as is true with almost every single week now, there's been a lot happening in the world of AI. Starting with the fact that once again we have a new king of the AI large language models. This week, Anthropic released Cloud 3.5 Sonnet. And if you look at this little graph that they put here on their website, here's where Cloud 3 Opus, their previous state-of-the-art large language model, landed. Well, the new one we can see here is much more intelligent and also much cheaper to use than Cloud 3 Opus. It's about the same cost as the previous version of Cloud 3 Sonnet, but way more powerful than Opus. It also operates at twice the speed of Cloud 3 Opus. And if we take a peek at the benchmark tests here, again, I don't always remember which benchmark test is best for which thing. I just kind of look at it overall. And as we can see, in every single area, Cloud 3.5 Sonnet beats out Cloud 3 Opus and in almost

every single Benchmark it beats out GPT 40. With GPT 40 seemingly slightly better at math, the vision model in Claude is also much improved. And as we can see here, once again in the benchmarks for testing Vision models, Claud 3.5 Sonet beats out Opus and Gemini 1.5 in every single Benchmark. And GPT 40 only slightly edges it out in this specific visual question answering Benchmark. And I mean it's by a small margin. We're talking 6.91% to 68.3%. Claude ALS also released a new feature, which they call artifacts. Artifacts appear in a dedicated window alongside their conversation. It creates a dynamic workspace where they can see, edit, and build upon Claud's creations in real time, seamlessly integrating AI-generated content into their projects and workflows. And the coolest part is this new model of Claud is available for free on cloud. So you don't even need the paid plan to use this. They are rolling out version 3.5s of Haiku

Opus later this year, so the premium member will get it update. But right now, their best model is available for free, and there's been a lot of demos on Twitter already of how good this model is. Here's Ali Miller talking about how in just 25 seconds, Claude 3.5 Sonnet coded a fully functional Mancala web app for me. We can see she uploaded a screenshot attachment here, gave it a single prompt about what she's looking to do, told it to think step by step. It opens up the little artifact sidebar to show it actually writing the code, and it presented a playable game right out of that single prompt that she put in. Ethan Mollik here also did the same thing. He said, This is a real-time video of me creating a playable game and editing it with Claude. We can see he gave it the prompt, Make me a game about crabs and let me play it. And then the artifact sidebar opens. We get the code generation. Didn't quite work the first time, so he said, Not crabby enough. And this time

around they actually look like crabs he says can you make it fill the whole screen and be even crabbi once again generates codes in the sidebar and generates a much more impressive looking crab game. Ethan says can you make it scarier now. The game is much darker and I guess scarier. Testing catalog news posted on X Cloud 3.5 just generated react jsx code with a simple contact form and managed to run it in the artifacts playground. If we open their screenshot here, we can see that they had it code a simple contact form. It showed the contact form in the right sidebar and was actually usable. It looks like you can switch between preview and code here. Dennis Sharv here says Cloud 3.5 Sonet is the first model to recreate the 3D scene data flow from the movie hackers on the first try. Great job anthropic. We can see what that looks like here. Petro Chirano here says Cloud 3.5 Sonet plus Maestro equals Sparks of AGI. I asked to make a Mario.

clone using just geometric shapes, and the wildest part is that it gave the character animations as well. The shapes seem like novel concepts. It took 3 minutes. Once the code was generated, we can now see exactly what the game looks like on the first try. This looks like a game I tried to make about a year and a half ago using AI to help me code. That took me, I don't know, probably 50 prompts in 5 hours to generate. This time, it looks like he was able to do it with just a single prompt. This new version of Claud Sonet is really impressive. Now, if you want to use it, head on over to cloud. Come up here, click on your name, click on this little feature preview in the menu bar here, and then and then turn on artifacts. Let's ask it to make a simple tic-tac-toe game where the user plays against the computer. Make it playable in the artifacts window. We can see it opened up our artifacts window on the right and started to write some code.

now we've got a tic-tac-toe window from a single prompt that shows up in our right sidebar. Let's go ahead and see if we can play against the computer. So I put the X here, the computer put o here, and it appears the computer's not very smart, but it does work on the first try. I can reset the game and play again, but to me this feels a lot faster and also quite a bit smarter than what we're getting out of chat GPT 40. Now in fact, if you want my opinion, I actually feel like GPT 40 got a little bit dumber than the previous version of GPT at least in my use cases. It actually seems like it performs worse than it used to. I don't know if anybody else is experiencing it, but that's what I found. And now with Claude having its free version, have the best model available on the market for anybody to use is pretty cool, but it does kind of feel like some of these other companies are starting to pass.

open AI a little bit. I mean, we saw Lum's Dream Machine recently, which came out before anybody got access to Sora. We're seeing models from Claude come out that are outperforming open AI's best models. In my opinion, it feels a little bit like open AI is falling behind. Now, we don't know what they've got training behind the scenes. They may release a GPT-5 or something else that they call it, and it might just blow everything out of the water and make things like Claude's son it just look really basic and archaic. But I feel like they need to do that fast because a lot of these other companies are starting to eat up some of the mindshare of the general public, and it could only be a matter of time before everybody is using the term Claude as synonymous with AI instead of using chat GPT as synonymous with AI. But this isn't the only reason that I feel like open AI is starting to fall behind a little bit. They do feel like they're going through

a pretty rough patch at the moment. The news also came out this week that Sam Alman is sort of mulling around the idea of getting rid of the nonprofit portion of OpenAI and just making it a full-on for-profit company. Eventually, they can IPO and make as much money as they want. This is just speculation. Sam's just made some comments about how they might do that. So, not a ton of news here, but now that the OpenAI board seems to be more like Sam's people instead of people that were super focused on the safety of these models, it seems like he can do whatever he wants the company at this point. Also, at the end of last week, which I didn't catch in last week's video, OpenAI appoints retired US Army General Paul M. Nakasone to the board of directors. They brought him on board because of his cybersecurity experience. But if you're curious a little bit more about what his cybersecurity experience is, he was the longest-serving leader of US Cyber Command and also led the NSA where he was

charged with safeguarding the United States' digital infrastructure and advancing the country's cyber defense capabilities. So they hired somebody that was high up at the NSA, who notoriously spies on US citizens, to be part of the board for OpenAI. And of course, Edward Snowden, who got in a lot of trouble for exposing the fact that the NSA was spying on us, had some words about this. He said, Do not ever trust OpenAI. He said, There is only one reason for appointing an NSA director to your board. This is a willful calculated betrayal of the rights of every person on Earth. You have been warned. Another security expert named Matthew Green went on to say, I do think that the biggest application of AI is going to be mass population surveillance. So bringing the former head of the NSA into OpenAI has some solid logic behind it. Now OpenAI hasn't commented on this specific stuff yet, but they have claimed in their articles that the security of OpenAI systems, from protecting the large AI training supercomputers we operate to

securing our sensitive model weights and the data entrusted to us by customers is central to achieving our mission. So they're claiming that he's just there to keep things safe and secure, but he does have a reputation with the NSA, who again is known for spying on its own US citizens. Also, if you remember Ilia seser, one of the people who helped oust Sam Alman back in November of 2023 and then was unheard of for months and months and months, now went to X to say he's starting up a new company called SSI or Safe Super Intelligence Inc. If we take a look at the announcement tweet here, they claim that super intelligence is within reach. Building safe super intelligence is the most important technical problem of our time. So they've started SSI. They go on to say our singular focus means no distraction by management overhead or product cycles, and our business model means safety, security, and progress all insulated from short-term commercial pressures. They're based in Palo Alto and Tel Aviv, and they're currently

assembling a team to make these goals achievable. Now, I don't totally understand what's going on with SSI. It sounds to me like Ilia is going and essentially creating an open AI competitor to try to get to Super intelligence. OpenAI's goal is to get to artificial general intelligence. Apparently, this one is to get to Super intelligence, but do it safely, and they're going to have no customer-facing products, at least not in the beginning. So, I don't actually understand how they're going to raise money. I guess there are going to be some people out there that would be willing to invest just for the tech itself, just to see it advance, but I have a hard time finding major venture capitalist firms and things like that dumping a ton of money into a company that has no outward-facing products. Like, they're just going to be dumping money into a product that has no revenue generation. On the other side of it, I just don't understand how they're going to raise money. They obviously need to use a lot of compute.

to get to Super intelligence meaning they need data centers and huge GPU clusters so they either need the money to go and buy all of that stuff, which I don't quite understand how they're going to raise on a business model that has no revenue generation, or they need to use cloud infrastructure that already exist. They need to use, you know, Google's cloud or Amazon AWS or Microsoft Azure or one of these other providers of cloud GPUs. And if they do that, how do they ensure that it's safe because now they're operating in somebody else's playground. I don't know, we still need to learn more about what is going on with Ila's new company. I don't quite understand the business model behind it, but it will be interesting to follow along, and of course, I'll keep you in the loop as I learn more about it. We've also had a ton of AI video news coming out recently. Last week, we talked about Loom's Dream Machine, which was really impressive. Not great at text to video, but really, really good at image to video. Well, this week, we

got the announcement of another really impressive looking AI video generator. Albeit they haven't given us access to it yet, but they showed it off and it looks really really good. It's from Runway, and it's called a gen 3. And this is their text to video model, well, their new text to video model. And the demos of it actually look really really impressive, like better than Luma's Dream Machine and about on the same level as Sora. In fact, a lot of people think that the reason Luma's Dream Machine kind of came out so quickly was because they knew that Runway gen 3 was about to be announced. And that's why the servers weren't quite ready for the onslaught they got when they announced Dream Machine. If you remember, on the first day they opened it, people were waiting, including me, up to 7 hours to generate a single video. They weren't ready for that launch, and a lot of people believe it's because they knew this was coming and they wanted to get out in front of it.

but the generations that this can do from text to video, like an astronaut running through an alley in Rio de Janeiro, look really good. We were not getting anywhere near this level of text to video from Lum's Dream Machine. We're getting this level when we start with images, but definitely not text to video. Here's a handful more examples that are on Runway's website. I will link up to their announcement in the description below so you can check out more of these announcements, but this stuff is looking really, really impressive. Here, humans are looking pretty dang realistic in these videos. They shared a whole bunch of examples of humans that don't actually exist, but we've got some pretty dang good video generation around them. And one of the things that seems to be really impressive that nobody's really talking about is the ability to generate text inside of these videos. Like look at this video that says color in water, and then like a splash of color hits the letters. Or this one where there's water spilling and then it splashes up to

reveal the words drink water and all of the letters and all the words just look really good. I don't know why this isn't one of the features more people are talking about, but nonetheless, some of these videos generated by Runway gen 3 are almost indistinguishable from real-life videos. This is the first actual Sora level video generation that I've seen that is not actually Sora, and this brings me to another reason why I mentioned. I think OpenAI is kind of in trouble. They announced Sora, everybody got excited about it, they waited so long, and then it gave time for things like Cing and Dream Machine and gen 3 to all pop up and go look, we've got things pretty much on par with Sora now, making a lot of the AI world to believe that OpenAI teased it and then didn't release it fast enough. By the time we finally get access to Sora, who knows, maybe people just won't care anymore because Dream Machine and gen 3 do it just as good and everything they want to.

do with it. They're getting out of these tools. While we're on the topic of AI-generated video, there was an AI-generated movie premiere for a film called The Last Screenwriter. Well, apparently, a movie theater in London cancelled this premiere because over 200 people complained about the fact that it was an AI-written script. Now, this isn't an AI-generated video. AI generated this script, and then it was acted out by real people. But because the script was AI-generated and 200 people complained, the movie theater went, Nope, we're not showing it, which is a pretty crappy stance for a movie theater to take. If we're honest, it probably would have done okay at the theater because the people that all hate AI would have showed up so that they could go and talk about how much they hated the AI movie, and the people who love AI would show up because they want to be a part of seeing the first AI-generated script in theaters. But oh well, my guess is this news story will spread a little bit further and

actually have a little bit of a strain effect on this film where more people will actually learn that this film exists because this movie theater, and these 200 people didn't like the fact that it existed. Google Deep Mind also showed off some new research in this past week showing that we can generate audio from videos. So we give it a silent video clip and then it figures out what audio works best with that video clip. And there's a bunch of examples on their website. I will again link it up so you can check them out, but let's preview a couple of them. Here's like a horror scene of somebody walking through like a dark area. Remember the video was given, the audio was generated with AI by watching the [Music]. Here's another example of a drummer. [Applause] [Music] [Applause] [Music]. And they even did a wolf howling at the moon, which I'm a fan of because that's one of my test prompts that I always

[Music] Not bad at all. Along a similar vein, 11 Labs introduced a text-to-sound effects API, including the first video-to-sound effects app. And not only that, but they made the app available for free and fully open source. Just upload your video, we then analyze what's in the video, and then use our custom AI model to generate the perfect sound effect for it. Takes about 15 seconds. And so, whether it's for a bottle smashing, a motorcycle racing through the street, or bringing to life your favorite meme. [Music] If you go to videosoundeffects.com, this is where you can try it out completely for free. Here's actually one of the videos that I made for my music video that I put out the other day. Let's go ahead and throw this in there and see what kind of sound effects it generates for it. Let's drag and drop the video in. They do claim it takes about 15 seconds, and it generated four audio files. So let's take a listen. Not too bad. I was hoping to get a little

bit more of the like robot sounds as their heads moved like a but it is what it is. It's pretty cool, and it's totally free to play with, so don't mess with it. Tik Tok also rolled out a new AI feature this week. Now who knows if Tik Tok's still going to be around in the US a year from now, but they rolled out this Symphony avatars feature where you can use pre-built avatars or create custom avatars that can talk on video. So check this out. Hi there, my name is Adrian. Well, not Adrian, Adrien, but Tik Tok's latest Symphony digital Avatar. I am a custom avatar born from Tik Tok's latest generative AI technology. Basically, it looks like you're going to be able to upload a bunch of text, and it will either use one of their stock avatars, or you can create your own avatar like yourself and have it speak out whatever text you put in. I have mixed feelings about this. I don't personally use Tik Tok much, but I don't know if I'm a fan of scrolling through

Tik Tok, and just seeing a bunch of tweets converted into videos. I just don't know if that's going to be very entertaining, but what do I know as somebody who I'm not really on the Tik Tok train myself. We got a whole bunch of new AI model releases this week from companies like Apple, Microsoft, Meta, and even Nvidia. In fact, this week Apple put 20 machine learning models on Hugging Face. Taking a peek over at Apple's Hugging Face page, we can see its models like state-of-the-art depth estimation, semantic segmentation, image classification, Transformer-based language models, mobile-friendly image text models. There are a ton of Apple models now available for you to use over on Hugging Face. So let's do state-of-the-art depth estimation. I can click into this one. We can see there's two models here, depth estimation small and V2 small. When we click into one of the models, we have all of the files available here. So if we want to build or iterate off of them, we've got it available. I'm not seeing any spaces.

available where we can use it directly on Hugging Face. For this specific model, but I imagine that's only a matter of time. Microsoft also dropped a model called Florence 2, which is a model to handle a variety of vision tasks. They also made it available over on Hugging Face, and if we click into this one over on Hugging Face, we do have all of the files available for us to use, but there's also some spaces available to test out Florence too as well. So if we wanted to drop an image in here, I'll throw this old computer image here, and then we have a whole bunch of potential tasks that it can do like caption it, a detailed caption, and even more detailed caption, object detection, all sorts of other stuff. Let's do a more detailed caption and submit it and see what it gives us. The image shows an old-fashioned desktop computer with a monitor and keyboard on a wooden desk. The computer is beige in color, etc., etc., so it's got a pretty detailed breakdown of what it sees in the image, again you.

can find all that over at Hugging Face. I'll link up to Microsoft's article about it, so you can find it and play with it if you want. Meta also released some new research models, including Meta's Chameleon model. It's able to perform any combination of inleaf text and images in its input and output without the need for modality-specific modules. Meta Chameleon uses tokenization for text and images, making for a more unified approach. Now, most of this here from this Meta article looks like they released the research papers but not necessarily the code yet. And then we have Nvidia, who also released a new model this past week called Neotron 4,340 billion. These are models optimized for Nvidia's Nemo and Nvidia's TensorRT large language model. They also put their model available over on Hugging Face. The link again will be in the description, but this model seems much too large to actually run on your own computer. But large enterprises that have the compute power to be able to do this stuff can.

now work with this neotron model as well. It also is starting to look like Gro which is the X a model that you can find over available on X might be getting mid Journey integration now. Mid Journey doesn't have an API yet. It doesn't seem to be integrated with any other tools yet, but rumors were circulating recently that Grock and Mid Journey might team up, and Mid Journey could be available to generate images directly from Grock. And now testing catalog news here on X posted did a screenshot of some of the code showing that Mid Journey was mentioned in the code, leading to more speculation that we might get Mid Journey directly inside of Grock pretty soon. Now, if you live in the European Union, I've got some bad news for you. It is starting to look like a lot of the large language models and AI companies are going to start avoiding giving you guys some of these AI features due to a lot of the laws that are over there in the EU. Meta paused their AI models launch in Europe. The article here says

meta will not launch its meta AI models in Europe for now after the Irish privacy regulator told it to delay its plans to harness data from Facebook and Instagram users. The social media company said on Friday. Also, all those AI features we saw Apple announce, well, it also looks like those aren't going to be available in the European Union. Apple announced that Apple intelligence iPhone mirroring and shareplay screen sharing will not be available in the EU this year because of the digital markets act, which allegedly forces it to downgrade the security of its products and services. So, the ACT, which is designed to look out for members of the European Union, actually is trying to get companies to decrease their security to comply. And while we're on the topic of Apple, Apple intelligence will label AI-generated images in the metadata. So right now, if you generate images with OpenAI or Adobe or Google or Microsoft, all of them add metadata to the image so that social media platforms can tell whether an image was AI-generated or not. A lot of people have been complaining about this.

on Instagram where they make a teeny tiny tweak with something like Photoshop and post it to Instagram. 99.9% of the image was a real image but they maybe like fixed one little thing, and now it says generate with AI. So a lot of the social media platforms are trying to figure out how do we actually deal with this when the majority of the image is in AI, but we also want to make sure people know when something is generated with AI. Well, it sounds like Apple Intelligence is going to add that metadata, but it's still up to the social media platforms to decide how to use that metadata and report on the fact that something is AI or not. Perplexity got some updates this week. It now displays the results for temperature, currency conversion, and simple math, so you don't have to use Google. So if we go to Perplexity.doai and we simply type What is the weather in Austin, Texas, it just gives it to me on the screen right here without needing to show me like Google results or something. And some

rapid fire interesting news. McDonald's is ending its drive-through AI test. There was a lot of funny viral videos online where people were trying to order things from McDonald's using their AI ordert Taker, and it just totally messed it up. McDonald's is saying, All right, we're done with that. Now they're probably going to circle back around to it at some point. I think most fast food companies are. They just need the AI to improve a little bit better. This is something that I hadn't heard about until I came across Tim's video over on theoretically media where he shows off the first AI Cinema Camera. It's this giant box that looks like a really old school camera where you can film actual videos with it, and then it uses like animate diffusion in the background to convert the real videos into like cartoony videos. It has a little knob on it where you could sort of decide how real versus AI you want the final video to come out. And here's some examples from Tim's video of what they showed off. Uh

you know, there's a lot of inconsistencies and morphing happening with the characters in the foreground. I actually think the backgrounds look really cool and stylized. Again, I'm going to link up this video so you can watch Tim's entire video where he breaks down this camera. I just thought it was really cool, something I hadn't heard of until I came across Tim's video earlier today, and thought I would share that with you as well. And finally, here's a hard pill to swallow. It's a little robot pill that you could swallow, and it can examine the inside of your body using cameras and sensors. The pill is currently in clinical trials, and the CEO himself has played a vital role in testing, having swallowed 43 pill bots to date. I wonder if some of those are the same pillbot. I don't know. I don't want to think about that. Sorry. It's expected that this could be released as early as 2026, and I think biohackers are absolutely going to love this. I'd

probably actually try this if I could take a pill, and it can help me figure out certain foods that I'm sensitive to. Or why I get stomach aches sometimes, or why I get headaches sometimes. Or it can help me like diagnose the root causes of certain problems. I would probably try something like this. Would you? I'm curious. Let me know in the comments. Anyway, that's what I got for you today. I do think that, in my opinion, OpenAI is starting to kind of get past a little bit. We're seeing models come out that are better, large language models than what OpenAI has available with things like Claude. We're getting better video models, or at least models almost as good as Sora, that we're getting access to and can play with right now before Sora is even available. When it comes to image generation, I don't think many people use DALL-E 3 as their first choice. I think most people are going to either Mid-Journey or Firefly because it's built into Photoshop or using Stable Diffusion.

within tools like Leonardo or Playground AI. I just don't think Dolly is even the leader in AI image generation anymore right now. Although OpenAI is probably the most popular, most famous, most well-known company in the world of AI, and most of the sort of mainstream, normal people who aren't paying attention to AI every day sort of use the words AI and chat GPT interchangeably. The people that are in the know, that are really paying attention to AI, are seeing OpenAI fall further and further and further down, while other companies are getting more and more momentum and looking like they're going to really start to blow OpenAI out of the water. Either OpenAI is got to release something really interesting really soon that shows how far behind all these other players are, or I don't know. I personally think they could be getting themselves into trouble, but again that's what I got for you today. If you like staying in the loop with the latest AI news and tutorials and tools and things like that.

make sure you like this video, and subscribe to this channel. I'll make sure stuff like this keeps on showing up in your YouTube feed. Also, if you haven't already, check out Future Tools, where I create all the coolest AI tools, all the latest AI news, and I have a free newsletter. Join the free newsletter, and I will send you just the most important news and the coolest tools directly to your inbox. It's totally free, and by joining, you get access to the AI Income Database, a whole bunch of cool ways to make money with AI. Plus, I'm doing monthly giveaways for cool gadgets and software and stuff like that. You have to be subscribed to the channel and the newsletter to be eligible for those giveaways. So, if you're not subscribed in both places, I would really, really appreciate it if you did, and I will try to hook you up as best as I can. Thank you once again for tuning in. It's been a fun week. We got more exciting things coming up in the world of AI, and I can't

wait to share them with you. So, I'll see you in the next video. Again, really appreciate you. Bye-bye.

AI News: ChatGPT is in BIG Trouble

AI, No-Code, Tech, Futurism - I'm a tech nerd and talk about tech nerd stuff

AI News: ChatGPT is in BIG Trouble

AI News: ChatGPT is in BIG Trouble

스크립트