Summary
Notes
Transcript
Thank you all for staying. until the bitter end. We'll have some thoughts about the week, and then we'll have--The final stuff is a sponge, and then you guys can go have your tea. Also, US is planning Australia soon. I don't think I'll end by noon, but I also don't think it'll stretch before noon. somewhere that I have to be Like holy things that are happening.
All right, so I have a bunch of thoughts, and those thoughts kind of relate to The guest speaker is. Did I interact? They're basically all the guest speakers. Yep.
So everyone who I saw as a guest speaker, I have some thoughts about to give you my take ashow I sort of interpret what they said. And then we'll talk a thing or two about how I think you can use like a little bit of tips on how to use AI in the most fundamental way as opposed to the boring ways that sort of most people are using and then and then I'll get you guys to say something at any point in time, ask me whatever you want to ask, and that's the plan.
Make sense? Yes. When I ask a question that's rhetorical, I want to. All right, let's do it. So the first keeper that we had was Darshan.
You recall, He told us the story of how he basically had a life transformation and So now he basically helps other people engage in that transformation. So there's a bunch of-Things that he talked about that he suggested you do and-I wanted to give you my perspective on the Thank you.
why some of those things work. Um... And I don't think it's because of what you're measuring. So if you remember, he gave us a list of 10 different things.
The outside regime is this, the higher rate of variability is like a million things, he recommended the group. I like all this stuff. And the impression one can have when you're listening to that talk is that it's because you're doing the measurement that you get the advantage. And I think that that's, to me, that's an essential thing. There's no doubt that in some extreme cases, like I think, Jihad, you had an extreme case where kind of a measurement kind of alerts you to something that you kind of weren't aware of at all. It's like all kind of a warning sign. And so without a doubt, those things can occur. But that's different than longevity and wellness. That's more kind of, you know, should be caught in your annual checkup.
And the question is, is your annual checkup sufficiently comprehensive? But he's not telling the gay, He's saying, I'm going to measure all these things because that's going to be the only thing. One thing you should always remember about longevity is that you can only belong to one condition in the experiment. Which means that, you know, you don't know how many more days you've got. because of all these things you're measuring. You're just assuming you've got more days. And so, and so, and And that's okay, but I think it's important for people to understand what they're buying.
So what is it about these metrics that's actually useful and why does it work and what does it work on? And so here's the thing about psychology. People respond to things that are concrete and things that are not stress. And, um... And so what these metrics do is they give you a bunch of concrete pieces of information that you feel you can actually react to it. And-And not just that, but what technology actually does is technology enables you to see these things not once a year, By the way, if you go at all, it's not goingPeople skip their annual checkup.
Here's what happened in the checkup. I went to the checkup and I discovered I was a bull. Between the years, I save time all those years. But basically, by having something that's concrete and orderly, or in some cases concrete but daily, What it does is that it serves as an impetus for you to act. Because ultimately, if you paid attention when he was talking, there's the measurement and then there's--The cure is the cure as we've always known.
Ancient man walked around a lot. We need to walk around. We need to act like a sort of-The determinants of health are the same determinants of health even before people started podcasting about longevity, which is social support. Relationships. Physical activity. Healthy eating. All right. So in other words, have friends. Right? Don't be lazy and don't eat like an idiot. So those, you just want a short life.
So all those things remain the same, What, what, uh... What's important about what Darshan is doing is that he's getting people to actually act on it, act and do those things. The problem wasn't We didn't know what to do. The problem is how do we actually put the fire in us to actually do it? And that's, the problem is behavioral, the problem is not health, right? It's your bad behavior leads to that, you know, like leads you to maximize your health potential, right?
'Cause your genes are still your genes. And it's not, I think he used the, he said 90% is behavior and 10% is genes.
I don't know what empirical evidence one would have to collect in order to be able to say that 90/10.
My view is that genes are undefeated, and so the question is, what do you do, those genes can express themselves in different ways, and your behaviors interact with them. And really, it's a huge mystery, and we shouldn't be, we shouldn't, You should believe anything or anybody who's over your confidence. about exactly what's happening Because scientifically, we're not exactly sure that--In jeans, we have some sense of-But I think the value of metrics is important.
And, um... And if that helps you, that's great. General lesson here, which is people respond to concrete metrics and not to abstract metrics. So to the extent also that you're using-you want to motivate people to act. About things being organized. then the more you live in the concrete, the more likely we are to respond to it, the more you live in the abstract. Good act, boy. Don't hold it to you. Because what that concrete stuff does is it will at least like I can attach my behavior to something that's actually happening. So that was the first thing I wanted to point out. The second thing that I wanted to point outWe can talk about morning habits.
Somebody said there's this book called Morning Habits, you gotta do this in the morning. Um. So, so morning habits work not because they're morning habits. And it's not because you did them in the morning. They work for another reason. Which is, um... that one of the things that we're good at a bunch of things and we're horrendous at a bunch of others. And one of the things that we're horrendous at is being in the plant. Okay, there's something called the planning fallacy. It's very, very common.
Who's, anybody here, some of you are in construction? Right, so you know, you always think it's gonna take X, budget X, and it takes X squared. How many of you have renovated your house? Always took more time. I think Dino has been building in Dubai for as long as I've known him. Every year it was going to be next year. When I was at Columbia, where's Sergio? Is he here still? Yeah, Sergio, remember when I joined Columbia in 2003, we were getting a new building.
I left the Columbia in 2011. We still didn't have a new building. They entered it in 2014. We're really, really, really, really, really bad at playing. We don't have a good sense of how long activities take. And so there's two reasons for that. One of them is that Activities have components. and we only account for those components if we think about them. And to think about all the components of an activity in a day is just kind of a lot of things to keep in mind. And we just don't.
It's just not top of mind. And so when all those non-top of mind things start having to happen, everything starts taking longer. That's the first thing. The other thing is we're overly optimistic about our ability to get things completed in that time. That optimism is something that actually serves us well. State or council question. The optimism actually motivates us. Um... But the net result is that you say to yourself, okay, I'm gonna do my movement, my steps, my this, my workout, my, my, that was I'm trying to convince Dan to dead hang. My dead hang is today at 5. 5 rolls around, and you're like, ah, shit, I didn't finish whatever it is I was doing. That's going to happen to you every day of the week. And with five children.
Five times over. And so why do morning habits work? Morning habits don't work because of their morning habits. They work because You get them out of the way before the cascade of Um... Out. Fallacious plans start coming out of your head. You didn't anticipate, you thought you'd be done by five, you're not, and then you're actually done by six or seven. But if you did it at 7:00 a.m. or at 6:00 a.m., then, If you're done with whatever task you have to do for work at six or at seven, it doesn't matter. You don't edit out that one thing that contributes to your health. Okay, and so, and so, I remember I was talking to Benedict about it. He's like, I got five kids and mornings are tough. I'm like, I'm going to come in six.
I'm like, I have to wake up at five. And so, yeah, that totally sucks. By the way, I mean, so Benedict, just kind of judging by the crow's feet, we're a similar age. It's easier to wake up early in our age. Or just when you wake up to P&M all night, just stay up. And then you have multiple options when you want to do that at 1 and 3 and 5. Just pick a session and just extend it. So that's the reason morning habits work.
They don't work because they're morning habits. They work because they get around the planning policy. And in fact, one of the things that both points one and point two, I want to say about them is that, look, we are the people that we are. We can't beat what we are. What we have to do is... create a system for ourselves that takes account of how our mind works. Um. Sometimes what I laugh about these things, like on these podcasts, is it's like they think that you can actually beat yourself. You can't beat yourself. We've been this way for millennia and we're not about to change.
What we can change is the environment around us that comes with it. Swim downstream if you want to get places. Swim upstream for fitness. the downstream that you want to get places.
You try to do something that's the opposite of what your natural predilection is. You're not going to do it right. But if you create a system that takes advantage of your natural predilection, you're going to be sidestepped. The downside of your natural predilection then you're fine. Listen, if you can't lay off the cookies, don't buy cookies. Right? So, so, um... So formate rules for yourself that you can actually follow. I try not to eat pasta unless it's homemade.
Like where they actually, you know, old Italian nonna like kneaded the dough.
It turns out that you're like worried about carbs. You still get to have it, but you don't have to, you know, you don't create a system to have it. The third thing is this whole thing about biohacking. So we laughed a lot about peptides, peptides, injecting peptides with some of the drugs.
Oh. And people are casting multiples. Times, you know, like, you know, as I was approaching 50, and now that I'm crossing 50, Do you buy a house? Do you buy a house? Do you know Andrew Buberman? You're at Stanford, right? Do you know Andrew Buberman? Do you listen to his podcast? Do you do AG1? All this stuff, and I'm like... You know, like, first of all, no. Like, I don't listen to anything like this.
But But I also think that I think it's important. I don't know enough about biology to know about biohacking. I think, I think, I think, uh, I think We have to be humble. If biohacking gets you to behave in a healthy way, that's awesome. Why does Darshan's practice-what is this? One thing he said though, One thing I know for a fact is Everybody. Is field back. Which means that at the end of all that $12,000 and all the tests and all the da-da-da-da-da-da-da-da, what do you compare it with? A subjective sensation. Subjective sensation. Even so now, is that not valuable? Of course it's that. Right?
You have no idea if you're going to live the extra day or how well you're going to live it. And that has value. Then you have to ask yourself, okay, well, what is it about this thing that made me happy? And my interpretation, based on what he said, and also kind of knowing this industry a little bit, is that what happens is that Um. You know, when it comes to things like our life, our health, especially as we grow old, it feels like the rest of the world is determining what's happening to us and we don't have agency over it.
And when you're young and you're healthy, you don't think of it. But as you get older and every time you do something funky in the gym, now your shoulder hurts for six months and it just never goes away. Or you find out a heart attack or all of a sudden you got a mole and it's cancerous or heaven forbid somebody had cancer or whatever it is, your sense of... You know, your locus of control, you're so used to having it be internal, it becomes something that's external.
And for people that are type A and highly educated, all of you, and kind of myself included, that doesn't feel good. It feels like a loss of control. And honestly, when are people the least happy? When their kids are between the ages of five and 15, basically. You're no longer in charge. There's this other entity that's in charge, and they're paying the ass. And so the reason this initially works and making you happy because it restores your sense of control.
That's the key. It's not what you measure. Is that your measure? And the fact that you're measuring gives you a sense that, wait a second, these things have a way to act in that way. Now, I don't know if any, like our president of this school got kicked out of the job. Because he claimed there was some genetic thing that could cure Alzheimer's. You're about to get that. Freshman at this college got a tip, wrote a series of articles.
The kid eventually won the award, George Polk Award, which is an award for leading journalist of the year in the United States. And he got kicked out. He got kicked out because everybody's kind of looking for the key to Alzheimer's. I'd be very surprised knowing that you have the gene. All of a sudden now you can do a bunch of things that's actually been over for a while. What we know is anything that you do to improve your vasculature helps, but to reverse it, I don't know, like the guy that ran the school got kicked out.
Because everybody thought this would be secure. So I don't know that-I don't know that he would have had to have done what he did if it was already a waiver. I don't think that's... I don't think that's the issue. You cannot do something about your health, is something that's enormously valuable. Um... And so there's a lesson there that's a management lesson. which is the importance of giving people a sense of control.
And because a sense of control is motivated. Okay, and so when you delegate at work, delegate the ability to make a decision. Don't delegate a job.
Your job isn't to serve in the business. Your job is to make a decision.
decision about all of this. Because I make the decision about all the rivets, now I have a sense of control. And that's what makes us happy. and I'll do this in a second, Bernardo, and when we think about AI, a human in the loop, Right. All you need is to give people a little bit of sense of control over the output of the AI for them to be willing to adopt it.
It's the part of like the computer is doing everything, we're just gonna automate everything. I met this group from Germany and some guy's like, "Ah, we're just gonna have these agents "and these agents are gonna be doing everything "in the back end, they're gonna buy it for you." I was like, "Dude, that's not like, people want.
I want the stuff I don't like to do not to have to do it, but that doesn't mean that I want to abdicate all the decisions in my life. The need for control is not going away just because some technology allows us, you know, is going to allow us to make decisions faster, cheaper, better. I think that totally misses the point. So there's, I think, a bunch of lessons here that also are relevant to technology.
Bernard, you had a question. Yeah, I just want to get your take on this. So, and this is a relatively personal decision, right? Yeah. From a psychological point of view, and I know you have, a bunch of us probably did 23andMe after years back. Sure. I did not check the box. to see if I had a precondition for Alzheimer's. Yeah. I was-because I-If you can do something about it, then maybe you can come up with something to find out and maybe you can do stuff.
I was Right now after hearing this guy, I checked the box and I'm gonna download the report, but I said I'm coming out. Let's take the box. So what's your take?
My take is you should only get informed if the decisions about your behaviors are gonna differ. They're not going to differ. I'm not sure that there's-what was the name of the guy in Thebes in the Greek mythology who was sleeping with his mother?
Oedipus. Oedipus.
You know, Oedipus was a way happier guy when he didn't know he was sleeping with his mother.
And so we fetishize knowing.
But, but, Once you know, you don't know what's going to happen. So if knowing means you would change your behavior because of it, But if it's not gonna change your behavior, Once you know, you can't unknow. And so and so in the end like care careful, like there's the In chasing the objective part of your health, Don't forget that your life experience is subjective. And so in the end, what makes you happy is your interpretation of the world around you.
And what you need to do is you need to interpretation hack. And if I am hacking this into that grid, The goal is interpretation. If you live to 100, healthy like a bull, like you drop dead, but Otherwise your life sucks. I don't know. I mean, those are the kind of decisions we have to make or something. to decide, sort of, what are the kind of maximums. So kind of that's my view on it. And so--You know, I like to show this I like to show this paper towel to Spencer every year.
And I think it's an important reminder Because this paper, Talbis Gunzer, is actually a brilliant paper, Talbis Gunzer. And what's brilliant about it is it connects Behavior to its effect. And it connects it in the immediate moment. So I talk about making things concrete, not abstract. If you think about sustainability for a moment, let's go away from bio-machines, let's go to sustainability. The biggest problem with sustainability is anything that you Your behaviors now have an effect on the environment in the distant future and in the distant place most likely.
The plastic bottle that some of you probably doesn't have any, oh, this plastic, oh, this is probably compostable because it's Stanford, but it comes in a while. You know, some of those, like if you're drinking that disgusting sugary apple juice outside, which none of you do because you're all biohacking.
First on your product is if you don't have some form of immediate value that you provide, you're in trouble. Some dimension of value has to be. Or you're gonna be swimming upstream. That's one thing. The second thing is, Um... Also with AI, when you want to get it implemented, there has to be some level of benefit that people experience in the short run. Which is tricky with AI because AI is predicated on data.
AI-native, sorry, sorry, sorry. Not even AI-powered, you're native. Yeah, we all want to be natives. So, um... So it was interesting, like his thesis was that this is going to replace services. I think the thesis has a hole. And I told him about it afterwards. And we discussed it. He's not resistant to it. If you notice, everything he said was like, you know, the AI is going to do the parts that are function. and the savings.
But the people are still going to have to do the rest. This is still mediated by people. And then you say to yourself, Okay, so is AI doing the most important part? What's when I buy a service from McKinsey or from Bain or from whatever, What's driving my willingness to pay? What element of the product value proposition does it come from? This is also a triangle I show you guys every year. You shouldn't forget it.
And for more money. And by the way, every time I buy that phone, by the way, I also buy it for myself, it actually gets worse. I've never been in a product environment where the price goes up and the quality goes down. at the same rate, and yet I continue to finance. Why? It's because of psychological value. And so the tricky part about Ames is It's very good about pointing out the things about the function of the economics, but I think it misses the psychology.
Chinese videograph if you want, I don't really care. The iconic things. Providing value in this space. And so I think honestly, faster, cheaper, What is it called? Faster, cheaper, better, better, faster, cheaper? By the way, all the betters. just kind of live here. So it's really faster and cheaper. It's not even better. They just say better because things that come in threes are always more exciting.
And so, you know, faster, cheaper, better is the "oo-oo". It's necessary, but it's literally the most boring. And by the way, it's the place when you differentiate vertically. That works if you have some kind of structural cost advantage or some kind of engineering advantage or in AI's case some kind of data advantage that's sustainable. But if you don't have that, And that advantage isn't sustainable. You will get eroded.
Right? Now, if you're able to create value here in some form, this is way, so this is a horizontal differentiation. This becomes way harder to copy. Like, think of all of you who love your fancy brands. If somebody is copying some fancy brand, you're like, no, that's not right. Because emotions work as long as they're authentic, and they don't work if they're not authentic.
Now, on the subject of pains and the way Jake discussed it, He was saying, look, I have a service. The service performs job X. Now I have this AI doing the service. It performs a job X, but it costs me a quarter and it takes a 10%. But then, what you have to ask yourself is when you hire McKinsey, What do you hire? And in his explanation of his... of his-of his thesis. He didn't go through why I hired McKinsey.
I hired for McKinsey more than just for the analysis that they give me. I hired McKinsey so I want to fire people so I have someone to blame. By the way, so that lives here. And that's not part of aims. Now I also think it's, extraordinarily difficult. This is difficult value to deliver.
So he describes the services market The services market, that total, that huge column, Think of it as a willingness to pay. is composed of willingness to pay due to this, due to this, and due to this. The open question is how much of that willingness to pay is due to this? If this is two thirds of it, The TAM is still big. I'm not saying not do AI. I don't think he's wrong about the fact that it's things like the services, but the thing is, we want to have a more holistic view of this.
By the way, the other reason this is super important is because the adoption of AI in your organization by your customers, This doesn't block it and this doesn't block it. This doesn't block it because it really is that and a lot of things I do. This doesn't block it because really I have absolutely no idea how much it costs. So it feels free at this time.
Right? This blocks it.
So if you don't spend time on this, don't be surprised that the stuff isn't getting adopted inside of you.
So, I mean, I, I'm always wary of competing on price. Because-Because price is the easiest dimension to compute. And so if you're going to compete on, again, compete on price, you have structural cost advantage. But when two people can sit and make a $50 million company, just as much as 100% company can or 200%. Then all of a sudden it's like, where's the structural cost advantage? And that's the hardest part.
I used to be able to hire tons of engineers in India. I had 10,000 guys in Bangalore for 20 bucks a day. Whoo! Structural cost advantage because I had my India guys. Now, India is not so cheap, but also, I don't need 10,000 guys in India. I can just have two guys in LA subsisting on 7-Eleven. And do the same work. So I'm like, I'm always wary of that. I'm also wary of... Because we don't actually know how much this costs, like right now, we're living in a world of subsidy.
Don't think about what the work is, think about customer journey and obtaining that work and find those levers and then there's places where technology is going to help and places where technology won't. Listen, we still have pilots in cockpits, right? And so it's okay to have a pilot. I'm not saying get rid of the pilot.
The purser will make the announcement. None of us would do that. I had a guy yesterday in this Adrian, your group that came, a guy who was like, "Well, I know we didn't used to get into Waymo's." And now we get into Wemos. I think that there's, like, the catastrophe level actually is quite different. The interesting thing also about Wemos is Wemos still has a steering wheel. Just so we feel like...
We feel like it makes sense. A plane with no pilot doesn't make any sense. Um. You have to do the customer journey and you gotta look for the places where you can, how far out can I take a pilot and where should I actually leave a pilot? And I think in his description, It's He's still-It's still a little bit too much to pilot doing things many times. It still feels a little bit like you're going to jump.
There's a few things that I thought kind of, Psychologically, it was just thinking it came up to me. So Manolo also mentioned this thing about bigger, faster, better. And like the speed, I used to have to do this and this and this. And this is one of the best product people in the Valley, really. And he-You know, he had to, everything took forever. Now it doesn't take forever, it's super quick. Now, Here's the part that I think is psychologically challenging and organizationally challenging. Producing an attribute quickly is great.
But it's only useful if you You can also learn as quickly as you produce. So now, in order for that to be the case, you have to have some kind of Mechanism feedback loop, what was the feedback loop before? You didn't even need a feedback loop because everything was taking forever. Just by virtue of exposure, you could have smart people pick things up. But when everything is happening at the speed of sound, and then what?
Make one change to something in order to be able to to be able to make a causal claim. And so what Lean Startup did is it forced everybody to think about, wait a second, what's my product? My product is a series of understandings about an attribute and what it causes.
When I slap all the attributes together, and some behavior change happened, I don't know which attribute caused it. If I don't know which attribute caused it, I don't know what I should keep and what I should dump. Actually, I'm not on the phone. And so, and so...
That process I think requires first, you have to formulate the experiment and run it. and then you have to observe the results, and then you have to draw a conclusion from the results, and then you have to use that conclusion as an input to your next experiment. And one danger of the intoxication of speed is that you skip those steps, you skip those steps and you don't learn. And if you don't learn, then you know like, Even when you're producing a product with AI, you still need to learn what consumer response is. A product is a stimulus. A purchase is a response.
So just understand that if you want to benefit the speed of AI and product making, you have to be set up to be able to also create a feedback loop that allows people to learn. So that was the first thing.
So we have dollars and cents. And that's a concrete currency, which is semi-concrete currency. Now we're in the world of-And so, you know, there's a lot of research that shows that people spend much more money when they use credit cards than when they use cash. Why? 'Cause cash is concrete. The credit card is abstract. Right? It feels like I didn't do it. Whereas taking it out of my wallet introduces what's called pain to pain. From pain to pain, nobody likes pain.
You know, obviously, different people in the room have different thresholds. for what that thing is, but eventually all of us, you know, are experiencing it. In comes this world where it's not like a credit card it's like even worse it's even worse than Because, you know, when I spent, you know, a hundred bucks on a sweater, I know what it takes to get a hundred bucks. When I say, hey, You know, I go to Gemini or whatever, or Clog, or I say, hey, which, which, um...
Yes, a budget is a limit. but a limit of something that I don't actually to you. Understand what it means is not really a limit. I mean, it's a limit in an artificial sense, but it's not-why do we have limits? We have limits to regulate our behavior. That's why we have a budget, to regulate our behavior. But then the budget is concrete, and I translate that budget into-I use that to be an input into how I'm going to regulate my behavior.
But if I don't understand what the currency is, then how can I use it as a way to regulate my behavior? So I don't exactly know how that gets solved. And it actually is exacerbated because we're subset. So we're training-by the way, it's very hard when people get used to free.
So I don't even know, like, our MVP's gone. What is it going to mean to have an MVP? In theory, you could tell the AI, make me an MVP. But I think every-I can't imagine that most users won't just want to say, hey, make me the app. And so I don't know how, I don't know if that gets handled. I think ultimately what that's gonna lead to is a lot of just crappy stuff. And it's not crappy because AI is crappy.
It's crappy because Only one in some amount of stimuli changes behavior. And we have developed a methodology to figure out which of that one it is. And now, in some ways, this technology maintenance has scarred that methodology. So now we're back to square one to not understand it. Yeah. Then in the case of developing So cute. ...
It's like having a seamstress, I have a machine that makes sure. It's interesting, one of the steps, like going through the regions. Yeah, because listen, when I need, I think when you know that what you need is you need the spreadsheet filled out, you need these receipts copied into this spreadsheet. Here. There's no uncertainty about that.
When you're making a product and giving it to consumers, there's uncertainty, or they don't want it. It's not a stimulus in the response. It's just a mechanical set action. Mechanical set of actions, go live.
The last thing is, when Manolo said, so he and I, we want to like, We want to think of writing something on what in the product making is going to remain the purview of people versus machines. You know, he's talking about taste. Um... He said, "You still need taste. The taste comes from the person." And I really resonate with that because I kind of believe in it. I don't exactly know what taste is.
And in theory, you know, like the AI maximalist will tell you taste can be modeled. I suspect taste can be modeled on some things, but I don't know what-but I'm sure it can't It can be modeled if it's able to measure things that are relevant to this variable called taste. But there's things in taste that I don't know that we measure. And if we don't measure, then there's not going to be any AI. But to me, that's... That's an open question.