So I'm currently out in Boston for HubSpot inbound, and well, I'm kind of in a tiny hotel room with no desk, so I'm improvising and using what I got because I want to share all the latest news with you in the world of AI. Right now, we're sort of in the midst of conference season. This week alone, we had HubSpot inbound, we had Salesforce's Dreamforce event, Amazon had an event where they announced some new features, YouTube held an event where they showed off some new features, Snapchat held an event where they showed off some new features. Next week, we got Meta Connect, which I'm going to be flying out to that right after HubSpot inbound. Tons of conferences are going on, which means tons of announcements are happening, but I'm going to go ahead and jump in and start with an announcement that didn't come from one of these conferences. Let's start with talking about OpenAI. So this week, OpenAI put out this blog post here, an update on our Safety and Security.

practices now. This blog post talks about some of the new security measures they're implementing, as well as being more transparent. But probably the biggest news to come out of this specific article on the OpenAI blog here is this portion right here about establishing independent governance for safety and security. Now, when it breaks down the board here in this message, we can see the board is going to be chaired by Zico Couter and include Adam D'Angelo, Paul Nakasone, and Nicole Selman. One name that is absent from this board is Sam Altman. He is apparently stepping down from this Safety and Security committee and completely putting it in the hands of others. Now, the biggest news that's been going on in the AI world as of recently was last week's announcement of OpenAI's 01 model. It's a model that is much more effective at logic and reasoning and complex math problems and sort of STEM subjects. Well, this week, OpenAI announced that they've increased the rate limits for Plus and Team users by seven times. Now, admittedly, I have not

been home almost all month this month, so I haven't really had the opportunity to deep dive and play with OpenAI 01 myself. I'm super excited to get back to my house after Medic Con next week and really spend some time deep diving on this 01 model. That way, I can share with you some videos and tutorials and different ways that I'm using it and my own findings from playing around with it. But honestly, I just have not had the opportunity yet. However, I have been seeing some really, really cool stuff that people have been doing with OpenAI 01. For instance, my internet buddy here, Amar RI, actually managed to make a 3D version of the snake game in under a minute. So, check this out when I pull this up to full screen here. He managed to make this 3D version of snake that we're seeing right here on the screen in under a minute, and it actually looks really, really good. So, I mean, with coding tasks, my main go-to has been Claude Sonet 3.5 5 lately. I thought

that's done a really good job with coding, but I don't know, man. Open AI 01 seems to possibly have it beat with coding, so I can't wait to play around with it some more. And since Open AI 01 has come out, pretty much all the tools that have integrated the previous versions of GPT-4 have started integrating this new model. For instance, if you're a GitHub Copilot user, this week you can now use Open AI's 01 model directly inside of GitHub Copilot. If you're a fan of perplexity like I am, our event here announced that they've added a new reasoning focus on perplexity for pro users. So, if you're one of the paying Pro members of perplexity, you can now use the new Open AI 01 Mini model directly inside of perplexity. Now, it doesn't search the web yet and integrate that with your responses, and you only have 10 uses a day, as we can see right here with it. But exciting to see that perplexity, a model that I really enjoy.

using is adding all of the latest state-of-the-art models into their platform as well. Now there's been some weird stuff happening at OpenAI around this 01 model as well. For instance, people are claiming that OpenAI is threatening to ban them if they try to essentially jailbreak this new model. According to this article here, someone claimed that they received a warning email if they use the term reasoning Trace in conversation with 01. Others have claimed that the warning is triggered simply by asking ChatGPT about the model's reasoning at all. Here's a screenshot of an email that one person received: Hello, we're reaching out to you as a user of OpenAI's ChatGPT because some of the requests associated with your email have been flagged by our systems to be in violation of our policy against attempting to circumvent safeguards or safety mitigations in our services. Please halt this activity and ensure you are using ChatGPT in accordance with our terms of use and usage policies. Additional violations may

result in loss of access to GPT-40, etc. Now, if you have used this model at all, you'll notice that when you ask it a question, it sort of shows you the reasoning and the logic that it's thinking through as it's thinking through it. However, it's showing you like a summarized version of that logic. We're not actually seeing the complete chain of thought that it's going through; it's just like little snippets of the chain of thought. People are trying to really reverse engineer that and figure out what is this model actually sort of telling itself and going back and forth on. OpenAI is still trying to keep that hidden. I'm assuming it's because they're worried that other companies might sort of reverse engineer it. And apparently, if you do try to reverse engineer it, you could potentially get banned from using OpenAI products. Here's another really weird one. I've only seen one claim of this, so I'm not sure if this is fake news or not. It totally well could be.

but one person on Reddit claims that chat GPT actually initiated a conversation with them before they ever sent a message. We can actually see in this screenshot that they shared here. Chat GPT shared the first message here. How was your first week at high school? Did you settle in well? This person, Milo, here responded, Did you just message me first? Chat GPT says, Yes, I did. I just wanted to check in and see how things went with your first week of high school. If you'd rather initiate the conversation yourself, just let me know. They asked if it's a new update, and Chat GPT replied with, Yes, it's part of an update that allows me to check in or follow up on things we've talked about, like your first day of school. I can still wait for you to start the conversation if you prefer. Again, I've only seen this one claim of this happening. I feel like if this was happening a lot, we'd see it all over X, and we'd see other YouTube videos talking about this right here. This is

the only one I've seen, so I don't totally know how legitimate it is. I also don't understand when this would happen. Would it happen like when you open up the chat GPT window and start a new chat, and then the new chat just starts messaging you? I don't really quite understand how this would happen in the first place, which is sort of making my alarm bells trigger that this could be fake. Moving on to Google, there's been a handful of little Google AI updates this week, most of them actually on the YouTube side, but they did announce that they're going to start flagging AI-generated images directly in search later this year. Inside of search, Google Lens, and the circle to search feature on Android, if an image has metadata that states that the image was generated with AI, it's going to let people know as soon as that image pops up. Now, it has to have that metadata. It can't actually tell what's AI and what's not unless this metadata is actually sort of baked into

the image moving on to some of the YouTube announcements. There was a little YouTube event this week where they announced a bunch of new features rolling out. I'm going to focus on the AI features, including one that's coming out of Google Deep Mind. Google's rolling out some new video generation features directly inside of YouTube shorts. They're going to be integrating Google's Veo model directly into dream screen. You can see here they have a little create button at the top of the screen inside of the app here. When you click on create, it asks you to describe what you want. It gives you some sort of presets like vintage, watercolor, digital, etc. And in this example, they type cinematic underwater reveal of the Golden Gate bridge. And then they click on Create. And then what it does is it generates four different still images that it will use as the sort of starting frame of your video. So you can see it generated these four images here. You select one that you like, and then it does its thing and converts it into a video version of that.

image that you just generated. Here's another example of a dreamlike secret garden with vivid colors, etc. And here's a whole bunch of different examples of what it generated from that. But that's not the only AI features they're rolling out for YouTubers. They're actually adding an inspiration feature here, which can help curate suggestions that you can mold into fully-fledged projects. So they can type brainstorm video ideas, new types of art, and they click get ideas, and then it lets them explore video ideas, and then it helps them create the outline here, and then it starts generating thumbnail ideas for that video for you. So it really kind of looks like it's going to help across the board with ideation, outlining, thumbnail ideas, the entire gamut of the sort of creative process of ideation for YouTube. They also announced that they're bringing auto dubbing to YouTube, which is really cool because I can record a video like this in English, and it will automatically dub it and localize it to whoever's watching my video in whatever

location they're watching it in. That can open up so many more views for YouTubers because I record in English, but I could potentially have a huge audience in India or Japan or Germany or one of these places. And if you don't speak English, you're probably not going to watch my videos. But if this dubbing feature is in play now, it will automatically dub any of my videos to whatever language. That's really cool. Now, they announced a whole bunch of other features for YouTubers as well, but those are the main AI features. There's like a new hype button. They're changing the way YouTube videos look on Netflix, things like that. But for this channel, I'm kind of focused on the AI stuff for right now. Moving on to large language model news, the company Alibaba out of China just released more than 100 new open-source models. These come from the Quinn 2.5 family of models, ranging from a half a billion parameters all the way up to 72 billion parameters. They aim to cater to

a wide array of AI applications across various sectors, including automotive, gaming, and scientific research. They also unveiled a new text-to-video model as part of its Tongle Wang image generation family. I'm sure I butchered that. Now, I haven't actually seen this text-to-video model yet, but these large language models that they put out, specifically the 72 billion parameter Quinn model, is now apparently the best open-source model in the world, according to Bendu Ready. Here, she posted this screenshot with these various benchmarks, and we can see Llama 3.1 405B over here, Mistol Large 2, Quinn 272B, and this new Quinn 2.5 72B. In most of these benchmarks here, it's outperforming the rest of these open-source models. All right, I want to move on to AI video because there's been a lot that came out in the world of AI video this week as well, starting with this one that actually came last week, but I sort of missed it in last week's.

video 'cause I typically record these on Thursdays, release them on Fridays. Last week, I recorded on Friday but had to release it on Saturday due to my schedule. And this came out on Friday after I recorded that video, and that's the fact that Runway dropped a new video model. Now, Gen One was already sort of a video model, but this one is so much more improved over that model. I found this Instagram post here from AI searches that shows off some examples of what it's capable of. The top being the original video they uploaded, the bottom being the one that uses the prompt running on Mars while wearing a space suit. And you can see the original video is translated into this AI version of the video. You just upload a real video of yourself, give it a prompt, and it can swap it into this AI version. Here's a Porsche in a warehouse, but we can see here all of the different ways that this Porsche is now transformed into a new video. In fact, that initial video may even have been an

AI-generated video. I'm not sure. Here's somebody walking on a treadmill and all of the different ways that's translated with big ears translated into these various AI videos. All sorts of cool opportunities and things you can do. Again, we were doing this with Gen one a while ago, but this is so much better. This looks so much cleaner. Speaking of Runway, they did have some other news this week, including the fact that they made a deal with Lionsgate in the first team-up for an AI provider and a major movie studio. Now, if you're not familiar with Lionsgate, they're the company behind some really famous movies, movies like John Wick, The Hunger Games, and the TV network Stars. So, they have a very large database of high-quality films and TV shows, and Runway is using what they can get from Lionsgate to build a custom AI video production and editing model. It says here to create the custom model. Runway will train upon Lionsgate's library of more than 20,000 film and TV.

titles now. This is fascinating, especially considering some of the things that have recently changed California law around AI actors and AI video editing. I'm going to get to that a little bit later in this video. It's just really interesting the timing of this announcement with some of the recent bills that were just signed in California late last week. Runway also announced that they're opening up their API so developers can have direct access. So we're probably going to see a whole bunch of AI video tools coming out in the near future that just leverage Runway's API in the back end that probably try to claim to be a new video model, but they're just leveraging the Runway API. If you're not familiar with an API, it's basically a way for other software to access the abilities that something like Runway can do. So other companies can go and build their own tools to generate videos and use Runway's technology on the back end. Now the Runway API is still in Early Access. You have to apply for Access if you're a

developer, and I guess they're slowly rolling it out to people on the exact same day that Runway announced their API. Luma Labs' dream machine also announced they're making their API publicly available so that companies can build with their API and their video generator. One big difference about the Luma AI API is that, well, you can start right now. So, these AI video wars are heating up. These companies are obviously competing with each other and trying to constantly one-up each other. Runway says, Hey, we've got an API now; get on the waiting list. Luma says, Hey, we've got an API; go use it now. Again, I think all of this is just going to lead to a whole bunch of new video apps popping up running future tools. I see so many apps submitted. I'd say only about 10% of the apps that are submitted actually make the website because so many of them are just, you know, GPT-4, but somebody put their own user interface over it, or stable diffusion, but somebody put their own user interface over. They're the same things.

that already exists, but somebody's just putting a new website in front of it and saying, Hey, this is my new tool, and they submit it to Future tools, and I'm looking at it going, This is just something that already exists. These APIs are really valuable if they're sort of built into something that's more of a workflow where multiple steps happen, but they're not super valuable if you're just going to go and create a super thin wrapper around the API, essentially build a front-end user interface. But what the tool does is exactly what Runway or Luma does already. That's not valuable, but I think we're going to see an explosion of that pretty soon. The Chinese AI video model cing also got some updates this week. They put out an announcement video here on X. They released cing 1.5, and probably the coolest feature of this new model is this new motion brush feature, which I'll show you in just a sec here. But they've also improved the image quality, so it looks a little bit more

realistic here improved Dynamic quality with improved motion rationalization, such as eating and bird flights. Improved prompt relevance, so it can now handle more complex prompts. But like I mentioned, the motion brush feature is probably the coolest feature that they've just added. So, for example, they've got this image here of a moon, and we can see they highlight the Moon by selecting it, and then it says draw a path, and they draw an arrow, and we can see it generates a video of the Moon moving in the direction where they just drew that Arrow. We've got this image of a cat here. They highlight the cat, and then they draw a path that kind of goes up and then back down again for the cat, and then the video shows the cat leaping over the Bowl here. Here's some more examples I came across on this L scene X account here. Here's another one of a cat where they highlight the cat. You can see as they sort of move the mouse around, it highlights areas to help you select the right thing. So, they select the cat. They

make sure they grab the tail and everything, and then they draw a little line showing the cat leaping onto the table. And guess what? We get a video of a cat leaping onto a table. This isn't real-time 'cause I'm sort of scrolling, but if I press play, that's the speed that the video goes through right there. Here's another image of a soldier sitting in mud. They highlight just the soldier here, draw a little arrow to show what the soldier needs to do, and this is the video we get out of it: the soldier getting up and walking away. Now the soldier morphs a little, but it's still pretty dang impressive. And there's a whole bunch of other examples here on this X thread, like this car going left or right on the fork, another image of a soldier jumping. But I'll link this up in the description below so you can check out all of these various examples. This is really exciting. I'm hoping to see Luma Dream Machine and Runway ML bake some of these features in as well.

because this is really exciting. Also, this week, Amazon had a conference called the Amazon Accelerate Conference, I believe, and they announced a whole bunch of new stuff, including their own video generator. However, this video generator is just for creating product ads for your Amazon products. So, if we take a peek here in the back end of Amazon under this new campaign feature, they have the option to create a video ad. Here, you select video ad, and then it asks you which product you want to make the video around. It then gives you a preview of four different potential video options that you can hover over, and it makes a nice little product video showing off your product. You can then edit the video with headline and change the text on the screen. The only thing is, when you're using video, you're kind of using video on Amazon to stand out. If everybody has access to this feature, does anybody really stand out? Are all these videos going to sort of look the same? I don't know. We'll see how that plays out. Amazon also debuted project

Amilia, which is an AI assistant for sellers, and we can see some screenshots here where they ask questions like, What are the top things I need to do to prepare for the holiday season? and it gives them suggestions based on their products and their store. Here's another screenshot where it is actually pulling in data from their Amazon store. So, it says, How's my business doing? and then it shows off some key metrics right inside of the chat for that user. That seems like it'll be pretty handy if you're an Amazon seller and you want to make your life a little bit easier with AI. Snap, the company behind Snapchat, held their annual Snap Partner Summit this week, and they showed off a bunch of new AI stuff and cool new tech, starting with a new, well, you guessed it, AI video generation tool. It says here the tool will allow select creators to generate AI videos from text prompts and soon from image prompts. The tool will be available in beta on the web starting today. This is from September 17th here for a small

subset of creators, and I guess during the keynote, they didn't share too many details about this, just that it's coming and it's rolling out in beta right now. They also announced that Snapchat is going to be getting a Google Lens-like feature. We can see here in this screenshot somebody took a picture of a flower and then said at my AI, What type of flower is this? and then they get this response, That's a heliconia. Snapchat also showed off their new augmented reality glasses. Here's what they look like, and apparently, they have a large language model, they have a heads-up display inside of the glasses, auto-dimming lenses, they have hand tracking similar to the Apple Vision Pro, where you can navigate what you're seeing in your glasses with your fingers. They sound pretty dang cool. The biggest problem with them is, well, look at them. That's what they look like. We can see in this closer-up image that it looks like the batteries and all the processing are sort of behind the ear here, but it sounds like it'll be a fairly decent

blend between what you get out of like the meta Ray-Ban glasses and what you get out of something like the X real glasses, where you can watch videos and things like that directly in the glasses. Now, apparently, this is just like a beta version of them. They're not really ready for full rollout yet. They only have a 45-minute battery life at the moment, but ideally, if they can improve that battery life and, well, make them look a little bit better, these could be a hit. While we're talking about glasses, meta just extended their deal with the Ray-Ban smart glasses through 2030, so it sounds like we're going to continue to get new versions and new models of the meta Ray-Bans for at least six more years. And like I mentioned, next week is meta connect. I'll be at that event. Rumor has it they're going to show off a new augmented reality pair of sunglasses at that event. They're not expected to be released until 2026-2027, but we might get a sneak peek of what else meta is

working on in the world of AR and sort of smart glasses. I found this to be pretty interesting as well. The late great James Earl Jones, the voice of Darth Vader, actually gave permission to Lucas Films for them to continue to use his voice in future Star Wars films. So, we'll still get to hear James Earl Jones AI-generated Darth Vader voice in future Star Wars movies. Now, this article goes on to say that this has raised concerns among actors after last year's strike. Well, California has been doing a lot about that. In fact, this week Gavin Newsom signed eight new AI-related laws. He signed two laws that criminalize deep fake nudes. One making it illegal to create them, one that forces social media companies to establish channels for users to report them. Another bill, SB 942, makes it required that there are watermarks inside the metadata of AI-generated images. AB 2655 requires online platforms like Facebook and X to remove or label AI deep fakes related to elections, as well as create channels for people to report them. AB 2839 is for social media users.

who posts or reposts AI deep fakes that can deceive voters about upcoming elections. AB 2355 now requires AI-generated political advertisements to disclose that they were AI-generated. And you know, coming back to what I was talking about with James Earl Jones here, AB 262 requires studios to obtain permission from an actor before creating an AI-generated replica of their voice or likeness. And AB 1836 prohibits studios from creating digital replicas of deceased performers without consent from their estates. Which is interesting because in the Star Wars film Rogue One, they actually created an AI-generated version of one of the deceased actors supposedly without permission. If they had made that video today, it would have been illegal to do so. There's one major bill left that I've talked about in multiple videos. We've done a whole podcast episode about it, and that's the SB 1047 AI bill, which basically holds the model creators responsible for major catastrophic harm that might come from the model, even if they weren't involved in that catastrophic harm. As of the

recording of this video today. Gavin Newsom has two weeks to decide whether or not he's going to sign that bill into law or veto it, and it sounds like he's still kind of torn. But he did speak in an event, and he kind of almost implied that he might veto it. He said, There's one bill that is sort of outsized in terms of public discourse and consciousness. It's the SB 1047. What are the demonstrable risks in AI, and what are the hypothetical risks? I can't solve for anything. What can we solve for? And so that's the approach we're taking across the spectrum on this. He also said in that same conversation, We've been working over the last couple of years to come up with some rational regulation that supports risk-taking but not recklessness. That's challenging now in this space, particularly with SB 1047, because of the sort of outsized impact that legislation could have and the chilling effect, particularly in the open-source community. So he does sound concerned about the impact it will have.

on open source, possibly meaning that it'll get vetoed. We still have to wait and see. We'll know within the next two weeks for sure. Moving on to the event that I'm at right now, HubSpot Inbound. HubSpot just launched their new Breeze platform with a ton of AI agents and features to help you sort of manage your CRM for you. Like, the AI can basically run a lot of things for you. This Breeze platform includes four Breeze agents to get work done fast from start to finish, including content agent, social media agent, prospecting agent, and customer agent, plus 80 more features embedded across the platform. I actually saw the keynote where they announced Breeze here at HubSpot. I actually had the amazing opportunity to spend a little bit of time with the co-founder and CTO of HubSpot, Dharmesh Shah, where he actually showed me one-on-one some of the really cool features that they're working on. So, I'm actually really excited. This video is not sponsored by HubSpot, but I just was really, really excited to sit

down with him and see behind the scenes some of these cool new AI features that these guys are rolling out over at HubSpot. The chip startup Grock just made a deal with one of the largest companies on the planet, Aramco. They want to build the world's largest AI influencing center with 19,000 language processing units. Aramco is going to fund the development, and it's expected to cost in the order of nine figures. It blows me away, though, because it says the data center will be up and running by the end of this year and can later expand to include a total of 200,000 language processing units. Now, this is a direct competition with Nvidia, although this is more like Cloud compute. I don't think you can just go and buy a Grock GPU and run your own AI at home. You can go to the Grock website or use the Grock API and run the AI inference through their systems, and it's insanely fast. While Nvidia is more selling the actual hardware itself to the companies.

that are doing the AI both trying to accomplish the same things for the end users, just two different approaches. It'll be interesting to see which approach wins out in the long run. If you're a user of Slack, Slack AI will generate transcripts and notes from huddles. So, when you jump on Slack meetings, at the end of the meetings, you can get key takeaways and summaries of the meetings and things like that. There's been some frustration around LinkedIn. Apparently, LinkedIn has been training on people's data, and they don't make it super easy for you to opt out of LinkedIn training on the data. It says here, If you're on LinkedIn, then you should know that the social network has, without asking, opted accounts into training generative AI models. LinkedIn introduced a new privacy setting and opt-out form before rolling out an updated privacy policy, saying the data from the platform is being used to train AI models. If you're a fan of generating music with Suo, they just rolled out a new feature where you can exclude styles, so you can exclude specific instruments.

specific styles or even specific vocal styles such as male or female vocals. If you have an Apple Vision Pro, they just rolled out Vision OS 2 for your Apple Vision Pro. It's got a new feature where you can take a 2D image and have it sort of turn it into a 3D image inside of your Apple Vision Pro. There's some new hand gestures that they added into it. You can now rearrange icons on your home menu and a handful of little quality-of-life updates. So, anybody that bought the Apple Vision Pro, it might be worth, you know, pulling it off your shelf or pulling it out of that drawer, playing around with it for another week before putting it back in the drawer. And finally, in the last bit of Apple news, last week or a couple of weeks ago, whenever that was, Apple announced their iOS 18, which didn't have Apple intelligence but said that 18.1 was coming soon with some Apple intelligence. Well, this week, 18.1 did come out with some of the Apple intelligence features. Now, you do have to

have an iPhone 15 Pro or better model to use these. It says here users must manually enable the feature by going to settings, Apple intelligence, and Siri. Join the Apple intelligence waitlist. I don't know how quickly you get access to it once you get in on the waitlist, but apparently 18.1 is rolling out now, and people are starting to get access to it. And that's what I got for you today. I know it was a ton. We're in conference season. There's a lot happening right now. I know this is a bit different of a video. I am literally in like the world's smallest hotel room. I don't have a desk. I don't have anywhere to set my computer down. I've literally been holding my computer in my hand this entire video. I'm sitting on my bed in my hotel room. I got this headset because I was sick of holding my mic up like this when I was in hotel rooms recording these videos. So this one is totally different than normal. Hopefully, you got the same amount

of Val out of it. I'm just trying to keep myself up with the latest AI news, and if I'm keeping myself up with it, might as well flip on the camera and help you keep up with it as well. So, I hope I achieve that for you. If you like videos like this, make sure you like this one with the little thumbs-up like button and subscribe to this channel. That will really, really help me out, and it will ensure that more videos like this show up in your YouTube feed. I make a lot of AI news videos, a lot of AI tools videos, a lot of tutorials, things like that, that I try to keep you tapped into the AI world and also show you how to use all this stuff. Now, typically in these news videos, I actually like to demo a lot of the tools and show them off and actually use them myself. That's really, really hard when I don't have a desk and I'm working on a laptop from my bed in my hotel room. But I promise once I get back home after Meta Connect in a couple

weeks. It's going to be back to regular scheduled programming with a lot more tutorials and tool demos and what you're used to from my channel. I'm just almost done with conf season. I'm having a lot of fun doing it, but I'm having a harder time making as many videos as I'd like to. But there's a lot more coming, and I can't wait to test out more of these tools and show off what I learn and what I find out playing with them all. So again, I would really, really appreciate that like and subscribe to this channel. And if you want to know about all the latest and coolest AI tools and the latest AI news, check out futur tools. That's the site where I curate all of it, the tools, the news. I even have a newsletter that's totally free where I'll update you in your inbox with just the most important news and just the coolest tools, totally free, all available, all over at futur tools. Thank you so much for hanging out with me, nerding out.

with me talking AI news with me. So much fun stuff happening, so many fun toys for us tech nerds to play with, and I am so excited to share even more with you. So, really, really appreciate you. See you in the next video. Bye-bye.

AI News: AI Video Creation Explodes in 2024

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

AI News: AI Video Creation Explodes in 2024

AI News: AI Video Creation Explodes in 2024

스크립트