play a few seconds of it. Let me just go a little bit later so you could hear some of the lyrics here. Let me press play on it. Time changing. Okay, so you get what that sounds like. But this is the first AI platform that lets me swap out either the vocals or the other instruments in the song. So right here, if I press remix, it says stems, swap vocals, or swap sound. So stems, if you're not familiar with it, every song is made with multiple different tracks. So for example, vocals could be one track, but things like drums and piano, those are other tracks. And when you blend those, that makes up the entire song. But if I don't like the vocals, I could just swap the vocals. But maybe I really like the beat. I think it started the right way. It works well with the project that I'm working on. So I could just swap vocals here. So I could type my own here. I could use this ghostriter here. Make this whole song actually take place in London. It was taking place in New York City. I could swap it that way. I could generate every season brings a different crowd dissolve into a passing cloud. And a brand new upgrade they had that I think you're going to find really beneficial here is an option to use vibe as a prompt. So, let me show you exactly how that works. On the homepage, I found this song I really like, but I did like the lyrics that I ended up creating. I just want the vibe from this song instead of the one I generated. So, all I have to do is click on the song. And you could do this pretty much with all the songs here. You could upload any type of royalty-free, like if a different AI platform generated a song. And I'm going to say use vibe. And it's going to take me to the same prompt box I had before under compose. So all my lyrics are still here, right? But now it's added this section right here under sound, which is that sound as a vibe. And then you could generate directly from here that way.

Okay, that turned out awesome. I'm actually going to publish that one here on my account. And if you want to test out Refusion for yourself, I'll put a link in the description. Thanks for Refusion for sponsoring this video. Let's get to the next one. Okay, the next upgrade is Claude. And if you go to claude. ai, you'll notice it looks completely different. So, they completely overhauled what this platform looks like and they've made it kind of a lot easier to navigate. But the biggest upgrade, which came out a couple weeks ago, is the ability to have search turned on, which now makes Claude really one of my favorite AI apps, if not even more useful than chat GPT. Now, chat GPT has the image generator, it has other use cases, right? But for general use case, this was one of the reasons why I constantly had to go to chat GPT even though I like claude 3. 7 sonnis writing style more and I've trained it even with my own styles here and sometimes I use one of these default styles. A lot of times we'll start with a search pulling that relevant up-to-date information and then ask claw to rewrite what it pulled in from the web. And you can see it pulls in all the different sources here. a much better user experience. I thought for some reason Claude was not going to ever have this and they were just trying to make this more for business use case and maybe for their API and not make it very practical as far as search goes for regular users. But it's here now. It's finally here after a couple years. So I think it's worth a try and I really like this new UI too. Okay, the next upgrade is inside of Hen. Hen actually rolls out ton of new upgrades. So, for example, this new one is generate an avatar from a text prompt. So, instead of choosing from avatars, you could also clone yourself. I've made bunch of videos. This one, for example, that's not me. Here, I'll just show you actually a few seconds of this. This is not me. This is my clone. And it's going to speak a different language, too. Watch this. That's in Japanese. And you see the lip sync also was in Japanese. It literally creates a video of you speaking a different language and matches your lips. But now they have this option right here. And you could see other ones that have a little new sticker, but this one create usergenerated ads. Okay. So what you could do is right over here under this tab, user generated, you could see these look totally different. So these may be good for Tik Tok or Instagram posts or ads or things like that. So, if you click on someone and then I could create a portrait video here inside of the studio and on the left side. So, these are the different avatars. I could use my own clone, but I'm going to go to the script section. You just type in your script. I'll just leave it as this default script here. And here's that quick 5-second demo. Welcome to the new era of video creation with Hen. Simply type your script to get started. Okay, cool. We could just download this from here. Obviously, it was just a demo script, but very practical use case of this one. Now, the biggest upgrade that came out the last week of March was actually the release of Gemini 2. 5 and Gemini 2. 5 Pro, which completely beats every single large language model, including the best models from Chat GPT. It says it's best for coding and for complex prompts. And you could try it right now completely for free inside of Google AI Studio. And it's also available inside of Google Gemini if you upgraded your plan there. And I'm making a full dedicated video actually comparing some of the best models right now, which includes obviously this one. But Gemini 2. 5 Pro right now, this is the hardest exam that they've put together to test these models. It's called humanity's last exam. I've tested a couple of the questions here and pretty much every time I've tested it, none of the AI models could get it right because usually the best version of OpenAI 4. 5 right now is only at 6% accuracy for that. Deep CR1 is at 8%. But Gemini 2. 5 Pro is at 19% almost now. And you can see in pretty much all these other type of benchmark tests, it beats every single model. And they compared it even against 03 mini high. So, if you haven't used Google AI Studio before, let me show you where that is. It's currently available here. So, you could just use it down here. And if you click this dropdown, if you want to always test the latest models to see if it's worth actually upgrading maybe in the platforms where it's not free like Google Gemini, you could see all the new ones are available here. And they had a new model as well here, Gemini 2. 0 0 flash with image generation which I also was going to cover in this video where you could upload images and edit. So image editing was available there. So it says add some chocolate. He adds some chocolate to existing images. And I try

to do some stuff that you can't use chat GPT's image generation model. Like removing a background will actually recreate the person and it won't be exactly the same person. It will be a close look alike. Well, Gemini kind of messed up my eyes a little bit here and it didn't quite remove the background. I said make the background solid blue, but my eyes are getting a little bit worse, but it's definitely me. It's not a close resemblance to me. It's 100% me. Open AAI does have a filter where it just doesn't make exact copies of people by design. So, that's just so people can't do a lot of deep fakes and things like that. But, Gemini 2. 5 Pro, definitely practice with it here. really useful. And then if you like it, you could get it inside of the regular Gemini app with Gemini Advance, which is a $20 a month upgrade, but I'll cover that in full detail with all kinds of different prompts, testings, and things like that very shortly. And Grock 3 actually had a new update too inside of gro. com. So they have this new option here called deeper search. So they have something called deep search, which in other platforms is called deep research, but they have another level to that. So, for example, if I want to find out which AI companies recently raised any type of money and what their valuation is currently, instead of doing just a regular deep search, I could do the deeper research and it will go through a lot more sources here. So, it breaks down the prompt into something that is more tangible for it to go research. Then, it pulls up all these different links here. It goes through those links and it basically combines all that research into a very tangible output for you. And it took about 2 minutes and 35 seconds. So from all the different platforms that do deep research or deep search, this one even with the deeper search is one of the fastest. That's why I really like it. But some of them do actually do look into more websites. So Google Gemini has a deep research that sometimes looks in the same exact prompt like 150 websites. But I feel like after 37 website, you probably got what you need. And here is exactly what I asked for. So I wanted to see every single AI company, top AI companies here, how much they raised and the valuation. So a much more in-depth analysis here than I would get from just regular deep search or just regular search. And we also have a couple of big updates inside of our platform. One is we changed our free trial here where you could get access to all these courses that we have. We have about 20. So we changed the free trial from 3 days to now 7 days completely for free. So if you want to enroll in any of these course or you want to binge watch some of them, make sure it's a good fit for you. You could now do it for 7 days for free. And if it's a good fit, you could subscribe monthly or yearly. And I'll have a special link in the description below this video for you. And we just released a brand new course called AI powered presentations and that will teach you how to create presentations using apps like Gamma and using more advanced prompting technique inside of chat GPT to organize all your thoughts. So that is available right now that you could try completely for free with the free trial link below. Thanks so much for watching. Everything is linked below. And if you haven't watched my use case for the image gen inside of chat GPT, the new image gen, I'll link that video here. And I'll see you next time.