2024 AI : 10 Things Coming In 2024 (A.I In 2024 Major Predictions)

So today we're going to be taking a look at the year of 2024 in AI. Everybody knows just how crazy the year 2023 was. And as we wrap it up, we really do need to look forward with what we could potentially be seeing in the future. Now, as someone that has been in the AI space and as someone that has been looking widely at all of the trends and following closely what we're going to be seeing in AI, I think it's important to make sure that we know what could be coming in the future and, and how we can best adapt so that we aren't caught off guard by any new AI technologies or trends. But as always, predictions are only predictions. And these companies could come out with some crazy technologies and major breakthroughs that just like before, we didn't expect. So let's get straight into this. So one of the first things I think that we're going to see is the rise of AI influencers and verification techniques. So this is a Spanish influencer created entirely by AI that generates its modeling agency £9,000amonth with 200,000 followers. Now, I don't think that this in and of itself is actually an issue. I think if people want to create AI influencers and these kind of, I guess you could say, cartoon characters or figures on social media, it could just be a way that technology is developing. This for me isn't really an issue. I just find that this is an interesting way for people to market stuff or what is going to be an issue, though, is the rise of fake influencers in terms of what I'm about to show you in a couple of other slides. But I do think that this is going to be something that is far more popular than we think, which means that in the future, not only are there going to be influencers, AI influencers that are going to be trying to earn money, but I do think that we are going to be seeing something where a large majority of the Internet isn't going to be real. I mean, for example, if I did see this person on the Internet, I would assume that if they had 200,000 followers followers, that they are real. So because of the rise of AI influencers, where you have people that have around 200,000 followers and, you know, it isn't disclosed, I do believe that this is going to fundamentally change the Internet. So right now, on Instagram, on Twitter and on YouTube, you could be an AI influencer and you don't need to disclose absolutely anything at all. There's no legal requirements, there's no you know, European laws that say you do. There's no digital laws, nothing says that you need to disclose this information. But, but I do think that in the future this is going to change, which essentially means that for a lot of services that you do want to use, you're going to need to have many additional steps in order to verify your account. So, for example, YouTube actually is talking about disclosure requirements and new content labels. So they said, we believe it's in everyone's interest to maintain a healthy ecosystem of information on YouTube. We have long standing policies and that prohibit technically manipulated content that misleads viewers and may pose a serious risk of egregious harm. However, AI's powerful new forms of storytelling can also be used to generate content that has the potential to mislead viewers, particularly if they're unaware that the video has been altered or is synthetically generated. To address this concern, over the coming months, we'll introduce updates that inform viewers when the content they're seeing is synthetic, which is really, really good, because you won't know if the content is synthetic. A lot of the videos that you're watching, sometimes the voice is AI. And only when the AI messes up is when you're able to tell that it's synthetic. I mean, it says specifically, we'll require creators to disclose when they've created altered or synthetic content that is realistic, including AI tools. And when creators upload content, we'll now have options for them to select and indicate that includes realistic altered or synthetic material. For example, this could be an AI generated video that realistically depicts an event that never happened, or content showing someone or saying something that they didn't do. So this is really, really important because now if you go to Twitter, if you go to Instagram, remember the AI image of the Pope? There were so many things floating around on Instagram that we didn't know if it was real or if it was fake. And this is a real problem because for the average person who's not in the media spotlight, like you and me, we don't have an issue because our faces aren't plastered across and we aren't in the spotlight. But for people you know, who are famous, they continually get altered. Their likenesses are used in these AI systems and it's a real issue when these things are happening. Now, you might be thinking, okay, they're famous, they should have signed up for that. That's what they signed up for. That's that life. But think about this, okay, so imagine you're a normal person. You've got a small account. And then someone starts to use that technology to make you do something that you haven't done and make you say things that you haven't said. So this is why I'm saying that the technology isn't just affecting famous people where their faces have been trained by websites like Midjourney. This is going to affect a lot of regular people and this is going to be a real, real issue. Okay? This is a new AI video driven avatar animation. And you can see just how realistic this looks. Okay? And the problem is, remember, the technology is only getting better. And the next slide that I'm about to show you is going to show you just how crazy this is. Now, I know people haven't talked about this. This was supposed to be included in a video. I'm not sure if I uploaded the video yet, but take a look at this. Okay, so we have here AI influences and verification. So this is a project which essentially you just need a single image and a driving pose and you can essentially make someone dance. You can make them do pretty much anything. So this is why I say that in the future verification is going to be extremely important because imagine someone takes an image of you, they use a driving pose and then they essentially get that person to do anything. So I think that that is going to be a major, major, major issue. You can literally see like look at the start of this clip. Okay, let me show you guys the start of this clip. Okay. You can see that you're rotating the character and they can pretty much do anything. Like that is absolutely incredible. Like imagine seeing a picture of yourself being rotated and then manipulated to be doing some crazy stuff that would freak you out. It definitely freaks me out. But this, this is why I say in the future, I'm pretty sure all of these platforms are going to update their guidelines so that influencers and verification is something that is really, really important. Now on the topic of scams and verification, we do need to talk about real time text to speech. So what the problem is with current text to speech is that a lot of the times it is limited in terms of how it sounds and how fast it is. But these systems are getting faster, the APIs are getting better, and the latency is getting shorter and shorter. The so we're about to see here is going to be Play HT's Turbo, which is the fastest generative AI text to speech API, which essentially means that if someone back engineered this or if other systems are even close to this, we could have real time scams just being Absolutely crazy. So you're about to see that this is absolutely insane. So I'm going to get you guys to listen to it. But don't listen to how the quality of the voice sounds. Although it's good, what you need to listen to is of course, the latency in which it is played. And the latency essentially just means how fast it responds to your message. So let's take a listen to this.

Hello world. Hi, how are you? Hi there. I'm calling in regards to the purchase.

You made last week.

Hello, Play support speaking please. Hold on a sec, let me just pull up your details real quick. Can you tell me your account, email or your phone number?

So what you just heard there just goes to show you how crazy this is. We got like 100 milliseconds response time in a simple message. So if you combine that with a model that's able to immediately read your message and able to just be completely trained on thousands of persuasive scams in order to convince someone that's also trained on a voice of one of your relatives, that could be a serious, serious issue. And the problem is, remember, you only need a couple of seconds of someone's voice in order to train a model that is exactly like yours. Software like 11 Labs can literally take 3 seconds of someone voice and make them say absolutely anything. Now it's not always effective, but the times where it is effective is really incredible. So with that being said, this is going to be a real, real issue in the future and I'm not sure how we're going to combat this just yet. So this is a real issue and it's happened many times, many times. Now once again, this is another thing that we also need to talk about because Google did this in 2017 and I've brought this up many times before, but once again, I'm going to show you that scams are going to get supercharged. Okay, Longer are going to be the days where you receive a fishy email. These scammers are going to be using AI technologies and it's going to be there really, really convincingly, which is a huge problem because usually you think, okay, that's a dodgy email. But if you have a relative that messages you or or is talking to in real time on the phone and is saying you need to do this, you need to do that, how are you going to verify that? Maybe you're going to have to have some kind of secret code, some secret code word to be like, okay, this is our new code word that's Something that could work. But take a look at this. Okay. Because this was something that Google sat on for quite some time and they haven't really brought any updates. But I wouldn't be surprised if there's someone somewhere working on this kind of technology.

So let's go back to this example. Let's say you want to ask Google to make you a haircut appointment on Tuesday between 10 and noon. What happens is the Google Assistant makes the call seamlessly in the background for you. So what you're going to hear is the Google Assistant actually calling a real salon to schedule the appointment for you. Let's listen.

How can I help you? Hi, I'm calling to book a women's haircut for our clients. I'm looking for something on May 3rd, so I. Give me one second. Sure. What time are you looking for? Around at 12pm we do not have a 12pm available. The closest we have to that is a 1 15. Do you have anything between 10am and 12pm depending on what service she would like? What service is she looking for? Just a woman's haircut for now. Okay, we have a 10 o'. Clock. 10am is fine. Okay. What's her first name? The first name is Lisa. Okay, perfect. So I will see Lisa at 10:00 clock on May 3rd. Okay, great. Thanks. Great. Have a great day. Bye.

Remember, this is old technology. So will Google release this? I don't think they will because Google has a tendency to make amazing stuff and then just simply sit on it like they did with ChatGPT, the architecture. But yeah, so this is something that is worrying. Definitely, definitely worrying. And if you think, okay, they're not going to get me, I know, I, I know exactly what my friend's voices sounds like. Yada, yada yada. You have to remember that if they don't get you, they are going to get someone that you care about and they might get one of your friends. So that is of course a problem that does need to be solved. So yeah, I'm definitely going to be very, very skeptical of that. Now here we have advanced robotics. Now advanced humanoid robotics is certainly going to come. And in the last couple of days we did get a few updates regarding that with Demis Hassabis, the CEO of Google's DeepMind, stating that that future versions of Gemini will be combined with physical robots because that is how the next step in AI AGI is going to be because it needs tactile feedback and physical senses. So you can see right here that this, if you don't know what the robot on screen that you're looking at is this is essentially OpenAI's backed investment robot, essentially where this is going to be one of the robots that they combine with artificial intelligence. Now, I do know that robotics is by far one of the hardest things to do, and I know that because even previously what we saw was that this was rumored to be announced in summer. That's what they did state on the webpage, but that's since been removed and they've since said that this is just something that they're working on. So whatever demo that there was supposed to be in summer hasn't come and I don't think it's going to be coming anytime soon. But essentially, humanoid robotics are really cool. And I'm going to show you guys the web page to show you why this is going to be really, really interesting, because large language model, multimodal monocle. But what happens when you combine that with a robot that has a physical body and is able to see, feel, interact, smell and do all these kinds of crazy stuff. So this is the website for neo. Now, of course, you do know that there is of course, Tesla's humanoid robots. There are many other factories, but for now we're just going to focus on neo. So essentially this is safe, balanced and smart. NEO is your intelligent Android assistant. So coming soon. They said it's in development. It was supposed to be. They were supposed to do like some press release thing. They are doing something very soon though, so I would pay attention to that as well. And if you scroll down, you can see that. Although this is really, really cool, I'm going to show you guys something that's really, really cool. NEO is essentially combined using AI, so it says, using embodied artificial intelligence. And NEO will understand its environment deeper. Thanks to the fusion of their AI senses and their physical body, NEO continuously learns and improves, becoming smarter and more capable over time. The more you interact with them, it's going to feel more natural and it's going to be able to understand how to handle fragile items. Creating seamless collaboration and immersive human Android interaction. So this is far, far, like one of the most interesting things. And I'm not going to be sure which company is going to get to it first. I just know that a company is going to be likely to combine a robot with some kind of large language model. And there's probably going to be, you know, some kind of human robot that might be talking back, you know, walking around, kind of like a real human. If I see that in 2024, that wouldn't be surprising. We did see small scale versions of this where people had their own robots that had like APIs and stuff like that that were, you know, being controlled. But something like this, that's really, really professional is going to be super, super fascinating to see. So, yeah, I wonder when this robot does come out. And remember, this isn't a random company. The robots that they've done before are actually really, really, really cool and they're really effective. You can see a robot here that's packing some stuff in. And the robots that they had from like 2017 and 2018 was super impressive in terms of what they could do. The only problem was is that they were really, really expensive. So yeah, there's other companies as well, like a GR1 robot, which is a lot cheaper. But it's, it's creeping up on Boston Dynamics. Like Boston Dynamics is definitely by far in the lead, but a lot of these companies are creeping up. So I wouldn't be surprised if this does happen unless this does come out. And of course is of course Nvidia Eureka. And, and essentially what's probably going to happen is we're going to have advanced robots moving at an accelerated place because we have something called the Nvidia Omniverse, where essentially you can basically get synthetic data through millions and millions of tries. And then you can use that synthetic data to train the robot, whichever one it is, in a virtual universe. Then once it figures it out, you can then transform that and transfer that into the physical robot so it knows exactly what to do. And essentially they taught a robot hand how to spin a pencil, which is definitely a very complex task for a robot hand. So this is something that was called Nvidia Eureka. We did have something like that earlier this year. So yeah, this is of course GR1. The GR1 robot, like I said before, is crazy. It's crazy because this one as well, it's going to be really, really cheap. This is Unitree's robot. I'm not sure if this is J1, but this is unitreeze robot. But yeah, this is the Unitree humanoid robot. And I think this one is supposedly going to be at around $80,000. But a lot of these companies you're seeing right here, like he literally kicked the robot and it's still walking, it's still balancing. That is a truly amazing feat. I mean, a lot of these humanoid robots, it's actually much harder to do than it looks. And you're seeing he's giving it some real good kicks and it's able to be really, really sturdy. Imagine what we have in the next 10 years when you like just, you know, like try and even kick it and it like just dodges you, like just like, I don't know, just analyzes what you're able to do immediately. So I think it's going to be super, super interesting. You can see as well, you know, the gears, you can see the vision capabilities. If we sc. Scroll back, you can see like the eyes and stuff like that. So, yeah, it's gonna be absolutely incredible if advanced robotics does make some massive strides. Hopefully we do. But I do think with advanced humanoid robots, we're going to see one of them combined with a large language model. And of course this is going to lead to some real time conversations with these robots, which is going to be really incredible. So of course we did talk about how this does essentially mean, you know, scams and stuff like that. But one thing I really do want to do with an AI is I really want to have like a real quick conversation that has no latency. So for those of you who use chat GPT and you've used GPT4 and if you've used GPT4's feature where essentially you can have a conversation with it, you know that one of the problems which makes it not that great is that it does take a minute to generate the information and then it does take a minute to, you know, speak that back. It would be really cool if you can have an AI that you could speak in real time, like real time latency. So you instantly get that information back and then it instantly says it. So that would be something that's really cool. So I think. I'm not sure if they do it now, but as the message is coming out, then they're talking and it's just really real time. Because when you have a conversation with a real human, you do talk really quickly. You're not like, okay, give it a minute, you know, you don't like, wait like five seconds with Siri. And that's going to be something as well. Now the next one as well, you know, apart from real time conversation, is the first video game NPCs. And I really don't know how this is going to go down because I do believe that this is going to be just. This is going to be crazy. Like this is going to be one of the most insane things. If we see real agents in games with generative AI, I believe that this is going to be absolutely incredible. I don't know why Steam is banning AI games So that's definitely hard for some publishers, but I'm sure that there's going to be some kind of indie developer solo publisher that is going to do this. So what you're about to see is a demo of the first video game NPCs where I'm not sure what they combined it with, if it was ChatGPT's API or something, but it's crazy. And I'm gonna watch, I'm gonna show you guys this. So you can see why this is a little bit disturbing to say the least. But let me know your opinions and then I'm gonna continue this.

So why are you here? I'm not sure, but I have a feeling that I'm here to fall for some kind of destiny. Okay. Do you think that we live in some sort of simulation? Maybe. I just don't know for sure yet. Do you think that your real?

I'm real?

I'm not sure what to believe anymore. What if I told you you're in the matrix, buddy? That would explain a lot, but I'm not sure I want to believe it. What if I told you you're an AI? That would explain a lot. But why would I be here? Maybe because you're programmed to be here. That would make sense. But why was I programmed this way? I don't know. Maybe someone thought of. Maybe. Why would they think to create me like this? Well, I don't know. What kind of problems do you have? It's not that I had problems, it's just that I don't understand why I'm here. Well, I can explain why you're here. You're in a matrix. That makes some sense, but it still doesn't tell me what I am supposed to do. Now the thing that. Okay, so my story is basically that you're an AI.

Oh, what do you mean by that?

You're an NPC in a video game.

That sounds frightening. Is that the truth?

It's the sad truth.

I see. This is a lot to take in. Are there any other answers I should know?

Not really.

This is very overwhelming. I'm not sure what to think or feel right now.

Yeah, I mean, you're not really real.

That's hard to process. Does this mean I don't exist?

You exist. This sort of.

What does that mean? Am I real or not?

It means that, you know, after I close this game, you don't exist.

Oh, I see. So I'm only here for as long as the game is running. That's difficult to comprehend.

Yeah. And then once you walk past this border.

So, yeah, that was something that I found to be completely just the most. One of the most fascinating things I've ever seen because these AIs were having an existential crisis about not. So, yeah, it was definitely fascinating to see this video because these AIs were having a complete meltdown, slash, you know, existential crisis about the fact that they weren't real. And I'm extremely, extremely concerned as how this is going to go down, because are we going to see like these LLM agents or whatever kind of AGI agents in game? Are they going to do some crazy stuff? Are they going to be some, you know, experiments? Are people going to say that this is unethical to create, you know, these kind of beings? I mean, it's a whole nother, you know, topic. There's numerous examples of AIs with vision and API, you know, just talking and stuff like that and really cool. But we haven't seen them being able to do and lived in their own ways. We've seen certain papers about it. But running some kind of experiment where you can do that would be truly, truly incredible. So with that being said, I do think that someone out there is going to create the first video game. I don't think it's going to be on Steam. I don't know what platform it's going to be on. It's not going to be on Xbox or PlayStation, because the time to get those games through the processes is definitely going to take a while. So I do think it's going to be like a solo kind of developer where you can play like a simulated game. But either way, I think that that probably will come. And I think what will happen as well in terms of the wider AI industry, in terms of gaming, is that maybe one person is going to do it and then monetize it, and then that could spark the whole entire revolution. But it will be interesting because, as you know, if an AI does something that's unpredictable or something that the companies can't predict, there's no way to ensure, you know, any kind of safety. And I'm pretty sure video gamers have always figured out bugs and stuff like that. So if they get the AI to do something that's, you know, completely out of character, it might invoke some kind of legal trouble. But it also does, you know, make, you know, games like no Man's sky, which was supposed to be a completely generative game, it doesn't make stuff like that possible. So if, you know what no Man's sky is it was basically a game before that was essentially supposed to be a game that was. So this game was essentially supposed to be a game that was generated generatively in a sense that, you know, every time you explored an island or something, it was supposed to be completely new and everything was supposed to be new for the user and it was supposed to be impossible for you to get a duplicate world basically like this generative game. But I think it was just a little bit before its time. But now with generative technology coming into effect, I think that this game is definitely going to, you know, might move in that direction, you know, combining its, you know, the new generative capabilities with open world generation. I think that this is going to be something that happens, if not to this game, but, but to maybe a smaller indie developer. So seeing something like that would be super, super cool. And like I said, I think it's going to take one company to see massive success for everyone to sort of pile on and think, you know what, this actually does make sense. Because right now the only reason people aren't doing this is because Steam basically, you know, they banned it. I don't know for what reason. Xbox and Microsoft, you know, it's really hard to get a game on there and yeah, it's kind of hard to build that stuff right now. But if there is a generative AI game that is something that I 100% will be playing, because of course it's not been done, then of course we have autonomous agents. So this is a clip from Sam Altman and he's actually talking about autonomous agents and stuff like that. So here we have a clip of the MIT C Cell Director Daniela Ross and OpenAI CEO Sam Altman discussing the future of autonomous agents. Because autonomous agents are truly going to be the future of AI. And I do believe that if we can figure them out, then it's going to be truly, truly incredible. So let's take a look at this clip and then I'm going to further discuss what's going to happen.

Start slow and the answer will be like, start slow and you know, set the training wheels very low. And then as we understand the risk surface and where it's okay and where it's not, we can make thoughtful, risk adjusted decisions about how to move those training wheels up. But the upside here is going to be tremendous for people. If you think about having an agent that you can give a task to to help you with whatever you're trying to do and be confident it'll go do it, that's awesome.

Daniela I really like the idea of agents, whether embodied or not, that can help us with cognitive and physical work. The embodied agents would help us with cognitive work, sorry, with physical work. And the non embodied agents would be like our, our super assistants. And so the question is, how do we get to the point where these agents can actually deliver on the tasks that we have to do? Many of these tasks happen in the physical world and part of the, part of the capability that we don't quite have yet has to do with interaction, has to do with interaction with people and with the physical world. And so in the future I really imagine progress in connecting the tools that we have right now with more extensive capabilities that will give the agents the power to do better interaction with the physical world. That will give the agents the ability to do common sense reasoning and to do human understanding. Perhaps at some point we will even get to agents who can understand us emotionally. But there is a long way to go before we get there.

Yeah, that was the clip. And like I said, I think that of course agents are going to be really hard to do because the problem with agents is that not only are they autonomous, they need to make the right decisions consistently over a long period of time. So for example, if it's, you know, let's say for example, it's making a video, it needs to research, get the script, get a voiceover, edit it, upload it, you know, do the titles, all of that, and publish it and of course choose the right file and all, you know, like there could be bugs that could happen, you know, the computer could freeze like all of these crazy different things that could happen along that journey could stop an autonomous agent from being successful. So it's going to be interesting because of course it does need as well advanced capabilities. So it's going to be interesting to see how autonomous agents do work in the future and how we manage to tackle that problem. Because I do know that of course some companies, that's just strictly their focus. So whenever that breakthrough does happen in 2024, if it does happen, it's definitely going to change everything.

The answer will be like, not only.

Will it change that, essentially it's going to change the Internet. Because this was a demo, I'm sure I was sent this demo. So this is self operating computer, some kind of AGI autonomous agent. And you say go to Google Docs and write a poem about open source software. And then you can see it going ahead and you know, going to google.com. so I'm pretty sure that is just chilling. And then You've input it in and then it goes into Google. Then essentially it, you know, opens Google Docs and does, you know, exactly what it needs to do. But, but this is what I'm saying, okay? This is going to change the Internet fundamentally, because if you have autonomous agents going around on the Internet creating documents, doing stuff, doing work where normal people usually would, in terms of how is that going to work? Like, are we going to have AIs watching videos so they can learn to do other things? Are there going to be autonomous agents that, if you don't prompt it to do a task, maybe by itself it got bored, decided to go do a task, decided to watch some videos. I mean, how is that going to be. This is why I'm saying that autonomous agents is going to change the Internet in terms of especially AI around the Internet, because we're definitely going to see an invasion of a lot of bots. So if you thought you had bots on the Internet before, there's going to be real, real bots in the future. So, yeah, the autonomous agents thing, I think it's cool. Definitely, I do want to see it works. But I think in 2024, hopefully we do get that breakthrough. Then, of course, this one's pretty unfortunate, and I didn't really want to talk about this, but literally, as I was making this video, there was a kind of update that actually did happen, which was pretty crazy. And what we have, guys, is AI warfare. Now, I talked about this in a video a couple months ago. I didn't get that many views, but the US Air Force is moving pretty fast on AI piloted fighter jets. And, you know, since I know a decent amount about fighter jets, because it's just something that is, you know, is interesting to me, I just find it really cool. I know that for a fact that this would be something that is rather effective because the advantages do weigh out the, you know, disadvantages in terms of AI, you know, you know, fighter pilots. And this is an excerpt from the article, and it says here that once you get through the process of connecting an AI to a supersonic fighter, the resulting maneuvering is endlessly fascinating. We have seen things that make sense and completely surprising things that make no sense at all. Thanks to our safety systems, programmers are changing their models overnight and we're engaging them in the next morning. This is unheard of in flight control system development, much less experimentation with unpredictable AI agents. So the only thing about this is, of course, is that you don't want some AI to go rogue and then, of course, use the flight a jets kill people. That of course would be awful. But we do have the fact that every time a human life is on the line, that is something that we don't want. We don't want, you know, men and women, you know, being in these environments where they have to be on the battlefield. It would much be better if it was fought with AI. So a problem, okay with fighter jets is that of course it does have to be flown by human. Sometimes there are two people there, you know, you've got your wingman in there. And then essentially what we have of course is G lock. So if you don't know when you are flying a fighter jet, one of the things that can happen is, is G loc. And this is a fighter passing out. So this is a, you know, a fighter jet, fighter pilot passing out. And you can see loss of consciousness due to the acceleration of the body is better known as G lock. And it incurs when increased G force, you know, cursed force due to gravity is applied on the body causing a loss of consciousness. So the blood like drains from your head and you pass out. Okay. And of course this is not good because this is something that does happen to pilots who are flying at, you know, like if you're in a battle, if you're in a dog fight, you. You're flying around the G forces in order to maneuver. If you want to turn left or right really quickly, the G forces can hit you. And if you're not trained enough, you can pass out, you can die, you can crash the plane. Doesn't happen as much now. But imagine having an AI system that doesn't experience G lock because it's just an AI system. It's empty. It's not going to have fear, it's not going to have hesitation, it's going to know exactly what to do. And what's crazier about this is the fact that with this is the fact that I'm going to talk about the slide in a second is that imagine we could simulate. All we need is like the Nvidia Omniverse, which is like a one to one reputation of really good physical environments. So all we would need guys, is we would need just a robot, okay? Like a robot fighter jets. And we would just need it to fight itself in a bunch of different battles. And once it fights itself in a bunch of different battles and it figures out what strategy the enemy is going to use or how to dogfight in the best way. Just like we did with Alphago. And if you don't know what AlphaGo was. AlphaGo was essentially a board game in which the AI played itself millions and millions of times. And because it played itself millions and millions of times, it knew exactly what to do every single time and developed unique crazy strategies that we never would have thought. Do you not think an AI system, if they manage to keep training it just on flying jets perfectly, it's going to be just the best fighter jet on the planet. It's going to be absolutely crazy. There's no way that humans are going to be able to beat that. If it does become, you know, that effective. If there is some kind of reinforcement learning loop to where it gets super effective. I mean, it's basically going to be like an autonomous drone. And I think it's going to be cheaper as well because not only that, you do have to have, you know, extensive training, extensive safety, all these kind of courses and stuff like that to train a fighter pilot when you could literally just buy, let's say for Lockheed Martin, these jets are super expensive. I don't know, I think they're like 50 to 100 million. Imagine they just, you know, have one which you could buy 100 million. It's a fighter jet, autonomous, and it's the best in the world. Like, you know, that's crazy. That is, that is of course, like, you know, something that we haven't had to think of yet. So will be interesting to see how that happens. And while I was making this video, there was literally an autonomous reusable fighter jet that can blow up drones. And this one is absolutely incredible. And I literally have been watching this on repeat again and again and again. And I won't include the music from this video because I don't know if it's going to be copyright. And I would rather this video stay up so you can watch this, but I will leave a link to this in the description. But this is a really reusable, cool like thing that can blow up drones. So essentially you just deploy it. It's autonomous. You can have a network and essentially like a network of them. So you can have like a swarm of these flying around and it's ready to launch within seconds. You can of course redeploy as well. So it's reusable and essentially it deploys from this. It's autonomous. It finds out what's ever going on and then essentially it just flies off the target and, you know, intercept something. Now you might be thinking, how is that any different from any other system but this? Guys, it could be completely different. Imagine having this On a completely different level. So definitely check out the description of the interview because a lot, there's, there's a lot of information. There's like a 10, 15 minute video on Bloomberg that you do want to see. So yeah, this is, this is crazy. Like where we have this, you know, kind of fighter jets that's flying around autonomously and it does it. And then what's crazy about this is that after it finds the target, intercepts it, is that it literally just returns to base. Like, that is absolutely incredible. Guys, mission accomplished. Vertical takeoff landing enables reuse of the Roadrunner. I think future is going to be crazy and I'm trying to talk fast here because I don't really want to make this video so long, but now we have text to video. So text to video in 2024. You know, even before 2024, like as I was making this video pick and release this, and this is essentially picker 1.0 and I'm going to be doing a full video on this. But trust me when I say it is absolutely incredible. I was really surprised, like genuinely surprised. I always thought that, you know, text to video was going to be something really hard to do. But Picker, I don't know what technique they're using, but the consistency, the fidelity, the quality, the consistency of like just all the frames is just really, really, really good. So I wouldn't be surprised if in 2024 we have full on text to video because even Meta released emu, which was a new text to video. People haven't heard about it that much, but it's actually also really good as well. So this is going to be something that's absolutely crazy. And the video on this will be coming out soon. But like I said, 2024 is going to be absolutely insane in terms of text to video because text to image has pretty much basically been solved at this point. The only thing that's left is in terms of more accuracy and smaller things like that. But I think text to video will have its breakthrough moment and it probably already did, just before 2024 started. And remember, this video is going up in 2023 and the only reason that there wouldn't be any more breakthroughs is because of course we do have Christmas and New Year's and there is going to be some time off. So it will be interesting to see Q1, Q2 in the entire year and look back on this video and see where things do go. Then of course I did actually have next gen new AI architecture with Google Gemini. But of course Google literally released Gemini yesterday. So that is going to be something there. And of course we do have GPT5 and how that is going to be comparing to Gemini. Will these systems like OpenAI's new system, like Gemini or GPT5, I mean, Google's Gemini and OpenAI's GPT5, will those systems, okay, will those systems use new architecture in order to get what they need to get done? I mean, there was something recently, language agent tree search that. I don't know why anyone hasn't really talked about this, but this one did. Insane. Okay. On the. I can't remember which benchmark it was, but you can see that GPT4 with this method got a 94.4 and then GPT 3.5 went from 46 to 86.9. So these different frameworks and stuff like that for thinking are going to be revolutionary. I mean, we know that search and learning is going to be a new architecture that they're trying to include to these large language models. So, I mean, the future, I mean, how is that going to work? Are they going to completely just change the architecture of how things work? I do think though, internally that GPT5 or GPT4, whatever OpenAI is doing, I do think that they definitely have some secret sauce because many companies up until recently couldn't catch up, but of course now they have in fact caught up. So. But it will be interesting to see how this stuff does actually manage to go in the future because we know what Gemini is like. We haven't of course got it yet, but we know that Gemini is good. And then of course we know that GPT5 is currently being worked on and how that is going to happen. And then of course, another thing that we do have is of course we have OpenAI's Q star. Is this going to come out? Is this going to be real? We did get some recent leaks about that stuff, but it will be interesting to see how that comes as well. And another thing that I do think that will happen in 2024 is of course, you know, offline large language models. And funnily enough, they released Gemini before and you know, they basically said that there's an offline LLM on your phone, the Pixel 8 Pro. So that's something that's already happening before. So to conclude and basically to summarize, in 2024, some of the predictions is that AI influencers are going to go crazy, we're going to need new platforms, and platforms are already rolling out ways to essentially protect themselves from synthetic stuff. People are going to get, you know, catfish people are going to get, you know, there's going to be a whole load of copyright issues in terms of using someone to do AI, you know, identity theft and all these kind of things. In terms of verification is of course going to be an issue. Scams and verification is going to get worse. But that is something that is needing new technology to combat this. Of course we do have the calls and stuff like that. Advanced humanoid robotics is going to be definitely crazy, especially with reinforcement learning and synthetic data. We're essentially likely to see some kind of breakthrough. Video Games with NPCs is definitely going to get crazy as well because we are likely to see the first generative AI game with LLM agents, autonomous agents as well in 2024. We've seen companies work on these systems which are so far pretty, pretty decent. So I wouldn't be surprised if that happens. AI warfare unfortunately is likely going to be increasing. And here's the thing, you probably won't know about this. So it's not like the US army is going to be like, hey, we developed the number one fighter jet in the world. So that's something that even if it does happen, we're probably not going to know about it because if they do make a breakthrough, they're not going to tell everyone because there's no point alerting your enemies or potential combatants to an advancement that you may have. And of course text a video already making breakthroughs is going to have more breakthroughs in the future. And of course the next gen architectures like GPT5 and Gemini are going to have new architectures and new ways of processing information and we may even get stuff like Q learning and offline LLM. So with that being said, if you did enjoy this one, I know it was long, but this is a lot to cover and I think it's important to know where we're headed in 2024. I'll see you in the next one.