AI AGENTS EMERGENCY DEBATE: These Jobs Won't Exist In 24 Months! Containment Has Failed, We Must Prepare For What's Coming!

[0] I think a lot of people don't realize how massive the positive impact AI is going to have on their life.

[1] Well, I would argue that the idea that this AI disruption doesn't lead us to human catastrophe is...

[2] For example, people are going to be unemployed in huge numbers.

[3] You agree with that, don't you?

[4] Yes.

[5] If your job is as routine as it comes, it's gone in the next couple of years.

[6] But it's going to create new opportunities for wealth creation.

[7] Let me put it to you this way.

[8] We've created a new species and nobody on earth can predict what's going to happen.

[9] We are joined by three leading voices to debate the most disruptive shift in human history, the rise of AI.

[10] And they're answering the questions you're most cared about.

[11] This technology is going to get so much more powerful.

[12] And yes, we're going to go through a period of disruption.

[13] But at the other end, we're going to create a fair world.

[14] It's enabling people to run their businesses, make a lot of money.

[15] And you can solve meaningful problems, such as the breakthroughs in global health care and education will be phenomenal.

[16] And you can live an incredibly fulfilling existence.

[17] Well, I would just say on that front, this has always been a fantasy of technologists.

[18] To do marvelous things with our spare time, but we end up doom scrolling.

[19] Loneliness epidemic.

[20] Right, falling birth rates.

[21] So the potential for good years.

[22] infinite and the potential for bad is ten times.

[23] For example, there's more undetectable beat fakes and scams so people don't understand how many different ways they are going to be robbed.

[24] Look, I don't think blaming technology for all of it is the right thing.

[25] All these issues, they're already here.

[26] They're all fathers here.

[27] So what are you saying to your children?

[28] Well, first of all...

[29] I find it incredibly fascinating that when we look at the back end of Spotify and Apple and our audio channels, the majority of people that watch this podcast haven't yet hit the follow button or the subscribe button, wherever you're listening to this.

[30] I would like to make a deal with you.

[31] If you could do me a huge favour and hit that subscribe button, I will work tirelessly from now until forever to make the show better and better and better and better.

[32] I can't tell you how much it helps when you hit that subscribe button.

[33] The show gets bigger, which means we can expand the production, bring in all the guests you want to see and continue.

[34] to doing this thing we love if you could do me that small favor and hit the follow button wherever you're listening to this that would mean the world to me that is the only favor i will ever ask you thank you so much for your time the reason why i wanted to have this conversation with all of you is because the subject matter of ai but more specifically ai agents has occupied my free time for several weeks in a row And actually, Amjad, when I started using Replit, for me, it was a paradigm shift.

[35] There was two paradigm shifts in a row that happened about a week apart.

[36] ChatGPT released that image generation model where you could create any image.

[37] It was incredibly detailed with text and all those things.

[38] That was a huge paradigm shift.

[39] And then in the same week, I finally gave in to try and figure out what this term AI agent was that I was hearing all over the internet.

[40] I heard Vibe coding.

[41] I heard AI agent.

[42] I was like, I will give it a shot.

[43] And when I used Replit, 20 minutes in to using Replit, my mind was blown.

[44] And I think that night I stayed up till 3 or 4 a .m. in the morning.

[45] For anyone that doesn't know, Replit is a piece of software that allows you to create software and pretty much any software you want.

[46] So someone like me with absolutely no coding skills was able to build a website, build in Stripe, take payment, integrate AI into my website, add Google login to the front of my website.

[47] and do it within minutes.

[48] I then got the piece of software that I had built with no coding skills, sent it to my friends, and one of my friends put his credit card in and paid.

[49] Amazing.

[50] So I just launched a SaaS company with no coding skills.

[51] To demonstrate an AI agent in a very simple way, I used an online AI agent called Operator to order us all some water from a CVS around the corner.

[52] The AI agent did everything, end to end, and people will be watching it on the screen.

[53] It put my credit card details in, it picked the water for me, it gave the person a tip, it put some delivery notes in.

[54] At some point, a guy is going to walk in.

[55] He has not interacted with a human.

[56] He's interacted with my AI agent.

[57] And I just, the reason I use this as an example is, again, it was a paradigm shift moment for me when I heard about agents.

[58] about a month ago, and I went on and I ordered a bottle of electrolytes.

[59] And when my doorbell rang, I freaked out.

[60] I freaked out.

[61] But Amjad, who are you?

[62] And what are you doing?

[63] So I started programming at a very young age.

[64] You know, I built my first business when I was a teenager.

[65] I used to go to internet cafes and program there.

[66] And I realized that they don't have software to manage the business.

[67] I was like, oh, why didn't you create accounts?

[68] Why don't I have a server?

[69] It took me two years to build that piece of software.

[70] And that sort of embedded in my mind, this idea that, hey, there's a lot of people in the world with really amazing ideas, especially in the context where they live in, that allows them to build businesses.

[71] However, the main source of friction between an idea and software, or call it an idea, and wealth creation is infrastructure is physical infrastructure is meaning a computer in front of you it is an internet connection it is the set of tools and skills that you need to build that if we make it so that anyone who has ideas who wants to solve problems will be able to do it.

[72] I mean, imagine the kind of world that we could live in where no one can be, anyone who has merit, anyone who can think clearly, anyone who can generate a lot of ideas can generate wealth.

[73] I mean, that's an amazing world to live in, right?

[74] Anywhere in the world.

[75] So with Replit, the company that I started in 2016, The idea was like, okay, coding is difficult.

[76] How do we solve coding?

[77] And we built every part of the process, the hosting, the code editor.

[78] The only missing thing was the AI agent.

[79] And so over the past two years, we've been working on this AI agent that you can just, since ChatGPT, this revolution with Gen AI, and you can just speak your ideas into existence.

[80] I mean, this starts sounding religious.

[81] This is like the gods, the myths that humans have created.

[82] They used to imagine a world where you can be everywhere and anywhere at once.

[83] That's sort of the internet.

[84] And you can also speak your ideas into existence.

[85] And it's still early.

[86] I think ReplitAgent is a fantastic tool.

[87] And I think this technology is going to get so much more powerful.

[88] Specifically?

[89] What is an AI agent?

[90] I've got this graph actually here, which I don't need to pass to any of you to be able to see the growth of AI agents.

[91] But this graph is Google search trend data.

[92] This also resembles our revenue too.

[93] Oh, okay, right.

[94] The water has arrived.

[95] Hello.

[96] Thank you.

[97] You can come on in.

[98] Can I have a go, please?

[99] Yes, it's 3951.

[100] Thank you so much.

[101] Thank you.

[102] This is like a supernatural kind of power.

[103] You conjured water.

[104] I conjured water from my mind.

[105] And it's shown up here with us.

[106] And it clearly thinks we need a lot.

[107] But just to define the term AI agent for someone that's never heard the term before.

[108] Yeah.

[109] Yeah, so I assume most of the audience now are familiar with ChatGPT, right?

[110] You can go in and you can talk to an AI.

[111] It can search the web for you.

[112] It has a limited amount of tools.

[113] Maybe it can call a calculator to do some additional subtraction for you.

[114] But that's about it.

[115] It's a request -response style.

[116] Agents are when you give it a request.

[117] And they can work indefinitely until they achieve a goal or they run into an error and they need your help.

[118] It's an AI bot that has access to tools.

[119] Those tools are access to a web browser like operator, access to a programming environment, say like Replit, access to credit cards.

[120] The more tools you give the agent, the more powerful it is.

[121] Of course, there's all these considerations around security and safety and all of that stuff.

[122] But the most important thing is that AI agent will determine when it's finished executing.

[123] Today, AI agents can run for anywhere between 30 seconds to 30 minutes.

[124] There's a recent paper that came out that's showing that every seven months, the number of minutes that the agent can run for is doubling.

[125] So we're at like 30 minutes now.

[126] In seven months, we're going to be at an hour.

[127] Then, you know, two hours.

[128] Pretty soon, we're going to be at days.

[129] And at that point, you know, AI agent is doing labor, is doing kind of human -like labor.

[130] And actually, OpenAI's new model, O3, beat the expectation.

[131] So it sort of doubled coherence over long horizon tasks in just three or four months.

[132] So we're in this massive, and I mean, this looks.

[133] this exponential graph, you know, that shows you the massive trend we're on.

[134] Brett, give us a little bit of your background, but also I saw you writing some notes there.

[135] There was a couple of words used there that I thought were quite interesting, especially considering what I know about you.

[136] The word God was used a few times.

[137] Well, let me just say I'm an evolutionary biologist and probably for the purposes of this conversation.

[138] it would be best to think of me as a complex systems theorist.

[139] One of the things that I believe is true about AI is that this is the first time that we have built machines that have crossed the threshold from the highly complicated into the truly complex.

[140] And I will say I'm listening to this conversation with a mixture of profound hope and dread.

[141] because it seems to me that it is obvious that the potential good that can come from this technology is effectively infinite.

[142] But I would say that the harm is probably 10 times.

[143] It's a bigger infinity.

[144] And the question of how we are going to get to a place where we can leverage the obvious power that is here to do good and dodge the worst harms.

[145] I have no idea.

[146] I know we're not prepared.

[147] So I hear you talking about agents, and I think that's marvelous.

[148] We can all use such a thing right away, and the more powerful it is, the better.

[149] The idea of something that can solve problems on your behalf while you're doing something else is marvelous.

[150] But of course, that is the precondition for absolute devastation to arise out of a miscommunication.

[151] To have something acting autonomously to accomplish a goal, you damn well better understand what the goal really is and how to pull back the reins in the event that it starts accomplishing something that wasn't the goal.

[152] The potential for abuse is also utterly profound.

[153] You can imagine, just pick your dark mirror fantasy dystopia where Something has been told to hunt you down until you're dead, and it sees that as a technical challenge.

[154] So I don't know quite how to balance a discussion about all of the things that can clearly come from this that are utterly transcendent.

[155] I mean, I do think it is not inappropriate to be invoking God or biblical metaphors here.

[156] You're producing.

[157] water seemingly from thin air, I believe that does have an exact biblical parallel.

[158] So in any case, the power is here, but so too is the need for cautionary tales, which we don't have.

[159] That's the problem, is that there's no body of myth that will warn us properly of this tool because we've just crossed a threshold that is similar in its capacity to alter the world as the invention of writing.

[160] I really think that's where we are.

[161] We're talking about something that is going to fundamentally alter what humans are with no plan.

[162] Writing alters the world slowly because the number of people who can do it is tiny at first and remains so for thousands of years.

[163] This is changing things weekly.

[164] And that's an awful lot of power to just simply have dumped on a system that wasn't well regulated to begin with.

[165] Done.

[166] Yeah, so I'm an entrepreneur.

[167] I've been building businesses for the last 20 plus years.

[168] I'm completely well positioned between the two of you here.

[169] The excitement of the opportunity and the terror of what could go on.

[170] There's this image that I saw of New York City in 1900 and every single vehicle on the street is a horse and cart.

[171] And then 13 years later, the same...

[172] photo from the same vantage point and every single vehicle on the street is a car.

[173] And in 13 years, all the horses had been removed and cars had been put in place.

[174] And if you had have interviewed the horses in 1900 and said, how do you feel about your level of confidence in the world?

[175] The horses would have said, well, we've been part of humanity for, you know, horse in hoof, hand in hoof for many, many years, for thousands of years.

[176] there's one horse for every three humans.

[177] Like how bad could it be?

[178] You know, we'll always have a special place.

[179] We'll always be part of society.

[180] And little did the horses realise that that was not the case, that the horses were going to be put out of business very, very rapidly.

[181] And to reason through analogy, you know, there's a lot of us who are now sitting there going, hey, wait a second, does this make me a horse in 1900?

[182] I think a lot of people don't realise how massive these kind of technologies are going to have as an impact.

[183] You know, one minute we're ordering water and that's cute, and the next minute it can run for days.

[184] And in your words, it doesn't stop until it achieves its goal and it comes up with as many different ways as it could possibly come up with to achieve its goal.

[185] And in your words, it better know what that goal is.

[186] I'm thinking a lot, Amjad, as Daniel speaking about...

[187] the vast application of AI agents and where are the bounds?

[188] Because if this thing is going to get incrementally smarter, well, incrementally might be an understatement, it's going to get incredibly smart, incredibly quick.

[189] And we're seeing this AI race where all of these large language models are competing for intelligence with one another.

[190] And if it's able to traverse the internet and click things and order things and write things and create things, and all of our lives run off the internet today.

[191] What can't it do?

[192] It's going to be smarter than me. No doubt.

[193] It already is.

[194] And it's going to be able to take actions across the internet, which is pretty much where most of my professional life operates.

[195] It's like how I build my businesses.

[196] Even this podcast is an internet product at the end of the day.

[197] Because you can create, we've done experiments now and I can show you the graphs on my phone to make AI podcasts.

[198] And they have, we've just managed to get it to have the same retention as the driver CEO.

[199] Now with the image generation model.

[200] Retention as in viewer retention?

[201] the percentage of people that get to one hour is the same now.

[202] So we can make the video, we can publish it, we can script it, you can synthesize my voice.

[203] Sounds like me. So what is it going to be able to do?

[204] And can you give me the variety of use cases that the average person might not have intuitively conceived?

[205] So I tend to be an optimist.

[206] And part of the reason is because I try to understand the limits of the technology.

[207] What can it do is any sort of set of human data that we can train it on.

[208] What can it not do is anything that humans don't know what to do because we don't have the training data.

[209] Of course, it's super smart because it integrates massive amount of knowledge that you wouldn't be able to read, right?

[210] It also much faster.

[211] It can run through a massive amount of computation that your brain can't even comprehend.

[212] Because of all of that, they're smart.

[213] They can take actions.

[214] But we know the limits of what they can do because we train them.

[215] They're able to simulate what a human can do.

[216] So the reason you were able to order the water there is because it was trained NSF data.

[217] That includes clicking on DoorDash and ordering water.

[218] I applaud your optimism, and I like the way you think about these puzzles, but I think I see you making a mistake that we are about to discover is very commonplace.

[219] So we have several different categories of systems.

[220] We have simple systems, we have complicated systems, we have complex systems, and then we have complex adaptive systems.

[221] And to most of us...

[222] A highly complicated system appears like a complex system.

[223] We don't understand the distinction.

[224] Technologists often master highly complicated systems.

[225] And they know, you know, for example, a computer.

[226] It's a perfectly predictable system inside.

[227] It's deterministic.

[228] But to most of us, it functions.

[229] It is mysterious enough that it feels like a complex system.

[230] And if you're in...

[231] the position of having mastered highly complicated systems and you look at complex systems and you think it's a natural extension, you fail to anticipate just how unpredictable they are.

[232] So even if it is true that today there are limits to what these machines can do based on their training data, I think the problem is to see what's going to happen.

[233] You really want to start thinking of this as the evolution of a new species that will continue to evolve.

[234] It will partially be shaped by what we ask it to do, the direction we lead it.

[235] And it will partially be shaped by things we don't understand.

[236] So how does this computer that we have work?

[237] Well, one of the things that it does is we plug them into each other.

[238] Using language, it's almost as if you've plugged an Ethernet cable in between.

[239] human minds.

[240] And that means that the cognitive potential exceeds the sum of the individual minds in question.

[241] Your AIs are going to do that.

[242] And that means that our ability to say what they are capable of does not come down to, well, we didn't train it on that data.

[243] As they begin to interact, that feedback is going to take them to capabilities we don't anticipate and may not even recognize once they become present.

[244] That's one of my fears.

[245] This is an evolving creature, and it's not even an animal.

[246] If it were an animal, you could say something about what the limits of that capability are.

[247] But this is a new type of biological creature, and it will become capable of things that we don't even have names for.

[248] Even if it didn't do that, even if it just stayed within the boundaries that you're talking about, you mentioned about it having median level intelligence.

[249] Well, that by definition means 50 % of the people on the planet are less intelligent than AI.

[250] To a degree, it's almost as if we've just invented a new continent of remote workers.

[251] There's billions of them.

[252] They've all got a master's or a PhD.

[253] They all speak all the languages.

[254] Anything that you could call someone or ask someone over the internet to do.

[255] They're there 24 -7 and they're 25 cents an hour, if that.

[256] So like if that really happened, like if we really did just discover that there were a billion extra people on the planet who all had PhDs and were happy to work for almost free, that would have a massive disruptive impact on society.

[257] Like society would have to rethink how everyone lives and works and gets meaning.

[258] And that's if it just stays at a median level of intelligence.

[259] Like it's pretty profound.

[260] I still think it's a tool.

[261] This is power that is there to be harnessed by entrepreneurs.

[262] I think that the world is going to get disrupted, right?

[263] And this post -war world that we created, where you go through life, you go through 12 years of education, you get to college, and you just check the boxes, you get a job.

[264] We can already see.

[265] that fractures with that.

[266] This American dream is perhaps no longer there.

[267] And so I think the world has already changed.

[268] But what are the opportunities?

[269] Obviously, there are downsides.

[270] The opportunities is for the first time, access to opportunity is equal.

[271] And I do think there's going to be more inequality.

[272] And the reason for this inequality is because Actually, Steve Jobs made this analogy.

[273] It's like the best taxi driver in New York is like 20 % better than the average taxi driver.

[274] The best programmer can be 10x better.

[275] We say the 10x engineer.

[276] Now, the variance will be in the 1000x, right?

[277] Like the best entrepreneur that can leverage those agents.

[278] could be better.

[279] It could be a thousand times better than someone who doesn't have the grit, doesn't have the skill, doesn't have the ambition, right?

[280] So that will create a world.

[281] Yes, there's massive access to opportunity, but there are people who will take - Seize it and then there'll be people who don't.

[282] I imagine it almost like a marathon race.

[283] And AI has two superpowers.

[284] One superpower is to distract people, such as TikTok algorithm.

[285] That's right.

[286] And the other superpower is to make you hyper -creative, so you become a hyper -consumer or a hyper -creator.

[287] And in this marathon race, the vast majority of people have got their shoes tied together because AI is distracting them.

[288] Some people are running traditional race.

[289] Some people have got a bicycle and some people have got a Formula One vehicle.

[290] And it's going to be very confronting.

[291] when the results go on the scoreboard and you say, oh, wait a second, there's a few people who finished this marathon in about 30 minutes and there's a lot of us who finished in like 18 hours because we had our shoes tied together and I can't understand if we've got equal opportunity why there's so much disparity between how fast, you know, and I'm using an analogy but this idea that, you know, someone like a lot of people are going to start earning a million dollars a month.

[292] And a lot of people are going to say, hey, I can't even get a job for $15 an hour.

[293] There's going to be this kind of interesting wedge.

[294] Well, but I hear in what both of you are saying a kind of assumption that this will all be done on the up and up.

[295] And I do want to just – I am not a doomer.

[296] I agree that the doomers are likely incorrect, that their fears are misplaced.

[297] But I do think we have a question of a related rates problem.

[298] You know, I said the potential for good here is infinite and the potential for bad is 10 times, right?

[299] What I mean is there are lots of ways in which this obviously empowers people to do things that they were going to be otherwise stuck in the mundane process of learning to code and then...

[300] figuring out how to make the code work and bring it to market and all of that.

[301] And this solves a lot of those problems.

[302] And that's obviously a good thing.

[303] Really, what we should want is the wealth creation object as quickly as we can get there.

[304] But the problem is, you know, as much as it, that hyper creative individual is empowered to make wealth, the person who is interested in stealing may be even more empowered.

[305] And I'm concerned about that.

[306] at a pretty high level.

[307] The abuse cases may outnumber the use cases.

[308] And we don't have a plan for what to do about that.

[309] Can I give you a quick introduction here, like the optimistic view?

[310] OpenAI invented, the first version of GPT came out in 2019.

[311] 2020 was GPT -2.

[312] And so OpenAI, now they get a lot of criticism and lawsuit from Elon Musk.

[313] that they're no longer open source than they used to be.

[314] The reason is in GPT -2, they said, we are no longer going to open source this technology because it's going to create opportunities for abuse, such as influencing elections, stealing grandma's credit card, and so on and so forth.

[315] Wouldn't you say, Brett, that it is kind of surprising how little abuse we've seen so far?

[316] I don't know how much abuse we've seen so far.

[317] I don't know how any of us do.

[318] And I also, even the example that you suggest where chat GPT is no longer open source to prevent abuse, I'm taking their word for it that that's the motivation.

[319] Whereas a systems theorist, I would say, well, if you had a technology that was excellent at enhancing your capacity to wield power, then open sourcing it is a failure to capitalize on that and that the most remunerative use is to keep it private and then either sell the ability to manipulate elections to people who want to do so or sell the ability to have it kept off the table for people who don't.

[320] And I would expect that that's probably what's going on.

[321] If you have a technology as transformative as this, giving it away for free is counterintuitive, which leaves...

[322] those of us in the public, more or less at the mercy of the people who have it.

[323] So I don't see the reason for comfort.

[324] We are at the dawn of this radical transformation of humans that by its very nature as a truly complex and emergent innovation, nobody on earth can predict what's going to happen.

[325] We're on the event horizon of something.

[326] And the problem is, you know, we can talk about the obvious disruptions, the job disruption, and that's going to be massive.

[327] And does that lead some group of elites to decide, oh, well, suddenly we have a lot of useless eaters and what are we going to do about that?

[328] Because that conversation tends to lead somewhere very dark very quickly.

[329] But I think that's just the beginning of the...

[330] the various ways in which this could go wrong without the Doomer scenarios coming into play.

[331] This is an uncontrolled experiment in which all of humanity is downstream.

[332] So I was trying to make the point that OpenAI has been sort of wrong about how big of a potential for harm it is.

[333] I think we would have...

[334] Heard about it in the news, like the sort of how much harm it's done.

[335] Maybe, you know, some of it is working in the shadows.

[336] But like the few incidents that we've heard about where, you know, the cause of LLM's large language models, the technology that's powering ChatGPT, has been huge headliners on like New York Times.

[337] Talked about this kid that was.

[338] perhaps goaded by some kind of chat software that helps teenagers to be less lonely into suicide, which is tragic.

[339] And obviously, these are the kind of safety and abuse issues that we want to worry about.

[340] But these are kind of these isolated incidents.

[341] And we do have open source large language models.

[342] Obviously, the thing that everyone talks about is DeepSeek.

[343] DeepSeek is coming from China.

[344] So what is DeepSeek's incentive?

[345] Perhaps the incentive is to destroy the AI industry in the US.

[346] When they released DeepSeek, the market tanked, the market for NVIDIA, the market for AI, and all of that.

[347] But there is an incentive to open source.

[348] Meta is open sourcing Lama.

[349] Lama is another AI similar to ChatGPT.

[350] The reason they're open sourcing Lama, and Zuckerberg just says that out loud, is basically, They don't want to be beholden to open AI.

[351] They don't sell AI as a service.

[352] They use it to build products.

[353] And there's this concept in business called commoditize your complement.

[354] Because you need AI as a technology to run your service, the best strategy to do is to open source it.

[355] So these market forces are going to create conditions that I think are actually beneficial.

[356] So I'll give you a few examples.

[357] One is, first of all, the AI companies are motivated to create AI that is safe so that they can sell it.

[358] Second, there are security companies investing in AIs that allows them to protect from the sort of malicious acting of AI.

[359] And so you have the free market.

[360] We've always had that, you know.

[361] But generally, as humanity, we've been able to leverage the same technology.

[362] to protect against the abuse.

[363] So I don't really understand this.

[364] And maybe this is actually, this is the exact discussion that you would expect between somebody at the frontier of the highly complicated, staring at a complex system, and a biologist who comes from the land of the complex and is looking back at highly complicated systems.

[365] In game theory, we have something called a collective action problem.

[366] And in the market that you're describing, An individual company has no capacity to hold back the abuses of AI.

[367] The most you can do is not participate in them.

[368] You can't stop other people from programming LLMs in some dangerous way.

[369] And you can limit your own ability to earn.

[370] based on your own limitations of what you're willing to do.

[371] And then effectively what happens is the technology gets invented anyway, it's just that the dollars end up in somebody else's pocket.

[372] So the incentive is not to restrain yourself so that you can at least compete and participate in the market that's going to be opened.

[373] And so the number of ways in which you can abuse this technology, let's take a couple.

[374] What is to stop somebody?

[375] From training LLMs on an individual's creative output and then creating an LLM that can outcompete that individual, can effectively not only produce what they would naturally produce over the course of a lifetime, but can extrapolate from it and can even hybridize it with the insights of other people so that effectively those who have the LLM.

[376] can train it on the creativity of others, not cut them in on the use of that insight, you can effectively end up putting yourself out of business by putting your creative ideas in the world where they get sucked up as training data for future LLMs.

[377] That is unscrupulous, but it's effectively guaranteed.

[378] In fact, it's already happened.

[379] So that's a problem.

[380] And likewise, what would stop somebody from interacting with an individual and training an LLM?

[381] to become like a personalized con artist, something that would play exactly to your blind spots.

[382] Well, that does happen.

[383] That is starting to happen.

[384] People get phone calls and it sounds like their daughter and I've lost my phone and I'm borrowing a friend's phone and all of that sort of stuff.

[385] What's interesting is that I think you make a really good point.

[386] I worry about the impact on society, and yet when I look at every single individual who uses AI regularly, it almost has nothing but...

[387] profoundly positive impact on their life.

[388] I look at people like I was just spending some time with my parents -in -law who are in their 70s and early 80s and they use AI regularly for all sorts of things that they find incredibly valuable and that improves the quality of their life.

[389] I personally did an M &A, a mergers and acquisitions deal where I bought a company last year and the AI was so powerful at helping that process.

[390] The conversations were transcribed and they were turned into letters of intent and then press releases and legal documents.

[391] And we probably shaved $100 ,000 worth of costs and we sped up the whole process and it was pretty magical to see how it could happen.

[392] With that said...

[393] You know, there's all of these, like, well, $100 ,000 worth of lawyers didn't get paid, right?

[394] Well, what I want to know, yeah.

[395] There's a lot of people upset about that.

[396] But if we look back at the invention of the cell phone or the invention of the social media platforms, there would be every reason to have exactly the same perspective, right?

[397] I remember the beginning of Facebook, and I remember the idea that suddenly the process that...

[398] used to afflict people where you would just lose touch with most of the people who had been important to you.

[399] That was not something that needed to happen anymore.

[400] You could just retain them permanently as a part of a diffuse social grouping that just simply grew and value was added.

[401] There's no end to how much good that did.

[402] But what it did to us was profound and not evident in the first chapter.

[403] Say the same thing about the cell phones and the dopamine traps and the way this has disconnected us from each other, the way it has disconnected us from nature, the way it has altered the very patterns with which we think.

[404] It has altered every classroom.

[405] And those things, I think, are going to turn out to have been sort of minor foreshadowings of the disruption that AI will produce.

[406] So I agree with you today.

[407] The amount you can do with AI is a tremendous amount of good.

[408] There's a little bit of harm.

[409] Maybe that's something we need to worry about.

[410] But as this develops, as we get to peer over the edge of this cliff that we're headed to, I think we're going to discover that we can't yet detect the nature of the alteration that's coming.

[411] I just wanted to add some context to that because, Amjad, I saw the interview you did in a newsletter in 2023 where you said, I wouldn't prepare for AGI in the same way that I wouldn't prepare for the end of days.

[412] It's effectively the end of days if the vision of AGI that some of these companies have comes to bear because it's called the singularity moment because you can't really predict what happens after that.

[413] And so, like, how would you even prepare for that?

[414] And you want to prepare for the more likely world and that world that you can actually predict.

[415] a world where, yes, there's like a massive improvements of technology and there's like insane compounding effects of technology and it's pretty hard to keep up.

[416] From that, it appeared that in 2023, you were saying a similar thing to Brett in terms of, we can't see around the corner here because it is a singularity.

[417] So you also used AGI, artificial general intelligence.

[418] It'd be interesting to know what your definition of AGI is.

[419] So what I was saying there is, even if I'm wrong, that you can actually create a unbounded, seemingly conscious artificial intelligence that can entirely replace humans and can act autonomously in a way that...

[420] Even humans can't act and can't coordinate across different AIs, different data centers to take over the world.

[421] So the definition of AGI is artificial general intelligence, meaning that AI can acquire new skills efficiently in the same way that humans can acquire skills.

[422] Right now, AIs don't acquire skills efficiently.

[423] They acquire a massive amount of energy and compute, entire data center of compute to acquire these skills.

[424] And I think there's, again, a limit on how general intelligence can get.

[425] I think for most of the time, it will be lagging in terms of what humans are capable of doing.

[426] The singularity is based on this concept of intelligence explosion.

[427] So once you create an AGI, once you create an artificial general intelligence, that intelligence will be able to modify its own source code.

[428] and create the next version that is much more intelligent.

[429] And the next version creates the next version and the next, you know, for infinity, right?

[430] Within a week.

[431] Within a week.

[432] Perhaps within milliseconds at some point, right?

[433] Because it might invent new computing substrate and all of that.

[434] Perhaps it will use quantum computing.

[435] And so then you have this intelligence explosion.

[436] in a way that is impossible to predict how the world is going to be.

[437] And what I'm saying is, this is sort of like an end of time story.

[438] Like, how would you even prepare for that?

[439] So if that's coming, like, why would I spend my time preparing for, I think it's unlikely to happen.

[440] Can't see around the corner.

[441] Yeah, but I'd rather prepare what I was saying there.

[442] I'd rather prepare for the more likely.

[443] world in which we have access to tremendous power, but the world is not ending and humans are still important.

[444] I don't know why you say more likely.

[445] I mean, I think the structure of your argument is sound.

[446] You would prepare for the world that might happen for which you can prepare.

[447] There's literally no point in trying to prepare for a world you can't predict at all.

[448] The only thing you can do is just...

[449] sort of upgrade your own skills and pay attention.

[450] But if I have one message for the technologists, it's that your confidence about what this can and cannot do is misplaced because you have, without noticing, stepped into the realm of the truly complex.

[451] In the truly complex, your confidence should drop to near zero that you know what's going on.

[452] Are these things conscious?

[453] I don't know.

[454] But will they be?

[455] Highly likely they will become conscious and that we will not have a test to tell us whether that has happened.

[456] Elon Musk predicts that by 2029, we will have AI with us, AGI, that surpasses the combined intelligence of all humans.

[457] And Sam Altman actually wrote a blog three months ago that I read where he said, We are confident now, Sam Altman being the founder of OpenAI, which created ChatGPT, we are confident now that we know how to build AGI as we have traditionally understood it.

[458] When I put these things together, I go back to the central question of what role do humans have in this sort of professional output and GDP creation if it's smarter than all humans combined, if Elon Musk is correct there, and it's able to take actions across the internet.

[459] and continue to learn?

[460] This is like a central question that I'm hoping I can answer today, which is like, where do we go?

[461] Yeah.

[462] I mean, in my vision of the world, we're in the creative seat.

[463] We're sitting there where we are controlling swarms of intelligent being to do our job.

[464] You know, the way you run your business, for example, you're sitting at a computer, you have an hour to work, and you're going to launch like a thousand.

[465] SDR, sales representative, to go grab as many leads as possible.

[466] And you're generating a new update on Replit for your website here.

[467] And then on this side, you have an AI that's crunching data about your existing business to figure out how to improve it.

[468] And these AIs are kind of somehow all coordinating together.

[469] And I am trying to privilege This is my mission, is to build tools for people.

[470] I'm not building tools for agents, and agents are a tool.

[471] And so ultimately, not only do I think that humans have a privileged position in the world and in the universe, we don't know where consciousness is coming from.

[472] We don't really have the science to explain it.

[473] I think humans are special.

[474] That's one side is my belief that humans are special in the world.

[475] And another side, which I understand that the technology today, and I think for the foreseeable future, is going to be a function of us training data.

[476] So there was this whole idea of like, what if ChatGPT generates pathogens?

[477] Well, have you trained it on pathogens?

[478] They were doing that kind of stuff in Wuhan, no?

[479] I mean, a lot of the biotech companies are essentially, using artificial intelligence.

[480] Like I can think of Abcelera, I think it's Abcelera in Canada.

[481] Their whole business is using AI to create new vaccines, using artificial intelligence and bigger data sets than we've never had before.

[482] And I know because I was very close to one of the founders of people involved in Abcelera.

[483] So that work is going on anyway.

[484] And if you think about Wuhan, it's quite probably well known now that it came out of a lab and people working in a lab.

[485] And in that scenario, that had a huge impact and shut down the world.

[486] The central question I'd love to answer before I throw it back open to the room is what jobs, because I know that you have this perspective, what jobs are going to be made redundant in a world where I am sat here as a CEO with a thousand AI agents.

[487] I was thinking of all the names of the people in my company who are currently doing those jobs.

[488] I was thinking about my CFO when you talked about processing business data, my graphic designers, my video editors, et cetera.

[489] So what jobs are going to be impacted?

[490] Yeah, all of those.

[491] And what do they do?

[492] Maybe this is useful for the audience.

[493] I think if your job is as routine as it comes, your job is gone in the next couple of years.

[494] So meaning in those jobs, for example, quality assurance jobs, data entry jobs, you're sitting in front of a computer and you're supposed to click.

[495] and type things in a certain order, operator and those technologies are coming to the market really quickly and those are going to displace a lot of labor.

[496] Accountants?

[497] Accountants.

[498] Lawyers?

[499] Yes.

[500] I mean, I've just pulled a ligament in my foot and they did an MRI scan and I had to wait a couple of days for someone to look at the MRI scan and tell me what it meant.

[501] I'm guessing that's gone.

[502] Yeah, I think the healthcare ecosystem is hard to predict because of regulation.

[503] And again, there's so many limiting factors on how this technology can permeate the economy because of regulations and people's willingness to take it.

[504] But, you know, things, unregulated jobs that are purely text in, text out.

[505] If your job, you know, you get a message and you produce some kind of artifact that's like probably text or images.

[506] That job is at risk.

[507] So just to give you some stats here as well.

[508] About 50 % of Americans who have a college degree currently use AI.

[509] The stats are significantly lower for Americans without a college degree.

[510] So you can see how a splinter might emerge there and that crack will widen because people like us at this table are all messing around with it.

[511] But my mum and dad in Plymouth in the southwest, rural England, haven't got it.

[512] They just figured out iPhones.

[513] So I got them an iPhone and now they're texting me back.

[514] AI is a million miles away.

[515] And if I start running off with my AGI and my agents, that gap is going to widen.

[516] Women are disproportionately affected by automation, which is what you were talking about there, with about 80 % of working women in an at -risk job compared to just over 50 % of men, according to the Harvard Business Review.

[517] And jobs requiring only a high school diploma have an automation risk of 80%, while those requiring a bachelor's degree have an automation risk of just 20%.

[518] So we can see again how...

[519] how this will cause a sort of inequality.

[520] It's also a huge risk with business process outsourcing, which is essentially Western countries sending jobs to India, to the Philippines.

[521] Like at the moment, millions of people have been lifted out of poverty through the ability to do those kind of business process outsourcing jobs, and those are all going to go.

[522] But they're going to have 1 ,000 employees.

[523] But also, these people are actually already transitioning to training AIs.

[524] So there's going to be a massive industry around training AIs.

[525] Until they're trained.

[526] Well, no. You have to continuously acquire new skills.

[527] And this is what I'm talking about.

[528] Again, if AI is a function, if it's data, then you need...

[529] increasingly more data.

[530] And by the way, we ran out of internet data.

[531] I was actually thinking, interestingly, that this might not be great for the United States or the UK, the Western world, because it is going to be a leveler when our kid in India doesn't need a Silicon Valley office and $7 million in investment to throw up a software company, basically.

[532] Yeah, my belief is that, so I have a more broad definition of AGI and the singularity.

[533] And for me, AGI is do we have artificial general intelligence in terms of generally speaking, can AI just do stuff that humans used to be able to do?

[534] And we've already crossed that point.

[535] We have this general intelligence that we can now all access.

[536] And 800 million people a week are now using ChatGPT.

[537] It's exploded in the last three months.

[538] And then to me, a singularity, when the first tractor went out onto a farm.

[539] For me, that was a singularity moment because everyone who worked in farming, it used to take 100 people to plough a field and now a tractor comes along and two guys with the tractor can now plough the field in just as much time and now 98 people out of 100 are completely out of a job.

[540] We also always underestimate.

[541] a technology.

[542] If it does go on to change history, when you look back through cars, horses, planes, the Wright brothers just thought of a plane as being something that the army could use.

[543] They had no idea of the application.

[544] So someone said to me recently, they said, when it does change the world, we underestimate the impact that it will change the world.

[545] And I see people now with their estimations of AI agents already incredibly optimistic.

[546] And so if history holds here, We are undershooting the impact it's going to have.

[547] And I think this is the first time in my life where the industrial revolution analogies seem to fall a little bit short.

[548] Yeah.

[549] Because we've never seen intelligence.

[550] I could think of this as, I'm not an intelligent person on this, but I could see that as like the disruption of muscles.

[551] Whereas this is the disruption of intelligence.

[552] Intelligence.

[553] That's exactly the thing, is that what makes human beings special.

[554] is our cognitive capacity and very specifically our ability to plug our minds into each other so that the sum is or the whole is greater than the sum of the parts.

[555] That's what makes human cognition special.

[556] And what we are doing is we are creating something that can technologically surpass it without any of the preconditions that make that a safe process.

[557] So yes, we've revolutionized the world how many different times?

[558] It's innumerable, but we've made farming vastly more efficient.

[559] That's different than taking our core competency as a species and surpassing ourselves with the product of our labor.

[560] I think your question is a good one.

[561] Then what does become?

[562] We only have one thing left.

[563] We have our muscles, which we got rid of in the industrial revolution.

[564] And then we have our intellect, which is this digital revolution.

[565] Now we're left with emotions and agency.

[566] So we essentially, the agency idea, I think we used to judge people on IQ and now IQ is the big leveller.

[567] And now going forward for the next 10 years, we're going to look at are you a high agency person or a low agency person?

[568] Do you have the ability to get things done and coordinate agents?

[569] Do you have the ability to start businesses?

[570] give orders to digital armies, you know, and essentially these high agency people are going to thrive in this new world because they have this other thing that's been bubbling under the surface, which is really interesting.

[571] When you said agency is going to remain as an important thing, we're sat here talking about AI agents.

[572] And the crazy thing in a world of AI agents that have super intelligence is I can just tell my agent, listen, I'm going on holiday.

[573] Please build me a SaaS company that spots a market opportunity.

[574] throw up the website, post it on my social media channel, I'll be in Hawaii.

[575] And this new agentic world is stealing that too.

[576] Because now it can take action.

[577] In the same way that I can browse the internet, I can call Domino's Pizza, speak to their agentic agent, organize my pizza to be there before I even wake up.

[578] And in fact, predictability, you know, OpenAI now learns.

[579] And Sam Altman said that they've expanded the memory feature.

[580] So it's knowing more and more and more and more about me. It'll almost be able to predict what I want when I want it.

[581] It'll know Steve's calendar.

[582] He's arriving at the studio.

[583] Make sure his cadence is on the side.

[584] Make sure his iPad has the brief on it.

[585] Do the brief.

[586] Do the research for me. And everything else.

[587] So it'll remember Brett's birthday.

[588] So when I arrive, there'll be something.

[589] In fact, it's removing my need for any agency.

[590] Yes.

[591] And again, I don't know how to make this point so that it occurs to people what I'm really suggesting.

[592] Today, maybe it's not conscious, but, well, let me put it to you this way.

[593] If you're conscious, you started out as a child that wasn't.

[594] And although this may not fully encapsulate it, you are effectively an LLM, right?

[595] You go from an unconscious infant to a highly conscious adult.

[596] And the process by which you do that has a lot to do with being trained effectively on words and other things in an environment in exactly the way that we now train these AIs.

[597] So the idea that we can take consciousness off the table, it won't be there until we figure out how to program it in, and we're safe because we don't know how consciousness works, I take the opposite lesson.

[598] We've created the exact thing that will produce that phenomenon, and then we can have philosophers debate whether it's real consciousness or it just behaves exactly as if it were, and the answer is those aren't different.

[599] Doesn't matter.

[600] And the same thing is true for agency, especially if you've created an environment in which these AIs are de facto competitors.

[601] What you're effectively doing is creating an evolutionary environment in which they will evolve to fill whatever niches are there, and we didn't spell out the niches.

[602] I have the sense we have created something that truly is going to function like a new kind of life.

[603] And it's especially troubling because it speaks our language.

[604] So that leads us to believe it's more like us than it is.

[605] And it's actually potentially quite different.

[606] By the way, he's the optimist here.

[607] He's so optimistic about LLMs and how they're going to evolve.

[608] Yes, it's amazing.

[609] It's amazing technology.

[610] Like I think it raised global IQ, right?

[611] Like 800 million people.

[612] Like 800 million people are that much more intelligent and emotionally intelligent as well.

[613] Like I know people who previously were very coarse and they kind of robbed.

[614] people the wrong way.

[615] They would say things in not so polite way.

[616] And then suddenly they started putting what they're saying through chat GPT in order to kind of make it kinder and nicer.

[617] And they're more liked now.

[618] And so not only is it making us more intelligent, but also it allows us to be the best version of ourselves.

[619] And the scenario that you're talking about, I don't know what's wrong with that.

[620] I would want less agency in certain places.

[621] I would want something to help me not open up a peanut butter jar at night.

[622] There are places in my life where I need more control and I would rather cede it to some kind of entity that could help me make better choices.

[623] I mean, unfortunately, even if there is some Small group of elites that are able to go to Hawaii while something else does the mundane details of their business building.

[624] We are rather soon going to be faced with a world that has billions of people who do not have the skills to leverage AI.

[625] Some of them will be necessary for a time.

[626] You're going to need plumbers.

[627] But this is also...

[628] not a long -term solution because not only are there not enough of those jobs, but of course we have humanoid robots that once imbued with AI capacity will also be able to take, you know, they'll be able to crawl under your house into the crawl space and fix your plumbing.

[629] So what typically happens when you have a massive economic contraction that arises from the fact that a huge number of people are out of work, is that the elites start looking at those people and thinking, well, we don't really need them anyway.

[630] And so the idea that this AI disruption doesn't lead us to some very human catastrophe, I think, is overly optimistic and that we need to start preparing right now.

[631] What are the rights of a person who has had whatever it is that they've invested in completely erased from the list of needs?

[632] Is that person responsible for not having anticipated AI coming?

[633] And is it their problem that they are now starving and they're being eyed by others as a useless eater?

[634] I don't think so.

[635] How is it different than when the looming machine came and the textile workers, the result of the Luddite revolution?

[636] How is it different than any time in history when technology...

[637] automated, a lot of people out of jobs.

[638] I would say scale and speed.

[639] That's how it's different.

[640] And the scale and speed is going to result in an unprecedented catastrophe because the rate at which people are going to be simultaneously sidelined, not just in one industry but across every industry, is just simply new.

[641] And it also did actually happen.

[642] For the first 50 years of industrialization, from like late 1700s to early 1800s, You actually, the Charles Dickens novels are essentially people coming from the farms who were displaced, arriving in cities, kids living on the streets.

[643] The British decided to pick everyone up and send them over to Australia, which is where I came from.

[644] And, you know, there was this massive issue of displacement.

[645] I think we're going to go into a high -velocity economy where rather than this long arc of career that lasts 45 years, we're going to have these very fast careers that last 10 months to 36 months.

[646] And you invent something, you take it to market, you put together a team of 5 to 10 people who work together, you then get disrupted.

[647] Can I mention a story here?

[648] There's an entrepreneur that uses Replit.

[649] In a similar way, his name is Billy Howell.

[650] You can find him on YouTube, on the internet.

[651] He would go to Upwork, and he would find what people are asking for, different requests for certain apps, technologies.

[652] Then he would take what they're asking for, put it into Replit, make it an application, call them, tell them, I already have your application.

[653] Would you pay $10 ,000 for it?

[654] And so that's sort of an arbitrage opportunity that's there right now.

[655] That's not arbitrage.

[656] It's that.

[657] How is it?

[658] You have somebody who has an idea that can be brought to market and somebody else is cryptically detecting it and then selling back their own idea to them.

[659] Well, they're paying them to do that.

[660] They're saying, I will give you $500 if someone makes this for me. Right.

[661] But this is what I more or less think is going to happen across the whole economy is that, yes, from this perspective, we can see that everybody is suddenly empowered to build a great business.

[662] Well, what do we think about?

[663] the folks who are going to be displaced from the top, what are they going to think about all these people building all of these highly competitive businesses?

[664] And are they going to find a way to do what venture capital has done or what record producers have done?

[665] What they're going to do is they're going to take their superior position at the top, and they are going to take most of the wealth that is produced by all of these people who have these ideas that in a proper market would actually create businesses for them, and they're going to parasitize them.

[666] And with this introduction of AI and AI agents, old value has moved.

[667] And now it's not going to be the case that the idea itself is the moat.

[668] And it's not going to be the case that resources are the moat.

[669] So in such a scenario, you still have to figure out distribution.

[670] You still have to have, for example, like an audience.

[671] So if you're a podcaster now, you have a million followers on Twitter, you're in a prime position because you now have something that...

[672] The guy with a great idea with no audience has, you have inbuilt distribution.

[673] So I now think actually much of the game might be moving to like, yeah, still about taste and idea, but also the motive's distribution.

[674] Yeah, and speaking of adaptive systems, one of the adaptations that will happen is people will seek humans and will seek proof of humanity.

[675] Oh, I agree that authenticity is going to become the coin of the realm.

[676] And anything that can be faked or cheated is going to be devalued and things, you know, spontaneous jazz or, you know, comedy that is interactive enough that it couldn't possibly have been generated with the aid of AI.

[677] Those things are going to become prioritized.

[678] spontaneous oratory rather than speeches.

[679] That answers some of your questions.

[680] No, it answers my question for the tiny number of people who are in a position to do those things.

[681] Stephen, you used the word moat, which I think is a really important word for entrepreneurs.

[682] We have to have a moat.

[683] We think a lot about moats, and it's an industrial age.

[684] A lot of people don't even know what you mean by moat.

[685] I often think about this idea of what are the moats that are left.

[686] So to define how I define a moat, got a castle and it's got like a small circle of water around it.

[687] And once upon a time, that circle of water defended the castle from attack.

[688] And you can pull up the drawbridge so nobody can attack you very easily.

[689] It's a defense from something.

[690] So it's your shield.

[691] It's your defense.

[692] And once upon a time as an entrepreneur, you know, I've got a software company in San Francisco called Third Web, and we raised almost $30 million.

[693] We have a team of 50 great developers.

[694] And much of our moat was...

[695] You can't compete with us because you don't have the 50 developers and the $30 million in the bank.

[696] How much of that $30 million went to coding?

[697] The vast majority of it.

[698] I mean, what else are we going to do?

[699] This is a good thing.

[700] I think modes are a bad thing.

[701] Let me make the argument there.

[702] So everyone is looking for modes.

[703] For example, one of the more significant modes is network effects.

[704] So you can't compete with Facebook or Twitter because To move people from Facebook or Twitter, it's the collective action problem.

[705] You need to move them all at once because if one of them moves, then the network is not valuable.

[706] They'll go back.

[707] So you have this chicken and egg problem.

[708] Let's say that we have a more decentralized way of doing social networks.

[709] That will remove the power of Twitter.

[710] to kind of censor.

[711] And I think you're at the other end of censorship, right?

[712] And so part of my optimism about humanity is that generally there's self -correction.

[713] Democracy is a self -correcting system.

[714] Free markets are largely self -correcting systems.

[715] There are obvious problems with free markets that we can discuss.

[716] But take health.

[717] There is obesity epidemic, this period of time when companies ran loose, kind of making this sugary, salty, fatty kind of snacks, and everyone gorged on them, and everyone got very unhealthy.

[718] And now you have Whole Foods everywhere.

[719] Today, people in Silicon Valley, they don't go to bars at all.

[720] They go to running clubs.

[721] That's how you meet.

[722] That's how you go find a date.

[723] You go to running clubs.

[724] And so there was a shift that happened because there was a reaction.

[725] Obviously, cigarettes is another example.

[726] You know, you were talking about phones and our addiction to phones.

[727] And I see a shift right now.

[728] Like in my friend circle.

[729] Like people who are constantly kind of on their phones is already kind of frowned upon.

[730] And they don't want to hang out with you because you're constantly staring on your phone.

[731] So there's always these reactions.

[732] But the problem is you reference self -correction.

[733] And I agree that there's actually an automatic feature of the universe in which the self -correction happens.

[734] You can't have a positive feedback that isn't reined in by some outer negative feedback.

[735] But the corrections, the list of corrections involves things.

[736] like you point to, where people become enlightened and they realize that they're doing themselves harm with either the sugar that they're consuming or the dopamine traps on their phone and they get better.

[737] But also on the list of corrective patterns are genocide and war and, you know, parasitism.

[738] And the problem is these things are destructive of wealth.

[739] And so you allude to the superior fact of an open market without moats.

[740] Presumably the benefit of that is that more wealth gets created because people aren't kept from doing things that are productive.

[741] I see that.

[742] But then what is the product of all of this new wealth that is going to be generated by a world empowered by AI?

[743] Does it end up so highly concentrated that you have a tiny number of ultra elites and a huge number of people who are utterly dependent on them?

[744] What becomes of those people?

[745] The learning process, the self -correction process goes through harm in order to get to that more enlightened solution.

[746] There's nothing that protects us from the harm phase being so apocalyptically terrible that, you know, we get to the other side of it and we say, well, that was a hell of a correction.

[747] Or maybe there's nobody there to even say that.

[748] Those are also on the table.

[749] It reminds me of a mousetrap where you see the cheese and we're going, oh, my God, my grandmother is going to be able to do some research.

[750] And, oh, my God, my life is going to get easier.

[751] So you head closer and closer to the cheese.

[752] And then, snap.

[753] Historically, if we look at all of the last 10 ,000 years.

[754] It's a very small number of elites who own absolutely everything and a very large number of serfs and peasants who have a subsistence living.

[755] You know, if the elites are too greedy and they freeze out the peasants at too high a level and they try to use brutality.

[756] Pitchforks.

[757] Yeah, eventually it comes back to haunt them.

[758] And so what you get is a recognition that you need a system that does balance these things.

[759] And, you know, the West has the best system that we've ever seen.

[760] It's one in which.

[761] We agree on a level playing field.

[762] We never achieve it, but we agree that it's a desirable thing.

[763] And the closer we get to it, the more wealth we create.

[764] But again, if AI empowers those with ill intent at a higher rate that it empowers those who are...

[765] wealth creating and pro -social, we may be in for a massive regression in how fair the market that the West is.

[766] Is that your top concern versus economic displacement?

[767] I think they're the same thing.

[768] How are they the same thing?

[769] Because the economic displacement is going to start.

[770] I don't know how many million people are going to be displaced from their jobs in the US.

[771] Suddenly, we're going to have a question.

[772] about whether or not we have obligations to them.

[773] You agree with that, don't you?

[774] Yes, but again, it's no pain, no gain.

[775] I mean, we're going to go through a period of disruption.

[776] And I think at the other end, the old sort of oppressive systems will be broken and we're going to create perhaps a fair world, but it's going to have its own problems.

[777] And what's the scale of that disruption in your estimation?

[778] It's hard to say because...

[779] You know, there's this concept of limiting factors.

[780] Like, you know, there is regulation.

[781] There's the appetite of people today.

[782] For example, the healthcare system is very resistant to innovation because of regulation, you know?

[783] And that's a bad thing.

[784] On the regulation point, it's worth saying that when Trump came into power, he signed in a new...

[785] law, which is called removing barriers to American leadership in AI, which revokes previous AI policies that were deemed to be restrictive.

[786] And obviously, when you think about where the funding is going in AI, it's going to two places.

[787] It's going to America, and it's basically going to China.

[788] That's the vast majority of investment.

[789] So with those two in competition, any regulation that restricts AI in any way is actually self -sabotage.

[790] And this is, you know, I live in Europe some of the time.

[791] And it's already annoying to me that when Sam Altman and OpenAI released the O3 model, this new incredible model, it's not in Europe because Europe has a regulation which prevents it from coming to Europe.

[792] So we're now at a competitive disadvantage, which Sam Altman has spoken about.

[793] And more broadly on this point of disruption, I was quite unnerved when I heard that Sam Altman's other startup was called WorldCoin.

[794] And WorldCoin was conceived.

[795] with the goal of facilitating universal basic income i .e helping to create a system where people who don't have money are given money by the government just for being alive to help them cover their basic food and housing needs which suggests to me that the guy that has bought the biggest air company in the world can see something that a lot of us can't see, which is there's going to need to be a system to just hand out money to people because they're not going to be able to survive otherwise.

[796] I fundamentally disagree with that.

[797] Which part do you disagree with?

[798] I disagree that, first of all, that humans would be happy with UBI.

[799] I think that core value of humans, and be curious about the evolutionary reasons, is we want to be useful.

[800] It's really important to know that A lot of the jobs that are at risk are the most high status, highly paid jobs in the world.

[801] Let's take the highest paid job in America, which is an anesthesiologist.

[802] This is the highest paid job.

[803] Highest paid, salaried job.

[804] Salaried job, yeah.

[805] And the majority of that job is observing a patient, knowing which type of medication would work best with their body, giving them the exact right amount, monitoring the impact of that on the body.

[806] and then making slight adjustments, the right technology and any nurse will be able to do that job.

[807] And you might have one anesthesiologist on site supervising 10, 20, 30, 40 wards, and the technology is doing the job, but that one person is there just to kind of supervise if something went wrong or if there was an ethical dilemma.

[808] What's wrong with that?

[809] I mean, if the precision is better where they are.

[810] Oh, no, there's nothing wrong with that except for the fact that A lot of people, hundreds of thousands of people have spent their entire life training to be that.

[811] They get an enormous amount of purpose and satisfaction about the fact that that's their career, that's their job.

[812] They have mortgages, they have houses, they have status, and that's about to go away.

[813] Well, if it's highest paid jobs, maybe you should start saving.

[814] Well, I mean, but I hear you.

[815] But you're talking about people who have done vital work.

[816] highly specialized work and are therefore not in a great position to pivot based on the invention of a technology that they didn't see coming because frankly, I mean, in the abstract, maybe we all saw AI coming somewhere down the road, but we did not know that it was going to suddenly dawn.

[817] And we do have to figure out what to do with those people.

[818] It's not their fault that they've suddenly become obsolete.

[819] And it's inconceivable that people will accept this.

[820] It is not, it is.

[821] fundamentally incompatible with our nature.

[822] We have to have things to strive for.

[823] And you can sustain life that way, but you cannot sustain a meaningful existence.

[824] And so it's a short -term plan at best.

[825] Let's talk about meaning.

[826] On that point of job displacement, this is already happening.

[827] Klarna, CEO, who has been on this podcast before, a great guy, said to...

[828] on a blog post that they published on Klarna's website saying that they now have AI customer service agents handling 2 .3 million chats per month, which is equal to having to hire 700 full -time people to do that.

[829] So they've already been able to save on 700 customer service people by having AI agents to do that.

[830] And they actually got rid of those 700 jobs, right?

[831] I don't have that information in front of me, but I'll have a look.

[832] I'll throw it up on screen.

[833] for anyone that wants context on that.

[834] But that's already happening.

[835] This isn't an anesthetologist or something.

[836] And these aren't high -paid people in every case.

[837] We've done something similar, by the way.

[838] Internally, we've replaced that function for 70%.

[839] I mean, our company, we're 65 people.

[840] And, you know, we...

[841] We make millions per head.

[842] Are you going to need to hire more people to get up to?

[843] I think so, but we're hiring slowly.

[844] We're using customer support AI, and that meant that we need less customer support.

[845] We're trying to leverage AI as much as possible.

[846] The person in HR at Replit writes software using Replit.

[847] I'll give you an example.

[848] She needed Orchart software.

[849] And she looked at a bunch of them, got a lot of demos.

[850] And they're all very expensive.

[851] And they're missing the kind of features that she wants.

[852] For example, she wanted like version control.

[853] She wanted to know when something changed and to go back in history.

[854] She went into Replit and in three days, she got exactly the kind of software that she wants.

[855] And what was the cost?

[856] Perhaps $20, something like that.

[857] Once.

[858] Once, right?

[859] And how many employees in HR do we need?

[860] Right now, we have two.

[861] If they're highly levered like that, maybe we do not need a 20 HR team.

[862] On this point of meaning, I've heard so many billionaires in AI describe this as the age of abundance.

[863] And I'm not necessarily sure if abundance is always a great thing.

[864] Because, you know, when we look at mental health and we look at why...

[865] how people derive their meaning and their purpose in life.

[866] Much of it is having something to strive towards and some struggle in a meaningful direction to you.

[867] And this is maybe adjacent, but when there was a study done, I think it was in Australia, where they looked at suicide letters.

[868] And in the suicide letters, the sentiment of men in those suicide letters was they didn't feel worthy.

[869] They didn't feel like they were worth it.

[870] They didn't feel like they were needed by their families.

[871] And this is much of what caused their psychological state.

[872] And I wonder in a world of abundance where we, you know, a lot of these AI billionaires are telling us that we're going to have so much free time and we're not going to need to work.

[873] If there is at all going to be a crisis of meaning, a mental health problem.

[874] I mean, there already is.

[875] And it doesn't require AI and it's going to get worse.

[876] I don't know what to do about it because.

[877] Essentially, as human beings, we are built, like all organisms, to find opportunity and figure out how to exploit it.

[878] That's what we do.

[879] And the world you're describing is really the opposite of that.

[880] It's one where you're effectively having your biological needs at the physiological level satisfied, and there isn't an obvious place for your spare time, if that's what you end up with, to be...

[881] utilized in something that there's no place to strive.

[882] And I do imagine almost at best what would happen is you have people who are being sustained by a universal basic income and then parasitized.

[883] Whatever currency they have to spend, somebody will be targeting it and they will be targeting it with an AI augmented system that spots their defects of character.

[884] Again, we're already living in this world, but it will be that much worse when the AI is figuring out, you know, what kind of porn to target you with specifically.

[885] That's a nightmare scenario.

[886] And I do think it would be worth our time as a species to start considering if we are about to find ourselves in this situation and we find some way of dealing with the basic needs of the large number of people who are going to be sidelined.

[887] What would a world have to look like in order for them to have real meaning, not pseudo meaning, not something that, you know, superficially, you know, a video game is not meaning, even if it feels very meaningful in the moment.

[888] I think that would be a worthy investment for us to figure out how to produce it.

[889] But frankly, I'm not expecting us to either have that conversation or get very far down that road.

[890] I think it's much more likely that we will squander the wealth dividend that will be produced by.

[891] by AI.

[892] Interestingly, you also see in Western countries that when we get more abundance, we start having less kids.

[893] And we're already seeing this sort of population decline in the Western world, which is kind of scary.

[894] And I think it's often associated with affluence.

[895] The more money someone makes, the less likely they are to want to have children, the more they try and protect their freedoms.

[896] But also on this point of AI, relationships are hard.

[897] My girlfriend is happy sometimes and not happy other times.

[898] I have to like, you know, go through that struggle with her of like working on the relationship.

[899] Children are hard.

[900] And if we are optimizing ourselves and, you know, much of the reason that I sustain the struggle with my girlfriend is I'm sure from some evolutionary reason because I want to reproduce and I want to have kin.

[901] But if I didn't have to deal with the struggle that comes with human relationships, romantic or platonic, there's going to be a proportion of people that actually choose that outcome.

[902] And I wonder what's going to happen to...

[903] birth rates in such a scenario, because we're already struggling.

[904] We're already in a situation where we used to be having five children per woman in the 1950s to about two in 2021.

[905] And we're seeing a decline.

[906] If you look at South Korea, their fertility rate has fallen to 0 .72, the lowest recorded globally.

[907] And if this trend continues, the country's population could half by 2 ,100.

[908] So, yeah, relationships, connections.

[909] And also, I guess we've got to overlay that with the loneliness epidemic, which is they promised us social connection when social media came about.

[910] When we got Wi -Fi connections, the promise was that we would become more connected.

[911] But it's so clear that because we spend so long alone, isolated, having our needs met by Uber Eats drivers and social media and TikTok and the internet, that we're investing less in the very difficult thing of like going and making a friend and like going and finding a girlfriend.

[912] Young people are having sex less than ever before.

[913] Everything that is associated with the difficult job of making in real life connections seems to be falling away.

[914] I'll make the case that everything that we've discussed here, all the negative things around loneliness, around meaning, they're already here.

[915] And I don't think blaming technology for all of it is the right thing.

[916] I think there are a lot of things that happened because of existing human impulses and motivations.

[917] Well, I wanted to go back to where you started.

[918] Because I do think that this maybe is the fundamental question.

[919] Why is it that we are already living in a world that is not making us happy?

[920] And is that the responsibility of technology?

[921] And I don't think it's exactly technology.

[922] Human beings among our gifts are fundamentally technological, whether we're talking about quantum computing or flintknapping an arrowhead.

[923] What has happened to us that has created the growing, spreading, morphing dystopia is a process that Heather and I in our book, A Hunter -Gatherer's Guide to the 21st Century, call hypernovelty.

[924] Hypernovelty is the fact of the rate of change outpacing our capacity to adapt.

[925] to change.

[926] And we are already well past the threshold here where the world that we are young in is not the world that we are adults in.

[927] And that mismatch is making us sick across multiple different domains.

[928] So the question that I ask is, is the change that you're talking about going to reduce the rate of change?

[929] In which case, we could build a world that would start meeting human needs, better open opportunities for pursuing meaningful work?

[930] Or is it going to accelerate the rate of change, which is, in my opinion, guaranteed to make us worse off?

[931] So if it was a one -time shift, right?

[932] AI is going to dawn.

[933] It's going to open all sorts of new opportunities.

[934] There's going to be a tremendous amount of disruption.

[935] But from that, we'll be able to build a world.

[936] Is that world going to be stable?

[937] Or is it going to be just one event horizon after the next?

[938] If it's the latter, then...

[939] it effectively says what it does to the humans, which is it's going to dismantle us.

[940] When I look out at society, I go, okay, it's having a negative impact.

[941] When I look at individual use cases, it's having a profoundly positive impact, including for me, it's having a very positive impact.

[942] So it's one of these things where I wonder, what is it that we need to teach people at school?

[943] so that they understand the world that we're going into.

[944] Because one of the biggest issues that we're having is that we're sending kids to school with this blueprint, this template that they're going to have this long arc career that no longer exists, that essentially we're treating them like learning LLMs and we're saying, okay, we're going to prompt you, you're going to give us the right answer, you're going to hallucinate it if possible.

[945] And then we go, okay, now go off into the world.

[946] And they go, oh, but wait a second, I don't know how money works.

[947] I don't know how society works.

[948] I don't know how my brain works.

[949] I don't know how I'm meant to handle this novelty problem.

[950] I'm not sure how to approach someone in a social situation and ask if they want to go on a date.

[951] So all the important things that actually are the important milestones that people want to be able to hit and that technology can actually have an impact on.

[952] We get no user manual.

[953] So I think one of the biggest things that has to happen is we have to equip young people all through school to actually prepare them for the world that's coming or the world that's here.

[954] Well, on the one hand, I think you outlined the problem very well.

[955] Effectively, we have a model of what school is supposed to do that, you know, at best was sort of a match for the 50s or something like that.

[956] And it woefully...

[957] misses the mark with respect to preparing people for the world they actually face.

[958] If we were going to prepare them, I would argue that the only toolkit worth having at the moment is a highly general toolkit.

[959] The capacity to think on your feet and pivot as things change is the only game in town with respect to our ability to prepare you in advance.

[960] Maybe the other auxiliary component to that would be teaching you what we know, which is frankly not enough, about how to live a healthy life, right?

[961] If we could induce people into the kinds of habits of behavior and consumption of food and then train them to think on their feet, they might have a chance in the world that's coming.

[962] But the fly in the ointment is we don't have the teachers to do it.

[963] We don't have people who know.

[964] And that is the question, is could the AI actually be utilized in this manner to actually...

[965] induce the right habits of mind for people to live in that world.

[966] I spend a lot of time in education technology.

[967] One thing that is, as we say on the internet, a black pill about education in general, education intervention, is there's a lot of data that shows that there are very little interventions you can make in education to generate better outcomes.

[968] And so there's been a lot of experiment around pedagogy, around how to configure the classroom that have resulted in very marginal improvements.

[969] There's only one intervention, and this has been reproduced many times, that creates two sigma, two standard deviation positive outcomes in education, meaning you're better than 99 % of everyone else.

[970] One -on -one tutoring.

[971] I thought so.

[972] I was going to say smaller classrooms and personalization.

[973] One -on -one tutoring, yeah.

[974] And by the way, if you look, someone also did a survey of all the geniuses, the understanders of the world, and found that they all had one -on -one tutoring.

[975] They all had someone in their lives that took interest in them and tutored them.

[976] So what can create one -on -one tutoring opportunity for every child in the world?

[977] AI.

[978] My kids use it, and it's incredible.

[979] As in like...

[980] they're interacting and it's adapting to their speed and it's giving them different analogies to work with.

[981] So like my son was learning about division and it's asking him to smash glass and how many pieces he smashes it into with this hammer and it's saying things like, no, Xander, go for it, really smash it.

[982] And he's loving it, right?

[983] Is that synthesis?

[984] Yeah.

[985] Yeah, I'm an investor in this guy.

[986] It's great to watch that.

[987] simulated one -on -one tutoring because it's talking to him.

[988] It's asking him questions.

[989] Brett, you're an educator.

[990] You spent much of your life teaching people in universities.

[991] How do you receive all of this?

[992] Well, on the one hand, I agree that the closer to one -to -one you get, the better.

[993] But I also personally believe that zero -to -one is best.

[994] And what I mean by that is Part of what's gone wrong with our educational system is that it is done through abstraction.

[995] And effectively, the arbiter of whether you have succeeded or failed in learning the lesson is the person at the front of the room.

[996] And that's okay if the person at the front of the room is truly insightful.

[997] And it's terrible if the person at the front of the room is lackluster, which happens a lot.

[998] So what doesn't work that way?

[999] is interaction with the physical world in which nobody has to tell you whether you've succeeded or failed.

[1000] If you're faced with an engine that doesn't start, you can't argue it into starting.

[1001] You have to figure out what the thing is that has caused it to fail, and then there's a great reward when you alter that thing and suddenly it fires up.

[1002] So I'm a big fan of being as light -handed as possible and as concrete as possible in teaching.

[1003] In other words...

[1004] when I've done it, and not just with students, but with my own children, I like to say as little as possible.

[1005] And I like to let physical systems tell the person when they've succeeded or failed.

[1006] And that creates an understanding you can extrapolate from one system to the next.

[1007] And you know that you're not just extrapolating from one person's misunderstanding, you're extrapolating from the way things actually work.

[1008] So I don't know if AI can...

[1009] be leveraged in that context.

[1010] My sense is there's probably a way to do it, but one would have to be deliberate about it.

[1011] Especially with robotics and humanoid robots.

[1012] Actually, that is the place where you can do this is with robotics.

[1013] It seems to me, yeah.

[1014] Well, robotics will teach you the physical computing part of it.

[1015] And then the question is, how do you infuse this with AI so that it is, you know.

[1016] provokes you out of some eddy where you're caught and moves you into the ability to solve some next level problem that you wouldn't have found on your own.

[1017] What do you think should be taught in the classroom with everything that you now know?

[1018] Well, you're all fathers here.

[1019] You all have your own children.

[1020] So it's a good question for you.

[1021] How old are your kids?

[1022] How old are your kids?

[1023] Three and five.

[1024] 19 and 21.

[1025] And six, seven, and 10.

[1026] My children are very young, but we already do use AI.

[1027] And I sit down with them in front of Repl.

[1028] And we generate ideas and make games.

[1029] And I would say, you know, what Brett said about generality is very important.

[1030] The ability to pivot and kind of learn skills quickly.

[1031] Being generative is very, very important.

[1032] Having a, you know, a fast pace of generating ideas and iterating of those ideas.

[1033] We sit down in front of ChatGPT and my kid imagines scenario.

[1034] What if there's a cat on the moon?

[1035] And then what if the moon is made of cheese?

[1036] And what if there's a mouse inside it?

[1037] And so we keep generating these variations of these different ideas.

[1038] And I find that makes them more imaginative and creative.

[1039] Rule number one that I tell my kids is stay away from porn at all costs.

[1040] I'd rather you have a drug problem than a porn problem, and I actually mean that.

[1041] I think porn is more dangerous to the human being, as bad as a drug problem is.

[1042] But when we get to the question of how to confront the world and the things that you're going to be expected to do in the workplace and all of that, my point to them is you are facing the dawning of the age of complex systems.

[1043] that you are going to have to interact with.

[1044] And in the age of complex systems, you have to understand that you cannot blueprint a solution.

[1045] And you have to approach these systems with an upgraded toolkit of humility because the ability of the system to do something you don't predict is much greater than a highly complicated system.

[1046] So you have to anticipate that and be very sensitive to the fact that what you intended to happen is not what's going to happen.

[1047] So you have to monitor.

[1048] the unintended consequences of whatever your action is, and that there are really two tools which work, one of which you just mentioned, which is the prototyping.

[1049] You prototype things.

[1050] You don't imagine that I know the solution to this and I'm going to build it.

[1051] You imagine I think there's a solution down there.

[1052] I'm going to make a proof of concept, and then I'm going to discover what I don't know, and I'm going to make the next version, discover what I don't know, and eventually you may get to something that actually truly accomplishes the goal.

[1053] So prototyping is one thing.

[1054] And also, instead of using the blueprint as the metaphor in your mind, navigate.

[1055] You can navigate somewhere.

[1056] And, you know, the way I think of it is a surfer is in some ways mastering a complex system.

[1057] But they're not doing it by planning their days surfing down the waves.

[1058] You can't do that.

[1059] What you can do is you can...

[1060] be expert at absorbing feedback and navigating your way down the wave and that that's the right approach for a complex system.

[1061] Nothing else is going to work.

[1062] And so I guess the final piece is general tools, always no specialization.

[1063] This is the age of generalists and invest in those tools and they will pay.

[1064] So the guiding philosophy for me is to produce high agency generalists.

[1065] So ultimately, I want them to be motivated self -starters and have a wide general toolkit.

[1066] I imagine them very much what you imagine, which is instructing robots, instructing agents, coming up with ideas.

[1067] And I imagine them having a very high -velocity life where they may be writing a book, organising a festival, having a podcast, starting a business.

[1068] and being part of somebody else's business all at once.

[1069] As they are of the ADHD.

[1070] Yeah, right.

[1071] So the high agency generalist is the kind of guiding philosophy.

[1072] Some of the things that we do is like we do chess, we do Brazilian jiu -jitsu, we do dancing, we do acting classes, playing in nature, entrepreneurship, understanding that you can start a lemonade.

[1073] We just did lemonade stands, which was amazing.

[1074] We sold lots of lemonade on the street.

[1075] So those kind of...

[1076] things and jumping from one thing to the next thing, but also trying to avoid too many screens and forcing them into making stuff from what's going on around the house.

[1077] Some distinctions that we try and give them is the difference between creating and consuming, because I think AI has this superpower of making you a hyper consumer or a hyper creator.

[1078] And if you don't understand the distinction between creation and consumption, you end up falling into the consumption trap, whether it be porn or just news or things that feel like you're productive but you're actually just consuming stuff.

[1079] Wouldn't that be the most successful AI, the one that plays with my dopamine the most?

[1080] Yeah, and makes you think that you're achieving something when you're actually just consuming something.

[1081] So trying to give them the understanding that there is this difference in their life between creation and consumption and to be on the creation side.

[1082] I started my first business at 12 years old and I started more businesses at 14, 15, 16, 17 and 18.

[1083] And at that time, what I didn't realise is that being a founder with no money meant that I also had to be the marketeer, the sales rep, the finance team, customer service and the recruiter.

[1084] But if you're starting a business today, thankfully there's a tool that wears all of those hats for you.

[1085] Our sponsor today, which is Shopify.

[1086] Because of all of its AI integrations, using Shopify feels a bit like you've hired an entire growth team from day one, taking care of writing product descriptions, your website design, and enhancing your product's images.

[1087] Not to mention the bits you'd expect Shopify to handle, like the shipping, like the taxes, like the inventory.

[1088] And if you're looking to get your business started, go to shopify .com slash Bartlett and sign up for a $1 per month trial.

[1089] That's shopify .com slash Bartlett.

[1090] The thing that we, I think, all agree on is that this is inevitable.

[1091] Do you agree with that, Brett?

[1092] I think it's sad that it is inevitable, but at this point it is.

[1093] And what part of it do you find sad?

[1094] We have squandered a long period of productivity and peace in which we could have prepared for this moment.

[1095] And our narrow focus on...

[1096] Competition has created a fragile world that I'm afraid is not going to survive the disruption that's coming.

[1097] And it didn't have to be that way.

[1098] This was foreseeable.

[1099] I mean, frankly, the movie 2001, which came out the year before I was born, anticipates some of these problems.

[1100] And, you know, we treated it too much like education.

[1101] I mean, like.

[1102] entertainment and not enough like education so we are now you know we've had the ai era opened without a discussion about its implications for humanity there is now for game theoretic reasons no way to slow that pace because as you point out if we restrain ourselves we simply put the AI in the hands of our competitors.

[1103] That's not a solution, so I don't advocate it.

[1104] But there's a lot more preparation we could have done.

[1105] We could have recognized that there were a lot of people in jobs that were about to be obliterated, and we could have thought deeply about what the moral implications were and what the solutions at our disposal might have been.

[1106] And having not prepared, it's going to be a lot more carnage than it needed to be.

[1107] Amjad, I heard you say a second ago that what we should be talking about is how we deal with job displacement.

[1108] Do you have any theories?

[1109] If you were prime minister or president of the world and your job was to deal with job displacement, let's just say in the United States, how would you go about that?

[1110] The first thing I would do is teach people about these systems, whether it's programs on the TV or...

[1111] outreach or what have you, just trying to get people to understand how ChatGPT works, how these algorithms work.

[1112] And as the new jobs arrive, I think there's going to be an opportunity for people to be able to detect that this job requires this set of skills.

[1113] I have this kind of experience.

[1114] And although my experience are potentially outdated, I can repurpose that experience to do that job.

[1115] I'll give you an example.

[1116] A teacher, his name is Adil Khan.

[1117] He started using, at the time, GPT -3 and felt like he does amazing work as tools for teachers or even potentially a teacher itself.

[1118] So he learned a little bit of coding and he went on Replit and he built this company.

[1119] Just two years later, they're worth hundreds of millions of dollars.

[1120] Obviously, not everyone will be able to create businesses of that scale.

[1121] But because you have an experience in a certain domain, you'll be able to build the next iteration of that user technology.

[1122] So even if your job was displaced, you'll be able to figure out what potentially comes after that.

[1123] people's expertise that they built, I don't think they're all for waste, even if your job went away.

[1124] You can never really predict what jobs are coming.

[1125] I mean, I think of this crazy situation where I tell my grandfather, what is a personal fitness trainer?

[1126] And his mind would be blown by this idea that, well, okay, I don't really want to go to the gym, so I have to make an appointment and pay someone to go to the gym and meet with me there.

[1127] And then he stands there and tells me to lift heavy things that I don't really want to lift.

[1128] And then he counts them and tells me that I've done a good job.

[1129] And then I put the heavy things down.

[1130] And then at the end of that, I feel really good and I pay him a bunch of money.

[1131] My grandfather would be like, what on earth have you been scammed?

[1132] So we can never predict what this future of jobs would look like.

[1133] Even just 20, 30, 40 years apart, the jobs.

[1134] rapidly and convincingly just morph into something else.

[1135] I think it's very dangerous, the idea that we need to focus on skills.

[1136] I think the future is not in skills.

[1137] Skills are being replaced.

[1138] It's this idea that the education system has to stop being compartmentalised and has to be a lifelong learning approach.

[1139] The Department of Education needs to be seeing people as lifelong learners who are constantly disrupted and need re -education.

[1140] Interesting.

[1141] That's going to be a thing.

[1142] The Department of Education needs to start, as a kid and go right through to maybe 70.

[1143] Does the Department of Education have a role anymore at all?

[1144] Depends on your definition of education.

[1145] I think if you're trying to teach kids or if you're trying to teach kids to, you know, remember facts and figures from a history book, then no. But if it's about coaching, mentoring, being displaced, finding the next thing, and maybe if it's AI driven and all of those kind of things, then...

[1146] It's a different paradigm shift around what education is and what its purpose is.

[1147] And if we see it as a fluid thing where we wave into an opportunity and then wave back into education, spotting a new opportunity and then back here, if we're learning rather than skills but we're learning tools.

[1148] So it's a tools -based education as opposed to a skills -based education.

[1149] The purpose of education for most of human history was about virtue, about becoming a great person who had good judgment and who had good values, and we don't really do much of that anymore.

[1150] But I think if we essentially said if we get back to what is, if we ask the question what is the purpose of education and where does it fit in our lives and what time frame does it go for, and then we just trust that people are going to come up with.

[1151] weird and wonderful jobs.

[1152] You know, this sounds crazy, but also, and this is a weird analogy, my cat is incredibly happy.

[1153] How do you know?

[1154] It demonstrates all the characteristics of being a happy cat.

[1155] And it lives in a world of super intelligence as far as it's concerned.

[1156] So there's this house and food just magically happens.

[1157] It has no idea that there's this Google calendar that runs a lot of things that happen around it.

[1158] The food gets delivered.

[1159] Money is magically made by something that is inconceivably more intelligent than the cat.

[1160] And yet the cat has evolved to be living this life of purpose and meaning inside the house.

[1161] And as far as it's aware, it's got a great life.

[1162] But you have the power at any moment, if you're having a bad day, to do something not so pleasant to that cat.

[1163] And it can't really reciprocate that.

[1164] Exactly, but what's in it for me to hurt the cat?

[1165] Because in this analogy...

[1166] You might want to move house and the landlord doesn't allow cats, so you've got a decision to make.

[1167] There are things that the cat is highly disrupted by due to no fault of the cat.

[1168] I get it.

[1169] But as far as cat existence goes and the history of cats...

[1170] If you were to ask that cat, do you want to trade places with any of the other cats that came before you, it would probably say, I don't want to take the risk because all the other cats had to fend for themselves in a way that I don't have to.

[1171] It's very possible that we end up living in a life a lot like the house cat in the sense that from our perspective, we're extremely, like we're having very interesting lives and purpose meaning and just there's this massive higher intelligence that's just running stuff.

[1172] We don't know how it works, but it doesn't really matter how it works.

[1173] We are the beneficiaries of it and it's doing important things and we're enjoying being house cats in its life.

[1174] I have a few things to say about this.

[1175] One, I'm pretty sure your cat's not as impressed with your capacity as you are or as you think he is.

[1176] I just know cats well enough to be pretty sure of that.

[1177] It looks down on me. Yeah, you're right.

[1178] I think it's a fair – point that there is an existence.

[1179] And actually, you know, pets really do have it.

[1180] If they have loving owners, they really do have it pretty great.

[1181] And I would also point out that there's a way in which we already are this way.

[1182] Most of us do not understand the process that results in electricity coming out of the walls of our house or the water that comes out of the tap.

[1183] And we're pretty much okay with the fact that somebody takes care of that and we can busy ourselves with whatever it might be.

[1184] But the place that I find something...

[1185] troubling in your description, is that you say that the nature of what we do is to deal with the fact that jobs are always being upended.

[1186] That's a very new process.

[1187] That is the hyper novelty process.

[1188] It used to be that it was only very rarely that a population had a circumstance where you didn't effectively do exactly what your immediate ancestors did, right?

[1189] In general, you took what the jobs were, you Picked something that was suited to you and you did that thing.

[1190] Intergenerationally.

[1191] Intergenerationally.

[1192] And the point is we've now gotten to the point where.

[1193] Even within your lifetime, what is possible to get paid for is going to shift radically in ways that nobody can predict.

[1194] And that is a dangerous situation.

[1195] Like probably every two years, like two or three years.

[1196] Right.

[1197] And so maybe there's some model by which we can surf that wave and you can learn a generalist toolkit and, you know, that your survival doesn't depend on your being able to, you know, switch up every two years and never miss a beat.

[1198] Or maybe we can't.

[1199] But I do think it is worth asking the question, if the rate of technological change has taken us out of the normal human circumstance of being able to deduce what you might do for a living based on what your ancestors did and put us in a situation where what your ancestors did is going to be perfectly irrelevant no matter what, that that is effectively a choice that has been made for us.

[1200] And we could choose to slow the rate of change so that we would live in some kind of harmony where our developmental environment and our adult environment were a match.

[1201] Now, as a biologist, I would argue if we don't do something like that, this is a matter of time.

[1202] How would we slow the rate of change?

[1203] Well, I mean, you can be the Amish, right?

[1204] You can be the Amish and live in your own communities.

[1205] And I would assume some people would want that.

[1206] Well, you know, when Heather and I wrote our book, I wanted the first chapter to be, are the Amish right?

[1207] And the answer is they can't be exactly right because they picked.

[1208] an arbitrary moment to step off the escalator.

[1209] But are they right that there's something dangerous about this continuing pattern of technological change?

[1210] Clearly they are.

[1211] What do the Amish do for anyone that doesn't know?

[1212] The Amish live as if it was, what, 1880?

[1213] 1850 or something.

[1214] They don't use cars.

[1215] I think they do have phones, but they do not have electricity.

[1216] Basically, they voluntarily accept a technology.

[1217] They're basically a Luddite community, and they have turned out to fare surprisingly well against many of the things that have upended modern folks.

[1218] COVID, one of them, right?

[1219] Yeah, COVID.

[1220] They did beautifully.

[1221] Quite happy people.

[1222] Very low autism rates.

[1223] They have all sorts of advantages.

[1224] So anyway, I'm not arguing that we should live like the Amish.

[1225] I don't see that.

[1226] But I do think the idea that they had an insight, which was you need to step off that escalator because you're just going to keep making yourself sicker, is probably right.

[1227] Now, maybe this is a one -time shift.

[1228] We've stepped over the event horizon.

[1229] We are going to be living in the AI world.

[1230] And maybe if we're careful about it, we can figure out.

[1231] how to turn that landscape of infinite possibility that you're describing into a place that doesn't change, that you always have the opportunity to decide what needs to be done, but that living over that event horizon is not an ever -changing process.

[1232] It's just the next frontier.

[1233] I do want to also propose or ask the question, when we talk about a hyper -changing world, isn't it harder for older people to learn?

[1234] because of the way that the brain works in terms of processing speed and memory flexibility.

[1235] So I was wondering if you're going to get a situation where my father, because of his brain and the reduced memory flexibility and processing speed that happens when you're older, is going to struggle significantly more than my niece, who can seem to learn.

[1236] I mean, my niece knows five languages and she's seven.

[1237] or something crazy like that.

[1238] I mean, the brain is much more plastic, isn't it?

[1239] And that goes back to our evolutionary psychology, our evolutionary history, which you know much more than I do about, of we're meant to learn our lessons when we're young, use that information for a lifetime.

[1240] But if that information is changing quickly...

[1241] Well, this is exactly what I'm pointing to.

[1242] It is not normal for your developmental environment to...

[1243] fail to prepare you for your adult environment.

[1244] The normal thing is as a young person, you take on ever more of the responsibilities of the adult environment.

[1245] And then at some point, you know, in a properly functioning culture, there's a rite of passage.

[1246] You go into the bush for 10 days.

[1247] You come back with, you know, a large, you know, game animal.

[1248] And now you're an adult and you take that program that you've been building and you activate it.

[1249] And that is normal.

[1250] You're a lot happier person.

[1251] You're a lot more fulfilled if your life has that kind of continuity to it.

[1252] And I'm not against the idea that we have enabled ourselves to do things that can't be done if that's the limit, but we have also harmed ourselves gravely.

[1253] And I would like to somehow pry apart our ability to improve our well -being from are self -inflicted wounds that come from this never -ending pace of change.

[1254] And I don't know if it's possible, but I think it's a worthy goal.

[1255] Something amusing, I don't know if it's exactly a counterpoint, but during COVID especially and through the recent technological change, some people have started living closer.

[1256] to the more ancestral environment.

[1257] So people whose jobs are online.

[1258] Some of my friends went and built communities.

[1259] like collectives where they, you know, live and they create farms and they eat and they have like an email job.

[1260] They do their email jobs for five hours and go out and they all have children.

[1261] And it's a fascinating life.

[1262] And there was so much rethinking in Silicon Valley about how we live.

[1263] And there's a bunch of startups that are trying to create cities where they're like, okay, we know that we're suffering because our cities are not really walkable.

[1264] And there's so many reasons why we're suffering.

[1265] First, we're not getting the movement.

[1266] Second, there's a social aspect of walkable city where you're able to interact with people.

[1267] You'll make friends by just happening to be in the same place as others.

[1268] Let's actually build walkable cities.

[1269] And if we want to transport faster, we'll have these self -driving cars on the perimeter of the city that are going around.

[1270] And I think there are ways in which technology can afford us to live, in a way that reverses, I guess, in a more local way.

[1271] I like that vision.

[1272] But I also am aware that there's a different vision, right?

[1273] You see people in Palo Alto, for example, actually exerting very strong controls on how much their children are exposed to phones.

[1274] I live in Palo Alto.

[1275] Yeah.

[1276] So you see that.

[1277] On the other hand, what I am worried about is that the The elites of Palo Alto don't realize that what they're doing is they're figuring out how to reduce the harm to their own families as they're exporting the harm to the world of these technologies that for everybody else are unregulated.

[1278] And so the question is, can we bring everybody along if the AI revolution is going to alter our relationship?

[1279] to work and everything else.

[1280] Can we bring everybody along so that at the end of this process, instead of saying, well, you know, it's a shame that, you know, three billion people were sacrificed to this transition, but progress is progress.

[1281] We can really say, well, we figured it out.

[1282] And everybody now is living in a style that is closer to their programming and closer to the expectations of their physical bodies.

[1283] You know, if that were true, then I would love to be wrong.

[1284] in my fears about what's coming.

[1285] But unfortunately, the market is not going to solve this problem without our being deliberate about forcing it to.

[1286] What's your biggest fear?

[1287] Like when you say my fears about what's coming, what's the picture that comes in your mind?

[1288] It's a whole different topic, actually.

[1289] My fear stemming from technology and AI is that this is a runaway process, and that that runaway process is going to interface very badly with some latent human programs that, in effect, the need for workers largely disappears, and the people who are at the head of the processes that result in that elimination for the need for workers start talking about useless eaters.

[1290] Maybe they come up with a new term this time.

[1291] Thin the herd.

[1292] Yep, or they allow it to be thinned or something.

[1293] Brett, I've heard you talk about the five.

[1294] key concerns you have or the five key threats you have before.

[1295] Could you name those five?

[1296] So the first one is the one I worry least about.

[1297] I don't worry zero about it, but I worry least about it, which is the malevolent AI that the doomers are so focused on.

[1298] The second one is the idea that an AI can be misaligned, not because it has divergent interest, but because it just misunderstands what you've asked.

[1299] These autonomous agents, the famous example is you asked them to produce as many paperclips as possible, and they start liquidating the universe to make paperclips.

[1300] And, you know, it's a sorcerer's apprentice kind of issue.

[1301] The third one, I would say, is actually all of the remainder of them, I would say, are guaranteed.

[1302] And the third of them is the derangement of human intellect, that we are already living in a world where it's very difficult to know what the facts even mean, right?

[1303] The facts are so filtered and we are so persuaded by algorithms that it's, you know, our ability to be confident, even in the basic facts, even within our own discipline sometimes is at an all time low and it's getting worse.

[1304] And that problem takes a giant leap forward at the point that you have the ability to generate undetectable deep fakes, right?

[1305] That's going to alter the world very radically when the fact that you're looking at videotape of somebody robbing a bank doesn't mean that they robbed a bank or that a bank was even robbed.

[1306] So anyway, I call this.

[1307] We deal with this a lot, by the way.

[1308] I think every single week, every single week I send to my, I have a chat of people that just are now basically spending, I'd say 30 % of their time dealing with deep fakes of me. doing crypto scams, inviting people to Telegram groups, and then asking them for credit card details.

[1309] We had one on X. I think you probably saw it, Dan, didn't you, of me?

[1310] But someone was running deep fake ads on X of me. And it wasn't just one ad.

[1311] It was like swatting flies.

[1312] There was 10 of them.

[1313] messaged them to X. And then there was 10 more.

[1314] Then the day after there was 10 more.

[1315] Then the day after there was 10 more.

[1316] Then it started happening on Meta.

[1317] So it's a video of me basically asking you to come to a Telegram group where people are being scammed and audience members of mine are being scammed.

[1318] And when I send them to Meta, they thankfully remove them.

[1319] But then there's five more.

[1320] So then I went on LinkedIn yesterday and my DMs are, Steve, by the way, there's this new scam.

[1321] And I actually, at this point, I need someone full time.

[1322] Just sending this over to Meta?

[1323] I'm the same, but on a small scale.

[1324] Every week, it's, did you really message me on Facebook asking me for my crypto wallet and blah, blah, blah.

[1325] My least favorite one's when the single mother messages me saying that she just paid 500 pounds of her money and how devastated she is.

[1326] And I feel this moral obligation to give her her money back because she's fallen for some kind of scam.

[1327] That was me. It was my voice.

[1328] It was a video of me telling her something.

[1329] And I don't know how you deal with that, but sorry, do continue.

[1330] Well, I mean, that's actually on the list here.

[1331] The massive disruption to the way things function, both because people are going to be unemployed in huge numbers and because those who are not abiding by our social contract are going to find themselves empowered more than the people who do.

[1332] So in this case, not only is this poor woman, you know, now out 500 bucks for whatever the scam was, but you've also been robbed.

[1333] whether or not you pay her back for the thing that she thought she purchased, your credibility is being stolen by somebody and you have no capacity to prevent it.

[1334] This has happened to me also, and it is profoundly disturbing.

[1335] And it is only one of a dozen different ways that AI enables those who are absolutely willing to shrink the pie from which we all derive in order to enlarge their slice.

[1336] You know, there are innumerable ways that this can happen.

[1337] And I think people do not see it coming.

[1338] They don't understand how many different ways they are going to be robbed every bit as surely as if somebody was printing money.

[1339] And then the last one is that this just simply accelerates demographic processes that do potentially result in the unleashing of technologies that pre -existed AI.

[1340] This can easily result in an escalation into wars that turn nuclear.

[1341] So anyway, I think that list could probably be augmented at this point now that we've spent a little time in the AI era.

[1342] We can begin to put a little more flesh on the bones, both of what is possible in this era and what we should fear.

[1343] One of those, you mentioned the truth, the problem of truth.

[1344] Would you say, A thought experiment.

[1345] Someone today, like an average person, college educated person, are they more propagandized or led astray than someone in Soviet Russia?

[1346] Well, I don't know because I didn't live in Soviet Russia.

[1347] But my understanding from people who did was that there was a wide awareness.

[1348] that the propaganda wasn't true.

[1349] It doesn't mean they knew what to believe, but there was a cynicism, which is one of my fears here, is that you're really stuck choosing between two bad options.

[1350] In a world where you can't tell what is true, you can either be overly credulous and be a sucker all the time, or you can become a cynic and you can be paralyzed by the fact that you just don't believe anything.

[1351] But neither of those is a recipe for...

[1352] Do you think Google search first, and maybe now Chad Chippity, has helped people?

[1353] more or less to find truth.

[1354] I think, and it's not ChatGPT exactly, but all of the various AI engines have briefly enhanced our capacity to know what's true because in fact they allow us to see through the algorithmic manipulation because the AI is not well policed.

[1355] You can get it to recognize patterns that people will swear are not.

[1356] true.

[1357] And so anyway, a lot of us have found it useful in just simply unhooking the gaslighting.

[1358] So that's been very positive.

[1359] But I also remember the early days of search.

[1360] And search used to be a matter of There are some pages out there.

[1361] I don't know where they are.

[1362] Here's a mechanized something that's looked through this stuff and just point me at the direction of things that contain these words, right?

[1363] Before the algorithmic manipulation started steering us into believing pure nonsense because somebody who controlled these things decided it was useful for us to believe those things.

[1364] So my guess is at the moment, AI is enhancing our ability to see more clearly, but that really depends on some kind of agreement to protect that capacity that I'm not aware of us having.

[1365] Angela, you implying there that AI will protect us from the AI, i .e. the woman that got scammed in my audience, the platforms would have a tool built in which would be able to identify shortly that that is not me and the ad has been launched by someone in another country potentially.

[1366] And then also when she starts being asked for her credit card details in such a way on Telegram.

[1367] 10 minutes later, the system will able to understand there that this is probably a scam at that touch point too.

[1368] And it will also be the defense, not just the offense.

[1369] First thing, question, is meta incentivized to solve this problem?

[1370] Yes.

[1371] Yes.

[1372] And so meta is probably actively working on AIs.

[1373] And again, it's going to be a cat and mouse game, like every abuse that happens out there.

[1374] I think that the market will naturally respond to things like that.

[1375] And the same way that we install antiviruses, as annoying as they are, I think we'll install AIs on our computers that will allow us to at least help us sort the fake from the truth.

[1376] Well, but let's take the example you say.

[1377] Is meta incentivized to solve this problem?

[1378] Superficially, it seems that it should be.

[1379] How many times in recent history have we watched a corporation cannibalize its own business over what at best is the bizarre desires of its shareholders?

[1380] Why was X throwing off people with large accounts or Facebook or Google?

[1381] It would seem that you would expect, based on the market choosing search engines or social media sites, you would expect these companies to be...

[1382] absolutely mercenary and say, if Alex Jones has a big audience, who are we to say?

[1383] That's what I would have expected.

[1384] Instead, you had these companies policing the morality of thought, even though it reduced the size of the population using the platforms.

[1385] I have a hard time explaining why that happened, but I have every reason to expect the same thing will happen with AI.

[1386] What are you excited about with AI?

[1387] What's your...

[1388] Your optimistic take, because at the start of this conversation, you said that there's infinite ways that it could improve our lives and there's 10 times more ways that it could hurt our lives.

[1389] But let's investigate some of those ways that it could drastically improve our lives.

[1390] There's a couple of different ways.

[1391] One, we have, as we mentioned before, a dearth of competent teachers and professors.

[1392] And that is a...

[1393] problem that will take three generations at least to solve if what we're going to do is start tomorrow and start educating people in the right way that would make them competent to stand at the front of a room and educate.

[1394] But if we can augment that process, if we can leverage a tool like AI so that a small number of competent teachers can maybe reach a larger number of pupils, that's plausible, I think.

[1395] Second thing is we have a tremendous number of problems that are obstacles to us living well on this planet that AI might be able to manage that human intellect alone cannot, right?

[1396] Just in the same way that, you know, compute power can calculate things at a rate that human beings can't keep up.

[1397] And there are certain things you want calculated very well.

[1398] There are also some reasoning problems.

[1399] You could imagine that instead of having static laws, that govern behavior poorly because they get gamed, that you could have a dynamic interaction.

[1400] You could specify a...

[1401] an objective of something like a law.

[1402] And then you could monitor whether or not a particular intervention successfully moved you in the direction that you were hoping to go or did something paradoxical, which happens all the time.

[1403] And you could basically have governance that is targeted to navigation and prototyping rather than to specifying a blueprint for how we are to live.

[1404] So we wouldn't need politicians.

[1405] At the moment, we're stuck with...

[1406] you know, constitutional protections that are as good as has been constructed and still inadequate to modern realities.

[1407] Dan, what are you excited about with AI?

[1408] From an individual level, but also from a societal level?

[1409] Yeah, well, the big ones are healthcare and education.

[1410] I mean, it's ridiculous that you are sitting there in pain, having had an MRI, and there just hasn't been someone to look at that MRI yet and tell you what to do.

[1411] And that could...

[1412] easily be solved.

[1413] There's all sorts of healthcare issues where, and also not only that, throughout the entire world, there are places that just don't have general practitioners and they don't have, you know, medical advisors.

[1414] And, you know, the breakthroughs in global healthcare will be phenomenal.

[1415] And the breakthroughs in global education could be transformational on the planet.

[1416] I'm excited at an individual level.

[1417] that I think the industrial age created a bunch of jobs that are very dehumanizing and we've just kind of gotten used to them and put up with them.

[1418] The idea that work should be repetitive and, you know, you just repeat the same loop over and over and over again and over a 10 -year period of time you might get, you know, graduated up one year and all that kind of stuff.

[1419] I don't think that's very human.

[1420] The idea that you could be simultaneously writing a book, launching a business, running a team, launching a festival.

[1421] having an event that you could actually be doing this kind of like mini kingdom work where you've got this little, you know, ecosystem around you of fun things that you're involved in.

[1422] That is actually made possible for a vast majority of people if they embrace these kind of tools.

[1423] You can live an incredibly fulfilling and amazing and impactful existence.

[1424] I know that I do.

[1425] as a result of having these tools in my life.

[1426] Like I'm doing things that I could have only dreamed about as a kid.

[1427] And what would you say to entrepreneurs?

[1428] I know you work with thousands of entrepreneurs.

[1429] What are you telling them in terms of their current businesses or business opportunities that you're foreseeing?

[1430] So I think that small teams have infinite leverage now and that when you have a team of, say, Five to ten people who share an incredible passion for a meaningful problem in the world and they want to see that meaningful problem solved and they come together in the spirit of entrepreneurship to solve that problem.

[1431] That little five to ten person team armed with the technology that we now have available, you can have a big impact.

[1432] You can make a lot of money.

[1433] You can have a lot of fun.

[1434] You can solve meaningful problems in the world.

[1435] You can scale solutions.

[1436] you can probably do more in a three -year window than most people did in a 30 -year career.

[1437] And then that little band of five to 10 people could either go together onto a new meaningful problem or they could disband and, you know, work on other meaningful problems with different teams.

[1438] In such a world where you have this sort of infinite leverage, but everyone else has access to the same infinite leverage, what becomes the USP?

[1439] Going back to this idea of the moat, like what is the thing of value?

[1440] when we've all got access to $20 infinite leverage?

[1441] Well, first of all, the first thing you need to do, you need to understand, is that this moment of time is the least competitive moment.

[1442] Like, if you understand how to use these tools, you can start making money tomorrow.

[1443] Like, you know, I see countless examples of people making thousands of dollars for these hustles that I talked about.

[1444] or building businesses that generates millions of dollars in the first couple of months of existence.

[1445] So I would say start moving now.

[1446] Start building things.

[1447] So it's an unprecedented time of wealth creation.

[1448] Clearly, at some point, as the market gets more efficient, as more and more people understand how to use these tools, there's less potential for creating these massive businesses quickly.

[1449] And we've seen this.

[1450] The dawn of the internet or dawn of the web was a lot easier to create Facebook than it is now.

[1451] Then we had mobile.

[1452] And for three, four, five years, it was very easy to create massive businesses.

[1453] And then it became harder.

[1454] Being just at the edge of what's possible is going to be very, very important over the next couple of years.

[1455] And that gets me really excited because the entrepreneurs who are paying attention are going to...

[1456] are going to be having the most amount of fun, but they're also going to be able to make a lot of money.

[1457] How many applications have been built on Replit to date?

[1458] I can talk about the millions of things that have been built since we started the company, but just since September when we launched Replit Agent, there's been about 3 million applications built purely in natural language.

[1459] with no coding at all, purely natural language.

[1460] Of those, I think 300 ,000, 400 ,000 of them were deployed in a real, the site was deployed and it is having, people are using it, some kind of business, some kind of internal tool.

[1461] I built one last night, by the way.

[1462] An internal tool or?

[1463] I built an application to track how my kids earn.

[1464] Pocket money.

[1465] Amazing.

[1466] So I just told it that I wanted to track the tasks that are happening around the house and assign a value to them and I want to be able to at the end of the week push a button and get a summary of how much to pay each child for their pocket money.

[1467] We are so screwed.

[1468] And within 15 minutes it had created this application.

[1469] And it was amazing.

[1470] Like you could toggle between like here's the place where you have the kids and here's the weekly reports and here's how much per task and you can tick off the tasks or remove tasks or add tasks.

[1471] So then I now have this application for which took 15 minutes of just talking about what I wanted and now I have an application to run the pocket money situation in the house.

[1472] By the way, having run an IT agency years ago, that's something that we would have charged 5 ,000 to 10 ,000 pounds to create, 5 ,000 to 10 ,000 US dollars to create.

[1473] And how much time?

[1474] Probably talking something that would have been a three-, four -week project.

[1475] At the start of the S -curve now that you're describing, and it's already, if it's a $20, replets roughly $20 a month.

[1476] $25.

[1477] For the base case.

[1478] You did one day of usage.

[1479] Let's say it's a dollar.

[1480] Yeah.

[1481] It cost you, and it cost you minutes and a dollar now, and we're at the start of the S -curve.

[1482] And you talk to it like you're chatting to a developer.

[1483] So one of the things that slows down the development process is you have to send the information to a developer, and they need to understand it, and then they need to create something and then come back to you.

[1484] This just happens in front of your eyes while you're watching it, and it's actually showing you what's being built.

[1485] It's really wild.

[1486] This one change has transformed how my team and I move, train and think about our bodies.

[1487] When Dr. Daniel Lieberman came on the diary of a CEO, he explained how modern shoes, with their cushioning and support, are making our feet weaker and less capable of doing what nature intended them to do.

[1488] We've lost the natural strength and mobility in our feet, and this is leading to issues like back pain and knee pain.

[1489] I'd already purchased a pair of Viva Barefoot shoes, so I showed them to Daniel Lieberman, and he told me that they were exactly the type of shoes.

[1490] that would help me restore natural foot movement and rebuild my strength.

[1491] But I think it was plantar fasciitis that I had where suddenly my feet started hurting all the time.

[1492] And after that, I decided to start strengthening my own foot by using the Vivo Barefoots.

[1493] And research from Liverpool University has backed this up.

[1494] They've shown that wearing Vivo Barefoot shoes for six months can increase foot strength by up to 60%.

[1495] Visit vivobarefoot .com slash DOAC and use code DIARY20 from my sponsor for 20 % off.

[1496] A strong body starts with strong feet.

[1497] This has never been done before.

[1498] A newsletter that is ran by 100 of the world's top CEOs.

[1499] All the time people say to me, they say, can you mentor me?

[1500] Can you get this person to mentor me?

[1501] How do I find a mentor?

[1502] So here is what we're going to do.

[1503] you're going to send me a question.

[1504] And the most popular question you send me, I'm going to text it to 100 CEOs, some of which are the top CEOs in the world running $100 billion companies.

[1505] And then I'm going to reply to you via email with how they answered that question.

[1506] You might say, how do you hold on to a relationship when you're building a startup?

[1507] What is the most important thing if I've got an idea and don't know where to start?

[1508] We email it to the CEOs.

[1509] They email back.

[1510] We take the five, six top best answers.

[1511] We email it to you.

[1512] I was nervous because I thought the marketing might not match the reality.

[1513] But then I saw what the founders were replying with and their willingness to reply.

[1514] And I thought, actually, this is really good.

[1515] And all you've got to do is sign up completely free.

[1516] I don't think we've spent a lot of time talking about autonomous weapons.

[1517] This is the thing that really worries me. And the thing that...

[1518] worries people about AI is this idea is that it is this emergent system and there's no one thing behind it.

[1519] It can act in a way that's unpredictable and not really guided by humans.

[1520] I also think it's true of corporations, of governments.

[1521] And so I think individual people can often have the best intentions, but the collective can land on doing things in a way that's harmful.

[1522] or morally repugnant.

[1523] And I think we talked about China versus the US, and that creates a certain race dynamics where they're both incentivized to cut corners and potentially do harmful things.

[1524] And in the world of geopolitics and wars, what really scares me is autonomous weapons.

[1525] Why does it scare you?

[1526] You can imagine autonomous drones being trained on someone's face.

[1527] And you can send a swarm of drones and they can be this sort of autonomous killing assassination machine.

[1528] And it can sort of function as country versus country technology in the world of war, which is still crazy.

[1529] But it can also...

[1530] become a tool for governments to subjugate the citizens.

[1531] And people think we're safe in the West, but I think the experience with COVID showed that even the systems in the West can very quickly become draconian.

[1532] Yeah.

[1533] Apparently I've heard in Iran that they have face recognition cameras.

[1534] that detect whether women are wearing hijabs in their own cars and it automatically detains the car if you're driving and you're not wearing a hijab.

[1535] And certainly if you're walking down the street, it just picks that up and immediately you're in trouble.

[1536] It acts as a police officer and a judge and...

[1537] You know, a lawmaker, it's the judge, jury and executioner essentially.

[1538] And it just happens instantaneously.

[1539] What happened in Canada with the truckers sort of protests where they froze their bank account by virtue of just being there.

[1540] Just by being in that location.

[1541] And just to confirm that, Iran has implemented a comprehensive surveillance system to enforce its mandatory hijab laws utilizing various technologies, one of which is...

[1542] cameras and facial recognition.

[1543] So they've put cameras in public spaces to identify women who are not adhering to the hijab dress code.

[1544] And just on that, London has just put those facial recognition systems into London and also all throughout Wales, and they're being rolled out at speed.

[1545] And, like, all you would need is a change of government that wanted to implement something similar and all the base.

[1546] Liar technology is already in there.

[1547] It gets a little bit worse in Iran because they have this new app called the Nazar app, where the government has introduced the Nazar mobile application, which allows you as a citizen to report another citizen who is not wearing their hijab.

[1548] And it logs their location, the time when they weren't wearing it, and the vehicle license plate.

[1549] With the crowdsourced data, it can then go after that individual.

[1550] I would also just point out that I think we're not being imaginative enough.

[1551] I agree with you.

[1552] I have the same concern about these autonomous weapons.

[1553] this doesn't have to occur in the context of war or even governmental oppression, that it is perfectly conceivable that effectively this allows, this drops the price of an undetectable or an unprosecutable crime.

[1554] And maybe economic moats return in the form of people taking out their competitors or anybody who attempts to compete with them using an autonomous drone that can't be traced back to them, you know, that follows facial recognition.

[1555] You don't have to kill very many people for others to get the message that this is a zone that you shouldn't mess around in.

[1556] So I could imagine effectively a new high -tech organized crime that protects rackets and makes tons of money and subjugates people who haven't done anything wrong.

[1557] I had Mustafa Solomon on the podcast in 2023 when all of this stuff started kicking off, and he is the CEO of Microsoft AI.

[1558] You're familiar with Mustafa, of course.

[1559] And one of the things he said to me at the time was, one of my fears is a tiny group of people who wish to cause harm are going to have access to tools that can instantly destabilize our world.

[1560] That's the challenge.

[1561] How to stop something that can cause harm or potentially kill.

[1562] That's where we need containment.

[1563] And it sounds a little bit like what you're saying, Amjad, that we will now have these tools.

[1564] You were talking in the context of the military.

[1565] But as Brett said, they're even smaller groups of people.

[1566] that might have been, I don't know, cartels or gangs, can do similar harm.

[1567] And at the moment, in terms of autonomous weapons, both the US and China are investing heavily in AI -powered weapons, autonomous drones, and cyber warfare, because they're scared of the other one getting it first.

[1568] And we talked about how much of our lives run on the internet, but cyber weapons and cyber AI agents that could be deployed to take down China's X, Y, or Z, or vice versa, are of real concern.

[1569] Yeah, I think all of that is a real concern.

[1570] Unlike Mustafa, I don't think containment is possible.

[1571] Part of the reason why this game theoretic system of competition between the US, China, corporations, individuals, makes it so that this technology is already out.

[1572] Really hard to put it back in the...

[1573] I did ask him this question, and I remember the answer because it was such a stark moment for me. I said to Mustafa, do you think it's possible to contain it?

[1574] And he replied, we must.

[1575] So I asked him again, I said, do you think it's possible to contain it?

[1576] And he replied, we must.

[1577] And I asked him again, I said, do you think it's possible?

[1578] And he said, we must.

[1579] So the problem with that chain of thinking is that it might lead to an oppressive system.

[1580] There is one of the, say, doomers or philosophers of AI, which I respect his work.

[1581] His name is Nick Bostrom.

[1582] And he was trying to think of ways in which we can contain AI.

[1583] And the thing that he came up with is perhaps more oppressive than something that AI would come up with.

[1584] It's like total surveillance state.

[1585] You need total surveillance on compute, on people's computers, on people's ideas to not invent AI or AGI.

[1586] It's like taking the guns or something.

[1587] Right, exactly.

[1588] I mean, there's always this problem of containing any sort of technology is that you do need oppression and draconian policies to do that.

[1589] Are you scared of anything else or concerned about anything else as it relates to AI outside of autonomous weapons?

[1590] You know, we talked about the birth threat crisis.

[1591] And I think a more generalized problem there is creating virtualized environments.

[1592] via VR, where everyone is living in their own created universe.

[1593] And it's so enticing and even creates, simulates work and simulates struggle such that you don't really need to leave this world.

[1594] And so every one of us will be solipsistic, similar to The Matrix.

[1595] Ready Player One.

[1596] Ready Player One, we're all kind of plugging.

[1597] Even worse than Ready Player One, at least that's a massively networked environment.

[1598] I'm talking about AI simulating everything for us, and therefore you're literally in the matrix, you know?

[1599] Maybe this is...

[1600] I had that same thought.

[1601] I've enjoyed this great simulation.

[1602] Yes.

[1603] And so, I mean, are you familiar with the Fermi's Paradox?

[1604] No, I'm not.

[1605] So Fermi's Paradox is...

[1606] The question, the professor, his name is Fermi, he asked the question, if the universe is that vast, then where are the aliens?

[1607] The fact that humans exist, you can deduce that other civilizations exist.

[1608] And if they do exist, then why don't we see them?

[1609] And then that spurred a bunch of Fermi solutions.

[1610] I don't know, you can find hundreds of solutions on the internet.

[1611] One of them is the sort of house cat thought experiments where actually aliens exist, but they kind of put us in an environment like the Amish in a certain time and do not expose us to what's going on out there.

[1612] So we're pets.

[1613] Maybe they're watching us and kind of enjoying what we're doing.

[1614] Stopping us from hurting ourselves.

[1615] Stopping us from hurting ourselves.

[1616] There are so many things.

[1617] But one of the things that I think is...

[1618] Potentially a solution to the Fermus Paradox and one of the saddest outcomes is that civilizations progress until they invent technology that will lock us into infinite pleasure and infinite simulation such that we don't have the motivation to go into space, to seek.

[1619] out the exploration of potentially other alien civilizations.

[1620] And perhaps that is a determined outcome of humanity or like a highly likely outcome of any species like humanity.

[1621] We like pleasure.

[1622] Pleasure and pain is the main motivators.

[1623] And so if you create an infinite pleasure machine, does that mean that we're just at home in our VR environment with everything taking care for us and literally like the matrix?

[1624] And the world, the real world would suck in such a scenario.

[1625] Yes, it'd be terrible.

[1626] I mean, the other simpler explanation of the Fermi paradox is that you generate sufficient technology that you can end your species.

[1627] And it's only a matter of time from that point, which, you know, we can have that discussion about nuclear weapons.

[1628] We can have it about AI, but does some technology, if we stay on that escalator, does some technology that we generate ultimately - Whatever allows you to get off the planet, allows you to blow up the planet.

[1629] There you go.

[1630] I want to get everyone's closing thoughts and closing remarks.

[1631] And hopefully in your closing remarks, you can capture something actionable for the individual that's listening to this now on their commute to work or the single mother, the average person who maybe isn't as technologically advanced as many of us at this table, but is trying to navigate through this to figure out how to live a good life over the next 10, 20, 30 years.

[1632] Yeah.

[1633] And take as long as you need.

[1634] I think we live in the most interesting time in human history.

[1635] So for the single mother that's listening, for someone who wouldn't be the stereotype of a tech bro, don't assume that you can't do this stuff.

[1636] It's never been more accessible today within your work.

[1637] You can be an entrepreneur.

[1638] You don't have to take massive risk to go create a business, quit your job and go create a business.

[1639] There are countless examples.

[1640] We have a user who's a product manager at a larger real estate business.

[1641] And he built something that created 10 % lift and conversion rates, which generated millions and millions of dollars of that business.

[1642] And that person became a celebrity at that company and became someone who is lifting everyone else up and teaching them how to use these tools.

[1643] And obviously, that is like a really great for anyone's career.

[1644] And you're going to get a promotion.

[1645] Your example of building.

[1646] a piece of software for your family, for your kids to improve and to learn more, to be better kids, as an example of being an entrepreneur in your family.

[1647] So I really want people to break away from this concept of entrepreneurship being, this is your podcast, Diary of a CEO.

[1648] You started this podcast by talking to CEOs, I assume, right?

[1649] And over time, it changed to Everyone can be a CEO.

[1650] Everyone is some kind of CEO in their life.

[1651] And so I think that we have unprecedented access to tools for that vision to actually come to reality.

[1652] Well, it is obviously a moment of a kind of human face transition, something that I believe will be the equal of a discovery of farming or writing or electricity.

[1653] And the darkness that I think is valid in looking at all of the possible outcomes of this scenario is actually potentially part of a different story as well.

[1654] In evolutionary biology, we talk about an adaptive landscape in which a niche is represented as a peak.

[1655] And a higher niche, a better niche, is represented as a higher peak.

[1656] But to get from the lower niche to the higher niche, you have to cross through what we call an adaptive valley.

[1657] And there's no guarantee that you make it through the adaptive valley.

[1658] And in fact, the drawing that we put on the board, I think, is overly hopeful because it makes it in two dimensions.

[1659] It looks like you know exactly where to go to climb that next peak.

[1660] And in fact, it's more like the peaks are islands and an archipelago that is...

[1661] in fog where you can't figure out what direction that peak is and you have to reason out it's probably that way and you hope not to miss it by a few degrees.

[1662] But in any case, that darkness is exactly what you would expect if we were about to discover a better phase for humans.

[1663] And I think we should be very deliberate about it this time.

[1664] I think we should think carefully about how it is that we do not allow the combination of this brand new, extremely powerful technology and market forces to turn this into some new kind of enslavement.

[1665] And I don't think it has to be.

[1666] I think the potential here does allow us to refactor just about everything.

[1667] Maybe we have finally arrived at the place where mundane work doesn't need to exist anymore and the pursuit of meaning can replace it.

[1668] But that's not going to happen automatically if we don't figure out how to make it happen.

[1669] And I hope that we can recognize that the peril of this moment is best utilized if it motivates us to confront that question directly.

[1670] Each one of us has two parents, four grandparents, eight great -grandparents, 16, 32, 64.

[1671] You've got this long line of ancestors who all had to meet each other.

[1672] They all had to survive wars.

[1673] They all had to survive illness and disease.

[1674] Everything had to happen for us.

[1675] Each individual, all of this stuff had to happen for us to get here.

[1676] And if we think about all the people in those thousands and thousands of people, Every single one of them would trade places in a heartbeat if they had the opportunity to be alive at this particular moment.

[1677] They would say that their life was struggle, disease, that their life was a lot of mundane and meaningless work.

[1678] It was dangerous.

[1679] You know, every single one of us has probably got ancestors that were enslaved, probably got ancestors that died too young, probably got ancestors that worked.

[1680] horrific conditions.

[1681] We all have that.

[1682] And they would all just look at this moment and say, wow, so are you telling me that you have the ability to solve meaningful problems, to come up with adventures, to travel the world, to pick the brains of anyone on the planet that you want to pick the brains of?

[1683] You can just listen to a podcast.

[1684] You can just watch a video.

[1685] You can talk to an AI.

[1686] Like, are you telling me that you're alive at this particular moment?

[1687] Please make the most of that.

[1688] Like do something with that.

[1689] You know, you can sit around pontificating about society and how society might work, but ultimately it all boils down to what you do with this moment.

[1690] And solving meaningful problems, being brave, having fun, making your little dent in the universe, you know, that's what it's all about.

[1691] And I feel like there's an obligation to your ancestors to make the most of the moment.

[1692] Thank you so much to everybody for being here.

[1693] I've learned a lot and I've developed my thinking, which is much the reason why I wanted to bring this all together, because I know you all have different experiences, different backgrounds in education, you're doing different things, but together it helps me sort of pass through all of these ideas to figure out where I land.

[1694] And I ask a lot of questions, but I am actually a believer in humans.

[1695] I was thinking about this a second ago.

[1696] I was thinking, am I optimistic about humans' ability to navigate this just because I have no other choice?

[1697] Because as you said, the alternative actually isn't worth thinking about.

[1698] And so I do have a optimism towards how I think we're going to navigate this, in part because we're having these kinds of conversations.

[1699] And we, in history, haven't always had them at the birth of a new revolution.

[1700] When we think about social media and the implications that had, we're playing catch up with the downstream consequences.

[1701] I am hopeful.

[1702] Maybe that's the entrepreneur in me. I'm excited.

[1703] Maybe that's also the entrepreneur in me. But at the same time, to many of the points Brett's raised and Amjad's raised and Dan's raised, there are serious considerations as we swim from one island to another.

[1704] And because of the speed and scale of this transformation that Brett highlights.

[1705] And you look at the stats of the growth of this technology and how it's spreading like wildfire and how once I tried Replit, I walked straight out and I told Cozzy immediately, I was like, Cozzy, try this.

[1706] And she was on it and she was hooked.

[1707] And then I called my girlfriend in Bali, who's the breathwork practitioner.

[1708] And I was like, type this into your browser, R -E -P -L -I -T.

[1709] And then she's making these breathwork schedules with all of her clients' information ahead of the retreat she's about to do.

[1710] It's spreading like wildfire because we're internet native.

[1711] We were native to this technology.

[1712] So it's not a new technology.

[1713] It's something on top of something that's intuitive to us.

[1714] So that transition, as Brett describes it, from one peak to the other or one island to another, I think is going to be incredibly destabilizing.

[1715] And having interviewed so many leaders in this space, from Reid Hoffman, who's the founder of LinkedIn, to the CEO of Google, to Mustafa, who I mentioned, they don't agree on much.

[1716] But the thing that they all agree on and that Sam Altman agrees on is that the long -term future, the long -term way that our society functions is radically different.

[1717] People squabble over the short term.

[1718] They sometimes even squabble over the midterm or the timeline, but they all agree that the future is going to look completely different.

[1719] Amjad, thank you for doing what you're doing.

[1720] We didn't get to spend a lot of time on it today, and this is typically what I do here, but your story is incredibly inspiring.

[1721] Incredibly inspiring from where you came from, what you've done, what you're building, and you are democratizing and creating a level of playing field for entrepreneurs in Bangladesh to...

[1722] Cape Town to San Francisco to be able to turn their ideas into reality.

[1723] And I do think just on the surface that that's such a wonderful thing that, you know, I was born in Botswana in Africa and that I could have the same access to turn my imagination into something to change my life because of the work that you're doing at Replit.

[1724] And I highly recommend everybody go check it out.

[1725] You probably won't sleep that night because it's so, it's so, for someone like me, it was so addictive.

[1726] to get to be able to do that because it's been the barrier to creation my whole life.

[1727] I've always had to call someone to build something.

[1728] Dan, thank you again so much because you represent the voice of entrepreneurs and you've really become a titan as a thought leader for entrepreneurs in the UK.

[1729] And that perspective, that balance is incredibly important.

[1730] So I really, really appreciate you being here as always.

[1731] And you're a huge fan favorite of our show.

[1732] And Brett, thank you a gazillion times over for...

[1733] Being a human lens on complicated challenges and you do it with a fearlessness that I think is imperative for us finding the truth in these kind of situations where some of us can run off with optimism and we can be hurtling towards the mousetrap because we love cheese.

[1734] And I think you're an important counterbalance and voice in the world at this time.

[1735] So thank all of you for being here.

[1736] I really, really appreciate it.

[1737] And we shall see.

[1738] These things live forever.

[1739] I find it incredibly fascinating that when we look at the back end of Spotify and Apple and our audio channels, the majority of people that watch this podcast haven't yet hit the follow button or the subscribe button, wherever you're listening to this.

[1740] I would like to make a deal with you.

[1741] If you could do me a huge favour and hit that subscribe button, I will work tirelessly from now until forever to make the show better and better and better and better.

[1742] I can't tell you how much it helps when you hit that subscribe button.

[1743] The show gets bigger, which means we can expand the production, bring in all the guests you want to see and continue.

[1744] to doing this thing we love if you could do me that small favor and hit the follow button wherever you're listening to this that would mean the world to me that is the only favor i will ever ask you thank you so much for your time