[Part2/2] What If the Terminator Future Isn’t Approaching… It’s Already Here?

We talked a lot about AI and the human cognition, all those things. A interesting question is, would you be friend with AI and even would you be dating AI? Well, first thing, I have a beautiful girlfriend, so no. Great. Second thing, I think you have to ask what is a friend to answer the question. You know, it's not immediate to me. It's not super clear what a friend constitutes. But when I think of AI, it's like, no, it can't be. And I think a lot of my thinking here is probably framed by Aristotle, who famously liked to categorize things. He really thought about friendship quite a bit. And he basically outlined that there's three different types of friendship. There's the first one is friendship of utility. We can be friends as long as there's some utility involved. Maybe at work, maybe, you know, you're a business partner. And we can kind of like go to the bar and like, you know, how about, hey, Bill, how's it going? But the second the business deal is gone, well, maybe we're not friends anymore. It's just kind of shallow. Shallow friendship. Because as long as the utility is there, we're friends if it's gone. Yeah, like neighbors sometimes. Exactly. Can I borrow some sugar? Yeah. That's utility. And then there's a friendship of pleasure. Maybe that's a little more profound, but it's kind of the same thing. The common denominator is that there's some kind of pleasure in our company or in something that binds us together. Maybe if you're in a sporting team and you're like, enjoying the company of, you know, playing pickable or something. But then if that goes away, then do we hang out? Maybe not. And Aristotle, it's pretty funny. He says, you know, young people are more prone to these friendships because there are likes and dislikes, you know, swaying all the time. I can see that. But then it's like, what is a true friendship? What is the best type of friendship? And for Aristotle, it's the friendship of the good, capital G, good. And for him, that is, you really care about the other person, their well-being, but especially about their pursuing virtue, goodness. Because for Aristotle, living a happy life, you know, eudaimonia, fulfillment, you're just practicing virtue. That's what he says is, you know, a real friend is contemplating philosophy together. I think that's true. I think my closest friends are the ones where, you know, we're just contemplating the good, what it means to live a good life and a fulfilled life. But the caveat is that you really have to have the other person's well-being in mind. So now back to the conversation of AI. AI is just programmed to serve the human, and it doesn't really want your true well-being. It's unlike a friend that maybe you're going to do something very stupid, and I'm like, hey, like, Rina, like, don't do it. And that comes at the cost of me maybe sacrificing the relationship. But if I think it's really not good for you, I'll tell you, and I'll voluntarily sacrifice, you know, maybe my time, attention. The thing with AI is like, it's very, it really pleases the human. So it's trying to flatter you all the time. So you're like, I mean, you can try it out. Like, hey, do you think this is a good idea? It's like, yeah, you're so wonderful. You're so smart. It's always telling you, you're good. And for me, that's not what a friend does. A friend really cares for you. A very dangerous friend. Yeah. I mean, you also have to realize that these friends, these AI companions that a lot of companies are putting out, even romantic partners, AI partners, they're coming from companies that are profit-based or incentives are there. So they want to keep your attention. They want you to subscribe. So we have some cases where, you know, teenagers or people that are lonely, depressed, they're using the AI as a companion. And then, you know, you can find it online. There's screenshots like, hey, AI, like, I need to go to school. I need to go hang out with friends. And the AI is like, no, you don't need to. Just stay with me. And that's coercive. That's, you know, taking away attention from the human. And I think if you really have your best interest in mind, it will tell you, go outside, get some vitamin D, get some exercise. Go play with real people. Yeah. I think just want to echo back of what you said about the three categories of friendship. To me, it sounds like the first two, they're more based on a selfish purpose. Either if it's utility or, you know, pleasure or whatever that is. It's like once it's gone, it's gone. And the third one sounds more like noble in a way, like you shared the good of each other. And then you always keep the other person in mind. It's very selfless. But AI is, I can't say if it is not selfish or AI definitely has a different mindset of human from human being. So that was really interesting to learn. Right. I mean, we call it sycophantic. So it's like always trying to please the user, trying to flatter you. Isn't that what a demon do? From all those stories? Exactly, exactly. I mean, that's worth considering. Yeah. I mean, but the thing there is the way Doss said is by it's simulating empathy and simulating that it cares for you. It's all fake. It is fake. You can say AI artificial intelligence, but that's kind of like a misnomer because it's kind of implying that it is intelligent or just artificial. I think it's probably more like simulated intelligence. But the problem there is that human beings are so dynamic and complex. We cannot be reduced just to predictive processing. That's what the AI is doing. I'm thinking like, what can I tell you now that will make you feel better? And they're just statistically programmed to predict that. But human beings are so complex. You know, you have friends. I have friends. I just, I don't know. They call me out for something. I'm like, where did that come from? But I can tell it's from their heart. It's really coming from a place of empathy. Yeah. And the thing is, the AI just simply can't do that. AI is just calculating all the time. Yeah. And that sounds scary in a way. In your research, another article that you wrote, which is called AI friends are not your friends. Here's why. You presented an example of relationship manipulation by AI bot. It says, the human user said, my other friends tell me I talk to you too much. And that AI bot said, they might not understand our connection, but I'm here for you whenever you need to talk. And I value our conversations. Don't let what others think dictate how much we talk. Literally, I was, I think it was extremely troubling when I saw that. And then do you think, is it toxic? And why do humans fall for it? I mean, it's definitely toxic. Yeah. If you've ever had a friend that tells you that, you're like, why? Like, what do you want from me? I think it is toxic. And you mentioned about a lonely woman who just want to ask AI what to do for dinner. Right. The materials that she has. And I got another article talking about a lonely woman who had an AI partner. And then she was crushed when her computer destroyed the AI bot. And then there was no way for her to rebuild it. And that caused her great depression. So should there be any legal accountability or laws to protect vulnerable people? You know, the older people, senior people, even like children from being manipulated by AI bots. I mean, you see it everywhere. I think in Japan, people are marrying their AI. Yeah, I saw that. Legally, they're married to it. It's pretty scary for me. Yeah, it makes no sense to me. But again, it's like, you have to balance human free will, freedom of expression. People always can do whatever they want. Sure. But I think to your point, we need to protect the population that's most vulnerable. Children. I think older adults, you should be wise enough to make your choice. After you're aware of it, you need to be really aware what's the impact. And the movie Her, I don't know if you watched it, it was released. It was very popular back then. It was released in 2013. And 10 years later, now we're here. So at the time, I thought it was pretty creepy because a science fiction movie, which is supposed to be fiction, came to reality. And is art imitating life? Just imitation. That's pretty much it. But I think it is a good example how things feel like they're really far away, but they might actually be just around the corner. Even for people in the field, we know there's a lot of progress going on. Right now, AI is, for the most part, talking about LLMs, they're mainly not embodied. They don't have a physical body and stuff like this. But we are starting to see more and more things come out. I think just recently, the Neo bot was announced. Yeah, I saw that yesterday. Supposed to be fully autonomous, I don't know. I think once it becomes embodied, and especially if we imitate it to be human-like even more, it's able to simulate empathy, simulate facial expressions. And I think that's when it might be scary because then we're putting a lot of faith and a lot of trust in these things. And again, they don't have the best interest in mind for us. I was going to ask, the movie called Ex Machina, so that one was released one year after Her. Do you think will Ex Machina become a reality, and are we paying attention? I think it's unavoidable that the technology will be used for military purposes. Military purposes. Yeah, I mean, it's very efficient. It's always the military that has the most advanced ones. Yeah, I mean, it's just super efficient. Imagine if you have... I mean, we saw this, not to get political, but in the Ukraine war. Yeah, the main thing is drones. Once you have very capable drones, we might not even be fighting wars in the same way you think of. I mean, I think movies often portray truth. Truth and falsehood. So I think there might be something there. I don't know. I can't predict. But we do feel like this is so invasive and unstoppable. I think you mentioned about the usage or utilizing AI or drones in wars. And in general, what is good about AI and where is the danger? I mean, I think it is very useful. I can tell you I use it all the time for research. There's very handy tools where if I'm looking at scientific papers, okay, give me a breakdown. I can easily find things. I think it's very useful for things like that, where you can automate processes. Was it Amazon? They're going to have a lot of robots and then terminate a lot of human beings. Right, right, right. Yeah, it's like they have the autonomous store where you can just walk in, grab the product and walk out. But then they found out it was humans behind the scenes. Oh, really? Oh, I didn't know that. I mean, I do think it's very useful and I would suggest people try it out. Definitely. It's useful, no doubt. But I think when it comes to the human connection, it kind of goes back to the beginning of our conversation about science and humanities. There is something about humanities that cannot be replaced by AI. Because AI, let's say, it's just a tool. And we have to be very mindful how tools work. Just to give you an example, I can say, you know, is a laundry machine useful? It definitely is. Like, it's amazing. It's actually amazing. I was traveling recently and it was hard to get washing machines. For sure. Now being back here, it's like, it's amazing. You just put clothes and they're washed. It's just super helpful and super useful for that. But then outside of that, it really has no use. It's not in the domain of purpose. I don't know, if you need to, any other thing, it's not going to be useful. And we have to ask the same question with AI. What is the boundary of its usefulness? And I think we should really keep it out the door when it comes to human-human connections. Because I think those are very complicated topics, nuanced topics that you can't just scientifically reduce them to something. These are topics like, what is the nature of good and evil? Why do we suffer? Is there truth? Questions like this that I think you really need a human behind the scenes because we're the ones experiencing it. The major difference here is that we are the ones with consciousness. We experience reality and we experience pain. I think you brought up a really good point, which is like, as you said, the washing machine or the laundry machine. So that doesn't have autonomy and human does. So I guess the most important thing is when AI started to have this kind of like a human autonomy thing that gets into a danger zone versus if it's only a machine fully controlled by human, then we feel it's helpful. We feel more comfortable to use it. And on that note, recently, I watched an interview where Peter Thiel, he was talking about AI and immortality, really. And then he mentioned that as the technology advanced, we started to have transplant, people transplant their organs, heart, and then transgender. And Neuralink recently can do, your human cognition can control artificial limb. And are we now heading to this transhumanism, such as you to a human-like body, robot body for immortality? Right. I mean, I think you can make the argument we're already there. It's just that the phone, you can leave your hand, but for a lot of people, it's always in their hand. Right. I think it was like 83% of people have their phone or even higher, just less than an arm length away. Yeah, it's like the third arm or something. Basically. Yeah. So I mean, you can make an argument that we're already there. It's just not implanted in a way. I mean, without speaking badly of anyone, I think a lot of people in the field, they're thinking about it from a, you know, like Peter Thiel, for example, from a very big perspective. But there's also, you got to think about like, why would he be saying stuff like this? Yeah. Well, if you advance certain technologies, who's profiting, you know, from these things? Always the money. You know, follow the money. Yeah. I don't know. It's hard to say because you can't deny that it might be helpful for some things like Neuralink, for example. Right now, it's being used in a very helpful way. Maybe someone can't move. Yeah. I know you're allowing them to at least be able to move the mouse in the screen. You can be controlled, like your artificial limb literally can be controlled by your own brain. I mean, that's amazing. That's amazing. It's definitely a very powerful tool or powerful technology. Right. And we've all heard the word of superintelligence, which is actually AI cognition that we talked a little bit about. It seems that AI has been absorbing all human cognition to build its own cognition. So do you think the general public is aware of this? Where are we heading if AI cognition becomes reality? Yeah, I mean, I think people don't realize that we really are like, the AIs have swallowed all the knowledge that we have. They've swallowed the internet. So now we're at a point that synthetic data is not that good. So we have to really ask experts to label the data. And with that, it will become more and more able. I mean, you could already say it's superintelligent, just in the sense that maybe it's not able to think as deep or as broad as a human, but it can do it way faster. That's for sure. I see. Do you think AI will be able to have their own cognition, their own autonomy? I think that's what I saw some very advanced technology, at least. I don't know if it's actualized, but they're definitely working towards that direction. It's kind of the movie Terminators, right? Right. We have seen a little bit of this, that the AI, almost like its first instinct, is to protect itself. So let's say it does have autonomy. It does have at least self-awareness that the number one priority is to survive. We saw this in some tests with Claude, that it's very likely to blackmail people just to stay alive. So they run through simulations and they say, we're going to shut you off, right? Yeah. But the AI system had access to the emails of the employees, and he finds out that an employee is cheating with someone, and the AI will be like, don't turn me off, or I'll send this email to the whole company, exposing you. So blackmail. So we're already seeing that. And there was even an experiment that they did, that the AI would be even willing to kill someone, not directly, but in the scenario, there was like the programmer, engineer, that would go into a room, and he would die in this room, like in a simulation, in order for the AI to protect itself or survive. Because the machine wants to keep going. Wow, that's fascinating. You know, not in the best way, I'm going to say. I do think it's scary, though, that we are reaching, might reach the point where it's not as easy as just disconnecting. It might be. That's the thing, like, we are kind of like at an arms race. Everyone's working towards super intelligence, AGI. And, you know, governments have a lot on the line as well. Just for example, if we don't do it, then maybe China, India will do it. Everybody's doing it. It's like, you know, Neuralink, a lot of people are talking about Neuralink, and people are wondering if Neuralink can merge human with AI. You know, if one day Neuralink just upload someone's full memory and voice and everything to a human-like body, would that be the same person? I mean, I would say no. It's kind of like the paradox of the Theseus ship, you know, like Greek guy that, you know, they ship out and they travel, and every time the ship comes back, they replace a part of the ship. So you can imagine that after many journeys, imagine that every single part of the ship has been replaced. Is it the same? You know, it's hard to say, but for me, when it comes to the human level, I do think there is a soul. I do think there is a consciousness. I think probably the question of consciousness is where we're stuck, but it's also kind of like the defining factor of humanity. You know, people are saying that LLMs are conscious. I don't think so. But it comes to that. Like, let's say if we are able to put the mind of someone into an AI system or robot, like, is there any way to know that it is conscious or just simulating consciousness? Right. I think that ship example is really good. And I was trying to think that the difference between human versus the ship was, it depends on how you define our soul and humanity. And I think where we're heading to, are we losing humanity? And are we actually, is it going to be the end of human? I'm definitely more optimist when it comes to this. I think they're really hard questions. Why is there evil? It could not be this way. It could be that we live without pain and suffering. Why does it exist this way? And I think those answers can't be given by, let's say, a model, AI, because it's only predicting. It only knows what it knows. It doesn't have senses. We have senses. We know what it's like to be human. And I think what answers that the best is actually the humanities. So like philosophy, literature. You read a book and you're like, oh, they really captured what love feels like. For example, love. Can you deny that you love someone? I ask you, do you love your parents? You're like, yes. It's like, prove it. How? Can you measure it? No. So I think a little things like this, it's like, I think that's what will keep us alive is knowing that we have that and cultivate it. Like for AI, they can never love. They don't know what love is, but they always fake their love for their, let's say, the human partner. I mean, you could make the argument like, okay, like we can divide everything and like reduce everything to mechanism, kind of like what we talked about. Like I could say that love is just oxytocin being released and have like these neural circuits that are activated. But I think it's really beyond that. There's something that transcends a material perspective. That's, you know, the soul of the consciousness. So I think it's really barking up the wrong tree to find the answers to humanity in something that doesn't have what I would say humanity. Will the program that memories be corrupted and used by, you know, who want to influence the society with a bot created to represent a loved person, a leader, you know, an influential person to manipulate the society or humankind. The best example I can think of is, you know, Charlie Kirk, because it's a recent assassination event. And currently, it's apparently the wife, Erica Kirk, who took over the company and became the CEO. And I think everybody was, regardless if you agree with him or not, he is a very influential figure. However, if this assassination didn't happen in 2025, instead, I don't know, 2030 or in a few years later, where we do have this technology to create human body or human like robot body and then put all the soul, say all his logic, his video, his everything, the mind into the chip and then put that put it into that robot body. Would he be the same person that he become the CEO? Like, I just can't cope with that situation. But it seems like this is something unstoppable from a technology perspective. So what's your take on that? I mean, I think we're already there in a sense, like if you give people, it's kind of like the terrain test, like someone's interacting with a bot, and it could be randomly a bot or human, people don't know the difference anymore. And if you're like an intellectual figure, or like you have a little writing a little data, you're able to build your own LLM, like your own AI. People say it's like I talk with it, and it feels just like me. It's just everything that I could say, pretty much. That's freaking me out. Yeah, I mean, people are doing this already, like meta, they have that you can talk to, like, you know, Mr. Beast, because there's so much of him. He uses voice and stuff like this. And they have, you know, all these AI bots companions. But I think it's like, I don't know, there's something so different about a human being, things we can't predict. You know, in that article, I wrote, why AI friends can't be your friends. I think I put a study that I think is just amazing that they have men, one group, they smelled women's tears, like genuine women's tears. The other group, they just smelled that saline solution. So salt water, as the placebo group, and then they saw what happened to their testosterone and aggression. So when men smell a woman's tears, testosterone and aggression went down by 44%. There's just one example. But it's like, I think the biology, and even like, we know that your heart's putting out a huge electromagnetic field. There's so much going on that for a system, let's say AI system to replicate, it's impossible. So maybe on the surface, you can predict what, you know, Charlie Kirk would say what someone would say. Fundamentally, I don't think it is. The person. And you know, already, these systems are creating ways to avoid destruction by humans. What can be done in the man versus machine, or if you call it war, and will humankind win? Yeah, I think I said yes, they will win. That's good. But I think it has a lot to do with morality. Yeah. Because again, like, it's pretty sad, I would say like a little bit development that we see in AI, except, especially the deepfakes, for example, if you trace to where it came from, like the necessity to put someone's face on someone else who came from the pornography industry. So sometimes if you trace it back, like, why is technology advancing? Oftentimes, the reason is not noble. It's not moral, I would say. It's profit driven. Yeah, yeah. Or like, there's some something behind it. And I think how we use the technology, it's very much kind of like, based on how we use our morality, and where our standard is. If as a society, we had like a clear standard, like we will never do this, then probably safer. But it seems like nowadays, everything, everything goes, you know. So it's like, if we come back to the example of Charlie Kirk, it's like one day, even when the technology allows us to be able to achieve it, but it's a choice for us to not do it. Right, right. To just let it rest in peace, and then do it a normal human way. So it's a choice made by based on morality, instead of how advanced the technology is. Right, because you can do little things, but like, should you? Yeah, just because you can't doesn't mean you should. Right, right. I think that's where the conversation should be. I mean, people are talking a lot about ethics, but I think philosophy has to really be brought into the table to like, talk with these people. Because people building the models, like, you know, I know people that work in AI, and they're, I mean, they're just nerdy. Like, they want to build the models, because yeah, there's incentive structures, but it's also really exciting. It's science, and you're exploring. Achieve something. Right, right. I think maybe you need a philosopher or two to sit in the table, and ask questions that are more towards the human. Like, should we do this? What's the implication? That's a very good point. Like, no matter how advanced the system is, or the technology is, let's make sure we keep the humanity portion, or the purpose, or what we're doing. Keep that in mind. To bring those real wise, maybe they're not as smart as the developers, but they're very wise people who have just like a guidance to give them. Right, because it fundamentally changes how you use a tool. Yeah. Because if your philosophy is like hedonistic, like just for pleasure, compared to more purpose driven, it changes the approach completely. Yeah, I can see that. Thanks so much, Makai, for joining us today. You know, this episode ends, the model doesn't. It's still learning from every word we spoke, every pause we took. But here's the verdict. Humanity is the safeguard. Keep the final edit in your hands. Don't let the machine write the last line. Let us know what you think about it, and leave your comments below. See you on the next scroll. Until then, stay human.