Meta's New AI Is Freaking Everyone Out...

So Meta's new AI chatbot is something that we've never seen before and it's freaking everyone out and of course we have to talk about it. So there was essentially this post from Business Insider that talks about Meta's new AI that basically says death isn't the end. Allegedly, and it's not really allegedly because it's true, Meta patented an AI that lets you keep posting from beyond the grave. And this is pretty creepy, but we need to dive into exactly what's going on here because this feels like a sci-fi future from Black Mirror. So, the patent is super interesting because it's called simulation of a user of a social networking system using a language model. So, I think you can easily see where this is going. And it says, "A social networking system simulates a user using a language model trained using training data generated from user actions performed by said user. And it says that the language model may be used for simulating the user when the user is absent from the social networking system. For example, when the user takes a long break or if the user is deceased. Now, I think you can understand why this has gone so viral. It's because that is one of the creepiest things if an AI version of you lives on as you pass away. I'm sure most people have seen crazy sci-fi futures where dead people are brought back to life with AI. But the fact that this is a patent that Meta filed and they've actually been granted this patent, it now means that maybe we're in the end stages. Now, you might wonder why they are actually doing this. And I think it's really interesting to read the fine print of the patent. So when you read the fine print here, it basically talks about the fact that social media allows, you know, people to connect. And it talks about the fact that when you do connect, you do things such as liking, commenting, and forwarding and performing other interactions on that specific social networking system. And it says that several users may be connected to a particular user or receive content generated by that user. And it says that if that specific user is absent from the social networking platform, the users connected to that user do not receive any content from said user during that user's absent. And a user may be absent from the social networking platform for long periods of time. Therefore, and although it says thereby affecting the user experience of several other users of that same social networking system. And it says that this impact on the users is much more severe and permanent if that user is deceased and can never return to the social networking platform. Essentially what they're saying is when people leave the platform for whatever reason it may be if that person dies or if they go on holiday, they're saying that and I'm using virtual quotation marks here that is an overall bad experience for everyone else. And I guess whilst that could be true, I think this is a really interesting reason to actually want to patent this. And so because people leave social media platforms all the time, I'm guessing they're saying that look, we don't really want people to leave and we don't want the platform's engagement to go down. So what we want to do is have a situation where the user can keep posting still as if they're alive until the social networking platform continues to exist in its way. And I mean that's super strange because I think part of understanding you're interacting with someone online is the fact that it is actually the real person. I'm not sure if you guys have realized this. I mean Meta launched something and I'm going to show you guys. I'm going to do a quick detour here, but it is particularly relevant to this video and it isn't about dead people, but it's about AI chat bots and when you talk to them. And so most of you will not have seen this video. Like most people don't even realize this was a thing, but Meta had AI personalities that were essentially celebrities that cosplayed as different people. You can see Tom Brady as Brew, the sports brain. It was actually pretty strange. Then you had Naomi Osaka as the manga master. I mean, essentially what you had was original celebrities that basically decided to, you know, brand themselves as different AI personalities. They of course got paid millions of dollars for this. And I mean it was not really the most successful thing and this was because users found the alternative personalities to be pretty weird and creepy. So Meta actually pulled the plug on this for less than a year after the launch. And the fundamental issue that I think Meta didn't realize here is that the problem is that most humans value social media for the human connection. It is called social media. Now remember the thing is that bots don't fit that value. So now the reason that this actually failed, okay, and the reason that the AI

chat bots actually just completely had no human interaction at all, the problem is that I don't think Meta realized that what social media actually is social media. Okay? It's human connection that people actually want. And when you actually think about it, bots don't fit that. Okay? And the value of these platforms, the value of places like Twitter and YouTube and of course places like Instagram is the fact that they actually connect you to the real human world. And whilst you know bots are cool and funny and they are able to do different things, you have to understand that these things basically highlight how far away you are from interacting with a real person. They just lack authenticity and they use language that doesn't feel right. it might, you know, seem a little bit right, like 80%, but it's just not right to be able to connect with that on any meaningful level. So, when you, you know, have an LLM that's trained on someone's post, it's not really going to be able to capture their personality, it's pattern matching, just their word choices. So, if people didn't want to talk to a fake Snoop Dogg who's still alive, while they want to talk to a fake version of him, or their grandmother who's, you know, not going to be able to consent or correct what the AI says on their behalf. So, the reason I'm talking about this because the patent and everyone on social media is essentially saying that this is really creepy. The reason I'm saying that is because you have to understand that we've already been down this route before. like we've done it with celebrities that are alive and when we've seen that this thing hasn't really worked already. And when you think about the fact that and no offense by this, my social media is pretty terrible, but the average person's social media history is often not that dense and it's, you know, not that datari, meaning that the simulation would probably be pretty bad. When you think about that, it's probably going to be pretty bad across the board. now you know maybe 10 years from now it's going to be super amazing or just based on a few things but you have to understand that trying to you know replicate human connection I think it's going to be a very hard thing to do even for tech conglomerates like Meta now Meta in their defense they did say that they have no plans to move forward you know with this example and the thing about this patent that you might not know is that it was filed in 2023 so the patent being granted you know two years later. It isn't indicative of them, you know, actually trying to achieve that now. I'm pretty sure if they were, we would have heard about it. But I think Meta may have understood that this wouldn't work on such a global scale. I do think that when they did launch that previous thing, they were maybe caught up with the AI hype and just wanted to launch absolutely everything to get on board. But clear testing shows that humans value social media. Now, it doesn't mean that, you know, people won't get completely attached to AI bots. Of course, that is a real thing and of course it does happen. But across the board for most users, this isn't what they use social media for. Now, I know we talking about this insane thing and I did find this and I can't really verify this information, but it was floating around the internet this thing called Project Lazarus. So, there was essentially an anonymous post on I think it was 4chan and it talks about the fact that someone from Meta posted saying that they were building an AI that can take over a deceased person's social media and continue making relevant posts as if they were alive. This includes age progress photos, interacting with other people's content, and everything else needed so that they can continue on the digital realm after physical death. And essentially they said that things were getting weird at the time and this was actually before Meta did file the patent and it did say and this is super interesting by the way. So all of this information take it with a grain of salt because it's just a 4chan screenshot. It could be completely fake but I have seen it a ton of different times even before this article was posted recently. And it does say that the AI is extremely capable of impersonating people. It doesn't take as much initial input as one might think to train the AI on how a certain person interacts with the digital world. It is very convincing. An entire island of people could go missing with little to no downtime and the AI could take over all their accounts and the world wouldn't even have a queue clue that life was just continuing as usual. And a lot of the project is becoming more compartmentalized and things have taken a dark turn. I feel like they forbidden communication between people working on different things. Something isn't right and I don't know what to do. I'm not going to post any personal identifiable information, but I will try to answer questions. So, I mean, this is written like a, you know, kind of conspiracy theory thing, but I mean, this was before they did file the patent in I think late 2023. So, maybe there is some truth to this. I really can't verify, but the fact is that it is a little bit weird. I mean, when you do look at this article from Business

Insider from 2023, this is where we get, you know, Mark Zuckerberg, who was actually talking in an interview with Lex Freriedman. And this is where they actually say that yes, they agree that it could be possible to bring people back, you know, and use the metaverse to bring people back from the dead virtually. So, you have to understand that this isn't some crazy kind of, you know, patent that, you know, someone at Meta just filed and it's completely random. Mark Zuckerberg did actually talk about the fact that they might want to offer this service for individuals who are struggling with grief. And I think it's particularly interesting that this is even a thing. I mean, the full discussion was actually prompted by Lex Freedman himself. He was actually wondering, is that something that you could do? I mean, you guys can take a listen to this. I mean genuinely in this interview I do have to give credit where credit's due. It doesn't sound like some creepy dystopian thing at least in this part. It just sounds like a simple service that they were offering. — Well this might be a bit of a complicated and a dark question but one of the first feelings I had experiencing this is I would love to talk to loved ones. And the next question I have is I would love to talk to people who are no longer here that are loved ones. So like if you look into the future, is that something you think about who people pass away but they can still exist in the metaverse and you can still have you know talk to your father, talk to your grandfather and grandmother and mother once they pass away. The power of that experience is one of the first things my mind jumped to cuz it's like this is so real. — Yeah. I think that there are a lot of norms and things that people have to figure out around that. There's probably some balance where, you know, if someone is has lost a loved one and is grieving there may be ways in which, you know, being able to interact or relive certain memories could be helpful, but then there's also probably an extent to which it could become unhealthy. And I mean, I I'm not an expert in that. So, I think we'd have to study that and understand it in more detail. And I think the actual use case of this was probably mainly to talk to, I guess you could say, famous individuals who have a super, you know, just really, really wide range of knowledge and, you know, I guess you could say unique perspective on things like I don't know, there's just a ton of historical figures that maybe people would want to speak to again. And I guess that's the kind of vibe that they were going for. I do know that whilst that is probably a little bit light-hearted, that it can easily turn into something that is a little bit dystopian if everyone has, you know, AI versions, AI clones of themselves that are just continuing to post on social media even after you're gone. And I really hope that doesn't happen, but part of me thinks that some other company will do it anyways. Now, if we're talking about other companies, um Microsoft are another company that have actually done another patent. And so, in 2021, they were actually very early to this idea. So, they decided to patent an AI chatbot that could simulate deceased people as well as fictional characters or celebrities. And so from Microsoft's one, it was super interesting because it could gather information on, you know, images, voice data, behavior information, text messages, and it could even take on a physical presence, the document show, which include plans for two and threedimensional recreations of people gathered through photos and videos of the individual. And while Microsoft's patent doesn't specifically say that the product would be used to bring deceased people back to life digitally, the plan does use dead loved ones as an example of how the chatbot would or could be implemented to recreate intimate to recreate and imitate individuals from the real world. So I mean this is super interesting because you have a situation where other companies have already filed patents on this thing and this was even before generative technology was as good as it is now. So I mean I think companies have already seen that maybe they may want to pursue this in the future. And I mean look I have to be honest I think that some people may want this product. I'm not speaking for myself here, but I do think that you may be surprised with as to how many people might want this. And I don't think a lot of people do, but maybe those who do want this are willing to pay the high price to be able to talk to someone that they've lost if they're unable to process the grief efficiently. I mean, not everyone does deal with the grief, you know, in a normal way. Some people may want to use something like this if they're unable to do so. Now, of course, there is obviously the legal trouble where even if the use of a deceased individual's likeness is tried, that could cause legal trouble because 23 states in the United States recognize postmortem rights, which allow a deceased person anywhere from 10 to 100 years of protection from unauthorized use of their identity, including their name, their voice, their image, and

likeness from it being used for profit. I think if you ever do sign up to one of these platforms, maybe there will be some kind of, you know, terms of service where you potentially agree to maybe waving those rights because of course in certain countries this isn't going to be allowed. We know that just because America's, you know, land of the free, different countries, even different states, different jurisdictions are going to have different rules on what is of course allowed about their users. Now, this is where we start to get into the incentives. So, I think this one's super interesting because like I said already, you know, the vast majority of people will not want this. This is very dystopian. I don't think anyone wants dead people online posting pictures as if they're still alive. It is very, very weird for a social media platform to do that. However, um you can see here that it says that it's more engagement, more content, more data, more data for the current and future AI. And I can see the business incentive for that. I'm just curious to how see they would when and if they will implement this innovation. Now I do agree because it's pretty hard to sort of market that to anyone and I mean I'm not sure how they would do that without facing a big backlash. Now you can see it says Joseph Davis a sociology professor at the University of Virginia said he was concerned with the impact that tools like a metabolic could have on how humans experience grief. One of the tasks grief is to actually face the loss. And yes, that is particularly true. One of the things that happens when dealing with grief is to of course come to terms with the fact that you've lost something. And humans, I don't think we're naturally wired to be able to get over things. You know, especially people that are close to us or even pets. Of course, you when you share a bond with someone, it is very hard to lose that. So, I think one of the things here is that if this does become a societal norm or if it does start to become something that people will start to use, it's going to make grief harder because of course part of, you know, moving on is accepting the fact that they're gone, honoring their death, and of course, moving on in a healthy way. Now, we can all debate what's the healthiest way to move on. But, of course, this certainly is going to introduce some complexities that we can't foresee. I mean, I don't know anyone that uses this. I mean, you've seen some Black Mirror episodes that have been pretty dystopian, but I do think that this is going to add to the complexity level of how individuals may move on in the future. Now, it does talk about, you know, if this was true, the clone was supposed to be responding to other people's content by liking, commenting, or responding to DMs. And for influencers who make their livelihood on meta platforms or who need to take a break from social media, such a tool could be useful. Now, I think this is probably maybe something that is useful. people right now currently do use AI tools already to post on social media. Although they're not very good. I can see them being good in maybe 5 to 10 years. So, you know, for example, an automatic DM feature, that is something that is normal. I guess you could say it's AI or templated. But if you have an AI that's actively going out there and commenting on other people's content, I think that's where you fall into the AI clone feature. Now, for social media pages and brands and all these businesses that want to just reach people and stuff like that, yep, 100% it does make sense. And, you know, I guess we're going to have to see how social media co-evolves. Are people going to like the fact that, you know, Mr. Beast or someone with 100 million followers is actually using an AI version of themsel to respond to DMs, or are people going to say, "Nope, I'd rather just not have the real version talk to me. " I think that is going to be where it sort of merges because I don't think most people have really explored that. I know I do have an AI voice that I very rarely use unless I'm really ill and I'm unable to perform the voiceovers. But I know that some people just completely have AI voices in their content. And there are complete AI avatars online on social media that do get content. And so of course you can see here in the article it says to fill that void meta would essentially create a digital clone of your social media presence training a model on user specific data including historical platform activity such as comments likes or content to understand how you would or how you rather did behave which is of course super interesting potentially dystopian I mean I do think we will see it in our lifetime I just wonder how it's going to be implemented and the one of the you know most interesting things here it says that meta's patent also references technology that would allow the LLM to simulate video or audio calls of users. And I think that one is uh pretty interesting because I think like I said already, you know, messages is one thing. Audio does take it to a new level, but video that is a completely different ball game. I mean, it was only recently that we got to see some AI real-time avatars being developed. And I think they're still in that, you know, uncanny valley where they don't respond 100% instantly. But I'm pretty sure the rate at which AI is developing, it won't be like that forever. If there's one

thing you can count on, it's the fact that AI development will continue to go in this direction. And so, of course, you can see some of the comments here saying, and this is just pretty hilarious, comments from Reddit, saying that, "Oh boy, I can't wait for a corporation to put words in the mouths of my dead loved ones. " And ads. Someone says dead loved ones spouting off ads is inevitable. So, I mean, it's pretty interesting. You have the large majority of social media saying that this is pretty strange and pretty weird that they're filing a patent. But remember, Microsoft filed this in 2021. I think it's just now more striking considering that we are in the AI age. We see AI every day. We know what it can do. We know how it affects social media. So, Meta introducing AIs that live on as dead personalities doesn't seem that far-fetched or that far off. And I do wonder how this entire thing will merge because you know there are platforms I'm not sure of the exact name right now. There are platforms where you can literally talk to AI versions of celebrities. Of course it's not very good. It's not got their personality but people do find it a bit gimmicky and sometimes they find it fun. I mean we do have to be careful with how we introduce these air relationships. But one thing's for certain, the future is very unpredictable. But let me know what you guys think about this and if you would ever use this.