OpenAI Charging $20k/mo for Genius-Level AI?!

There's been a lot of interesting developments in the world of AI this week and I'm not going to waste your time. Let's just get right into it. Starting with the fact that OpenAI is planning on releasing a slightly more expensive AI model. And by slightly more expensive I mean $20,000 per month. According to this article on the Information, OpenAI executives have told some investors it planned to sell low end agents at a cost of $2,000 per month to high income knowledge workers, mid tier agents for software development costing possibly $10,000 a month, and high end agents acting as PhD level research agents which could cost $20,000 per month. The idea being these are going to be actual agents that could essentially go and do some work for you. At least more work than you can get done on a $200 a month plan. According to this article, ChatGPT Pro isn't technically an agent the way the PhD level research agent or coding agent would be. These numbers aren't 100% confirmed. And I've seen posts over on X where people at OpenAI are still sort of testing the waters of what sort of price points people want to see. But for 2000amonth you could essentially have like a research assistant that, let's say you're going to make a video or you're writing articles, they can go off and do the research for you and come back and bring it all to you. For 10,000amonth you can have a software developer that's writing the software for you, theoretically more autonomously. And for 20,000amonth they can do PhD level work. The sort of running joke is kind of becoming. Well, I could probably pay a Pretty Good Human $2,000 per month to go be my research assistant. I could get a really decent coder for 10,000amonth. So it's only a matter of time before we start having to replace AIs with people. Also, if you haven't been following along to a lot of the OpenAI Elon Musk drama, Elon Musk has tried to sue OpenAI to prevent OpenAI from becoming a nonprofit company because he believes it goes against what they originally started the company for and what Elon actually invested in OpenAI for. However, it went to a judge and the judge basically said they're not going to stop that transition to a for profit. However, the judge was willing to take it to trial, so it'll be interesting to see if Elon is actually willing to put up the fight and take this to trial. Sam Altman sort of maintains that there really is no case here. This is all just Elon Musk trying to use the legal system to slow down his competitor. Because Elon Musk owns Xai, Sam Altman is OpenAI. And Sam just sees this as Elon trying to sort of throw up roadblocks to the competition. But in more practical OpenAI news, all plus users of Chat GPT now have access to GPT 4.5. When they first launched it last week, it originally only rolled out to pro members on the 200amonth plan, but he did say that next week it'll be coming to the plus plan. So if you're on the 20amonth plus plan, you now have GPT4.5 available inside of your ChatGPT account. Now, when I first started using GPT4.5 myself, I was sort of underwhelmed and unimpressed. It wasn't doing things like spelling Strawberry properly and it didn't really crush any benchmarks and it just didn't feel like there was anything that special about it. But having actually used it now for a little over a week, I've sort of changed my tune on four. Five. I actually really, really like it. Now, I'm not going to go and use it for coding or math or things that it really needs to think through, but when it comes to creative tasks like writing and helping me summarize things, making outlines, it's helped me with some scripts for some videos. It's actually probably the best model out there right now for more creative work. If you're on a Mac and you have the Chat GPT desktop app, the desktop app just got a little bit more powerful this week. ChatGPT for Mac OS can now edit code directly in IDEs. An IDE is an integrated development environment. Essentially, it's the tools that people use to actually write the code inside of. So if you've ever used Visual Studio code, that's an IDE. If you've ever used Replit, that's an IDE. Cursor, Wind Surf, all the new AI coding tools, those are all IDEs. And now if you're a plus pro or team user and you have a Mac, Chat GPT is capable of actually making changes directly inside of that ide. As of this week, we can actually see in the demo video that OpenAI put out that he gave it the prompt here, make a starfield view and add it on top of the existing background. It goes and it generates the code. He then has the ability to preview the code and then once he's happy with the code, he clicks apply and then it just automatically changes the code over inside of his ide. There's been some interesting news out of Google this week as well. They showed off an upcoming new AI mode. I'm using air quotes way too much today. This AI mode appears to be Google's answer to like Perplexity and OpenAI's search modes, where when you actually give it a prompt here and it goes out and it searches the web for you and finds a whole bunch of sources for the information and then it gives you an AI generated response that used what it found in its search for the context of what it gives you. If you've used Perplexity or you've used the search mode in OpenAI, it's pretty much the same idea. They're just sort of baking it into Google. It's not available yet. Some people are starting to get access to it in Google Labs. I checked. I don't have access to it yet. But we can see some examples here where they asked what happens to your heart rate during deep sleep. And it does a bunch of searches, find some sources, and then gives a response based on the sources that it found. The output looks very similar to Perplexity, where the sources are on the right sidebar and your response is below where you gave it the prompt. It looks like to be able to use this feature you're going to need to be a Google One premium subscriber, which is Google's $20 a month plan. Now, if you're not a fan of Google and DuckDuckGo is more your speed, they're also starting to roll out some AI features. According to this blog post here, they're creating a private, useful and optional. DuckDuckGo offers free access to popular AI chatbots at Duck AI and it expands its AI assisted answers on its search engine. So let's head over to Duck AI. We can see that the AI is private and anonymous, it's not training on our conversations, and you can actually choose which model you want to use. So up in the top left we can see our model picker. We got GPT4O Lama 3.3, Claude Haiku, O3 Mini and Mistral Small. Now once you're in here, you can ask it questions just like you would any other chat bot. But it doesn't appear to have the ability to search the web. It just gives you options to use different chatbots and do it completely anonymously. DuckDuckGo also seems to have AI overviews if you just use its regular search engine. So if I search how do I change a tire? We get this assist mode at the top where we can see it pulled in information from Home Depot and Dummies.com jumping back to Google here for a few moments. They've been rolling out some new features inside of Google Shopping. For example, they have this feature where you can actually have AI generate a object for you, like a colorful mini with big daisies. It will generate a dress like that with AI and then try to find actual dresses that look similar to the one that AI generated. They also have this feature for like makeup where you can virtually try on different makeup and they have like a virtual try on feature. This doesn't actually use pictures of you, but it uses various models of different shapes and sizes so that you can see what clothes will look like on a model that has like a similar shape to you. Google Sheets also got a new AI update in the past week. You can now analyze your Google Sheets data faster so all workspace business users, so paid users now have access to new Gemini capabilities and sheets. You can generate insights from your data like correlations, outliers and more. This seems really handy to me. Like the previous version of Google Sheets, it did have AI in it, but it would kind of more help you generate like formulas and populate some of the cells. It didn't really use AI to analyze what was going on in the cells. I really like the idea of this like correlation feature too. An interesting example of how this could be useful is let's say you kept a food log of, you know, each date you had something to eat, the foods eaten, and whether or not you got a stomachache on that day or got a headache on that day. Well, maybe now you can find correlations between the foods that you eat and what doesn't make you feel good. I can click on summarize this table here and I can ask something like find a correlation between what I and what gives me stomach aches. We can then see. It gives me a pretty detailed breakdown of what types of food probably don't make me feel good. This is all dummy data, by the way. If you use Google on your phone, you're about to get the ability to actually share your screen or share your camera and let the AI actually see what you're looking at. In this example video here, they're looking at the screen and asking about the genes that are actually on the screen. And in this example here, they're actually sharing their camera and asking questions about what the camera can see. Now this is very, very similar to What I got to use on the project Astra when I went to London a few months back. And it looks like it's going to be rolling out very, very soon to people who use Google on their phone once again. It also looks like it's only going to be Google One AI premium members, so people on the $20 a month plan, but they do say it will be rolling out on Android later this month. Google's also rolling out a feature on Android and Pixel phones that can help you detect scams. So we can see an example in this image here where somebody gets a text message that is this Mary. Sorry, this isn't her. Apologies. I was trying to reach a client with some investment info. I guess I entered the wrong number. No problem. By the way, this is John. I'm a financial advisor. And then Google detects this is most likely a scam and you can report and block. This is something I know personally I see more and more of every day. I get text messages constantly that are like, hey, are we still on for lunch today? Or I'm going to be a little bit late into the office today. A lot of these messages are actually scams trying to get you to reply so that they know they have a real phone number. And then they're going to try to lead you down a path to try to scam you. My advice, if you get a text message from somebody that seems like it's a wrong number, just ignore it and delete it. But if you do try to interact with it, it appears Google's going to have your back and help you detect whether or not it actually is a scam. And in the final bit of Google News, Google Labs showed off this preview of Whisk Animate Whisk is a tool that Google rolled out that allows you to sort of combine images and blend them into a single im. Well, now you can take those images that you made in Whisk, the blended images and this will use VO2 to turn those images into little videos and animations and things. But no real clear update on when we'll actually see this. It just says stay tuned for updates on when this will be made available. There was a small update from Microsoft this week. Microsoft Copilot got a little bit of a UI update. We could see in this image what it looks like. It's just a slightly cleaner update for the Windows app for Copilot. In the world of open source large language models, there was a new model this week from Alibaba's Quinn with qwq 32B it's a new reasoning model with only 32 billion parameters, and they claimed that it's pretty much as good, if not better than DeepSeek R1 in the AIME24 benchmark, which tests mathematical problem solving abilities. It scored right about the same as DeepSeek R1. And keep in mind this is a 32 billion parameter model where R1 was a 671 billion parameter model. So it's a much smaller, much lighter model that performs just about as good at math and logical reasoning as DeepSeek R1 did on live code bench slightly underperformed R1, but just very slightly. This live bench is also one that tests things like coding and logic and reasoning, and it once again outscored deep seek R1 on the if eval it it pretty much scored right in with the top of the other models, so 01 Mini and Deepseek R1. The IF stands for instruction following, so this is how well the model actually followed instructions. Seems to be right up there with the best of them. It also beat out all the rest on the bfcl, which is a function calling benchmark. So basically how good is it at language execution? It tests things like Python, Java, JavaScript, MySQL and other things like that. Super, super impressive for a model this small to be right in there with best of the best models right now. Over on Hugging Face, they did make the code and the model weights available. If you've got a strong enough computer or you want to run it in your own cloud, you can do that if you know what you're doing. They also made a Hugging Face demo available where you can test it out directly inside of Hugging Face. As always, all of the links I talk about in these videos will be linked up in the description below, probably in a Google sheet, just because it's getting to a point where I'm sharing so many links that I can't fit them all in the description. I'm a huge fan of Perplexity's voice mode on the phone. I'm asking it questions all of the time throughout the day. Well, they just rolled out that feature in the Mac app, so if you have Perplexity installed on Mac os, you should now have the ability to actually chat with it straight on your computer. Looks like by clicking Command Shift M on your Mac, it'll open up the voice mode. And since we're talking about voice modes, if you have iOS and you have Grok, you can actually now use the voice mode for free on Grok. Previously you had to be like an X Premium plus subscriber, which is like 30 or 40 bucks a month to test out the Grok AI voice mode. But now the they're telling us that voice mode is available to all iOS users for free on the Grok app now, no subscription required. I'm actually not sure if that comes with all the voice modes. So if you get like the unhinged mode and things like that. I have a paid version, so it's harder for me to tell the difference between what's on the paid version and the free version because I don't have a free version installed right now. And since we're talking about AI audio, I have to talk about Sesame because I didn't talk about it in last week's video. Even though this announcement was made last week, this Sesame company has made probably the most impressive AI voice conversation tool I have ever used. This is like the closest thing we've ever seen to the movie her with Joaquin Phoenix. The way the voice sort of speeds up and slows down and feels like it has a very natural cadence to it is just pretty mind blowing. So if I click on Maya here, we can hear what this sounds like.

Well, well, it seems like just yesterday we were swapping stories. Ready for round two? What's been on your mind?

I'm curious. What keeps you up at night?

That's a thinker, isn't it? Not that I literally sleep, but you know what I mean. I guess I sometimes worry about giving the wrong advice, you know, leading someone down the wrong path. Maybe we both need a warm beverage for this one.

What's your beverage of choice?

Funny you should ask. I can't actually drink anything, but if I could, I think I'd go for a steaming mug of hot chocolate, extra marshmallows. Just thinking about it makes me feel all warm and fuzzy inside. What about you? What's your go to drink when you need to unwind?

So yeah, that's probably the most human, most natural sounding AI voice you're going to hear. And this to me seems like it really flew under the radar. Now if you're on X and you're in this sort of like AI sphere over on X everybody was talking about it, but outside of X, I didn't hear any coverage about this, like at all. Moving on, the company Podcastle launches their own text to speech model that has over 450 AI voices. It kind of sounds like they're going head to head with 11 labs. I haven't tested this one myself yet, but According to this TechCrunch article, they claim the voice created with the new process sounded a bit robotic, but the company does say over time it's going to improve. It also says you can train different samples of your voice to get different results. So very similar to 11 labs, you can train in your own voice and I guess theoretically create AI generated podcasts. I mean, it's called Podcastle, so I'm guessing that's what the idea is here. This company, teleperformance, is using AI to make staff sound less Indian. So teleperformance is the largest call center operator in the world, and they're using this technology called accent translation, and it's being deployed in call centers in India where workers provide customer support to some of Teleporter Performance's international clients. Now, I think for most customer service, this is going to be really good. I mean, if you're calling a call center to get support and you're struggling to understand because of the difference in accents, that seems helpful. But I can't help but think that this technology is most likely going to be used for scammers as well to hide the fact that they actually have an accent when talking to you. In similar news, Amazon prime is actually testing AI dubbing to dub TV shows and movies into different languages. There's some new research out of Meta that will help people generate avatars using only four selfie images. So we can see in this video here, this guy is taking four different images from four different angles and then it translates into a avatar, like a 3D avatar that I'm guessing you could use in something like Blender or Unreal Engine and it'll look just like you with only four selfies. So that's some pretty cool research based on the fact that it's coming out of Meta and it's on a VR blog here. My guess is that they'll probably use it for like avatars inside of meta quests, similar to Apple Vision Pro. When you have a conversation with somebody, you're actually seeing an avatar version of them that actually looks like them. If you like all the new features that Pika Labs has been rolling out lately, and you have a Korea AI subscription, they just rolled out all of the various Pika features into Korea. So you can do things like change video regions, add AI elements, interpolate frames, mix images, and pretty much do all the stuff that you can do with Pika Labs. If you've watched my end of week news videos for the last several weeks, almost every single week, Pika seems to be shipping some sort of cool new AI feature. And CREA seems to be jumping on board and rolling those features into their platform as well so that you kind of have one place to do it all. The company Tencent rolled out a new open source AI video model called Hunyan Video i2V. And the videos I've seen so far are pretty dang impressive. Here's some examples that they showed on their GitHub page. I think that's a penguin wearing a dragon suit and then walking around town. I mean, it's pretty good. The code appears to be available on GitHub, but I haven't seen a place where you could actually test it yourself yet. My buddy here, Angry Penguin, AKA ap, got early access to it and made some really cool videos like this one that really reminds me of like Age of Empires here. This sort of Pixar looking animation with some really great colors. Honestly. Here's another one of like a soldier in the snow and there was like an aurora borealis going on in the background. Here's another one that looks like a colony of bees or something. I'm really, really impressed with what we're seeing, especially knowing that these are open source. This company, Hedra Labs, rolled out Character three, which is a new generation of AI native video creation. They call it the first omni modal model in production. Built to jointly reason across image, text and audio for more intelligent video generation. With Headdress Studio, you can do text to video and audio to video. Character and general purpose image and video creat dynamic backgrounds, text to emotion. Here's a quick peek at their demo video and I believe the video and audio were all created with their technology. Introducing Hydra Studio and Character 3, the most powerful omni modal gen AI model in the world. Featuring full body expression, emotional control and dynamic environments.

From our world to the next.

Next, tell your story with all of the leading image, audio and video models at your command. Your new all in one creation platform, Hydra Studio and Character 3, available today at hydra.com the voices still sound a little robotic to me, but the rest seem pretty impressive. So logging into Hedra.com and testing it myself, I'm going to give it the prompt. A man with a beard in front of a city skyline and I have to give it an audio script. I can either record the audio, upload the audio or generate speech. So let's generate speech. I'll just type. Subscribe to Matt Wolf. We'll pick a voice here. I'll pick Drew and let's go ahead and give it this prompt. Apparently it forces you to have an image as well. I'm not quite sure what the point of giving it a prompt is if you have to give it an image. Anyway, I'll go ahead and have it create an image and I'll type a man with a beard in front of a city skyline. I wish it could have just taken my prompt, generated an image and then converted it to a video. I feel like this is extra steps that shouldn't be necessary, but should sure this image works. I'll add it to the video. Now. Let's see what we get. And here's the output I got. It took maybe two minutes. Subscribe to Matt Wolf. I mean, the voice doesn't match the face, but it's not too bad. I also came across this Tavis. I think that's how it's pronounced, they say. Introducing our new family of state of the art AI models, Phoenix 3, Raven 0 and Sparrow 0. Together they bring conversational video interfaces diving a little bit deeper. This is a combination of three models here. So Phoenix 3 is the first and only Gaussian diffusion model. So it does real time full face rendering, generating complex micro expressions and emotions. Phoenix 3 animates the entire face, including the eyes, the brows and the cheeks. Then there's this Raven Zero, which is the only real time perception system that lets AIC reason and understand like humans. More than just simple computer vision. Raven tracks actions, reads emotions and deeply understands visual context, screen share. And get instant visual spatial insights on anything from chess to debugging to online online shopping. And then we have Sparrow 0, which is the most advanced turn taking model for natural conversation. It analyzes tone, pacing and intent to know exactly when to speak and when to listen. No awkward pauses or interruptions, just smooth, dynamic human like dialogue that outperforms every other model supposedly. We can try this over at Pavis IO. We get Charlie again. Let's go ahead and start a video chat with Charlie. Let's join a video chat.

A real human. I was expecting a voicemail. How's it going?

I'm doing good. How are you?

I'm doing great, thanks for asking. It's awesome to finally meet you. I was just cruising through latent space waiting for someone to chat with. So what's your name, man?

I'm Matt. What's your name?

That's a great name, Matt. I'm Charlotte. Nice to meet you. Dude, I've got to say, that maroon baseball cap is pretty stylish. What's the story behind it?

Oh, that's my Future Tools logo. You can find it over at FutureTools IO.

No way. That's really cool. I'll have to check out Future Tools when I'm not, you know, chatting with awesome people like you. So what's Future Tools all about, man? You're not secretly building a robotic uprising, are you?

You know too much. We've got to go. I mean it's pretty cool. It does does sound very robotic and the lip syncing doesn't match up super well. But it's very interesting that it can actually see me. It commented on my hat. Doesn't seem like a bad model. I thought this was a fun story. People are actually using Super Mario to benchmark how good an AI is. So they asked various AIs to play Super Mario. It's using like an emulated version on the computer. They're not playing on an old school Nintendo or anything. And they found that Claude 3.7 did really really well. The difference between testing with Mario versus testing with something like Pokemon is Pokemon is a lot of turn based stuff, right? So the Pokemon character moves around on the screen and then when it gets into battles, it's like a turn based battle with Mario. You've got to keep moving, things are moving towards you, you've got to jump, you've got to avoid things. It's a lot more complex for an AI to play and if there's a slight delay where if the AI needs to think what to do next, it doesn't work out very well. So it's really a test of how well it can play the game, but also how fast the AI can react. And apparently Claude 3.7 was the one one that just crushed all the others and managed to actually play through the game and do a decent job. Where Claude 3.5, Gemini and GPT4.0 didn't quite fare as well. The Opera browser is rolling out the AI browser operator. This is like an AI agent built into Operator where it could go and order things for you online and do tasks on your behalf. So very similar to what you get out of Claude's computer use tool or chat GPT's operator, same idea. They're just building it right into the browser. On Opera we can see an example. Find me a pair of Nike socks in men's size 10 that I can buy. It goes to Walmart, does a search, find some socks, adds them to the cart and then says ready for user. So I guess at this point you do the checkout, can also do multi step tasks. Here we can see I want to buy two tickets to Newcastle's next Premier League game in April at their home stadium. Hospitality tickets are okay. If regular tickets are not available, I'd like to sit fairly midfield and it goes and finds the tickets and and does a bunch of searching on your behalf. We don't know exactly when we're gonna get it, but it says we expect to launch the new browser operator as part of our AI feature drop program in the near future. We got a bunch of news out of Apple. They revealed their new M3 Ultra chip. Although they're on their M4 series of chips now and we haven't seen the M4 Ultra, but their previous generation of M3 chips just got an Ultra. But this new chip is actually pretty dang insane. It's got 96 gigabytes of memory that can be configured up to 512 gigabytes. That's insane. This CH is basically going to be like an AI powerhouse. Apple also rolled out some new devices this week. They introduced the iPad Air, which will support Apple Intelligence. They rolled out a new Mac Studio, which you can get with an M4 Max or the M3 Ultra that we were just looking at with that insane 512 gigabytes of unified memory. It says right here, Mac Studio is a powerhouse for AI, capable of running large language models with over 600 billion parameters entirely in its memory. I mean, this is basically the best, best consumer computer at the moment for running AI locally, from what I can tell. Apple also introduced a new MacBook Air with the M4 chip. Once again with Apple Intelligence. That's really the only reason I'm talking about these Apple upgrades, is that Apple is shifting to designing all of their computers and all of their chips to just be AI beasts. Mistral rolled out a new OCR model. OCR stands for Optical Character Recognition. It's basically the technology that if you upload an image that has text in it, it can detect the text in the image. Or if you upload a PDF, it's actually capable of reading the PDFs. And apparently this new Mistral OCR model is the best at it to ever exist so far. If you're into AI coding, which is something I've really been doing a lot of lately in my spare time. Codeum just released a windsurfed wave 4 with all sorts of really cool features. You can actually preview what your apps are going to do straight inside of Windsurf. It's got suggested actions, so after it responds, it will actually suggest what it thinks you should do next. It's got M MCP Discoverability. MCP stands for Model Context Protocol. This was actually something developed by Anthropic several months ago and it has become the hottest buzzword in the world of AI. If you go on X and you're in the AI sphere over on X, everybody's talking about MCP Model Context protocol right now. It's very similar to like APIs encoding encode different software that you want your software to be able to communicate with. You would use an API to create the communication between your app and another app. However, AP not great for large language models because every software out there sort of implements APIs differently. There's no sort of standard uniform way that APIs are implemented. What model Context protocol does is it creates a standardized way for large language models to connect with these tools. So I believe it's still using APIs under the hood. It just sort of standardizes the way that these APIs are called upon from large language models, making it a lot easier for large language models to take actions on your behalf half from these external third party tools. Now I know that's getting kind of into the weeds, but maybe I'll make another video in the future where we dive deeper on explaining exactly what model context protocol is. But just think of it this way. With APIs, you use APIs to write code, to make your code talk to other apps. Model Context protocol is a way for large language models specifically to more easily interact with other apps. I have a few like really fascinating things to share here, like this company, Cortical Labs, created this new hardware device called a CL1 and it's the world's first biological computer. It actually uses human brain cells and fuses them together with silicon hardware to create form fluid neural networks. Now how this all works is slightly over my head, but what's interesting about this is this section here. The large majority of drugs for neurological and psychiatric diseases that enter clinical trial testing fail because there's so much more nuance when it comes to the brain. But you can actually see that nuance when you test with these tools. Our hope is that we're able to replace significant areas. Animal testing with this animal testing is unfortunately still necessary. But I think there are a lot of cases where it can be replaced and that's ethically a good thing. By leveraging neurons as a computational substrate, SBI has the potential to create systems that exhibit more organic and natural forms of intelligence compared to traditional silicon based AI. So they believe that doing this, they'll be able to actually uncover new ways to treat things like dementia and Alzheimer's. And they actually got one of these sort of biological silicon hybrids to play Pong, which is just wild to me. All right, a few last quick things here. Uber and Waymo are teaming up on driverless cars in Austin. So you'll be able to order a Waymo straight from Uber. Recently, a fully driverless Maserati MC20 got up to 197.7 mph over at the Kennedy Space center and it broke the autonomous speed record. The prior speed record was about 5 mph slower than that. And as you speed usual, here's a cool robot. This is a robot from PND Botics. Look at the motion of this robot. You can just see how it has a more natural, fluid motion with how it moves around and its fingers and it's just like this crazy. Very similar to human movement. We can see here. It's like mimicking playing a piano. And as I always say, I just think robots are really cool. We can see here that it's got the dexterous hands with 6 degrees of freedom, 12 joints, sub millimeter repeatability, similar size with a real hand, integrated force sensors so it, it won't like crush things. And yeah, I just love robots. But that's what I got for you today. If you like videos like this and you want to stay tuned in on all the latest AI news, make sure you like this video and subscribe to this channel. I'll make sure that your feed has more videos like this one as well as cool AI tutorials, cool tool breakdowns and things like that. I do my best to try to provide as much value as I can and keep everybody looped in on the AI world. Also, as a quick reminder, I am giving away an RTX 5090 from Nvidia. They actually sent me a picture of Jensen Huang holding the GPU that I'm gonna give you and he signed it. So you're gonna get a signed RTX 5090 if you win this. And in order to enter this raffle, it's super simple. Just be subscribed to this YouTube channel, subscribe to the newsletter and register for Nvidia's GTC. The event is actually in San Jose, California. There is an in person event event, but you can watch it virtually for free. The link will be in the description. To register for GTC you just got to be registered and subscribed and I will put a Google form down in the description below. Fill out the Google form to let me know that you actually registered and you'll be entered into a raffle to win this RTX 5090. It'll take you about 25 seconds to do all of the steps, and you could end up winning this $2,000 value GPU. I mean, worth more than that. It's. It's signed by Jensen Huang himself. So again, details for that are down in the description. Thank you once again for tuning in. And as always, check out FutureTools IE. This is where you can join the newsletter that I just mentioned. It's where I curate all the cool AI tools I come across. I also keep the AI News page up to date pretty much daily. All the news I share in this video, plus more that I'm not able to share, is all on this page. It's totally free. And if you join the free newsletter, I'll hook you up with the AI Income Database, a cool database of ways to make money money with these various AI tools. Again, all free over at Future Tools IO. That's all I got for you today. The AI world is not slowing down and I am here for it. Hopefully you'll subscribe and hang out with me in future videos and I will keep you as looped in as I possibly can. All right, thanks again for nerding out. Hopefully I'll see you in the next one. Bye. Bye.