← Big Technology Podcast
Big Technology Podcast
She Wants AI To Help You Thrive, Not Just Keep You Company — Y-Lan Boureau
01:04:05
Loading summary
Transcript107 lines
- [00:00]
A
Is the AI industry's obsession with companionship and therapy misguided? And should artificial intelligence products really be about something else, be about thriving? We're joined today by a brilliant artificial intelligence researcher and founder and CEO, Ilan Boro, who is the founder and CEO of ThrivePal, joining us here in studio. Ilan, great to see you. Welcome to the show.
- [00:24]
B
I'm so happy to be here.
- [00:26]
A
Let me tell everybody where we met because there's an interesting story, right, and you have very fascinating perspectives on where things are heading in AI and where they should be heading. And that was also the case about a decade ago when we met, when you were at what was then Facebook, researching whether if you could have AI, basically go in the comments and take a heated conversation, inject some facts and maybe diffuse some of the polarization we were seeing in society. So talk a little bit about that experiment and then give us your perspective on what happened there.
- [01:01]
B
So for me, working on AI has always been about helping people live better lives. And so at the time that was like, I think back in 2015 or 2014, I don't remember exactly, but I started talking about joining Facebook in 2014 and I joined in 2015, so probably 2015 or 2016. But at the time, one problem that I thought was something that was making a lot of people unhappy in their interaction online was how they were always fighting about stuff. And I had this like very common sense knee jerk reaction that, well, if you just show people the facts then they won't be disagreeing so much and you can just show them, no, this is, you know, this is the research that's very solid and if you make it very obvious and you really show reliable sources that are trusted, then people are going to make it much find it much easier to agree. So first is reliable and trusted is something that does a lot of work. And unfortunately that's the pieces that have been going down and down. The trust in institutions, the trust in these debunking services. Basically either people already trust these services and then usually they don't need the debunking or it's just kind of like minor things or they don't. And then you can do all the debunking you want. You are always going to get an answer which is like, well, I don't trust these services anyways. They're all like deep state or something. Or like they don't trust trust them anyways. So what I do when I'm interested in a problem is always I try to see if there's Research on something. And for these topics of like misinformation, there was a lot of research that had been done by very good people who looked at basically the effect of doing that kind of common sense responses, I'm going to show facts. And the effect is not very encouraging. Basically, people dig their heels and they're even more convinced of what they thought before than before you tried to reason them. If they're already like so far gone that they're very committed to their position. And a lot of Facebook would like.
- [02:55]
A
Come in and interject and be like, actually this is the truth in the debate that you're having, it would make them not only not trust that source, but even get further entrenched in whatever they believe that was based off something that was true.
- [03:09]
B
It's not just Facebook, Facebook or another institution that they don't trust. The one kind of constituency who has a chance of making a difference is their friends, people they already trust. So people who are already in their circles, people who have this relationship of which is usually pre existing or it's because they're more aligned that they already trust that person. But you cannot just build that like out of nowhere saying, hey, I'm going to just inject myself here and give you facts. So basically it's often around belonging and being part of a community and being part of circle you trust. And you cannot just say, oh no, we're in service of truth and finding facts. And these are the same facts for everyone. If you remember the funny phrase alternative facts, that I think was a brilliant phrase. It is funny. But in practice it's what lots of people kind of live with. Just because often reality is complicated and it's often too complex to fit into small sound bites or three bullet points. And often people are like, no, no, but I don't want your 400 pages, give me three bullet points. So if you have to really, really boil down something into three bullet points, you're always going to have to make choices. They're always going to be somewhat inaccurate. So you're kind of never telling the full truth anyways. So there's a lot of room for interpretation, alternative fact and stuff. So even of course some things can be like categorically untrue, but there's lots of room for building these relationships of trust, showing how you've been lied to before. So it's all very murky and it's often more about relationships, community, showing that you are part of a group, showing allegiance to a group, signaling to this group that you're a Good member. It's like, hey, that person posted this, I'm defending them. It's more important to you. If I'm part of a tribe and someone I really look up to say something often I'm going to, my first instinct is going to defend them. Right. So there's all these dynamics that come into play. So it's really rarely about is it true or is it false? So that was a very long answer. But that's why for me, at the end of the day I thought this is not a technical problem, it's a people problem and it's a psychological needs problem of how do you make sure that people feel like they have communities that they can trust. They feel like they can understand the world. They feel like the world is not a place where they don't have agency, where they don't have any control. These kinds of things.
- [05:35]
A
Right. So basically what you were going to do, you're thinking about doing is the end of your research would be something within a social network could use natural language to determine whether there was some sort of political disagreement that was veering off the rails and, and then have the network come in and share some fact based research. But what you found was that people were more interested in supporting their team versus actually getting to the root of what was actually going on.
- [06:04]
B
So I was never going to do something specific. I thought the fact checking wouldn't hurt. It wasn't the end goal, but I thought it wouldn't hurt. There were already lots of initiatives around this and it was very clear from the research that right now it doesn't work. So I was not going to push for that. If it doesn't work and you're like, oh, but is that the right thing to do? Anyway, that's not how I function. I'm a very pragmatic person. It's like even if the intention is correct, but it doesn't get you the outcome that you think is good. It's like perhaps don't do this. The other things I thought might be a good idea was rather than injecting just facts, thinking, okay, if people are having a position that perhaps a bit too extreme, should we perhaps show viewpoints that are a bit more moderate and like do kind of gradient descent in opinion space. So like you take whoever is somewhere and try to see can we show viewpoints are not so extreme? Because we know that doesn't work but that are like closer but same thing. It doesn't. It's a lot of work when at the end of the day often People, it's more about groups and communities. And the other thing I always wanted to do, which is how I ended up focusing all my work in the end, was understanding what makes people happy. I mean, happy or like, feel like their life is fulfilling. Because often when people go online, what are they looking for? Do you think they go online to find facts? Do they go online to be entertained? Do they go online to punch and release some anger? Do they go online to huddle and find others like them? It's like the motivations are very, very varied and they're rarely going to be about fact finding. Often you want to look something up, so you want to be shocked, you want to find something that you're going to be outraged. I remember a friend of mine who said one of the.
- [07:48]
A
Maybe it's just you want to feel something.
- [07:50]
B
Yeah, you want to feel something, but not sometimes you want to be. To feel community, like not just a meaning, but find others, find that you're not alone, find that other people view the world like you. So it's lots and lots of motivations. And often you cannot just boil this down. It's like trying to find the truth.
- [08:08]
A
Exactly. And so you mentioned that you were trying to figure out what people were, what makes people happy online. And that is, I think, a theme in your broader work is how you can use artificial intelligence to make people find happiness. And it does seem like we often online are going after things that we think will make us happy but don't. Things like fighting with other people. So talk a little bit about the power of the Internet to shape our ability to find happiness.
- [08:34]
B
Yes, but first I want to qualify. I was not. My whole life I've been interested about how people can live fulfilling lives. Not especially online. It's just, if they're online, what do you do online? But perhaps they should be less online. And so perhaps it's like, how can we give them stuff online so that they get more offline? But in general, I think lots of people think that online dynamics are very different. But a lot of what goes wrong in online dynamics was already existing offline. So, for example, something that I've always found very, very dishearten happening is if you go to the supermarket and you stand in line like before groceries were delivered, actually. But so, because when they're delivered, you're kind of like, you know what you're getting, you're not getting anything more. But if you go, like when I was a kid, you go to the supermarket and you're standing online and you're in line. Sorry, I keep bumping that thing, it's going to make noises. So you're in line and then you have all these like sweets and chocolates and temptations and they're putting them there when you're stuck there being tempted and it's like they're like stroking your self defeating urges. And then by the time you get to cash register you have all these extra sweets and you shot yourself in the foot. Never happened. Yeah, right. And I remember even as a kid like walking down the street in Paris where I grew up and there would be this huge billboard with this piece of melting chocolate. The brand was such a. And it was like chocolate Rocher and it was something, it says le s' est vousante which means like indulge or be tempted. And I'm thinking how dare they just like enter people's thing. And they know lots of people are struggling with their weight and a sweet tooth and stuff and they're just like pushing that on you, which is not very hard. Pushing people to do something they really want to do is not very hard. So how do you help people go towards what they're trying to do? That's harder. Like be a good person, deal with your emotion, regulate your emotions, do long term work, exercise, moving off this kind of thing that takes some effort rather than saying okay, how do we get people to yield to their short term urges more often? Which is not very hard, they want to do it anyways. And so online you see a lot of that too. Online is like the immediate low hanging fruit is how do you get people, how do you capture those short term urges? And that's really nice because they go super fast. If like the great late Danny Kahneman had this book, thinking fast and slow about how in our brain we have this way of taking decisions very fast and this way that's slower, more deliberate, takes more time. So if it's kind of a race, the faster one is going to win every time. And so that also means that you can have a lot more fast feedback. If you use something where it's like every second you're going to get a hit. I feel like all the systems that depend on capturing engagement, on trying to see what people do, if you try to capture that trend of the immediate reward, you're going to get a lot of it. And so it's very easy to train if you try to say no, I really want to help people do stuff that's going to really help them. In studying years, like what you get one data point every 30 years. Good luck with that. It's much harder. So that's what I want to do, but it's much harder. So for me, the problem with tapping into people's immediate urges is not just online. It's just that online, it's like I don't even have to wait for you to be at the supermarket. I can get you like all the time. So it's very tempting. But it's just a amplification of something that was already very present offline.
- [12:23]
A
And amplification is the right word because there's scale online. Right.
- [12:26]
B
Well, I mean, especially for me, it's not just online. It's because, like, what, I'm 40 now, so 40 something. So I grew up with the beginning of the Internet when I would go on like netscale, like net scale and have a model maybe like, and then you would like start to load an exact page and then you would like go back, have a snack, come back like 20 minutes later and you had like two thirds of the page loaded. It's like that was online at the beginning. Then we had a bunch of things and you had to be at home. Right. So I think for me what really changed is that a mobile phone, because suddenly you're always online. So it's not just online versus offline, is always versus in boxes.
- [13:07]
A
Right. And the point I'm making, I was making with the last point was you can try to influence every supermarket and that's going to take you a long time. But with online and always online, you can do things at scale and that might end up pointing people back to fulfilling things or to achieve better fulfillment. But it's probably the hardest problem on. I mean, just go back to the fact that you work for Facebook. You know, Facebook is, I would say, the number one company on capitalizing on that system. One need for instant gratification. I mean, they do it in an incredible way.
- [13:40]
B
Yeah.
- [13:41]
A
With 3 billion users.
- [13:42]
B
So for me, what really drew me to Facebook? Because for three years I did not go into big tech, even though I finished my pitching in 2012. So that was a good moment to go into big tech. But instead I want to do my postdoc in the neuroscience department with a great guy, Nathaniel Dahl, working on self control and decision making. And I stayed there for three years. And the reason I ended up going to Facebook was at some point I'm like, if I do all my research, I'm going to write a book. How many people are going to read that? And Danny Kahneman already writes book. Right. So if that was enough, we'd know. Whereas there's this huge megaphone with billions of people that Facebook has. So if I can go there, put little bots to teach people how to handle their own lives better, then there's a more chance of actually making a difference. So that's why I ended up going to Facebook.
- [14:30]
A
And it wasn't just about the political stuff.
- [14:32]
B
Oh, no. It was primarily about trying to like what I wanted to do from the start. But it was. There was no direct way to do it because at the time in 2015, I don't know if you remember how bad the chatbots were, but they were really not very good. So for me, a lot of like, I wanted to already build bots to teach people stuff like how to breathe, how to meditate, how to exercise, teach. There's like 40, 50 techniques that are very well established. Decades of research show that they work. And often out of these 40, 50, there's going to be 2, 3, 4 that appeal to a given person. You have to find the right one for that person. So we kind of know a lot of things that work very well, but how to deliver that in a format that's interactive, that's not boring. So I thought chatbots would be a very good way to do this, but they have to not be boring. And at the time, the chatbots were not very good. So that's why that part of what I wanted to do in practice, it translated into me doing research on conversational AI to try to make the bots better. So that was one of the main things I wanted to do. But for all the parts about misinformation and polarization, that's something where I felt like, oh, I can do something right away. And then I looked at the research, I'm like, oh, actually, no, not really, but so that was already part of what I wanted to do early for Facebook. So you're saying they're tapping a lot into the short term urges, but they're also tapping into something that's very important for flourishing, which is a connection to other humans, because they allow lots of people to stay in touch. Sometimes people forget that about Facebook. It was originally really a social network to give everyone a voice to connect people, and it still plays that role. It allows people to stay in touch in a way that they weren't. So there's a lot of good stuff there. The part that makes it a bit trickier is indeed that it's very easy to try to optimize on the signal you get right away. And you're going to get more of that signal when it's those short term urges and itches rather than long term. And again, I'm very against saying something like, oh, those big companies are just taking advantage of people because it's what people want. No, no, I know you didn't say that, but I often hear that. And for me it's simplifying the problem a bit too much because the battle often is in people's own head. It's between what I want now and what I want tomorrow, or what I want now and what I would want if I was stopping for five minutes and really thought about this. And like, do I really want to be doing this right now? And so often it's like all ourselves battling like the short term reward and the long term efforts and the fact that whatever I want to do for the rest of my life, do I really have to start today? Can I start like tomorrow?
- [17:19]
A
I mean, this is life, right? Cheeseburger or salad?
- [17:21]
B
No, exactly, exactly. And so it's not often a lot of the technologies lowering friction and lowering friction helps the short term urges because whatever friction was stopping that, it goes away. So you're going to get those very quickly. But it can also help lower friction for the harder things. So what I'm trying to do is how can we help technology lower friction for the harder things?
- [17:47]
A
Right. And so you mentioned that you ran to a wall at Facebook, which was maybe not Facebook technology, but just broader technology. And that was. The chatbots weren't good. 2015, 2016. I mean, Facebook built this test chatbot called M, which I had access to, which I think is the reason why we ended up meeting in the first place. And it was all humans on the back end that were answering queries. Queries.
- [18:10]
B
We had some. So the way I worked on the M team. So no, there was, was a hybrid thing. Like the AI would make suggestions and the humans were just to make sure that it was still a good quality. Which a lot of systems are still like that. Like chatbots are going to make those AI systems are going to make suggestion and then there's a human overseer that.
- [18:30]
A
Okay, but you can at least admit that when you were asking these bots to do research or to draw a picture on the other end, there were humans doing research and humans drawing pictures.
- [18:39]
B
For me, I was just working on the AI system, so I like for sure. But it's. For me, it's not so much like my main problem with the chatbots at the time is they were really not Just not very good at, like anything, they would not. They were not able to answer questions in a cogent way. They did not. All the things that basically now people take for granted when they talk to ChatGPT or Meta or any of the chatbots, they're all very good now. It's like we were so far away from that in 2015. They were very, very stupid. Could do some stuff, but not much.
- [19:15]
A
And then we have ChatGPT shows up in 2022. GPT models continue to improve, and you say to yourself, wait a second, if I use this better chat technology, I can start to make something that people are going to want to interact with that can help them live more fulfilling lives. That's the theory behind your company.
- [19:37]
B
Yeah. So for me, the chatbots are good enough for what I want to do. What I want you to do is there's a thing that hasn't changed so much in 10,000 years. Technology has changed a lot, but humans have not changed that much. So a lot of our wiring is super well adapted to a world of scarcity. And if we were starving half days and we had to run around all the time, which we don't really have to do anymore, but so a lot of our impulses, the things that we find easier are yet toward trying to save energy and towards comfort. And if we get given that comfort, we kind of tend to want too much of it. And then we end up more like the. The humans of Wall E. I don't know if you remember Wall E. Sure do. Okay, so in Wall E, the humans are like these big blobs in. In chairs, and they're waited on my AI and they're just being entertained all day long at the end, like this epic, like taking three steps kind of. If, if you just. And this happened over several generations, you see, at the beginning, when they get stuck, you see, like each captain gets like big and less mobile.
- [20:43]
A
Never forget the big sodas that they hold in their.
- [20:47]
B
And the robots do all the things right. And so for me, that's kind of. Because all the instincts we have around what we want came from a very long time ago and have not really adapted to the world and a world of abundance. And so what's great about this is that these challenges have faced humans for decades and centuries. And so lots of people have thought about this. Philosophers and poets and psychologists. And so we have lots of techniques that work. People always look for a new technique, but we have old techniques. And I would say they're still the best we have. So AI is not going to. I mean, I'm not looking for something that suddenly makes it easy. It's always going to be a bit hard. But we already know what works. So what I'm trying to do is use AI to transmit those techniques that we already know and help people navigate them and help people learn about them in a way that's more entertaining, more graded, more personalized, more tailored to what they want to do. And also something that's very important to me is help people who already have part of the thing figure out, like someone who's good at exercising or someone who's good at something to share that with their friends. So often if you. That goes back to this idea of you often trust the people you already have a relationship with. So. So if some of my friends have a challenge with, like building muscles, for example, often they're going to ask me for advice because when they go online, there's like 5 million different influencers who have different advice. You're going to see some things like, oh, you need to eat carnivore and only eat meat. And someone else is going to say, no, you have to eat no meat whatsoever and only plants. And they're both going to have lots of followers, lots of people saying, no, that's the right way. And it's very easy to get lost. So often when you don't know where to turn, you turn to people in your circles that you know and trust. And so can we help people who have some stuff figured out help their friends and say, hey, I've read the literature for you and I made a little course for you on this and we're going to like, work through this together. Because at the end of the day, thriving for me is really a team sport in the sense that I don't want a world where I've taken super good care of myself and I live to be 150 and all my friends are dead. I mean, that's not fun. You want to bring your village around, like with you. And so that means how do you help other people? How do we share knowledge? How do we make sure that it's not just a single player mode, but it's more sharing learnings and taking advantage of existing relationship with friends and other people.
- [23:26]
A
Yeah, I mean, you could be 150 and still have Netflix, but I guess that's not the path to happiness.
- [23:31]
B
Well, I do believe that there's a lot of joy in actually interacting with other humans with all their flaws and stuff. And also there's something else which is we are ruthless adaptation machines. There's this thing where your whole brain is looking for homeostasis. It doesn't really care about you always being happy or fulfilled or like overdroid. If you're overdried, your brain's going to think, oh, there's a lot of these droid chemicals. Let's tune it down a little. We don't need that many receptors because there's so much stuff flooding right now. So let's down regulate those receptors. I keep bumping, it won't pick up.
- [24:11]
A
Don't worry about it.
- [24:12]
B
It's just like going to downregulate those receptors so that you get back to your baseline and your new overdrive state is now your baseline. So it means if you never get any adversity or anything that's negative, you're just going to take this for granted and you'll just be exactly as happy as before. Except now it will require a lot more to just get you to that baseline. So in a way, we need to constantly get outside our comfort zone. Otherwise our brain cannot handle it. There's lots of examples like this that are very, very concrete. Even if you think this great western invention of sitting in chairs, we used to have to squat down to go to the bathroom. It's great for hips. And then status symbol, now you don't, you can just sit. Our hip mobility is terrible. If you go to people who do like Asian squat, for lack of a better term, like the full squat, where your leg really like this, have much better hip mobility, they can do this into their 70s. It's like lots of westerns have lost that ability because they never go into that position. Is that really progress? Just the fact that you never visit those position of discomfort anymore, Your body thinks, no need to maintain them. I'm just going to lose that ability and adapt to my new range. And so it's the same for everything, including our emotional range. If we just always see things a certain way, we don't get stretched and the body just. And the brains, they just adapt to that and then you lose that ability. So in a way, I think we cannot wish for a thing where we only have positive things and we only have good interactions. Everything's great. Never challenge for anything because it's just going to shrink what we're able to take and we become like very non resilient humans.
- [25:54]
A
Okay, so you said a couple things I want to touch on. First, that you want friends to actually help others as opposed to the bots help the people themselves. And then you talked about there's a number of things that, you know that we can do to become more fulfilled. So I want to cover both of those when we come back right after this.
- [26:17]
B
Hey, you.
- [26:17]
A
I'm Andrew Seaman. Do you want a new job or do you want to move forward in your career?
- [26:23]
B
Well, you should listen to my weekly.
- [26:25]
A
Show called Get Hired with Andrew Seaman. We talk about it all and it's waiting for you. Yes, you, wherever you get your podcasts. Hi, I'm Kwame Christian, CEO of the American Negotiation Institute, and I have a quick question for you. When was the last time you had a difficult conversation? These conversations happen all the time. And that's exactly why you should listen to Negotiate Anything, the number one negotiation podcast in the world World. We produce episodes every single day to help you lead, persuade and resolve conflicts both at work and at home. So level up your negotiation skills by making Negotiate Anything part of your daily routine. And we're back here on Big Technology Podcast with ilan Boro, the CEO and founder of ThrivePal. And we stopped at the break talking a little bit about how the plan for you is to have people use AI to sort of inspire them to go help others. It's very interesting that that is your perspective because we just saw some research come out that shows that the number one use case of chatbots today has gone from idea generation to companionship and therapy. And it seems like you're not taking that path and you're trying to have an alternative where you're trying to spark human connection as opposed to have the bot be the connector itself. Why is that?
- [27:58]
B
So it's first is I'm completely okay with just about teaching you lots of techniques. I think it can work very well. But I think a lot if you want to enlarge that first circle of people who can just learn from first thing, lots of people can learn just from books. Those people have already learned those techniques if they're interested in it. I read so many books on this. If you read all the books, you know a lot of things already, but that's going to be a narrow set. Then you have some people who will learn just from about a little game. A thing like you can think duolingo style. Okay, perhaps if something is more like bite sized, I can learn and I still enjoy doing it single player mode. And the thing that has been shown time and again in research is that you need some kind of accountability. And accountability works better if you have a human reference in the loop, someone who keeps you to your commitments, someone who checks in on you. And someone who, like, for example, if I have an appointment with a doctor or a personal trainer or a friend, I don't just not show up. And that's okay. If I don't show up, I have to warn them or I have to give some kind of excuse or warning or something with a bard. If you don't show up, it's like nothing happens. You don't have this whole social expectation thing. So to leverage all these kind of things, you need a human in the loop who's going to be to tell you, hey, you know, we were supposed to go running and you didn't come. Something like this, or we were supposed to lift weight together and you're canceling on me again. And they remind you first they know you in a way that like this whole premise of upload your entire life to the bot so the bot knows you the way your friend would is like, hey, you already have friends, they already do know you, they already know your leaders.
- [29:40]
A
We forget that in the comments.
- [29:41]
B
No, that's the thing. It's like, why not ask your friends? And your friends have your best interest at heart often. And you know that you trust that they do. There's not. And plus you don't have this thing where suddenly the startup that you were entrusting all this. I say this as a founder startup myself, but it's like if that startup goes under and disappears, what happens? Your friends, you kind of have more of a sense that they've known you for a very long time, they aren't going to get bought by something else. And suddenly there's going to be a pivot and suddenly your data is going to be used for something else. So there's more of a sense that there's a shared flourishing objective, that they want you to thrive and that you trust them. Which I think is very hard to replicate with a machine. And my question is not just that it's hard, it's like, why would you when one of the biggest source of thriving is to help others? Like whenever I help people and whenever someone tells me that I've helped them improve their lives, it always brings tears to my eyes. It's just so beautiful and it's like so deep and so touching and it's like this kind of deep emotional fulfillment that you need to feel useful and so why take that away from people when it's both more effective and kind of a win win situation? Often people forget that a lot of situation in life there is a win win solution. Not always, but that's one case where there really is also because a lot of what makes it hard for us to thrive is that we don't have the infrastructure so much for thriving in the sense that, for example, if you want to go running, it's much easier if there's a running club in your building with your friends that if you have to start it from scratch, that's an example. If you want to eat healthy and you go out and you see five fast food restaurants, it's harder than if you go out and you see delicious healthy food with all good salads. That's all very healthy, cheap. And like, that's what I mean by infrastructure for thriving. Now, the default thing when you want to socialize, what it is, go get wasted and drink. Why don't we have stuff where like people go have tea. I don't drink. So it's very frustrating for me that the default socialization thing is always to go to a bar. And especially in New York City, it's so noisy you can barely hear people. It's like, I want to have tea parties. This kind of thing is like, why don't we make it easier to do the right thing? And so when I say the right thing, I mean the healthy thing. Long term, in a way, that would still be fun right now. There's often an opposition between being a cool person that's fun and doing something that's healthy for you. I don't think that's a necessity. I think we can have lots of things which would be fun and still be social, but for that you need to bring everyone. And so that's why I think doing that as a team thing where someone who's helping you discover the pleasure of tea drinking as opposed to just getting drunk. It's like after that you can all socialize as a group, eating healthy food and drinking tea. I sound like Brian Johnson, but I mean, he has a point sometimes for that. So for me, that's one of big reasons. Even if you could get just everyone to have that with a bot, it wouldn't fix the thing that you need your group to want to socialize in this way so that you're not the only one saying, oh, no, but it's my bedtime, I need to sleep and I don't drink and I do these kind of things. So building it as a team.
- [33:15]
A
So then I want to know what you think about the fact that everybody seems to be gravitating towards the bots as companions or even lovers. I mean, I had an experience a couple months ago, where I was walking on the streets in Brooklyn and these two friends were talking and this woman was talking about her conversation with ChatGPT about the type of man that she should have in her life. And I was like, wait a second, you have your friend right there. Like, why are you speaking with ChatGPT and not your friend about this very kind of very important decision. And I, I don't think she's the only one. I think this is very common.
- [33:48]
B
Now, first thing, I think there really is a place for ChatGPT and friends. When I say friends, I mean ChatGPT's friends. I don't know if they're friends, but you know, meta AI like all the other bots, like pick your own bot.
- [33:58]
A
I think people really do view them as friends.
- [34:00]
B
Yeah, but they don't really talk to each other. Well, I mean, they're friends to you. But what I mean is the friends of ChatGPT like ChatGPT and friends like all the bots. So because you can use them in a very different way, they have lots of strength in terms of you can rehash the same thing 5,000 times and not be using up their mental capacity because they're always there. You can talk to them at 5am, which you cannot always do that for your friends. Sometimes you shouldn't be doing this. Like, people have boundaries, which is very healthy. All these things.
- [34:32]
A
That is one nice thing about chatbots is you can just go over and over and over again where a friend would just lose patience with you.
- [34:38]
B
Exactly.
- [34:38]
A
And it's there.
- [34:39]
B
And it's like also, I mean, they have this huge access to body of knowledge, for example, even outside the realm of companionship. But if I want to do some kind of deep research or something, if I do it myself, I know how to do this, it was my job. But it's going to take me a long time. I cannot just instantly scan the Internet and summarize it. I cannot. It would take me time. So it's a big time saver and just it also gives you a good sense of what I call the distribution. So if I want often there for lots of cutting edge questions, there isn't a consensus, you have a distribution of opinions. And if I want to get a sense of what that is, I can ask, okay, what's the mainstream opinion on this, how established this. And an AI system can go look at everything and kind of say, okay, this is a fraction of people who believe that. There's this other fraction that believes that, and this is how they interact. So you get a sense of like a meta view of the opinions on that question, which is very hard to get from just one person because it takes a lot of time. You need to not just read one paper, but read everything and get a sense of how they interact. So that's super useful for forming an idea. So for me, I view more like the bot as part of the pod, but you have to have your friends in the pod. And so it's very tempting to have the bot because the bot has all these very nice qualities. So for me, it's not unlike video games. Video games have been around for a long time and they're very tempting because they give you this instant sense that you matter. I mean, often you are the person saving the world. I don't know if you've seen that movie. What's his name? Free Guy, I think.
- [36:20]
A
Oh, yes.
- [36:21]
B
Free Guy. Yeah.
- [36:22]
A
You know, that's my father's favorite movie.
- [36:23]
B
Oh, really? I didn't know that.
- [36:25]
A
He's watched it probably more than 10 times.
- [36:27]
B
Okay, so I love that movie.
- [36:28]
A
That's our Saturday night activity at home is we put the put on Free Guy and we watch Ryan Reynolds. Wake up. Video game atmosphere.
- [36:34]
B
Yeah. So I love it. So do you remember that scene with Channing Tatum?
- [36:38]
A
Tell me more about it.
- [36:39]
B
Okay, so it's the one who. He has a part. No, he's a streamer. Streamer, Right. And so he's live streaming and he has. In the game, he's this Channing Tatum character and he has this catchphrase and then suddenly he runs into Free Guy and while they're having this big scene and he wants him to do a catchphrase and then you hear his mom saying, hey, can I do the vacuum? Or something like this? Like she's in the back. Because he lives and then you see the real world and he's like this. He's just stuck in his gaming chair and he lives with his mom and she's trying to vacuum. And so it's like juxtaposition of real life where he feels like he cannot be a hero, and a video game where everyone is a hero and you matter and you're super powerful and it fulfills all these things that life did not. So it's a shortcut, but it's a simulacrum that at the end of the day is not really fulfilling because you don't get the real thing. So the question is, are we going to get to a point where it's so good that it actually fulfilled completely and then you get brand new world, the soma kind of. So soma is, I've read that a very, very long time ago, but like the perfect drop drug, you just, you blissed out in your bed and everything's great and you don't do anything or matrix the metrics, you get good stick and stuff. So is that a problem?
- [37:51]
A
One could argue this already exists.
- [37:53]
B
Yeah, well, we don't have a perfect drug yet, but you get some kind of that. And, and so of course if you, if you look for that in technology, you're going to find something that gives you a lot of things that are very tempting. Because first. And a lot of these systems, they're very, very attuned to your needs. They always think you're great, they always give you lots of compliments, always tell you what you want to hear. They're always available and they don't have needs. And if you get used to this as what a great interaction should be, then it makes you less and less able to interact with humans who have flaws, have their own needs. It's more of a give and take. And it's very healthy to learn that give and take. So for me that's why it's very tempting. You can have a bit of it but not get hooked on it so much because you need to be able to again stretch your mind, be resilient to the world, sometimes pushing back. And also because if you want the whole world to thrive, which I do, you cannot aim for a world where there's no configuration networks. So for example, I think I give you that example last time we worked together. But my English teacher used to say back in high school or I don't remember, used to say, oh, it would be great if everyone was in the top 10. It was a joke, but it's kind of the idea. It's like you cannot have everyone be in the top 10. So if everyone wants to be the top person, the only way this works is if you have a multiverse that's completely splintered but where everyone has their fantasy world where they're the top person. So you can have that. And that's what virtual reality gives you. Like everyone has their own virtual world where they're, whoever the superhero is or you change what you want and you're like, I don't really need my fulfillment is not just me beating every, like getting the best of everything. But it's more having my community that is happy and everyone can thrive and we can do something together. And there's lots of ways to find that that. Right. And so trying to move people towards this is more important to me rather than trying to indulge people into something kind of fantasy where they can get that. But it's fake.
- [40:09]
A
Yes, so, so you have a company that is attempting to make this happen, to sort of nudge people into moments that or into habits that would make them more fulfilled. It's funded by open AI, I think that's public. Right, we can say that. Tell us a little bit about what you're trying to do with ThrivePal.
- [40:26]
B
So it's funded by OpenAI and General Catalyst mostly and a bunch of other people. What we're trying to do is to give people the tools that are very science backed, that we know, work and help people discover them and experiment with them and find what works for them. So those kind of tools, they've existed for a long time, but not everyone's going to be receptive to the same one. So a very well known one is meditation, for example. If you look at meditation, you're going to find lots and lots of research that shows that it can do great things for people. It doesn't mean that everyone is going to find that meditation is the answer to all their problems. It means it's worth trying, it's worth learning about, it's worth playing with and experimenting with. If it works for it, for you, you keep it. If it doesn't, you try something else. So there's a bunch of these things that are powerful and work for some people. People and some things that even if they don't work for you, you kind of have to find a way to do that, otherwise you're not going to function so well. Because until we get uploaded to cloud or whatever, we live in bodies. How to take off your body, for example, exercise. Physical things we all need. Yeah, exactly. So how do we get people to find something where they can move in a way that's fulfilling for them? It's going to look very different for different people. Some people like running, some people like boxing, some people give them a sense of, okay, this is what you need to do. You need to do some cardio, you need to do resistance training. That one is really important to me, especially for women because for lots of women, they think they need to be toned, which is kind of a way to say that they don't want to get too bulky. I used to also not want to get too bulky, but actually you need to do resistance training because that's the only way to train strong bones. And also because muscle have a really strong Role in metabolism for dealing with glucose. So for lots of health reason people need to do resistance training, not just run. Like exercise means lifting or you can do body weight stuff you can do. There's many ways to do resistance training. But trying to get that message out because we know it's true, the research is very solid on this and somehow lots of people don't know this. And you know also one of my personal thing is people wearing foot shaped shoes. This kind of things where people wear shoes that are terrible for their feet, that way too narrow. These kind of things which I think is really common sense but somehow the message is not super well known. So what I'm trying to do too is to get those messages that I think are really important out in a way that can speak to people. And what's great with AI is you can iterate a lot. You can try lots of different ways, you can try lots of different voices, lots of different packaging for these kind of things and see what works and get some kind of feedback on this. So that's also what I'm trying to do.
- [43:13]
A
Right. But it's not easy because these are eat your Wheaties type messages, for lack of a better term. Right. It's like I, it's going to take me like I have a resistance. Even though I love exercising, I have a resistance to doing it. Getting to the gym is the hardest part. Once you're at the gym, it's actually kind of fun.
- [43:31]
B
Yeah.
- [43:31]
A
And so for me, like I just opening an app that's going to tell me things that I need to do that don't bring me that instant gratification. There's like a pain to it even.
- [43:41]
B
Yeah. Yeah.
- [43:42]
A
So how are you going to do this?
- [43:43]
B
So first thing is, I think there's several levels of learning. The first one is even knowing that it's a thing. So for example, if I tell you I have groundbreaking information, if you smoke, you might increase your chance of cancer. You're like, oh, I know that. Everyone knows that. That wasn't the case a few decades ago. So first getting the information out. There's some things like this thing around the role of muscle for metabolism where I think the information is not out yet. So that's level one is like putting the information out for some of these things. Like even before you finding it's painful to do something, knowing what it is that you should do and aim to do. So like yeah, step one. So this for the microbiome. I think there's lots of things where people are not aware of how important it is to take care of the bugs in their. In their gut. Like the fact that 70% of your immune system is in your gut and that it reacts so much to what you eat. And that perhaps if you have a choice between taking this antibiotic and not. And you're just taking, like to be sure, like, just wait a bit, know that there's a downside that you're going to completely wipe out all the bugs in your gut and it can cause a lot of downstream problems. So I think there's a lot of information that still needs to get out at level one. Level two is knowing once you know, kind of like where you want to be, knowing the techniques that can take you there. I'm still not at the level of you actually doing it, but at least being aware of what they are. And then you choose whether you do them or not. But learning that mapping of in this context, if I want to get there, I should do this. And so the way I'm trying to do this, we're going to build a game. So hopefully it's going to be fun and entertaining just to familiarize people with that mapping of these techniques exist and you use them in this context to get into that outcome and then how you decide to actually implement them. That's the hard part. It's still hard. There's no way to. But there's still many ways that are more like what I call meta techniques, techniques which are about how you build a habit, how things that work for setting goals. For example, you know, you have lots of things around this. Researchers at NYU who have developed the WHOOP framework, which is wish outcome Obstacle plan. So showing that if you set a goal, it works better to not just visualize the outcome, but also think about all the obstacles and make a plan for what you're going to do when you run into these obstacles. So lots of techniques that people can learn that are about how to make those changes in your behavior to get to the goal you want. And something that's very, very important to me too is often people haven't thought all that much about their values in the sense of what they value in life, what they want out of life. And often, I think with AI, it's especially important. People often talk about the algorithms and what you're trying to learn. But with any algorithm, you have a sense of like this objective function, what it is that you're trying to achieve. For lots of humans, it's not very clear. They can tell you, oh, I want to make more money or something. But it's very important for people to take the time to think what it is that really resonates with you, what it is that you want to get out of life. Which often I mention this thing about helping other people. It's like I myself, I have a list of things that make me cry for, like things that I find deeply, deeply moving. For example, there's this video of Sir Nicholas Winton. Do you know who that is?
- [47:02]
A
No.
- [47:03]
B
Okay, so Sir Nicholas Winton was a. A British person. I'm going to butcher his story, but more or less a British person who was a, I think 20 something just before World War II. And he went on vacation in the Czech Republic for two weeks. And when he was there, that was before the Nazis arrived in Czech Republic. And he realized that all these children, the Jewish children, were going to get like, wiped out, out. And so he by himself organized all these trains, eight trains, I believe, to find families for all these kids. And he did it like a mail order catalog. Like, he would take pictures of the children and then send to families in England. And people could pick the kids they liked, like a mail order. And he saved something like 700 kids, a lot, A lot of kids who would completely have died without him. And he just did that, like himself. And he didn't really talk to anyone about it. And they just really. People randomly found out, like when he was 80 or something like this. And there was this BBC documentary about him where. Which is the part that makes me cry that you see it's online on YouTube where there's this BBC show, I think it's called, that's Life. And he's sitting there in front. They just ask him to come. And then the presenter ask him, tell him his story. And he says, well, you see, sitting next to one of these children and it's this lady. And then the present. I'm going to cry again. But the presenter asked them, like, is there someone else who owes their life to Sir Nicholas Winton? And then the entire room raises.
- [48:39]
A
Oh, man.
- [48:39]
B
And it's all these people that he had no idea where they had become. They had gathered all of them. And they said it was like all these people had kids and grandkids. He saved like 15,000 people. This kind of thing, it's so moving. And it's this thing that really shows people like, what, what connecting what you want to do with meaning is. So that's a big story, but there's another one, just a silly commercial for kettlebells, right? Kettlebells exercise. So you see a guy, an old guy, and, and he, and he just.
- [49:08]
A
First, it's gonna be any different from story number one, but go ahead.
- [49:12]
B
No, I, I, I know, but it's like just about connection to humans.
- [49:15]
A
Yeah.
- [49:15]
B
I don't know.
- [49:15]
A
I wanna hear it.
- [49:16]
B
So he goes to, he's like, he's this old dude and he like walks and goes to his like closet and he takes this kettlebell and he likes it. He can barely raise it. And then he tries and he barely does it. And you see him like go back and every day he can raise it a bit bigger, a bit higher. And in the end he managed to raise it. And at the end, and you wonder, and his neighbors think he's crazy and they call the cops and whatever. And then at the end, you see it's Christmas and he all dressed up and he picks up his grandkid and he gives her a star and he raises her that's beautiful. So he can, she can put the star. And he's like. So it's like, that's the thing, it's like connecting what you do, remembering why you do things. And often the end goal is going to be something connected to people you care about or to lives. And it can be your family, like the second case, or it can be complete strangers, kids like Nicholas Winton, which he lived to be 105. And he's like, you should look his interviews when he's, I think there's one where he's 1, 103 and just saying, he's just talking about, well, I don't like to talk about the past so much. People talk about the past too much. I want to worry about what we do now. And he's like so humble. He's like, like I, I like him very much, but it's very heartwarming. So like remembering why we do things, why, why do we care? Why do we want to be alive? And trying to build that connection. So figuring out what it is that gives people energy, it's going to be different for different people. People care about different things. They have different values. But lots of people have deep values that mean something to them. It's not going to be the same for me as for you, but there is something. So building that connection. And what I think back to AI and stuff where AI can help is to help like build a step by step process for how you figure out what it is that you value. Remind you to connect that to how it translates every day. If you want to be there to raise your granddaughter on the tree in 40 years. What do you need to do today and every day so that you can get there? So these kind of things.
- [51:13]
A
So what you're building is not. When I first downloaded the app, I was like, okay, maybe this is because it's an app. You can go get it. You go to your website now if you want to try it out, right?
- [51:22]
B
Yeah.
- [51:22]
A
On thrivepal.com and thrivepal.com thrivepal.com folks. And so what you do now is you download it and you sort of decide what you're interested in. Some sort of physical health thing or some sort of other goal. And I thought maybe this is going to be like a therapist or a companion. But really what it is is it seems like it's more of a coach to help you.
- [51:43]
B
Yeah, it' really more of a coach. And like, what we're trying. So we're building it little by little, of course. But yeah, my. My goal is to build this kind of roadmap for thriving. Like, you have a blueprint for how to thrive in your life, how to build these systems so that it's an upward spiral rather than just following what it is that you want to do right now, which can very quickly lead into more of a downward spiral where you're tempted, you do more of these things, then you eat too many sweets and you don't move enough. This kind of thing where you just chase the temptation instead of saying, okay, I can actually build the system so that my life improves little by little and. And you get to where you want to be. There's also we. We. I started, I just started like two weeks ago, our Facebook page on ThrivePal. Not Instagram, because Meta suspended my Instagram page. So for now it's like tiny. But Facebook, we still have this page.
- [52:34]
A
Wait, why do they suspend your Instagram?
- [52:36]
B
I have no idea. The systems are, you know, you don't have any poll.
- [52:39]
A
You can't DM market Mark. All right.
- [52:44]
B
I contacted one of my meta colleagues, so we're working on it. But the Facebook page, we have a bunch of posts around this, trying to convey these concepts of what people should do too, because I think for lots of people, they are very strong sources of motivation for them that can help them achieve these goals. Goals. But we need to find them. And it takes a lot of time. And that's where AI can help, because AI has a lot of time.
- [53:11]
A
So do you think that the technology itself, like AI technology, is going to get so good that people are going to want to do things because want to do these Things I feel like anytime someone's tried to build an app with these type of purposes, like even Duolingo has all these addictive dark patterns built into them. So are you going to be able to do it on the strength of the AI technology itself or are you going to have to resort to things like, like, I don't know, you have streaks. But like Duolingo has the most passive aggressive notifications, like, your grandma's going to die if you don't do Spanish today. So how are you going to get people to use this?
- [53:52]
B
Well, so. Well, there's also, you know that app Finch?
- [53:55]
A
No, I haven't used it before.
- [53:56]
B
Okay, so if you know Finch, it's trying to push self care. It's very encouraging to me because they have streaks too, but they're very like polite. They're never pushing you, they're never pressuring you. It's very supportive and people love it. And it's kind of in the same mind space as what we're trying to do at ThrivePal, which is really trying to support people to build better habits in a healthy way and not kind of take advantage. Of course there's this thing that we still need to survive at a company and it's much harder to push grilled broccoli than it is to push soda. So yeah, that's the thing is like, like how, how do we manage to do that? But there's lots of things that can work because it's a thing that we. So for example, there's one of the posts I did on the Facebook page is about temptation bundling. Right. It's the idea of trying to harness the things that you want to do for which you're already motivated to get to your long term goals. Because sometimes you can find a way to trick yourself into doing something by using some of the things that you like doing. So for example, let's say I want to watch a TV show. It's one of my personal things. Sometimes when I start watching a TV show, I cannot stop. I can always try thinking, oh, I'll just watch one episode and then I'll watch the entire seven seasons and I haven't moved from my car for two days. You know, it's kind of things. So there's several things I can do. Either I say I don't watch a show myself alone. Like someone else has to have the remote control. But something else that works very well for me is I'm only going to watch shows when I'm like at the gym. And so that kind of built this nice partnership where the thing that's painful limits the other one and the other one pulls the first thing. So trying to find this balance. And so I think for a lot of the what you mentioned, dark pattern and stuff, there's ways to leverage something that are going to be tapping into short term motivation, but it's aligned with what you want to do long term. So an example that actually comes up a lot is when we talk about the game design, because I mentioned we also want to do a game which is not the app. Like the game is not out yet, but we want to make a game. It's going to be more engaging because you have some game stuff where you have in game currency that you try to accumulate. But something that we're not trying to do is to have it be to really trigger immersion when we try to keep people in there, because that's not a goal of the game. We want them to come back every day to learn new stuff, but to not spend two hours, that's not. And actually that's a lot. That's very true for the gym. Too often it works better if you were to do five sessions that are shorter than one gigantic session every month. It's just repeated short things. It's more like brushing your teeth. Can we get to that brushing your teeth part? And we don't need people to stay long trying to find how can we leverage these things Just because even the way your brain is wired, you need dopamine to do things. Otherwise, you know what happens when you don't have dopamine? You're like you have Parkinson's, you're not moving. So to get people to move, you have to tap into that short term motivation. So either you find a connection with deep motivations or you ensure that, hey, I'm going to use it in a way like with temptation bundling, where it's like horses that draw a card that I want drawn, you know, so that's what we're trying to do.
- [57:20]
A
Okay, I think this is going to be my last question for you.
- [57:22]
B
You're.
- [57:23]
A
You're creating a AI startup that basically people talk to this coach that helps them flourish. The big question is, is the value going to be within the big chatbots themselves or in. I don't, I don't know what you call it. An AI wrapper or is that demeaning?
- [57:41]
B
Sorry, I don't understand the question.
- [57:42]
A
AI wrapper. Like a. The term AI rapper is, you know the term AI rapper? Yeah, yeah, yeah, right. An app built on top of a model, I guess like to put this question a different way. Is there a chance that people will just go into ChatGPT and say hey, I want to thrive help me?
- [58:00]
B
Oh yeah.
- [58:00]
A
And so how do you, how do you anticipate that and how do you plan for that as someone who's building an app?
- [58:06]
B
Yeah, no, no for sure. So first thing is, knows lots of good things. But ChatGPT is very permissive and not so opinionated necessarily about what's good. And so we give a lot more structure as to like okay, these are the things that you need to think about. This is how you organize your thing. These are good techniques we give a lot more of. We have a strong stance on what the good directions are. So that's kind of way more curated and same especially for the game where these are all going to be what we provide. And if you go into our app there's this thing called like instant insight which are little techniques that you can learn about. I decide what's in there. So it's more like this choice, this editorial choice. So it's like editorialized ChatGPT and also the way I really ask it to be science backed and I check like all the literature myself because I've read so many papers, I've published in neuroscience and like theoretical neuroscience, these kind of things about how the brain works. So making sure that this is reliable, which ChatGPT is very happy giving you bous information sometimes because some people believe it's true and science is complicated. So I trust myself way more. So I do that work to ensure that what we put in there is in the right direction and that's defensible. Long term, you mean? Yeah, well I mean I would be very happy if ChatGPT starts becoming very reliable for sure. Now there's this thing, it's like there's no necessary very consensusly you're always going to have lots of people thinking different things are more important. So I have my own way of what I think is important and I think, I think my way is good.
- [59:51]
A
Yeah.
- [59:52]
B
So that's kind of what we have there. And this idea of having more of a community multiplayer thing for me is, is what's good. And this whole thing around the game also is for me is going to be very like unique, not at all like ChatGPT. So for me that's very different. But I think it's trying to be at the same time it's very broad because about thriving. It's way more narrower than what ChatGPT is trying to do to accommodate so many different things.
- [60:19]
A
So, okay, this is really the last question. When we met, we met a couple weeks ago after, I think we met like maybe eight years beforehand. So you asked me a very interesting question towards the end, which is what my hope was for the world. I just want to end with that. What's your hope for the world?
- [60:34]
B
World. So you know what I mentioned about how people have like deep motivation to see others thrive. And it's like, I really hope we can get to a world where people don't think as much in terms of zero sum thing, where it's all a competition and if you take something, it has to come out of somebody else. But that we leverage the higher space of possibility to encourage more human flourishing as a community, as a village, as a group of people, rather than continuing to think, how do we optimize in a very narrow way that's often short sighted. So for people to feel like their life has a lot of meaning, that they know what that means for them and for their friends, and that we can find these ways of being where they're not as influenced by that whole past of scarcity that we had for millions of years as humans. And then we kind of adapt to the new technologies we have and we are like aware of the fact that we often are our worst enemy and we shoot ourselves in the foot ourselves. And we're going to get bigger guns, like metaphorically with all this AI stuff, which is going to make it very easy to have this tempting stuff coming at us. And so I really want people to have all the tools so that they can be more competent and autonomous and make the right choice for them for what really allows them to thrive as a, as a whole species. You know, long term, I want everyone to be happy around me and everyone to be thriving and my friends to be healthy and people to know how to regulate their emotions so that we can have civilized conversation and debate stuff. And everyone is kind of making music and dancing and running around these kind of things.
- [62:19]
A
Yeah, no, I love it. It's full circle, right? We started talking, talking a little bit about the polarization and you can bring it back now that people should, you know, not view the world as zero sum and not allow some of the things that are trying to rip them apart do that. And if AI can help us get there, I think that's a much better future. So thank you for sharing your vision with us.
- [62:38]
B
Thank you so much.
- [62:39]
A
All right, everybody, thank you for watching. We'll be back on Friday. Breaking down the week's News with Ron John Roya. Until then, we'll see you next time on Big Technology Podcast. Sam.