Zack Kass: Should You Be Scared of AI? Former OpenAI Leader Reveals the Truth | Artificial Intelligence | E405

Indeed AT&T Business, Pro-Lon, Cash-Up, and Northwest Registered Agent.

“Shopify is the global commerce platform that helps”

you grow your business. Start your $1 per month trial at Shopify.com/profeting. Quote is an AI-powered phone system that brings your calls, text, and context together in one place.

Try Quote for free.

Plus get 20% off your first six months

when you go to Quote.com/profeting. Indeed helps you attract interview and hire all in one place. Get a $75 sponsored job credit to boost your jobs visibility at indie.com/podcast.

AT&T Business delivers reliable business-grade connectivity that gives your team a competitive advantage. Switch to AT&T Business at business.att.com. Pro-Lon helps you reset your body

and reverse your age with a five-day plant-based program that mimics fasting while still nourishing you. Go to Pro-LonLife.com/profeting for 15% off site-wide plus a $40 bonus gift when you subscribe.

If you've been curious about Bitcoin but haven't made the jump yet, cash-app makes it easy. For a limited time, new customers can get $10 added to their balance.

Just use Code Cash App 10 when you sign up. North West Registered Agent gives you the tools and guidance you need to build a complete business identity.

“Visit North West Registered Agent.com/yap-free”

and start using free resources

to build something amazing.

As always, you can find all of our incredible deals in the show notes or a young and profiting.com/deals. More foyer, more intrigued the dragon keren Zurück.

The absolute mart is the sound of a knife. Your Reich would unbezzwing what's a reneera. Stream the new Staffel, House of the Dragon, up 22nd Juni with wow.

Freightly outside them of Staffel 1 and 2, the following series and weitere highlights. There is incredible negativity towards AI. Housing healthcare and education

have gotten 280% more expensive over the last 30 years. These three things are not technological failures. They're expensive because of policy failures. We are governed by a pernicious anti-social,

geriatric group of parasites. Most people hear AI and immediately think, "What is this going to take for me?" My guests today's act cast wants us

to ask a much better question. What could AI make? Possible. Most people are really excited

“for everyone else's jobs to automate just not their own.”

Everyone wants stuff to arrive a little bit faster, a little bit cheaper on Amazon without realizing it means automating the ports. I heard Elon Musk say something recently

that work one day will be more like a hobby. How do you feel about the future of work? First, the reason the future of work is such a ridiculous discussion right now is because none of the economists understand AI

and none of the technologists understand labor economics. Second problem and no one's really talking about this is. Why are communities pushing back against data centers? There's lots of people who are really upset about

the environmental impact of AI. If you really care about the environment, the most environmentally efficient thing you can do is. Zach, welcome to Young and Profiting Podcast. Thanks for having me.

We are just going to jump right in. So I was reading your book and a big premise of it is that people are scared of AI. And not just because AI is dangerous, but because we're wired to be scared of change

and things that are new. So help us understand why it's our natural instinct to just be afraid of AI. That is a good question that none of the people ask about everything

that they're afraid of actually. We are wired as humans and the disclosure viewers I'm not a behavioral psychologist, but I'm going to tell you what many behavioral psychologists would say. Most of what any human believes is.

They are predisposed to believe. So most of actually what you experience and this is this is a little controversial. Actually comes from genetic predispositions. Your ancestors passed you a bunch of beliefs

and biases that are inherent in your DNA and then you have lived experiences that compound those things. And a lot of people actually from a very young age, their lived experiences either confirm or slightly tweak what it is they are predisposed to believe.

But of the 180 cognitive biases that we have,

there are a few that are really powerful.

One of them is a confirmation bias. So we often see things that we already believed and they reinforce what we believe. So they confirm or reinforce her ideas. One of them is a negativity bias.

And the negativity bias served us a long, long time ago and actually to some extent, can still serve some people today,

The negativity bias tells us to look at something

“as though it can harm you even if the chances are low.”

And the negativity bias is the thing that plagues a lot of the people who are most cynical in your life as it is or most pessimistic. And then we have a bias that we call rosy retrospection,

which is one of these really incredible biases.

We tell ourselves that the past was way better than it actually was. We tell ourselves these really romantic stories of a bygone time with a sense of nostalgia. And we've now created the sepia lens

that our brains view the past through. And I don't know if you're a 90s baby. I was a 90s baby and they, you know, the Instagram is spilled with these like sort of nostalgic images and videos of what it was like

as a 90s baby. And of course, what those videos do not remind us is that being a child was far more dangerous than the 90s than it was today because of kidnapping and other forms of violence.

And also that the 90s themselves were plagued with all sorts of other problems that we don't experience today, but we remember them fondly. And so we do this and it compounds in these interesting ways. Now, that in and of itself makes us

afraid of change. It makes us afraid of sudden movements. And it generally makes us wary of things that feel like we can't, you know, shape them. AI presents a whole other slew of issues.

And we'll get to these. But the principal issues that I call it in the book, I think, are born from a context window of anyone's most recent relationship with technology. And so in places like the United States

and in particular in cities in the United States that have been hauled out economically or spiritually culturally,

there is incredible negativity

bend in view towards AI. And you don't have to really try to figure out why. The internet and the digital economy promised to make everyone's lives better and a lot of lives got better,

but if you live in the rest belt or if you live in a city that was forgotten in a digital economy, you're going to look at this with an enormous amount of skepticism and trepidation.

Moreover, if you're a young person and you were told that the world would continue to get better. And instead, you were fed unmetered pornography, gambling, comparison, violence, outrage, right Jonathan Heitwrights about all this in the English

generation. And you observed that housing healthcare and education have gotten 280% more expensive over the last 30 years. You're going to look at AI and go great. A whole lot more of that garbage.

And so we sit in this incredible moment

where humans have this incredibly proud relationship with technology. For so much of our history, the history of human progress has been told as the history of technology.

“Outside of the good guys winning the important wars,”

everything we celebrate is us discovering something. Basically, you don't hate air conditioning. You don't hate electricity. You don't hate refrigeration. You don't hate antibiotics or planes, trains and automobiles.

You hate your relationship with the screen often. And you hate the fact that technology is not made housing healthcare and education less expensive over the last 30 years. And so many people look at this moment with AI and they go cool.

The detect giants are going to make my lives worse as they did for the last 30 years. And the best comparison and the reason why not everyone hates AI. You go to China. 80% of Chinese people believe that AI is going to be better

for them or make their lives better. And you don't have to do a lot of research to figure out why. They have an incredibly healthy, recent relationship with technology. The average life for the average Chinese person

has improved meaningfully over the last 30 years because of very, very explicit efforts by the CCP to diffuse the value of electricity in the internet to the benefit of the average person. Yeah.

And that explains it. And it actually isn't much more complicated. And so the book basically tries to spell out why this moment is not that unlike other technological moments. So what we need to do and reconcile the fact that the kids have

every right to be pissed because they're staring down the barrel of prohibitive living and addiction to some of the worst vices we've ever invented. But what do you say that's a recent CBS if we're kind of

“comparing how life has transformed over the last 30 years?”

And when we step back, the arc of human progress has actually been exceptional. Well, now you could write the forward to my book. I mean, yeah, totally. And also typically a recent CBS, you and I would,

we could go to a next game and suddenly we're a mixed family and we were in spurs territory. We could go to spurs game and then suddenly we're spurs fans. That's a recent CBS. That's a recent CBS, most people can make sense of.

At some point, you actually have to consider something a bit of a trend. And it's true that we should look at the arc of human history to actually examine, which is the case I make, that I think this will go well.

to be really wary of in this moment. Yeah.

“And especially those who feel like they are,”

and I say this quite freely, wealthier in so many ways than all of their ancestors combined and yet spiritually depraved.

And for whom hope has never felt further in grasp,

despite all the promises that have been made by, by all of the technology companies for the last 30 years. That's not a recent CBS. That's, that's an honest, very reasonable concern and something that we have to acknowledge as we talk about the world

that, that is going to get better. Yeah, Pam, I'll admit it. The first time I heard somebody explain Bitcoin, I felt like I had accidentally walked into a finance bro Ted Talk. Charts everywhere acronyms flying around, one guy saying decentralized,

like it was a personality trait. I nodded along and understood absolutely nothing. And then I just didn't touch it. It didn't invest in Bitcoin for years. But here's what I've learned in business.

Sometimes things are not actually complicated.

“They're just explained in the most complicated way possible.”

If you've been curious about Bitcoin, but haven't made the jump yet, Cashap makes it easy. You can set up automatic purchases with zero fees or by larger amounts also with zero fees.

Start small or go bigger. It's designed to be simple either way. For a limited time, new customers can get ten dollars added to their balance. Just use code Cashap 10 when you sign up and don't forget this part,

send at least five dollars to a friend in the first two weeks. Terms apply. Cashap is a financial services platform, not a bank. Banking services provided by Cashap's bank partner, Bitcoin services provided by Block Incorporated Brandt.

For additional information, see the Bitcoin disclosures at cash.app/legal/podcast. Yeah, fam, I built an app in 15 minutes. And before you ask, I didn't suddenly become a software engineer.

In fact, I've never written a line of code in my life.

But now, trying to now apps like it's my day job. And that's because I learned how to do it through mind stone. It's an AI transformation company that helps close the gap between having access to AI and actually getting value from it. So this all started when I attended their breakthrough AI weekend.

I learned how to build apps. And I also learned about their platform rebel, which basically acts as a second brain. And it actually helps to use AI in a way that completely transforms the way that you do your work.

It changed my life. I left that weekend thinking that I have to roll out mind stone to my entire team. And we did. So we started with the four week AI competency program.

On line, you don't need code. It's made for non technical professionals. And it's really affordable. And so I sent 60 people on my team to take this training. And with that training, you get access to this platform called rebel,

where you can ingest your email, your slack messages, your fireflies, all your drive resources. And then you can basically use it as a coach, as a tool, as a thought partner every day. Before I hop on a meeting, I ask rebel,

hey, what do I need to know for this meeting? What do I need to bring up? And it will scan slack and even send me things that I didn't realize. But it was going on in my company. You can create skills, which is created a process

that might have taken two or three people and going into different apps. And I can do that on your behalf. It is amazing what you can do with this platform. It is drastically improved our efficiency at yet media so much.

So that we're pausing hiring on a lot of roles. I recommend you start with their four week AI competency program. You can get access and get 10% off at experience.mindstone.com/yap. That's experience.mindstone, M-i-n-d-s-t-o-n-e.com/yap for 10% off their four week AI competency program.

“Yapping, do you want to look and feel your best this summer?”

And I don't just mean looking good on the outside. I mean feeling amazing from the inside all the way down to your cells. Yep, your cells. If you do, I'm going to put you on to a reset that I've been absolutely obsessed with. It's ProLon's five-day fasting mimicking diet.

ProLon is a plant-based nutrition program with soup, snacks, and beverages that keep your body in a fasting state while still nourishing you. It is the only nutrition program clinically proven to trigger a toffogy. Your body's natural cellular cleanup mechanism. And a toffogy can support longevity and it can even reverse

your biological age. For the last couple of years, I've known about it.

But I was always scared to actually do a fast because I work out a lot.

I didn't want to lose muscle, and honestly, not eating for days sounds way too hard. But then with my birthday coming up, I wanted a reset. I wanted to lose like five pounds, I wanted more energy, and to look my best. And that's how I came across ProLon.

And everything comes pre-packaged, label day by day.

And I loved it so much that now I'm planning to do it three times a year.

“With ProLon, I'll be young and profiting forever because with three resets,”

you can actually reverse your age by 2.5 years, which is insane.

That means every year, I'll basically be getting younger.

If you're ready for your own reset for a limited time, ProLon is offering young and profiting listeners 15% off-site wide, plus a $40 bonus gift when you subscribe to their five-day program. Just visit prolonlife.com/propheting. That's P-R-O-L-O-N-L-I-F.com/propheting. Prolonlife.com/propheting.

To claim you're 15% off discount and your bonus gift, prolonlife.com/propheting. So I know that you're history major. You want school for history? Yeah, I guess we still. I guess I still am.

I still haven't read the past. I agree in history and computer science from Berkeley. So you saw AI from a different lens. You saw it from a lens of history. And I was reading your book and one of the things that you say.

Well, thank you for doing that. You didn't have to, but I appreciate you doing that. Well, I feel like it could be a better conversation if I understood. And it's a recent book. You just put it out in January. It's right. So basically, you say that in the late Middle Ages, people thought the world was ending.

And you said there's some similarities to what's going on with the release of AI. Like, people just are looking at it. Like, it's it's doomsday. But out of, you know, this, this doomsday came to Renaissance the first Renaissance. Right. So talk just about what happened in the late Middle Ages. Why do you think it's similar to now?

Unless you think we're out of that phase. So you've read the chapter. The chapter is called Welcome to the Late Middle Ages. And it's kind of tongue and cheek. Like, I don't actually propose that we are in the late Middle Ages. What I propose is that if you had lived during the late Middle Ages,

you would have had every right to think that the world was ending. Half of the human population was decimated by disease and some more during the late Middle Ages. We basically stagnated on a technological progress standpoint. Because we started to observe the average life expectancy decline for a couple hundred years. So people stopped inventing things because they didn't have time to actually invent stuff.

And they had to simply survive. And a lot of the people who were inventing things died while they were inventing them. So progress was halted for a couple generations. And at the end of the late Middle Ages, you have Gutenberg's printing press. So you have this invention that actually then sparks the Renaissance.

Now, I have never met someone in all of my walk of life that tells me with a straight face.

They'd rather live during the late Middle Ages than now. And actually, I've only met a few people that can confidently tell me they'd rather live at another point in history. And I've never met a woman who actually thinks that they'd rather be alive at any other point than today. For all sorts of obvious reasons. Well, you say that.

But a lot of people are willing to tell me very straight face that it's that we're that we're fucked. A lot of people are willing to tell me very seriously, this is it. This is the peak of civilization or actually we peek some time ago. People go, I missed the 90s. Now, let's talk about that.

But my whole argument is if you actually believe that this is a bad period of time. Let me remind you how bad it once was.

And also what came from that that actually hard times do produce a bunch of incredible moments.

And I'll look at all the stuff that's fundamentally broken with you and stare at it.

“And also say, here are all the reasons why I think a lot of these things can get a whole lot better.”

And you couldn't have really done that in the late Middle Ages. You needed a lot of hope. And you know what else? A lot of those people died without it. And I say this to people so much of the world we live in today with, you know, truly unmeted information.

Is he just no too much. And a lot of people have a window to the world that is actually too big. I really think so the cross that especially the young people bear today is that we gave them access to all the horrors, all the atrocities of the world without the brain that can metabolize it. And so of course they graduate college thinking that the world is ending because when we graduated college,

we were like, I mean, I remember exactly what I was thinking about when I graduated college. And it wasn't the world ending. It was trying to find a job and find a girlfriend and party and become a man, which is probably what a male brain at 21 should be thinking about in a developed world. Like we solved a bunch of problems so that I had that opportunity.

Yeah. And young people have lost that. So I remind everyone, if you really want to be really despondent, let me tell you, all the other times in human history, where people were despondent and actually, it hope was very clearly on the horizon.

“Yeah. Well, you're actually an AI optimist. Would you say that?”

I've stopped using that term. I used to use that term because it like told people what they should expect. And now I actually just say, now I just grab myself as something entirely differently, which is I work in the arena.

And I'm at this point probably best described as an AI advisor. But I also having studied history and computer science and having studied this technology very closely, consider myself a very accomplished and credible realist. Yeah. And it just so happens that it's very hard to refute a lot of progress that is currently being made. Well, talk to us about some of the progress, health care progress, education progress.

“Like, what are some of the things that maybe people don't know about yet that is really cool in the AI world?”

Well, I think before we talk about the progress in AI, which we can, I think we should just talk about sort of like, we should set the stage for what we think, what I think actually is most broken and it needs to be fixed. Yeah, because there is a really remarkable thing about the world we live in, which is this TV screen was once $10,000. Now it's a hundred bucks. Yeah.

And this is like a very nice dress that would have been prohibitively expensive for any of your grandparents to own, for example. And there are almost everything in our lives that can come down and cost has plummeted and cost. And then there are sort of three exceptions to that housing health care and education. And not only have they gotten more expensive over the last 30 years, by about 280%. They've gotten worse, right, that the actual outcomes are worse.

And I have to remind everyone that these three things are not technological failures. They're not expensive because they must be. It's not that we haven't figured out how to make them cheap. They're expensive because of policy failures. They're expensive because we've chosen to make them, though.

Yeah, corruption, basically.

And bluntly, and actually plainly, we are governed by a pernicious anti-social, geriatric group of parasites. They are all really excited to watch us fight with each other and blame the technologists.

“Now look, I think that Mark Zuckerberg might be a public enemy at this point.”

Like, I'm actually, I watched the Congressional hearings. I watched him in the head of Instagram, tell American households that being that a child staring at Instagram for nine hours a day doesn't actually constitute addiction. And if it is addiction, it's okay because they're learning. And I, in that moment, decided that both of them could be tried in the Hague, as far as I'm concerned. Like, that's probably, if you profit here when my child stares at a screen, I probably don't want you to do well.

And actually, I probably want you to do poorly. But we should also acknowledge that on the whole, the technologists are trying really hard to move stuff forward. Yeah. And that it continues to be the politicians and the policymakers who have not. And the reason we can observe this is that TV screen was $10,000 now.

It's $100,000 thanks to the technologists. Yeah. And thanks to the technologists, you can drive across Los Angeles for $15 in an autonomous vehicle. Thanks to the politicians, a trip five blocks away in an ambulance might actually bankrupt you. That's crazy.

The kids have every right to be pissed. I just think they're pissed with the wrong people. In California, in particular, we, you know, our, our governor is a hero. I don't know why among, among so many people, like, I can't figure it out. I don't know why any of them are heroes.

Like, I don't know why any politician, except those who are talking about cost of living and the ability for technology to deflate that cost are heroes. Because everything else is a lie.

Your national policymakers are not your heroes, they will never be.

And only the ones who recognize the simplest and best way to make your life less expensive is by inventing tools to do so. Everyone else is lying to you. Yeah. So with that table set, what I, what I basically acknowledge is that there's only so much technology

hand you, right? We've created a world where actually technology has made so much dirt cheap. You need gas station sushi. Like, it's kind of wild. I mean, truly a delicacy.

Like, you can now go to a gas station and get half decent sushi for like six bucks.

“Why the hell do you have to work till you're 80 to afford your parents house?”

Hmm. That's housing policy. It's not because the building materials are too expensive. It's not because we haven't figured out robotics. In fairness, we haven't really done a lot with the robotics, but we would have done a lot more.

We would have done what China has done by now, which is how to figure out how to build a lot of infrastructure if we were building more infrastructure.

We built 18 million new homes in 1980s.

We built 17 million new homes in the 1990s. We built six million in the 2000s. Hmm. Nimbi policy. Nancy Pelosi and Chuck Schumer took power and they said,

Let's keep it just the way it is. Let's pass a bunch of policy to our exclusive benefit. And then let's blame everyone else. And they've shot us all in the foot. And they're holding a smoking gun being like, who did this?

Yeah. So all that being said, there are huge problems in the society that almost no amount of technology can fix.

And some of those exceptions include drug discoveries.

“So I tell everyone that AI could truly save itself.”

If it can figure out how to cure cancer and neurodegenerative disease, because even the policy makers can't stop that. And the conspiracy theorists will pile on and they'll say, they'll stop it. They can't. If a drug manufacturer actually figures out the cure for these diseases,

they will, they will sell them. They will, they will ship them. And you can't stand in the way that politically speaking. Outside of that, it's actually pretty hard to think of a technology that will change the current course of our cost of living. The given the political restrictions.

And I say this because we actually could observe a world where AI continues to improve at a breakneck speed, where the frontier as it has continues. And you can, you can show it chart continues to improve logarithmicly. And it won't matter. And I'm actually about to publish a paper called the diminishing model returns theory,

which says at some point the frontier will not matter to anyone. The next model, the next great model. Because we were too busy digesting the last model or the one before it. And a lot of that digestion is going to be political. A lot of that digestion is going to be,

it's not going to be enterprise updating. It's actually going to be the policy maker saying you can automate this. I wrote another paper that's very unpopular called illegal automation,

which documents the 1.5 million jobs in the United States that you cannot legally automate.

There are 7 million in Europe, by the way. Now, when we talk about illegal automation, people go, it's the policy makers protecting workers. Yeah. Totally fair. Except they're protecting jobs to a collective detriment.

Like everyone gets their cut until you turn around and you go into second, we've passed so much anti-social policy that you were glasses or context. If you've ever wondered why your context is so expensive, and you've ever wondered by going and getting a new eye prescription is so expensive, it's because the optimal just won it that way.

And they have passed incredible amounts of policy. It's a cartel that has passed a ton of policy that makes it really hard to automate a lot of these procedures and therefore drive down the cost. And they did it to protect their jobs and they did it to protect their little unit. And then you look and everyone else is trying to do it.

And there's this collective bargaining that is actually anti-social inherently. So all this is to say, the models are getting a lot better, and they will continue to get better. There is a really upside-down world where in two years, and I think it's quite possible.

We have exceptionally high quality videos of Donald Trump making out with Taylor Swift and still no great drug discoveries. For what it's worth, I don't think this would be the case.

“For what it's worth, I think what's going to happen is we're going to start to diffuse the value of this technology”

to the benefit of the average person. And we're seeing it in education. So in Austin, Texas, McKinsey, Price, Alpha School is building a private education system where kids can do two hours of very high quality,

basically one-on-one tutoring with the machine

and then they spend the other six hours a day outside. And it's having exceptional benefit. It's too expensive for most people to do, but she's proving a model. And that model, as long as the teacher's unions allow it,

will diffuse to a public school within two years, is my bet. And we will see a massive rebound in our current degradation in literacy and motor function. You know, the average gen Z is less likely to read, write a bike and swim than a millennial. Wow.

I did not know that. I didn't know the urge. And so then you look at, you're like, yeah, we did you dirty. Like, we gave you a phone and we said, go hang on in your room. You don't need to do anything.

We took you out of school during COVID and we said, figure it out. And they did nothing in many cases. So there is a huge bounce back coming. And some of it is going to be that the technology itself gets a lot better.

And some of it is going to be that the policy makers die. And we elect new ones. The other thing that's going to happen is we're going to discover that AI isn't actually an app. So many people stare at AI and they're like, oh, it's an app. No, it's it's cognitive infrastructure.

And the sooner that AI becomes invisible, the sooner that people are going to realize it's benefits.

“Because today you have to stare at this thing.”

And it actually just automates a little bit. But in many cases, it has this insatiable appetite for more of your attention. I mean, the really messed up thing about CHCBT and Claude is you actually have to stare at it to use it. What we want is a more agentic world where all of the activity happens in the background. And that also is moving in a pretty interesting direction where the infrastructure.

Underneath will actually start to automate because we're powering it like we did with with electricity. What's up, young and profitors? When you start a business, nobody warns you that you're about to become the creator, the marketer, the finance team, the customer support team. And the person who's chat GB team, why checkout is not working.

That's exactly why I always say start with Shopify.

Shopify has been a real business partner for me as I've grown yapp Academy and launched products like my LinkedIn secrets masterclass.

Whether you're a household name or just getting started, Shopify has got you covered.

“Because let's be real, your business should not require 17 tabs, five logins and a minor emotional breakdown just to sell a product.”

Shopify lets you build your store, market your products, manage payments, track analytics, and handle shipping returns and inventory all in one place. Shopify does this all so you don't have to do it alone. Start your business today with the industry's best business partner Shopify and start hearing. Sign up for your $1 per month trial today at Shopify.com/Proffeting.

Go to Shopify.com/Proffeting. That Shopify.com/Proffeting. Hold your money back, take your time with your business partner. Our application for your podcast is fresh and delicious from Aldi.

Always good. Always good. Always good for you.

Cut and say fresh for Aldi. For Aldi price. This week, Mini-Watter Million. The kilo for only €1.29. Or nectarines. The 1 kilo sale for only €1.80.

And today there are many other Angebotes in Diner Aldi, not Filialle. And Viter Gates, Einfach Laoschen and Genießen. Aldi. Good to see you all. So you're talking about how different regions of the world have different policies around AI.

So I'd love to understand from you, like, what is the EU doing differently than the US or China and like, who do you think has it right?

“Why am I confident we're going to do a lot in drug discovery?”

We discovered at first antibiotic in 60 years very recently because of AI.

We split HIV out of DNA very recently because of AI. Exactly one year ago last week, the first infant ever in the United States received a custom gene therapy, a single shot that cured him of a previously deadly disease. And thanks to CRISPR and AI. So we're seeing now a rapid improvement in the speed and the rate at which we are discovering

drugs and other things. It's not limited to bile and life sciences either. There are basically three world models right now for how to govern AI. Europe has taken a regulate first, figured out later. Agenda and the EUA act basically says it's very punitive policy and very aggressive policy

that makes it pretty hard to build or leverage AI safely as far as the government is concerned. That's having a very nasty economic consequence. So they're more like, like, tight with it. There's more restrictions around AI and what you can build. Far more.

It's actually, so you hit the nail and head. It's actually what you can build and what you can deploy. But it's having a particularly bad effect on what you can build.

So if you, they basically have decided that the model manufacturer will be liable.

And it sounds smart. Like it sounds nice. So like we're going to hold open AI and anthropic accountable. Then you realize that it is going to keep anyone in open source from building models. It's actually going to, it's going to make sure that open AI and anthropic are the only ones.

Of course, mistural has some, has some French special provisions. But there is no serious lab in Europe. Because everyone is now terrified to build models that, that do something wrong. It seems so silly because they're going to lose out on so much money eventually. Well, Europe also did the same thing with chemical manufacturers.

And by the way, you can, you can make the argument that chemical manufacturers should be held responsible when they create chemicals that harm people. This is of course the Aaron Rockabitch case. Yeah. Actually, you probably want to make the argument that people who are using the chemicals in appropriately

should be held responsible because if you don't, then you actually keep anyone from experimenting with chemicals. And it's not that you want to let people build chemicals that are bad for humans in the world. It's actually that you want to make sure that people feel safe to try to experiment and create new things. Yeah. So anyway, Europe is taking this very anti-development approach and it's having, you know, the expected consequence.

But Europe has been a museum for a while. So I don't think anyone there is surprised and I don't even think Brussels is a surprise.

“I think they've just sort of, they seem resigned to be, you know, the world's vacation destination.”

Asia is broken into two categories. Japan, Korea, Singapore have taken really aggressive, have invested very aggressively. And you've seen in the Korean and Taiwanese stock markets in the last, the doubled in the last six months. And they're rip-roaring because they're trying to develop and infuse or digest.

And they're like, build, will figure it out.

“China is taking a very interesting approach.”

The Chinese approach is to create a very open, build environment and actually to regulate the ingestion. So the CCP has taken very explicit measure to figure out how this technology will actually be infused into society. And which it has the benefit of because it has central planning. But it also has built a ton of infrastructure to accommodate it. So China has, I'm actually going to China in a week.

It'll be my tenth time in China in four years. I'm fascinated by the country. I really like Chinese people. And the CCP has gone to aggressive measures to make sure that the average person sees the benefit of technology. When you say ingest, what do you mean?

They have built 45,000 miles of high speed rail.

They have built three times the amount of electrical output capacities in the United States.

They have built in the last decade. They have built 100% more housing units than the United States. So they're building a bunch of infrastructure. Now they have a population problem. But they're building a bunch of infrastructure.

They're building roadways that are forward compatible with autonomous vehicles. So everything they're doing they're trying to build for a future society. Where people can do a whole lot more, where there's a lot more basically mobility, physical, etc. And probably economic. And they are also mandating the adoption of AI in systems like schools, although they're controlling it in systems like social media.

So hospitals, if you run a hospital in China, you're basically required to make that hospital better faster cheaper year over year. And if you don't, you get fired. Let's get out. It's awesome.

It would be pretty amazing if we did something similar with fire.

Every hospital administrator on Earth probably. And this is coming from the son of two doctors and the grandson of two as well. She did this thing 10 years ago.

“And if you have Chinese friends that they'll remember this, he made private tutors illegal.”

Because he said, all the wealthy parents are getting tutors. I mean, all the poor kids are getting a massive disadvantage. Because we know empirically the most helpful thing in a child's development is one on one teaching and tutoring. And the degree to which I allow wealthy people to outpace for people is going to actually compromise the fabric of society. And so you said private tutoring is illegal.

And he created this massive gray market that's actually wild in China. So like you literally like have to like sneak around and get tutors. He is like this incredible, the private tutoring is really expensive and they are very hush-hush. And they operate with a degree of risk and danger because you can be fined a ton of money if you're found out to be hiring or being one. But now she has realized suddenly everyone can have a private tutor.

And actually what they're now doing is diffusing the power of AI into one on one education that allows every child in China to have this benefit. All while making sure that doying and buy-do and their and buy-dance and their other social medias have no violence pornography gambling. Now, Chinese have a very strict understanding. I mean, I want to be very careful before I spouse how China treats their people and also the freedoms of a Chinese person. Yeah.

So let me be very clear. I love living in the United States. I really want to be American. But we are social media looks very different than theirs. And that is what Central Planning looks like with regard to ingestion.

Right, so they've said to the model manufacturers build open source models. We're okay not building the frontier. So they're specifically saying we're not going to build the best model. We're going to build the next generation. We're going to make it dirt cheap.

And we, the CCP are going to help guide how it's diffused into infrastructure so that the average Chinese person sees it as a net benefit. Yeah. This reminds me of something in your book that you were talking about. It's a very useful framework, which is technological threshold versus society, societal threshold. Yeah.

Tell us about that. Extra points to you. Thanks for, okay, this is one of my favorite frameworks. I wish, so more people, it didn't, it didn't catch on quite like I thought it would. Sticky IP is hard.

It is what it is.

“But a technological threshold asked the question, what kind of machine do?”

It's all that. A societal threshold asked the question, what do we want a machine to do? There's a very different things that turns out. And the best example and the one I drawn in the book, assuming that a lot of listeners won't read, so I can spoil it, is the autonomous vehicle. Now, so much is said about the autonomous vehicle.

Yeah. People hate him. He's not amazing though. So by the way, I'm sure you also read the movie. I hate him personally.

Do you?

Tell me more. Well, we have way modes in Austin and they're just really dumb. Like they're just really dumb.

Like they're always about to hit people.

“And I know that you might say that's like a bias that I have that I think that the machine is more dumb.”

Maybe. I'm not. I'm not judging you. I hope you don't see anything. No, no, no, no.

But yeah, I don't know. I just don't, I wouldn't trust it. I don't trust it. Do you trust your Uber drivers? No, not really.

I hate Uber drivers too. But I guess I trust them more because if I, when I have a choice, I don't go for a way. Well, I go for Uber. You obviously grab driving, taking taxis. Did you trust your taxi drivers?

Yeah. Okay. Let's come back to this. There's a lot here. Let me come back to this because to explain societal thresholds.

Let me quickly tell a story because I think it's worthwhile. Okay. In 1854, Elijah Otis introduces at the world's fair, a people mover. And the people mover is a prior to the people mover. We moved men, people up and down mine shafts and tall buildings using ropes and police systems.

And they were very dangerous because of the rope or pulley broke.

Everyone plummeted to their death basically.

And Elijah goes, you know, we should, we could build a system that, the, the, that catches the rope or pulley. Such that there's a fail safe. And actually then create more of a mechanized means by which we move people up and down. And it worked empirically. It worked really well.

And almost one invention fundamentally changed the odds of someone who was working in a mine shaft or tall building. And a bunch of tall building owners start buying these people movers. And by 1855, a bunch of these tall building owners start calling up Otis Elevator. Now, aptly named. One of the two great elevator companies in the world.

Shit Shinler Electric being the other. And called Otis Elevator and said, hey, no one's writing these people movers. Can we return them? These building owners say, can we return them? Of course, they, Otis goes, no, you cannot return these elevators.

But he's got a problem now because he solved the technological threshold. He solved the empirical problem of moving people up and down the building safely. But he hasn't solved the societal threshold. So he goes back to the drawing board. He gathers this executive team together.

“Because how do we convince people they should write in these elevators?”

Because if we don't, there no one's going to buy him anymore. It doesn't matter if they work. And they come up with three ways they think they can convince people to ride in the elevator. And people thought it wasn't safe. That's why they weren't writing.

That's exactly right. The feedback was, thank you for calling that out. The feedback was, no one thinks they're safe. And actually, more importantly, people are terrified. They didn't even understand them.

They're these mechanical contraptions that with steel cages on them, that are actually just terrifying in many ways. So no one wants to get in them. They go back to the drawing board and they come up with three ways. They think they can convince people to ride in the elevator.

The first, they put a mirror in the back of every elevator.

So when you walk in, you'd be distracted by your own image. It's not. Very smart. Then they put music in the elevator. So then when you walk in, you'd be a little entertained.

The third is to put an operator in every elevator. Such that when you walk in, you're given the sense of human agency. Someone to control a lever, even though that person was not at all necessary. And those three things work. They move the societal threshold.

Everyone starts riding elevators. They don't make the elevators safer. But everyone feels safer and more welcome. And the rest of the history. Now, I tell the story and people chuckle.

How could you not? Right?

“I think we all ride elevators if you were to come up with.”

I think about it. And if you're in the top of the entire state building, you've ridden the birch cleave. You've ridden a very tall wall. You can understand why they were scared though. They're going in like a metal waxing really fast.

I mean, I can think of a lot of moments in history where I would not have been the first person to do something that I will inevitably do. I don't need to go to the moon. Right now. Do what you go to the moon now?

I wouldn't either. One day, you and I probably will go to the moon. Realistically. I mean, probably. At the current rate, the technology is progressing and the rate it was stuff to fuses.

I certainly would not have been the first person to fly in a plane. I am blown away by the Wright brothers. Yeah. And frankly, anyone who flew planes early on. Now, I don't think twice.

Yeah. So I'm with you. But it does also afford us the opportunity to make the question what is our elevator. And we have many, as you pointed out. We have so many, it's actually hard to count.

And the reason for that is that we are moving technological thresholds faster than any generation prior. The one though that I like to talk about is the autonomous vehicle. Now, what I am particularly fascinated by relative to the two, all the other things people talk about. Because people talk about autonomous vehicles, add nausea. The debate has gotten so stupid in my opinion.

What is most interesting on the marquee, the most interesting thing, is that the roadways for you and me are probably the most dangerous place on earth.

Neither of us are at meaningful risk of a cardiac event.

“We both live in cities where we are physically very safe.”

But the roads are very dangerous for you and me. And 45,000 people in the United States alone die on the roads every year, of which, by the way, 7,000 are children. And despite that, there is one obvious solution, the autonomous vehicle. And the obvious solution pulls at about 35% approval. 35% of people are excited about autonomous vehicles.

That alone blows my mind. Because it's not that I need someone to want an autonomous vehicle. It's not that I want someone to want taxi drivers to be out of business. It's not that I want someone to not want to see someone driving their car. It's that I am so shocked that people don't want 45,000.

By the way, 1.3 million in the world.

Well, it goes to the point that we forgive ourselves, right, for having accidents and failures. Now you're getting hit. So I wouldn't study this. I became fascinated by this gap. I actually don't care at all how good autonomous vehicles are.

At any given moment, I don't care.

“I'm like, why do people hate these things that are so clearly going to save literally 7,000 children's lives when we, when we roll them out?”

Okay. Reason one, humans love control. We love control. To understand this, I went to Disneyland. Have you ever been to Disneyland?

Yeah. Okay. I went as a kid and I went with researchers and I stared at Autopia, the ride Autopia. It's one of the original 7 rides in Disneyland. It is a cement track on which children age about 3, 3, 8, right around in a go cart. A gas powered go cart at about 2 miles an hour. And they have about this much leeway.

So you don't actually have to turn the wheel and the pedal is Boolean. So it's either on or off and you can only go 2 miles. And it is actually a real embarrassment to Disney. It's a blight. It's a heat this ride.

“It smells bad and actually reeks of fumes.”

The workers complain of headaches so they have to rotate them out. It has very little foliage. It has no Disney branding. And it sounds terrible. It's like a noise problem.

And despite that, the reason that they cannot change this ride and the reason that it is only now going to go into construction. After 70 years in operation is that it has been the most popular ride for children age 3, 3, 3, 8, since its exception. 1954, Utopia has been the most popular ride at Disneyland. Because you do not need to explain to a child how fun it is to step on a pedal and turn a wheel. There is something thrilling about controlling a machine.

BMW sells it to you in every commercial. We don't know how to describe it.

It is so awesome to control something that is way more powerful than us.

Getting your driver's license is equal parts thrilling because of the freedom and thrilling because of the control of this machine. And as a result, you see it everywhere people insist on controlling machines that are bad for them, harming them, killing them, killing others. Machines that they heat using. Because machines that are inefficient because they want to control these machines.

And you see this all over the place. And it is one of these problems that is actually just going to be really tricky as we try to adopt machines. The second reason is most people have no idea how good autonomous vehicles are. Technological thresholds are moving way too quickly when I started doing the study autonomous vehicles were actually pretty bad. Now they are amazing.

Basically no one knows unless you work at one of these companies unless you have been in a waymo. And even then you are like maybe I got lucky. No one is in the idea that waymo's are empirically multiple standard deviation safer than most drivers. And certainly many standard deviation safer than senior drivers who are somehow still a lot to drive. And that's just a huge problem because now more and more the technological thresholds are moving so quickly that no one has any idea what they are actually supposed to adopt.

But the third reason is the one that you call that which is that humans have an exceptional tolerance for human failure and none for machine failure. And I stare at this all day. I stare at I watch 1.3 million people die in the roadways. And I watch waymo's sort of into the wrong lane and there is an international outcry to shut the whole thing down.

I don't not get it. It serves us. It's why this building will never collapse. It's why planes don't fall out of the sky.

But it will have this strange buttress, this strange like artificial limit to what we adopt because we are so willing to watch each other fuck up. And we're so unwilling to watch a machine do it because we assume that it's systematic and foundational failure. And that's just going to present all over the place. It's going to present everywhere all the time. And it's going to be why we move pretty slowly through a lot of systems that could really meaningfully improve lives. Because we are afraid to, you know, for that point zero one percent mechanical failure rate.

Well since we're talking about autonomous cars, I let first to spend some time on where you think the future of AI is going to bring us.

Do you still align to those three stages? And how do you imagine the world to be in 10, 20 years with AI?

“Totally. The stages are enhanced applications, agentic or autonomous agents and natural language operating system.”

Enhanced applications is exactly what it sounds like. It's, you know, apps that we've always used. Now, super charged with AI.

It's funny because we had to go through this. Magical AI. Well, it's an messenger. That is what ChatGbT is. We had to do it in order to bring people up to speed, but it's not actually. We needed the interface. We needed the thing that everyone that comported to the way in which people wanted to work. Yeah. And what were used to already kind of? Totally. And they're like, cool. It's magical. And I've been doing it for 30 years, which is why we get away with marginal productivity gains, because despite all the promises of AI.

We'll go in. I can still, I still log in to email. I still do these things, but it's scratching the surface of where we're going. Second phase is agentic AI. Agentic AI, by the way, has entered its teen sex phase. Everyone is talking about it. It is not clear who is doing it. And it turns out very few, although some people are doing it a lot, parents of teenagers may not like that one, but it's true. And the thing about agentic AI is, loss of stuff is going to change. And two things in particular are going to change a lot that I don't think people are talking about.

So the autonomous agent phase, which I think begins in earnest later this year, and then probably last the next decade, maybe more, maybe less, is going to force us to rewrite the fabric of the internet. When we built the internet, we originally built it as a library.

“So a bunch of nerds built the internet. And so rightfully, they were like, a library is the optimal use thing, because that's what they were thinking about.”

And so for a period of time, the smartest people on earth put their best ideas onto this into the space. And it proliferated all the universities and lots of corners of the world, were smart people gathered, turned the internet into this really great repository of information. And then one day a marketer probably was walking by and stared at the nerds huddled around their machines and was like, the optimal state of that internet thing is not a library. It's a shopping mall. And they built the browserable HTML internet.

And it's the first corruption of the internet, the second being social media.

And it's a corruption. It's a fausty in bargain that we would make 10 times out of 10. We would always, always trade.

The world's best practices for trillions of billboards in exchange for GDP, e-commerce. I mean, you can buy anything from anyone anywhere, but it totally ruined browserable internet. It's actually totally ruined the nature of the internet. And you don't have to, I don't need to prove this. You can know this every time you go to a website. Yeah, or in Google or wherever.

Anywhere.

“Obviously, present companies included and everyone in all of your listeners, your websites are great, but most websites are shit.”

And the reason I can, I know this is because I log into my banks website once a week to get my account statement sends my account. And as soon as I log in, I'm served an ad for a credit card that I've owned for 13 years. It's not because they're evil, and you're going to make the arguments, whatever. It's not because they're bad designers, and it's not because they're stupid. It's because everyone has to fiduciary responsibility to sell everyone everything all the time.

As soon as you're with an eye shot of someone on the internet, you've got to sell them that thing. So what I realized in all this is as soon as an agent can reliably get my account statement and send it to my account.

And I am never going to go to my bank's website again.

Ever. It's not because I don't like my bank. I actually do like them. It's because there's nothing forming at that site. I don't want to be there.

I don't want to be on the internet for the most part. I want my accountant to have my account statement. And if I don't have to do anything in it, great. Yeah. And then you realize most people actually don't want to be on most of the websites they go to.

Most of the internet is not a place you want to be. It serves some practical purpose. You have to do a thing. And when agents can reliably achieve most of what you want in the internet. You're going to stop browsing to most web pages.

And we are going to slowly and then very suddenly rewrite the fabric of the internet. And the optimal state of the internet is not in the HTML browser. But internet meant for our eyes. It's a TXT XML internet meant for our agents. And I can basically prove technically that the optimal state of the internet is a bunch of text files.

And that's going to be pretty amazing. That's so interesting. So no longer will like they won't be able to sell to us because it will be agents who are on getting the advertisements and they don't need anything. Well, now you've, okay, hold that thought. Because now you've introduced the next one of the consequences of a TXT XML internet.

There are many consequences. There are tons of first and second order consequences. Some of them are really good. One of the ones that's neutral and certainly strange is that we're going to discover pretty quickly how many preferences people actually have.

Because everyone's like, oh, it's an economic problem.

I was like, fuck no, it's not. If you actually automate everything to the hilt, the world gets way wealthier. You cannot convince me otherwise. You can't convince an economist otherwise. Worst thing that could happen actually is that we stop automating things.

And you live in Cuba or North Korea. It just places without technology. But there's a limit to automation that actually stops serving us. And that I introduced by asking a question. I started asking people at dinner tables.

“If you could automate everything in a life where would you stop?”

I wouldn't want to automate exercise. I wouldn't want to automate cooking. Or, you know, things I enjoy like that. I wouldn't want to automate any sort of experience that I enjoy doing. That's a good characteristic in one that a lot of people arrive at.

On my list, there are a few things you mentioned that I would automate. But we'll come back to that.

When I ask people this question, I realized that it actually goes to this amazing question of values.

And frankly, what it means to be human. And the list, by the way, is amazing. People reflexively, the most common thing reflexively people say is cooking. Which is really incredible in me because I would automate cooking. Yeah, I know that I'm thinking about it.

It's like maybe I would like sometimes to be able to automate it. By the way, you get to change your mind. I'm not the genium. It's not the last time you'll get to do this.

“But the next thing that people talk about is exercise.”

And a lot of people talk about runs on the beach. That's one of these weird things that comes up. People talk about time in their kids. They're kids in the bed reading their kids stories.

And it won't automate these things.

When I was asking people about that time, that that's that stroller. That AI stroller was out or the AI rocker was out. That weird thing that rock kids to. Yeah, I was like, yeah. And to be able to like, I want to do that job.

Now, you say that. And then I'm like, what if it's at three a.m. Anyway, what you realize by the way, one of the answers that people are willing to get to. Pretty quickly is very crude. People, there's a lot of things that a lot of people agree.

We just should not amate. And one of them is sex. And what's really fascinating is you realize like there's something very simple about the question that goes to what it means to just be human and to love life. And there's a lot of things that people would not amate. It costs nothing.

Which is also really exciting to think about. A place is, you know, it's time with friends and family and places of worship. It's a very, very common answer, which fills me with hope. I, to understand this better actually try to go look at like the atomic unit of automation, which I decided was people grocery shopping. The easiest way that I could explore automation and the most atomic unit is by watching people shop for groceries.

And I live in Santa Barbara. And there's a Bristol farms in Santa Barbara. And if you're in a Bristol farm. No. It's a very fancy grocery store.

Okay. I really like it. But it's, if you've been to it, it's silly. And I watched the 20 people in Santa Barbara walk up and down, Bristol, Bristol farms, Isles, and consider bunch of groceries.

I didn't tell them anything and when we got out. I asked each of these people about three groceries that almost all of them had considered or purchased. Water, toothpaste, tortilla chips. And I would say then we got outside and say, "Why do you buy that water?" Or that toothpaste and tortilla chips. Why do you buy that? And they say, "Well, it's my favorite."

80% of the people's my favorite. Say, "How many others have you tried?" Let's say on just that one. In Santa Barbara's Bristol farms, there are 50 varieties of tortilla chips. Over 60 varieties of toothpaste and over 85 varieties of water.

Oh my God.

“How could you possibly know your favorite of any of those things?”

Then you look at all the things. We purchased and you realize that you and I, and actually most, the average person in the developed world, or in the Western world, makes 125 purchased decisions a day, which is 124 more than our great grandparents. Capitalism is awesome.

It brought us abundance, abundance brought us optionality, and optionality is properly breaking our brains. And you realize and talking to people, as you start to examine the atomic purchase decisions, the simplest things in your life that even the wealthiest people often have to decide to purchase. Most people make purchase decisions not out of brand loyalty.

They make them out of preservation for computational load. People cannot possibly consider all the things they buy, so they do it out of inertia or randomness. And when you explore this, you realize that when the average person, or even especially the wealthy person, is offered an opportunity to let someone else or something else fill their grocery cart.

And they're going to say yes in large part because there are so many things to choose from, and so many things to buy. And when someone says or an agent says, "Do you want me to plan you a dinner date?" It might be like, "Yeah, I don't want to investigate every new restaurant in Austin,

four open last week, and they're all supposed to be amazing."

And by the way, I choose the restaurant based on menu items anyways. You know my taste, pick the menu item. What about a vacation? Awesome. I would love that.

Make it sunny, direct slides only. And you realize that the automation boundary, we pick these big things, but actually along the way, there's a bunch that we're going to have to seriously consider. And rewriting the fabric of the internet makes it particularly easy for us to automate way more than we realize today.

“And one of the challenges I think of the agente AI moment and the internet, and the next 10 years,”

is figuring out what we have to preserve for our own humanity and dignity, and what we have to preserve for a collective humanity and dignity. What it actually means to be alive. There is a world in which we live and simply subsist because we have automated so much, and you don't have to look that far to see some of it today.

We stopped average parent in the developed world, became very uncomfortable with their own child's discomfort, and automated most of that discomfort away. And so the conclusion that I draw is actually in discovering what I call virtuous friction. So much of the automation boundary goes to this idea of what work serves us and what work does not.

Virtuous versus vicious friction, walking to miles for clean water is clearly vicious friction. No one should have to do that. Drilling a well with your parent in your backyard to see that water can come out of the ground is virtuous. Emotional abuse of any kind, but especially of a child in high school is terrible. Being broken up with as a high schooler, pretty good experience.

Good to know what it feels like so that when you get into the real world, you can feel a sense of rejection and you're observing a generation right now,

Gen Z that have never been broken up with because they don't date.

And so the feeling of rejection is terrifying. All of this is going to be presented slowly and then suddenly. And the agentic internet presents one more weird outcome, which you kind of touched on, which is, there is a, I worry about these dystopian outcomes that I don't think are talked about enough because they're more insidious. Like everyone's like, oh, is there going to kill us all? And I don't think so.

And there's no real scientific basis to that argument. But like what about a world in which one company makes all of our purchase decisions? And people like, yeah, that's crazy. And then it's like, yeah, Google kind of controlled most of the internet for the last 25 years. And they did it in a black box way.

They didn't have to tell us how the algorithm worked. They didn't have to tell us how they were serving ads. And if Google suddenly decided tomorrow, they said, you know what? Leave it to us. We're going to buy everything for you.

We're going to, we're going to schedule you. We're going to buy your clothes. We're going to get you styled. A lot of people might be like, that's awesome. Yeah, they might do it.

And we should acknowledge that there's a insidious dystopia that doesn't feel dystopian because it's uncanny. It's uncannily similar to the world we live in, except that we are now controlled by an even more narrow apparatus. By the way, there's a solution to this. Don't lose hope. There's a solution, which is an open internet.

And people are working really hard for this.

“An open internet protocol that you should advocate for, where the internet itself operates like it does.”

It's an open internet. And everyone brings their own agent. So it's agents are open sourced. And everyone brings their own agent. And that agent is beholden to you and operates on an open internet protocol.

You can tell your agent what you wanted to do. And at the end of the day, it retains your data. It doesn't sell at the highest bidder and it makes decisions exclusively on your to your benefit. I want to kind of like back up a bit. I don't think everybody really understands what an agent is.

I was talking to my friend. I was going on a walk. And the teen sex era. Yeah. Exactly.

Like you do because you're an expert, but people listening in. They probably don't quite understand what an agent is. I was taking a walk with my friend who started an AI company and helping people build. AI agents. And he explained it really in a really easy way to understand.

He basically said an agent can go from app to app to app.

And you don't need a human. To facilitate it whereas like, you know, you can use a chat to P.T. That's not obviously that's not an agent. But an agent can go from all different apps and like accomplish what a human can do. How would you explain it?

It is a machine that can execute tasks and goals across browser data lake. An application environment. It is quite simply in other words, a Siri that works.

“That is the best way I know how to describe an agent to anyone.”

By the way, I made that joke all up in Chicago three months ago at a talk.

In the back of the room sits Adam Chier, the founder of Siri.

He's so fast. He wasn't happy. Now, half the room thought it was really funny. And the other half is half didn't. And he hasn't responded to any of my apology emails.

In my defense, he should have built the technology that worked. And so now we can talk about Siri as a false dawn. But it's exactly that. It's a thing that can navigate systems the way that you and I navigate systems, except far more seamlessly.

And the major breakthrough, and if you're like, well, Siri sucked. Why is this different complex reasoning? So we're as Siri could only go forward. And it couldn't really turn left or right. And it actually couldn't turn around at a dead end.

So if it reached a place, even if it was wrong, it was like cool. We're going to do this now. Yeah.

So it's like always, you know, it always opened your garage door.

And it always called your answer something. But the modern agents are capable of complex reasoning. So they actually can go side to side. They can explore this app and say that's not right. They can explore this app.

Say that's not right. And they can move much more freely. And that allows them to do really complex things on an open internet or inside of environments that you give them access to, like your email or other applications.

Would you say that agents are going to be always like on our screen? Or like, how do you imagine AI evolving with agents? Well, so the third phase that you alluded to, or that I guess I wrote about the book that you just alluded to, is we call the natural language operating system.

So I think what I think happens, right now we're in this enhanced application phase. So we're still really nascent in AI. And if you use AI, when people say to you use AI, they literally assume that everyone has to.

“In order to use AI, you have to go into your device.”

Now, of course, most of us are beneficiaries of AI in very, sort of, discrete ways or like non-active ways, where the grid that is powering this building runs on AI. Like there's a bunch of stuff that happens auto pilot in planes, right? But that doesn't actually change the fact that for most people,

they're sort of understanding the AI and the benefit to them. Happens when they use it. They have to be using the app. They have to press buttons. Yeah. So it's just another thing that demands our attention.

Agents will help people, especially as they work better and start to become more self-evident. They will help people use AI without actually, or benefit, I should say, benefit from AI without actually using it, without actually interfacing with it. So what comes next is a term that's used is ambient compute. And ambient compute is this, or ambient AI is this idea of machines that,

well, one, we're going to go from carrying our devices, laptops and cell phones to wearing them. Watches, rings, glasses, wristwatches, et cetera. While it's pens, people are like, oh, that sounds dystopian.

“How could it get more dystopian than staring at your screen?”

Yeah. Like, it cannot. If you told me that my daughter in when she grows up in 15 years is going to have to stare at a screen like this in order to use a computer. I would tell you that we have regrets, like something horrible has happened. A meteor is hit the earth or something.

We cannot possibly want this world.

We were never supposed to live this close to screens and captured by them.

They have an unsatiable appetite for our attention, they are demons, they are destroying our lives. We are spiritually devoid because of these things. We are further from God, not closer to because of screens. So machines that can relieve us of the attention starvation that screens are producing. Good.

But also, they are going to be embedded with telemetry and audio and video sensors that actually start sort of passively working on our behalf. Which also relieves us of the need to constantly press buttons. And that is also to our collective benefit because so much of the value that we will get from machines will start to become invisible. Which is exactly how the internet could and should feel to exactly how electricity feels. And it turns out all great technology at the limit moves into the background.

Because the purpose of living is actually this, without anything else. Like the purpose of living is time and friends and family in physical community places of worship and outside dining room tables. That's it.

“That actually I can feel very confident saying if we live for another 50,000 years will continue to be the most important thing that humans do.”

Because it has been for the last 200,000 years. And that brings us closer to God and closest to each other. Machines that work passively do another thing which is that they dissolve the digital divide. So many of our friends today don't feel like they use AI because they don't feel like they know how. And these are the same people who in 20 years would feel like they had totally missed out on economy and society.

And you see this with lots of old people. There are older people in our lives that can't use face time.

They barely can use their phones so they show up at the airport four hours early to get their, you know, physical boarding passes.

I mean, all these things where the digital divide is pronounced because machines became so complicated to use. Should in fact become obsolete. The ultimate achievement of AI will end up being moving to the background so that the value is actually distributed fairly evenly. Yeah. And this is the natural language operating system that I described.

“And I think it arrives at some point in the next 10 to 15 years where so much of the,”

the the stuff that we organized today actually becomes. Yeah.

And is this what you call unmetered intelligence where basically just becomes like electricity.

Like we don't even notice that it's there. Unmeted intelligence is a, is it the array proposed in 2021, which proposes that at some point intelligence would behave like a resource. Water, electricity, internet, food, stuff in the developed world. It would go from very scarce to very abundant.

“And that arrived sometime between now and natural language operating system.”

Where you could argue we're sort of there like you and I can do a whole lot more than we've ever been able to. What's interesting about unmetered intelligence is it doesn't propose universal brilliance. I'm not proposing everyone to be smart. I'm proposing that everyone will have access to brilliance in the same way that literacy doesn't mean you read and the same way the internet doesn't mean you do research in the same way that a gym membership doesn't mean you're fit. Unmeted intelligence doesn't mean you'll be brilliant.

It means that you and we and collectively and individually will have access to it and what we do with it will define everything. Okay, where I want to go next is talking about the human experience relationships and how you think that's going to evolve with AI. It sounds like you have a very positive outlook considering that you believe that screens are going to become less common or needed to get our daily actions done.

“How else do you feel like it's going to change the way that we interact with each other?”

The screens have already become less necessary to living life. They've just become more necessary or they've become more dictating in life, but so much of my hope actually comes from watching. So this is this is a little strange, but watching Gen Z is really hard. I mean honestly, like it's actually like heartbreaking to watch an entire population that has so much opportunity and feels so broken spiritually emotionally.

We should never have given kids smartphones. We should most people shouldn't have social media. I believe I'm terrified of it.

The smartphone is a portal and unmitigated portal into pornography, depravity, dereliction, violence, comparison, the list is so long. And we 16-year-olds don't have the ability to metabolize that. And they also don't have the ability to know what they should spend their time on and not. You know, left to their own devices, most 16-year-olds would get terrible tattoos and a lot of them would get pregnant. And that's okay, your kid, you're supposed to have your parents there and a lot of parents gave their kids phones and said figure it out.

You never drop your kid off of the tattoo parlour and say figure it out. We were so protective of kids in the physical world. And we were so unabashed in the digital one. And we're paying a serious consequence. Many standard deviation, higher rates of depression and anxiety and reclusion and isolation. And especially in the developed world places where kids have plenty, have it all and feel like they have nothing. And I, my heart breaks. I mean, there's nothing else to say. And like I, I want to fix it and the, and the cool thing, the promising thing I should say is that they've raised their hand and they said, we're not happy.

They don't want to live this way. And Jen Alpha has watched them and said there's no fucking way we want to look like this. Millennial parents are raising Jen Alpha very different than Jen X raised Jen Z and Jen Alpha has basically gotten up and said, I don't want to be this and you're watching social media addiction decline. It's subtle now and it will be pronounced soon. You're watching people return outside consumer confidence keeps waning. Concert tickets for Benson Boone keep getting more expensive people want to go outside like people don't like drinking so bars are struggling, but parks are exploding museum attendance all time highs in a lot of towns local theater all time highs.

Broadway roaring back physical world is on fire right now in a really good way and it doesn't actually take a genius to figure out why even though consumers are like I'm worried about my savings even though they're they're they're they're taking on more credit card debt than they ought to. They're also saying I want to be with other people I want to be doing real things and Jen Alpha in particular like shopping physically. So all these mall experiences I mean like rigorous built the growth in LA and all and you did like these really cool mall experiences they've become like cultural touchstones in ways that we never expected because Jen Alpha's like I want to be with my parents physically touching all this stuff.

Thousands of years to figure out that women shouldn't smoke while they were smoking drink while they were pregnant.

They've got thousands of years okay all of our basically everyone in our in our lineage smoke and drink while they were pregnant that we had access to tobacco nicotine and alcohol.

Except for our moms like truly except for once who were like people like they don't do this it took us 20 years to figure out that device addiction was really bad. That's super promising.

“Look at this and say well the poison proliferated really quickly it did because of the internet and the pollution is bad it is but also you have to find a lot of hope and promise and the fact that the next generation.”

Watching this is like no no no no I don't want that I want my soul and spirit. They are willing to admit we sacrifice the soul and spirit of Jen Z at the altar of a digital display and they're like don't do that to me. And so what I propose now is we're just graduating problems it's not that there won't have experienced their own of course they will then every generation inevitably does but the fact that. Each generation can now observe and and counter correct is pretty impressive. Yeah and so all of this boils down to something that I believe since I was a child I mean I was born with a predisposition to loving time and humans I could sit here with you for a long time.

Which is that the most wonderful thing that we can do on this earth is spend time with friends and family and physical community places of worship and the dining room table ideally outside. That's it.

“It's why we're here it brings us closer to God it brings us closer to each other it makes us happy it explains why when you go to Costa Rica people are so happy despite having absolutely nothing.”

Or buton it is a very simple heuristic that makes sense to the world it explains why the wealthiest people are often not the happiest and in many cases often sometimes the most miserable because they are the most devoid of this time. And if you use this heuristic to explore the purpose of technology then eventually you will either believe that technology is going to destroy us or you realize that we we have and will continue to figure out that the purpose of technology is to free us to do this. Which I believe yeah and so I'm very optimistic for our ability to figure out time and time again to reconnect to what I call our happiness function to reconnect to this idea that the purpose of technology is to free us all the awful vicious friction.

So that we can spend time doing the virtuous. Well a lot of us like to spend time working a lot of people who are listening to the show are very successful entrepreneurs who really identify with their work and. Love to work in general and I heard Elon Musk say something recently where he was saying that work one day will be more like a hobby will be like gardening you can choose to work you could choose not to work. How do you feel about the future of work one I don't like to make a habit of disagreeing with Elon it's not it's not it's not a good investment strategy I also tend to agree with a lot of what he says I disagree with this point.

But I don't disagree with it for the reason he is making the statement so first the reason the future of work is such a ridiculous discussion right now is because. None of the economists understand AI and none of the technologists understand labor cannot. So like technologists are like. And of course the labor economists go AI wasn't good when I wrote this paper six months ago and so this is and of course then all the technology have to go buddy you are very out of date. That's the first problem the second problem.

“And no one's really talking about this but it's important to talk about is political protection.”

Legal automation protects 1.5 million jobs in the US 7 million in Europe. And is the reason that it is very hard to actually accurately predict which job is going to automate.

Wait so this illegal automation you brought it up before I will I've interviewed every AI expert and nobody has ever brought this up I've never heard of this phrase before.

With respect. Technologists a high experts are narrowly often narrowly brilliant people. It is why so many conversations that relate to AI are so devoid of any material nuance. Because technologists were never actually in a great position to apply on for example the meaning of life. And especially the AI safety folks but most technologists don't live lives that you are I find appealing.

They are obsessed with work in ways that I don't think is very becoming and they are often spiritually devoid bluntly.

And I always say I was joke if someone has bad personal hygiene I don't give a shit what they think about the future humanity.

Elegal automation is bluntly and very simply political protection. It explains why your contact lenses are expensive.

“It explains why if you ever hosted an event you have to set up a certain way and you have to hire a certain group of people.”

We can't carry the sand on it or someone has to carry this in for you literally unions have made it illegal to automate the you know dropping off things inside of a warehouse because they have to be the ones who take it in there it's protection for their work. And the list of this stuff is long. It's also why the asymmetry of stuff that can automate is really wonky.

You look for example you're like wait a second. Why is it that that you can go into a doctor's office and you can you know do a bunch of stuff on an iPad and you can register into all these things.

But as soon as you get into the OR there are nine people in there. Oh, the hospital has ensured that there have to be all those people in there in large parts of that they can build those people.

“Yeah, and the life is filled with this stuff when people are like software engineers are going to automate first I go yeah that actually might be true, but it's not because cloud code is great.”

Cloud code is good because we're allowed to do it's new jobs because you don't have a union. It's all the intellectual jobs like lawyers well lawyers are going to make damn sure that they cannot be able to be kidding they write the policy. Yeah, like I tell everyone there's no way the law gets automated but I feel like it is already. You may get to do work that doesn't require a lawyer. I can guarantee you if the lawyers feel threatened existentially you are going to see an exceptional amount of policy get written by lawyers to protect themselves the reason.

The software engineers are a decent bed and I don't even think they're it's right for reasons I can explain is that software engineers don't have a labor union they never organized because they literally we invented software engineering and they got rich overnight and no one ever bothered to be like we should unionize we should have it we should have national associations. So now is the time I guess to start to do the same. I've seen some grumblings but the other thing I'll say is even if they do unionize you'd have to be a very brave politician to drag Jenny and Scott to Microsoft engineers who make 500 thousand dollars a year down congressional steps and tell the American people that these jobs must be protected.

No one is going to have sympathy.

“Truck drivers have a lot of sympathy congress will certainly say we cannot let these truck drivers lose their jobs these are very important jobs that feed families fine.”

Need I remind you how many people they kill and by the way I'm not I don't want to sit here and moralize every job I just want to point out that.

Political protection is the reason that there's so much asymmetry in jobs that do exist in jobs that don't when technology can can solve them the third reason is new work.

You and I this moment would not have made sense to someone 50 years ago yeah let alone maybe even 30 years ago and most jobs that exist in modern economies were very recently invented and we have no humility for that we have no backward humility we celebrate constantly. We don't do work that our ancestors did without actually acknowledging that inevitably our kids will do work that we don't do because we are so sure that these machines are going to do everything and I can bet that's not true. And the fourth reason is quite simply that the future of work matters a whole lot less than the future of living I actually am kind of exhausted talking about the future work because we live in a moment where even if you design the perfect job even if you design all the perfect jobs everyone in the economies can have a job you're still looking at a world where the average college grad needs to work until they're 80 just to afford their parents house.

But difference does it make if we keep talking about the future of work then we plan the politicians hands because then they can protect work and they can say we saved you we did it. Shut the fuck up just make shit cheap housing healthcare and education we can do it technologically allow us to do it. Any discussion by a policy maker about anything other than allowing technology to diffuse to the benefit of the average person is annoying and a distraction.

You failed. Yeah, right. By the way, the last reason is identity displacement and I talk about this in the book and we don't have to talk about this about it right now.

I am sure that job displacement is going to have a profound economic impact good one like I'm willing to bet that and that'll sound callous until I remind you that we are all descendants of people whose jobs were automated to our collective economic benefit and we never think twice about them the reason that everything is so cheap is that we automated most of our ancestors out of work. And you wonder now the you realize that most people wonder the earth average asking when is this good or service going to be better faster cheaper without realizing what they're asking is when is a human going to be extricated from the manufacturing of this good or service.

“Which we don't do because we're jerks we do because we are rational economic actors condition to believe the world should get better faster cheaper all the time.”

And it occurs to one most people are really excited for everyone else's jobs to automate just not their own.

Everyone wants stuff to to arrive a little bit faster a little bit cheaper on Amazon without realizing it means automating the ports.

And the long Sherman who run the ports want their taxes to be free not realizing that it means or maybe realizing that it means automating the accounts. And the accountants really and the list goes on. Yeah. The hardest part in all this I don't think is going to be managing the economic spoils. It's going to be figuring out how to manage the emotional cost of our work changing so frequently and so much because where I do think Elon is right that work can become voluntary.

“I don't think it simply becomes a means of pleasure I think it will have to be I think it will have to serve as a means of purpose.”

And until we have a spiritual awakening until we reconnect with God which I think is coming I really think people are going to need work to find out why they're here. And two of the 6 most common last names in the United States are Miller and Smith because we named ourselves after our jobs for thousands of years. So as work changes frequently and a lot even if the the economic spoils are amazing and even if we continue to toil for money. A lot of people are going to be like I don't know what I am anymore I had this career path I studied this thing I had this credential I had you know.

First question every adult asked another adult what do you do hmm it's a mark of status if we if that if that ladder fractures it's going to be really weird.

Yeah and that to me is like there's no other way to say it we need to we need to acknowledge that that this is going to be really tough and that it is emotional crisis we are facing a spiritual crisis. Last thing on this. I'm not the first person to say this not only am I not the first person John Miner keen said this in 1930 he was traveling in the world. The father of modern macro economics traveling the world talking about how everything was getting better and people are like John you're a little aloof.

I mean look people are literally dying in the streets of starvation what was the time period 1930. 1929 he's traveling Europe giving this lecture series and people are like your tone deaf bud and then he goes home and he doubles down he writes his paper the economic possibilities for our grandchildren in which he writes and I quote. I must now disembarrass myself to take flight into a future that I will not live to see one in which humans may have solved the economic problem and be faced with something more profound.

The father of modern macro economics proposing that at some point we would close care for feed house everyone and then humans would face our greatest struggle.

“To anyone who will listen I really believe that we have been in the midst of a spiritual battle for a while.”

We bought all the stuff we thought we would ever want and then some and we're not happier. We are actually less happy in the developed world than we have been in quite some time and it is plain to see at this point that we need to reconnect with what it actually means to be human and I think on the other side of this whenever whatever this is. That is a reckoning for humans with with our actual purpose that we have distracted ourselves from for quite some time now on that note by the way there's thing called an adaptability trap.

Everyone now right now thinks they have to be adaptable and so I did this fun study it was a fast one I asked about a hundred people to describe their least favorite friend. That I knew what I was looking for and so I got some confirmation bias but it actually turned out. So think about your least favorite friend in your mind and think about a couple so. Not your enemy, but someone you love who you just don't like very much.

Very unreliable or exceptionally hard to pin down on any idea. And for many people their least favorite friend about 70% of the population fits some category of a sick of fans or an infinitely adaptable person. They say all the things that they want someone to think they believe or they think someone wants them to say they are desperate for approval they are desperate for like ability. And this moment serves those people really well you know this. I can think of this unfortunately I can think of the same person when I go to this place.

And they love AI because it allows them to like be lots of stuff and they're constantly talking about all the things they're doing and the new businesses they're starting are the crazy idea that they have. These are the people who post on LinkedIn. The most in trait sounding AI slop garbage that gets a bunch of likes because people like oh that sounds smart and then they get a bunch of approval from it.

“You love this person, but you hate them and you hate the fact that they don't stand for anything and there is never been I think a more interesting time to stand for something that is unpopular.”

And in particular especially among the younger population that are so eager to do group think so eager to tell each other how much they hate a group that's unpopular how much they love a thing that's popular right everything's performative. There is so much value now in believing in a set of mission vision and values and talking about them especially when it doesn't serve you obviously in a public sphere.

Because your ability to achieve those things has never been better so your ability to prove a group wrong has actually never been greater.

And it's never been a better time to be contrary and it's never been a better time to have a big idea that doesn't sound good. And so I tell everyone adaptability is actually this weird trap. It's this infinite adaptability trap that's super fucking annoying and I'm like no no no stand for something stand for something especially if it's you know in popular and then. Be adaptable to how you get there be adaptable to the tools you use be adaptable to the technology you adopt be adaptable to the to the time that it might take.

“But pick a castle on the hill describe a world that you want to live in and then work really hard to achieve it. I totally agree I feel like focus is so important and being an expert I think in this world where you can learn anything like.”

High level right actually knowing and going deep because you'll know more than the AI and it is deeply rewarding it turns out. Yeah, like it's mastery we do not teach kids to sense a mastery in high school we mile wide inch deep because the American high school experience fill your resume look awesome. Accomplish is you know effectively nothing and unless by the way you're an elite athlete or an elite musician in high school.

You never taste mastery. So we graduate all these people that have never really actually worked hard on a very specific thing.

And we put them in a world where they can be general experts. I don't like it's never been a better time to get really good at a thing. Yep, every really really successful person right now is just good at one thing. And they usually got really good at something before that they usually like figured out like wind surfing or an instrument or a sport something taught them how to taste mastery so that they had some you know pattern matching muscle memory for what it feels like to work really hard at something.

Yeah, I want to end this conversation with headlines. I felt like this would be a really great way to just like bring up any sort of ideas anything new that you want to talk about.

And basically these are like big trends happening in AI. These are real headlines. Okay.

All right. So the first headline is Challenger says AI isn't a job apocalypse yet, but companies are citing it the most when announcing layoffs. So basically they're saying that companies are using AI as an excuse for layoffs and it's not really the AI automating jobs.

“As of time of recording, the job numbers I think come out on Monday next week for last month.”

Really estimate suggests we will have added the American economy a hundred thousand jobs way over our forecast at the outset of the month. And a lot of people are starting to realize that maybe AI is not going to eliminate jobs the way we thought and actually. It may start adding jobs pretty soon as it has in AI and other fields. Especially in the US, I think. Actually, because like I find myself. Well, just like this might be, I don't know if you've ever thought about this, but so I've got a team of 60 and I have a lot of international employees.

And a lot of my employees are in Nigeria, the Philippines, India. And now I'm like, when I'm thinking about new hires, I'm like, well, I want a US person because before it would have been advantageous to kind of hire internationally, but now a US person can do the work of five people and it makes more sense to hire a US person.

I generally think most labor numbers are going to look good except for in economies really uniquely exposed to like customer service.

“But even that, it seems like customer service numbers are still actually net growing.”

Will Gidara who ran 11 Madison when he onboarded open table to 11 Madison was the first three Michelin star restaurant to move reservations online. He took all of his reservation is and said I want you to call every single person and learn about them once they make the reservation. Figure out why they're coming and then we're going to do something special when they come to the restaurant. AIwashing is an inevitable outcome, you know, Jack Dorsey spooked everyone when he announced that block was going to fire, I don't know, 5,000 or 10,000 of its employees.

And because of AI and I kind of want to remind people that the smartest thing that Jack Dorsey could do if he thinks his company has over hired is to blame a technology that he could not have predicted well as the reason that he over hired. Because traditionally layoffs were met with, you know, an inquisition. The board would call the CEO and say why did you hire 5 to 10,000 too many people. Now Jack Dorsey can say well, this thing happened, who could have predicted how good AI would be we have to redesign the company around this.

When actually you might be willing to make the argument that block should never have grown to 30,000 employees,

which did during COVID at a time where hiring was considered a mark of success. And literally companies were being rewarded in the public markets for adding head count because head count was relatively cheap. We'd closed offices and everyone was sewing going so long on software, you just wanted to show heft. And so he did the rational economic thing during COVID, he hired a bunch of people. Now he's going to do the rational economic thing and say a hired too many people, but he's not going to say that he did it because of a failure of planning.

That would be beneath him. He's going to say he did it because AI has now, he's laying these people off. They've done good work, but AI has compelled him to do this. Now, I obviously don't work at block. I haven't looked at the filings.

I can't make an accurate assessment. It seems pretty clear that Mark did the same thing at meta. They hired way too many people. He's probably exhausted by the number of people. It's well known that Mark is pretty exhausted by all of the people inside of meta.

The culture that he built and probably is relish at this moment to say AI changed the game. When in fact, it's probably just mismanagement. So interesting. Okay, next one. AI generated work-slop is destroying productivity.

So essentially, a lot of articles are talking about how AI can save us time, but it's actually just giving more work to the person you're delivering that asset to. That a lot of managers are receiving AI work-slop. Someone made this argument the other day that it is actually immoral or immoral to send someone AI-generated work because you are potentially making them review. It cost you nothing to create a bunch of work for someone else. Like, the argument which I liked is that it is equivalent to smashing a plate in someone's home and walking out.

You've created a mess that they have to grab or decorating someone's home.

And for a party they never asked for.

And I tend to agree. I agree.

“I also have a slightly weird take on this, which is, I think most workers bullshit.”

I think most emails are bullshit. I think most things that people write are shit. And like, I don't love my book. I mean, I like my book. I'm super glad I wrote it. I'm proud of the fact that it exists. I wrote it to find conviction and clarity in a bunch of ideas.

And I wrote it to serve as counter narrative to prevailing. I mean, I wrote it as a reasonable argument to make that constructive responsibility is a reasonable alternative to ambient doom. And there are a bunch of chapters. And like, I wish I had written this better. The amount of shit that people have been producing for since long before AI is actually quite ridiculous.

And AI is simply shining a light on the fact that people are very comfortable producing garbage work.

“Like, I think the only difference now is that the cost actually producing the work has gone as zero.”

Not that people are inclined to produce it. Yeah. There's a sense of aesthetic that some people have. There is a sense of propriety that some people have when they produce work. And there's a sense that some people are devoid of.

And AI is simply shining a light on the people who were slop cannons before.

It's so true. It's, uh, I love this idea that you had in your book about the intellectual K curve. Yeah. Sure that was really good. Yeah.

I mean, basically, are you, is it like, if you were going to do really well before AI, you're probably going to do really well now.

And if you weren't going to do anything before AI, you're probably going to do even less now. And my whole argument is like, actually, we should celebrate this. I, like, this is people are going to get maybe a little angry here, but I'm like.

“We've created a world of economic abundance where if you want to do absolutely nothing, you can actually do it.”

Like consider the fact that we crawled out of caves once where if you did nothing, you were like left for dead by the village or the village died because you needed to do something. Where for most of human history people had kids because the kids needed to work. And then we graduated to a place where kids didn't have to work anymore. And for a little period of time, basically when you were in high school and our parents. People did jobs to learn a sense of responsibility and we blew straight through that chapter.

And parents said to their kids in the developed world, you don't need to work anymore. And a lot of kids did nothing. Everyone credits the current cognitive decline of Gen Z with AI and I'm like, no, it predates to 2012 when we gave kids phones. It becomes pronounced in 2015 when the average high school in the developed world stopped having a summer job. And then it pronounces again in 2020 when we take kids out of the physical school and make classrooms online.

And I'm like, look, this is an economic condition, not a technological one. We gave kids the choice to do anything and they a lot of them did nothing. And what you observe also is a world where kids who want to do a lot or doing a lot more than ever. Gen Z while it appears on average less smart than Millennials has in your standard deviation higher occurrence of savant. So we're watching Gen Z come out of the woodwork around the world with arts, sciences, culture, sport, overperformance against expected average like we couldn't have imagined.

And the third fastest, two mile in the world is a high schooler in the United States, really United States, you're trying to self to run on YouTube. Gen from Collier, this multi-gramming award-winning artist taught himself to play like every instrument on YouTube. And more chess masters under the age of 20 now than above the age of 50. Like you're watching or under the age of 25 than above the age of 50.

“You're watching a world where if you want to do nothing, you can do nothing.”

You can order a door to ask, you can do scroll, you can subsist, and you can blame others that you are doing this, but it is your choice. It is an economic outcome that actually I think is worth celebrating. We've created a world where there's so much freedom, you can do nothing and get away with it. And also if you want to do a bunch, you can do it. And this cake curve is pronounced not by how well to your parents are, not by where you go to college, not even by where you grow up.

But by how badly you want it, whatever it is.

I think that's kind of amazing.

Yeah, it is. I also think we should quickly redesign schools and childhood education. Like I think one of the detriment in the developed world and in Asia for that matter, in particular is that we've created. We've made school a place where you graduate gates of knowledge. You have to know a thing and then you can get to the next grade.

When in fact, the purpose of childhood development and childhood education is probably not knowing about your facts. It's gaining the ability to explore more ideas on your own. And this idea of designing a world that inspires agency, designing a world that inspires personal responsibility, designing a world and shot a development that inspires morality so that we are graduating people who see the world as a place of beauty and wonder, who want to make the world a better place and a really eager to figure out what they are capable of.

Instead of insisting has a society that kids know a set of facts, which are obviously going to become commoditized anyways.

“Okay, last headline for you, why are communities pushing back against data centers?”

So this is about the environment. There's lots of people who are really upset about the environmental impact of AI. This is, I don't know if I, I guess they're going to be listeners just stuck around or like that kind of ideas are interesting. Then they're going to be listeners stuck around like, I hate this guy. One awful thing is you're going to say next, here you go.

This is this is that if you hated me already, you're going to hate me even more now. The most popular question I speak when I travel now, I speak a lot. I'm the most popular question I get asked by a girl in high school. The boys asked weird questions. The girls are all really smart.

The question is particularly interesting if it gets asked in California because in California for whatever reason, and they're probably in the US, but particularly in California, there's a obsession with water consumption. You probably heard this. Yeah. So the question sometimes will be framed with what about AI's consumption of water.

Now, California, my retort is really simple. What do you think is California's number one source of water consumption?

What do you think the answer is by the way?

Showering? No. Cows. Cows number two, beef is two. Almonds.

Almonds are one. I mean, they are so far out in the front.

I love California. I'm raised in living California. It is a very corrupt state, and it's more corrupt than people realize. And actually, if they realized how corrupt it is, they would get even more flat. It is a deeply corrupt state, and the corruption is actually embedded in the state's water rights.

“To understand the history of California corruption, you need to understand the history of our water rights,”

which are owned by a few families who got tax subsidies passed for almond farmers, so that they can sell water to these farmers at a premium. And the farmers can make money on their almonds, right?

And it's basically to figure out how to get their highest per dollar gallon returns.

It's fucked up. Now, we don't talk about it for all sorts of reasons. In part, we don't talk about it. I actually think, because it's just not convenient. People love almonds.

They're like, ah, tell me something else. I keep data centers. Let's talk about that. People love almonds, and they're like eating beef, right? The vegans will talk about beef, but they don't even like to talk about almonds.

They're like, ah, I don't know where this is. I don't know where this is. I don't know where this is. I don't know where this is. And because the corruption is so embedded, because it's so old.

You know, there are people like, why don't I know this? Well, people make documentaries. People talk about it. I talk about it. But it serves Gavin that it's not in the headlines.

And it serves the politicians.

“And so we, we live our lives with about, I think it's like 67% of our water going to almonds.”

It's a ridiculous amount of our well water going to all night. It's easy. And it's fucked up.

Now, the third, by the way, on that list, is golf courses.

Mmm. Golf courses. And this is where I say that, I say the gross, if you do that golf. And of course, they're sheepish. And I don't like putting them on spot, but I'm like, look, I say these things.

Because I want to remind you that there are lots of things in your life that you do that are very energy inefficient. And you do them anyways, because they're convenient. And your dad, you don't just want to hear you talk about hating golf. He wants to golf. He wants to have a nice day.

And frankly, don't bother about it. I guess. But like the reason that you're telling me about is because you want to moralize something. You want to perform to your peers. It's convenient, I say, to talk about AI and water consumption.

Because it sounds good to the people around you. You look moral. You found a position that people can celebrate. People, by the way, sometimes they'll ask these questions. And they'll sit down and their teammates will go into applause.

Or their classmates will go into applause. Yeah. And they don't even look at me for the answer. Half these girls will sometimes look at their friends like I got them. And then I have to wait for them to look at me.

I'm like, I want to answer this question. But you've got to look at me. I don't want to let you just perform. It gets better. So then I say, OK, fine.

“So we agree, where do you think data centers ring some water consumption?”

Data centers ring some California number 11. Which AI consumes about 10%. Now, I don't like what about it. I'm not going to sit here and say, we should consume the amount of water we do. We should be far more efficient with water.

But I want to remind you that there's a bunch of stuff on that list that we could talk about that would be really helpful. That aren't going to bring as much benefit as AI law.

Ah, well, let me get to that in a second.

So I go, but then there's the energy problem, which you point out. Because the environment also, you know, there are other cost besides water. They're one of the missing energy carbon emissions and land, frankly. So then I say, do you know how much, and by the way, I'll ask you this, do you know how much energy data centers in the United States take up?

Maybe 20%, 10%, 3%. Man, it's so lapsided on the internet. You'd think that it's data centers is number one. Because people want their plenty of cases to make against AI. I mean, I will I will rattle them off.

They're making gambling more addictive. They're making pornography more addictive. They're stealing the souls and spirit of people. I mean, like, the work that people at the work that mothers have done in school districts to fight meta to create safe places for kids is God's work. I don't know any other way to say it.

They're they're we are rescuing children from crippling addiction. And we should talk about that all the time. We should talk about the rapid secularization of the United States. We found God in the worst place on the screen. And that is we should talk about that.

We can talk about autonomous warfare. I mean, that's that is definitely a discussion worth having. The environment is where I'm like, man, there is a ton of work we could do. I wouldn't start with data centers. I mean, I would start with subsidies for corn.

I'd go to subsidies for meat. And and and and by the way, and you said this earlier. We are building right now in data centers. It's true. We're building infrastructure to power videos of Donald Trump and Taylor Swift making out. I agree.

We are also building the future of medical discovery infrastructure and education infrastructure. I mean, data centers are going to power a lot of the stuff that people that get people access

They're also creating a ton of jobs. And whereas farming, no longer creates jobs. And farming cannot cure any diseases.

And it's not going to give us a major breakthrough, AI can.

And the last thing I'll say is this. And this is super important. If we're serious about the environment. And I think most people have asked this question and you say this should or not. I do not think they're serious about the environment.

And I realize that the present there may be present company excluded. Like there might be someone right now who was like I care a ton about the environment. And by the way, the most this is fucked up. If you really care about the environment, the most environmentally efficient thing you can do is not have kids sit inside all day and just be like just be quiet.

“And the next most efficient thing you can do if you have to do something is actually just scroll your device.”

As soon as you go outside or if you live and as soon as you start driving places, flying places, you are way more carbon inefficient than if you were just sitting inside on your phone. I hate that argument. I hate the person who says that. I fucking hate that.

Don't listen to me. But just acknowledge that like living is carbon inefficient. Yeah. And like that's just the function of it. What really bothers me is there are incredibly interesting solutions that we can be discussing.

Like building way more livable walkable communities. If we actually built town centers with protected bike lanes and sidewalks. And dense housing, which would solve home affordability, people would stop driving so much. And the same person who is like, we can't build any more data centers is like, oh and also we can't build any more homes. Housing density will ruin the fabric of my neighborhood.

Fuck you. Yeah. Actually, fuck you. Because you have more-alized one behavior and then you made it impossible for us to solve it. In any other means, except that which serves your worldview and this very, very specific way of living.

No data centers and also no housing. And dense urban centers or dense town centers, six stories the way Paris has designed their cities are super environmental efficient. If people can walk and ride their bike to places of worship, to where they work, to where they eat, to where they play. The world will, and the United States especially will consume way, way less energy. And we will bring our carbon emissions down overnight.

And this is my final crusade. We talked earlier about the national politicians. They are no one's heroes. They cannot be our heroes, sorry, but your local politicians can be. Your local politicians can actually be the gateway to resurrecting and rejuvenating the United States from the inside out.

Inside out, we can build dense city centers. We can rebuild, we can build parks where we build parking lots. We can build diners where we build drive-throughs. We can actually move highways. We can rebuild schools, we can rebuild town centers. And we can actually rejuvenate the United States from the inside out in a very economically efficient way and a very environmentally efficient way.

Building spaces that exist in Europe where people can walk to everything. Yeah. That is actually how we will meaningfully reduce our carbon footprints very quickly. And it's possible. Thanks to very effective local policy.

“And it's proving possible, but you have to care a lot about your local politicians.”

You have to hold them accountable. You have to elect people that are capable. And you have to hold them really accountable and you have to unelect them when they don't. But protected bike lanes inside walks are very easy to build. They do not require enormous bills.

They can be done in a year. And they can totally rewire fabric of communities. I love that. Have you ever thought about getting into politics? No.

I am really excited about this work. I think I can make. I think I'm also very eager. I will say I have thought about it.

And I'm one I would never put my wife and family through it.

The vitriol alone would be pretty tough. I don't have the stomach for it. But also I kind of want to prove that you don't need to be in politics to make a change. Like I actually don't want to.

“I think I want to be able to try to inspire people a lot.”

You are inspired without people for sure. Being a policy maker. I think it's way more effective. This has been such an awesome conversation. So I end my show with two questions I ask all of my guests.

You can be faster than what is one actionable thing. Our young and profiteers can do today to become more profitable tomorrow. The only piece of advice I consistently give any young person in young is relative that I speak to. Is fixate on problems you can solve.

I love my wife for so many reasons she is a truly remarkable, amazing woman.

One of the reasons that I adore her so much is she does not concern herself with problems that she cannot fix. She doesn't watch the news, she doesn't watch TV, she doesn't really know what's going on in the world. And I used to, it very early in my life, I would have seen it as a mark of someone who was a loof or worse dumb. Now I see it as a mark of an elevated person who cares so much about the community that she can care for, that she can fix. And not at all about the stuff that she cannot.

It makes her, it's she has high agency.

And it's taught her that she can make a huge impact.

If you are young, you want to learn how to make a change, go make a change. Go fix a thing that you can actually fix. Don't fixate on stuff that you cannot. Don't perform for the purposes of your friends and family. Go solve a problem in your backyard and you'll learn really quickly the value of agency and the fact that you can make a huge impact.

“And what would you say your secret you're profiting in life is?”

I have made a lot of hard decisions in service of my time with friends and family. And one of the reasons I live in Santa Barbara no longer work at Open AI is optimizing for time with my parents when I didn't know how much I would have left. I tell everyone, "I have a lot of creds in AI. I'm a serious AI thinker that whose ideas are taken seriously by business leaders. But I think the coolest and best cred that I have now is I'm a dad. And I can confidently say that I would give up all of my money to build a better future for my daughter.

And that has made me really successful. Like that I am certain that belief.

It has allowed me to make decisions that I think others wouldn't have made and basically build a life that feels richer than I just could not have imagined living this life when I was a kid.

That's beautiful. Well, Zach, thank you so much for this conversation. It was so inspiring, entertaining, educational. Thank you so much for your time and joining us on young and profiting podcasts. Thanks for having me.

Yeah, fam, this conversation with Zach really gave us a bigger way to think about AI because it's easy to look at AI and immediately go to fear.

“Is it coming from my job? Is it going to replace creativity? Is it going to make human less human?”

But Zach reminded us that fear is often how humans react to change, especially when a new technology feels bigger than our ability to control it.

And yet when you zoom out, so much of human progress has come from moments exactly like this. The printing press, electricity, the internet, and now AI. Every major shift has forced us to ask not just what a machine can do, but what we actually want that machine to do. And that question matters so much for us entrepreneurs, because the goal is not to automate everything just because we can. Zach's idea of virtuous friction really stuck with me.

Some friction drains us, slows us down, and keeps people stuck. That's the kind of friction we should absolutely use technology to remove, but some friction shapes us. It builds resilience, creativity, relationship, and purpose. That's the kind of friction that we need to protect. That's the human experience that we need to protect. AI agents may soon handle our emails, book our trips, compare our options, and move through the internet for us.

That may sound convenient, and in many ways it will be, but it also means we have to get very clear on our values. What decisions do we want hand off? What work still gives us meaning? What parts of business, leadership, and life should stay deeply human? And maybe that's the real message of this episode. AI is not just about the future of work. It's about the future of living. It can help us build faster, learn faster, and solve problems that once felt impossible.

But if we use it well, it should not pull us deeper into screens. It should free us to spend more time doing the things that actually make us feel alive, like building, creating, connecting, serving, and being present with the people that we love. So do not sit on the sidelines of the shift. Learn the tools, experiment with them, bring them into your business.

“But as you do, remember that the goal is not to become more machine-like.”

The goal is to become more fully unapologetically human. And if this episode made you think differently about AI, share it with somebody who's still scared of it. Maybe this conversation helps to move from fear to curiosity. Thanks for listening to Young Improveding Podcast. Every time you follow, subscribe, share the show, or leave us a five-star review,

you'll help us reach more people who are trying to learn and grow and build something meaningful. You can watch this episode on YouTube or Spotify video and come say hi to me on Instagram at Gapathala or LinkedIn by searching Hala Taha. Until next time, this is your host, Hala Taha aka the podcast princess signing off.