Moonshots / Peter Diamandis
Astro Teller: Inside Google X’s 2,000-Idea Moonshot Factory & How Google Brain Made the T in ChatGPT
🇬🇧 EN🇪🇸 ES
Ask the tutor about this video
Sign in with a verified email and ask the tutor anything about this video.
TL;DR
Research summary
▶ What makes a moonshot
After brief introductions, the interviewer asks how to create and launch moonshots. Teller defines one through three components: a large, clearly named problem; a science-fiction-sounding product or service that would address it; and breakthrough technology offering at least a small chance of making that product or service. Together these form a testable hypothesis, not a guarantee of success. He says teams need equal measures of audacity and humility: pursue unlikely ideas while recognizing they may fail, and learn quickly and cheaply enough to move on. He says ideas may be stopped for being too small, too likely, harmful, or economically implausible; first-principles checks of materials, weight, cost, and what customers would pay can expose problems early. In his view, purpose and profit can support each other, while a money-losing idea is unlikely to change the world.
▶ X’s funnel, AI, and Google Brain
The speakers discuss the pace of change in science and materials, Google’s growth, and X’s project funnel. Teller says X advances about 100–200 ideas a year far enough to receive code names, and about 5–6 years later graduates roughly two moonshots; the interviewer calls that a 2% hit rate. Teller expects AI to shorten the time needed to de-risk ideas and hopes it will let teams pursue more ambitious work. He describes Google Brain as an early X project begun about 15.5 years earlier, when neural networks were out of favor. Andrew Ng and Jeff Dean explored industrializing neural networks by making them tens of thousands of times larger. Teller attributes TPUs and the Transformer to Google Brain and characterizes its role as producing seed technologies that later change the world. He says innovations that seem to succeed overnight often take 15–20 years to develop.
▶ Unrealized moonshots and the stagnation question
Asked about projects he wishes had succeeded, Teller names clean water, energy shifting across time and location, and education. For water, he says changing the world would require an all-in cost near one cent per liter; he cites almost 3 billion water-stressed people and says climate change could worsen the problem. He says X’s work has reached about 10 cents per liter, so it stopped that effort and may revisit it. He says X has tried energy projects without finding results that excited the team, and plans to keep returning to education because he considers it fundamentally broken in both developed and developing countries. On the proposed half-century slowdown in moonshots, Teller says society lost some of its explorer spirit after the 1960s. He suggests wartime and Cold War spending tolerated low returns on investment, while later emphasis on returns reduced moonshot activity. X’s aim, he says, is to make moonshot development efficient enough to justify greater investment. He reports weekly approaches from companies or countries interested in creating moonshot factories and expects more such efforts.
▶ Protecting high-risk projects and building the right culture
The interviewer proposes keeping a moonshot group at the edge of an organization, focused on 10x outcomes and reporting to the CEO. Teller agrees. He illustrates the incentive problem with a choice between a guaranteed $1 million in value and a 1-in-100 chance at $1 billion: the latter has ten times the expected value, but he says leaders and boards often do not support employees who choose it. Teller says high-risk projects should be placed where they are wanted and their difficulties tolerated, comparing this to Lockheed’s Skunk Works and the Macintosh team working off campus. He rejects the idea that X’s results come from large budgets: teams are small, and Google Brain graduated at about 18 people. He says X seeks a breakthrough that avoids solving problems in the ordinary way and is willing to stop if one cannot be found. He describes its process as a mindset and points listeners to a masterclass certificate course on breakthrough innovation, estimated at about 20 hours with the work or 3.5 hours for videos; an evening session is announced for 7:00 on rapid prototyping.
▶ Moonshot costs, efficiency, and organizational limits
The speakers debate whether moonshots are becoming more expensive or cheaper. Teller says X tracks the cost of reaching graduates and that it has fallen by a factor of three over 16 years. He clarifies that this is the cost of getting an individual project through X’s process, not the value it produces; he estimates the cost is declining by roughly 10–20% year over year, partly through better practices, earlier graduation decisions, and AI. He says a Waymo’s value is very large relative to the cost of developing it and the projects that did not work. Asked why X does not produce far more companies, Teller says efficiency depends on a difficult-to-maintain culture of long-term thinking, intellectual honesty, humility, teamwork, and low ego. He says he knows how to cultivate that culture among a few hundred people but not a few thousand, calling scale a culture problem.
▶ Systematizing innovation and using AI with people
Teller calls systematizing radical innovation the meta-moonshot, and says copying moonshot factories is a further challenge he has not yet solved, though organizations are asking for help. In a discussion of AI, he says the people-and-agents mix is an implementation detail; X focuses on problems such as clean water, grid-scale storage, education, the electric grid, and circular material use, while trying to build them into businesses that can grow without excessive cost. Teller says X currently has a clean-water team and that its salary costs exceed its AI bill, though he does not know the difference. X is automating parts of its process to work faster and more efficiently. He does not expect innovation to be fully automated within the next decade: choosing problems, solving them in socially acceptable ways, selling through distributors, and working with communities require people, in his view, for at least the next decade or two. He says compute should be used when it produces benefit, not merely because deploying agents feels powerful.
▶ Ambition, business timelines, and project attrition
A panelist distinguishes moonshots aimed at pain points from aspirational ones such as Mars travel or solving aging. Teller says he wants both, provided teams can move an idea from “crazy” to “not crazy” in under a decade, then to an enduring business in under another decade. For a Mars project, he would ask how it could become an enduring business within two decades; he says he is not claiming it cannot. He explains that Alphabet funds X and needs to understand how its funding creates a useful growth cycle.
▶ Testing early, learning from failure, and saying no
Teller says about 2,000 projects received code names over 16 years, of which roughly 35–50 graduated, depending on how they are counted. He uses a monkey-and-pedestal analogy: test whether the monkey can perform the difficult task before spending time building the pedestal. At X, this means testing the core uncertainty before investing in supporting infrastructure. Teller says he does not seek failure for its own sake; he values it because it reveals when a model was wrong. He argues for destigmatizing failure and ending projects before they become long-running “zombies.” He mentions a current lab project that he says is almost certainly wrong, yet is working for reasons the team does not understand; after about five months, they are working hard to discover why they may be mistaken and stop it if necessary. He declines to identify it and says it makes him both uncomfortable and excited. Asked about regrets over rejected ideas, Teller says X is often wrong but does not regret saying no: he estimates a false positive can cost many tens of millions of dollars, whereas he considers the cost of a false negative zero if there are always more problems and solutions to pursue. Another speaker says X may not make decisions quickly enough.
▶ Materials and the hardest part to replicate
Teller says he is excited about new materials but stresses that a discovery is only the beginning of a business. Using steel as an example, he says refinement and industrialization took centuries. A room-temperature superconductor discovery might be worthy of a Nobel Prize, he says, but would represent only the first 5% of the work: production at thousands of tons per day and high ductility for uses such as winding wire would still need to be solved. He says the remaining scale-up work is central to a materials moonshot. Asked what other companies underestimate when trying to reproduce X, he points to leadership’s sustained effort to engineer a culture that enables radical innovation efficiently and creates protected space for explorers. He recommends X’s two podcast seasons, available through podcast platforms, as accounts from people who went through the process. The interviewer adds that Google’s leaders have kept X at the edge of the company rather than pulling it back into the organizational “immune system”; the session ends with applause for Teller.
Full transcript
[0:05] Good to see you again, Astro. >> Thanks for having me. >> Uh pleasure, pal. And we've known each other for God knows a couple of decades now. You know, I think most people don't know something about you. It's worth mentioning just for fun, since we're talking about genetics. Uh his maternal grandfather was a Nobel laureate in economics, and his paternal grandfather was Edward Teller, the creator of the hydrogen bomb. It's pretty interesting fact. >> No pressure. >> No pressure. >> It was It was It was a high-pressure
[0:35] childhood, yes. >> Yeah, I I guess so. So, um let me let me kick off the first question. So, you've been in the moonshot business for over 20 years. Uh walk us through your advice of what it actually takes to come up with a moonshot and launch a moonshot. You know, how do you make this real? We've got a lot of individuals here who are here because they believe in optimism, they believe in moonshots. You know, in 10 seconds. No, just
[1:05] kidding. Uh what's your advice? I mean, you've learned so much. >> Let Let me Let me give you two quick things. The first one is, let's just define moonshot, at least how we think of a moonshot. >> Which is how I think of it as well. >> Excellent. >> from you. >> So, three basic components. First, there has to be a huge problem with the world that you can name and you want to solve. If you can't name the huge problem, then it's arguably an academic exercise. Second, there has to be some kind of science-fiction sounding product or service that, no matter how unlikely it
[1:36] is you could make it, we could pre-agree that if you made that science-fiction sounding product or service, it would resolve that huge problem with the world. And then, three, there has to be some kind of breakthrough technology that gives us a prayer, at least a tiny chance, at minimum, of being able to make that science-fiction sounding product or service and resolving that huge problem with the world. When you have that, we would call that a moonshot story hypothesis. That does not mean you're going to win, but at least means it's testable.
[2:08] The second thing I would say is like one of the fundamental issues that we found at X is that you have to have two things in equal amounts in order to be a moonshot explorer, to be a moonshot team. The first one is you have to have very high audacity. You have to be willing to suspend your disbelief for non-stupid reasons and go on unlikely journeys that might actually profoundly help the world. But
[2:40] the second thing you have to have in equal measure is humility. You have to know from the first moment you set out on that unlikely journey that it's unlikely because if you don't, you will go ruinously far down that path before you find out that that is one of the 99% that just isn't going to work out. So, being able to say right at the beginning, "I'm proud that I'm trying this, but it probably won't work. Let me learn as fast and as cheaply as I can whether this is even possible." allows
[3:11] you to get onto the next moonshot when that moonshot's not the right one to do. >> So, your advice for folks, I mean, at with the challenge of picking a moonshot, I mean, right now the speed at which you can test things, create things, right? But what about the beginning, choosing the idea? Any advice there? Well, I mean, let me put it this way. What do you do to shut down an idea quickly? >> So, sometimes the idea is just not big enough. Sometimes the idea is too
[3:41] reasonable or likely to succeed. Sometimes it's just not good for the world or the potential for too many sort of secondary problems associated with it is just too high. Uh often it's techno-economics. I cannot tell you the number of moonshots that feel so good to people in the early days. Often these are very technical people and it's not until you say, "Really? Like let's go step by step. What is the best case scenario for the
[4:12] build of materials that you would make this from? What is like just the raw cost in weight of the stuff that would go into that? What are people really going to pay? Could this even plausibly zero >> uh first principle thinking in that regard. >> Right. First principles thinking about the techno-economics. You can often kill moonshots in the very early days because it's just not going to turn into an enduring business and at least what I believe is that purpose and profit can
[4:43] support each other and then if you're doing something that's going to lose money, it's probably not going to change the world. >> Yeah. >> Dave >> Yeah, I'm really curious about you know Alex and Peter have this solve everything book. It's phenomenal but the rate at which physics and material science chemistry is going to change just astronomically accelerating. So when you joined Google what 25 23 years ago? How long ago? >> Uh 16 and a half years ago is I co-founded X 16 and a half years
[5:13] >> 16 and a half years ago. Okay. And at the time it was about a 30 billion dollar company? >> Ish. Yes. It was It was very sweet. It was very small back then. >> Yeah. A little 30 billion dollar company. Now just about half a trillion dollar company. So the scale of R&D in a corporate environment like that has grown miraculously. Just absolutely tremendously. And I think you were saying backstage that the ideal target for a new idea 100 ideas a year two of them succeed. >> That's right. We we start one to 200 ideas a year that make it far enough
[5:43] that they end up with a code name. I who knows how many we actually look at but one to 200 ideas a year that are make it far enough we get a code name. Of those, about 5 to 6 years later, we graduate to moonshots out of X. >> Wow. So, 2% hit rate. >> And so then if you know if if AGI is imminent and the idea rate >> No, it's here. >> is astronomically >> Right. >> Does does that grow on that same exponential curve? Just that the number of things at the front of the funnel go from 100 to 1,000 to 10,000 to
[6:13] >> Oh, for sure. I mean, there's no shortage of problems with the world. There's no shortage of creative ideas for running at those problems. I think AI is going to somewhat shrink that time from a crazy idea to us having de-risked it to the point where we can say definitively here's evidence it's not a crazy idea anymore. So, that's getting shorter. I hope that AI is going to let us raise our audacity even higher, so we'll be shooting higher, and so we'll be able to do more in a shorter amount of time. But I'm But fundamentally, I
[6:44] think the structure will be >> May I think I think uh you know, if you asked me 3 years ago, you know, what are the big things that came out of out of X, uh Waymo is at the top of the list. Um Google Google Brain >> So, that's what I was going to ask about. So, Google Brain was probably like, "Ah, yeah, this is cool." Up until some point a couple years ago when it was like, "No, this is the entire future of Google now." >> guys just spent 80% of the last 3 hours talking about the stuff that arguably like Google Brain sort of birthed into the >> What What is Google Brain? Define it
[7:15] because we You know DeepMind, but we don't know Google Brain. >> So, Google Brain, uh which started at X 15 and a half years ago or so, was one of the earliest things that we started. At the time, there were a set of academics who were kind of out of vogue. This was like Andrew Ng, Yann LeCun, a few others who were saying, "I think scale might matter." And neural nets were absolutely dead in the field at the time. And they're saying, "I think scale might matter. It it actually bring neural networks back. And everyone else in the field said, Nope, you're crazy.
[7:46] Andrew Ng came to Google X at the time, and we set up what it was what was called Google Brain with Jeff Dean. So, the two of them paired together to say, What if we industrialized neural networks? What if we made them tens of thousands of times bigger than anyone had ever done before? Could that actually be the breakthrough that everyone's been waiting for? So, we were the first ones who really made what Fast Forward >> called deep learning. >> TPUs ultimately came from Google Brain.
[8:20] The Transformer, which is sort of underpins everything, it's literally the T in ChatGPT, that came from Google Brain. So, you know, our job is to make the seed crystals that ultimately go on to change the world, and I think Google Brain counts as one of them. >> But there must have been a long period of time there where that was kind of quaint and promising, and then a period of time just a recently where it's like, Holy crap, that was a lot more than quaint and promising. >> Of course, but I think the reality of any innovation is it takes a super long time. Every single innovation is like
[8:52] a overnight success that was 15 to 20 years in the making. And just because the general public only wakes up to it in year 18 of that slog, it feels like it happened overnight. >> Every overnight success after 11 years of hard work. >> For sure. And >> So, Astro, you you think I watched hundreds of moonshots, evaluated hundreds. What are a couple of examples of moonshots that you would like to have seen that didn't quite make it, didn't cut the mustard, whatever you would have loved to have had them come to fruition?
[9:24] >> I I mean, I have a very long list. I'll give you two examples, but >> Sure. >> we've been working on clean water over and over again. If you can make clean water, if you could pull it from the atmosphere, if you could desal for a tenth the price, you have to be able to get to like a penny a liter all-in cost for it to really change the world. But we have almost 3 billion people in the world who are water stressed, who don't have clean water to drink. That's going to get worse as climate change picks up. We're
[9:54] going to have hundreds of millions of climate refugees around the world, and it's often going to start because they don't have enough clean water. This is one of the biggest problems to solve. >> Mhm. >> But techno-economics we've done some great work in the space, but that it's going to be 10 cents a liter. >> Nope. >> Well, we'll stop doing that. Try it again. So, that's an example of something we keep coming back to. Um >> Two more. >> Two more, okay. Um being able to uh time
[10:26] shift and location shift energy. And I say it that way on purpose because if you say transmission lines and batteries, people tend to narrow their focus in. But being able to time shift and location shift energy is absolutely going to change the world. We've taken a bunch of runs at those. So far, you know, we haven't been super excited with what we've found. Education is another one. Education is not working. It's It's not working in the developed world. It's not working in the developing world.
[10:56] >> [applause] >> We're going to keep coming back to that one till we find something cuz that has got to be one of the sort of most fundamental moonshots that you could do. >> Got it. >> Love it. >> Alex, please. >> Yeah, I'd love to talk a bit about the history of moonshots. So, one of your grandfathers invented the hydrogen bomb. And then if if you look at history, and I I should add maybe humanity was in the process once upon a time of landing humans on the moon in actual moonshots. And then that stopped. Sometime one can quibble over the chronology, but
[11:26] arguably depending on which version of the great stagnation thesis one subscribes to, there was maybe up to a 50-year period, call it 1970 to 2020-ish. Again, one can quibble at the margins, when it would seem that there was, call it a reduced launch velocity of moonshots from humanity as all. We stopped landing humans on the moon. We stopped inventing, arguably, applied physical innovations, regardless of their warfare potential or not, like hydrogen bombs. And then, again,
[11:57] arguably, sometime around 2020 or so, plus or minus a few years, moonshots started again. We got superintelligence that actually works. We're getting robots that actually work. Physical intelligence, so-called, is seeming to work. We got, arguably, medical nanotechnology in the form of lipid nanoparticles that helped to treat a pandemic. So, a few related subquestions. One, do you believe, do you subscribe to any variant of a great stagnation hypothesis, that there was a half-century pause in moonshots? And to
[12:28] the extent you do, another question, how do we prevent that from happening again? >> Well, um the first thing is I know you didn't maybe intend it that way, but I'm going to take that as an enormous compliment and 11-year in the making joke that we just made. Since we started in 2010 and you said 2020 several times, exactly my point is we started literally a moonshot factory in 2010. >> you're actually you brought moonshots back. >> Well, yes, to some extent, I would like to
[12:59] think that I helped to lead we did. >> [applause] >> So, that's the first half of the answer. And and yes, I do think that I don't know I don't think anyone was planning on it being a stagnation, but I think functionally we had lost our way a little bit somewhere between, you know, you had your Star Trek stuff up here a little bit. >> Yes. Somewhere between like the 1960s and NASA and the inspiration that that created, there was a period where we sort of lost the explorer spirit.
[13:29] >> you think? What do do you have an ideology, a diagnosis? What do you think went wrong? >> I don't know, but but here is a partial answer to that, and then I want to get to the future. Here's Can I tell you all a a dirty little secret? It'll be our little secret. Taking moonshots is really, really easy. >> Huh. >> It's it's it's laughably easy if you don't care about efficiency. You just find some super energetic people who are sort of delusionally optimistic, pour a bunch of money on
[14:01] them, you will absolutely get some moonshots. It's just not a very good return on investment. And so, I think we had a period, mostly because of World War II and the Cold War, where people didn't care about the return on investment. >> Yeah. >> Then there was a period where people probably still wanted moonshots, but they cared about the return on investment, and so there was less of it. What X has been trying to practice is making a moonshot factory, being able to
[14:32] get at that audacity efficiently enough to systematize the process, not just to do it a lot, but to do it efficiently so it is rational for people to put a lot more money in. And to your point about the future, good news, I think, we get approached now on a weekly basis from some large company or some country saying they want to set up their own moonshot factory. I'm not sure exactly how we can help them with that, but there is a hunger to
[15:02] do more of this, and so good news, I expect you'll see a lot more of it in the future, not just for that. >> But Astro, take a second, and there are a number of people here in the audience that are running companies that do want to set up a moonshot factory, and and we have discussed a little bit about this, right? Which Let me paraphrase what I've learned from you. You're my moonshot mentor here, uh which is, you know, your core organization is responsible for delivering 10% profit year after year.
[15:32] And And Salim and I have written about this that your moonshot factory, if you would, needs to be on the edge of the organization where your crazy thinkers exist, right? And you Again, quoting you, you know, if you work on anything at 10%, you're fired. Focus on 10x, the 1,000%. And then have that side moonshot organization report directly to the CEO. Is that still the formulation? >> Absolutely. You know, I've run the following experiment. I apologize to Peter who's uh heard this a bunch.
[16:02] Salim's probably heard it a few times, too, but just bear with me for 45-50 seconds. Choice A, choice B. Choice A, you can give a a million dollars of value to your business this year, but it's And it's guaranteed. Or B, choice B, billion, you can give a billion dollars of value to your business this year, but it's not guaranteed. It's one chance in a hundred. So, A, million guaranteed, B, billion, one chance in a hundred. Who's choosing choice A?
[16:32] All right, almost nobody here, almost nobody in the world raises their hand. WHO'S CHOOSING CHOICE B? WOO! All right, so you all passed the math test. It cuz it has 10 times the expected utility. I've done this with CXOs the world over. And I've said, "Okay, now keep your hand up if on their best days, in your wildest dreams, your manager, your CEO, your board of directors actually supports you doing choice B." >> Yeah. >> And every hand in the room goes down.
[17:02] Yeah. And I say, "You don't need a lecture on innovation, you need a new manager." >> Yeah. >> This So, the formulation The formulation you just described is exactly right, and it's because of this thing. You have to sequester the choice B's to a place where they're actually wanted and and the mess that comes with them is tolerated. >> This is Lockheed Skunk Works off on the edge. This is Steve Jobs taking the Mac team off campus to build it. >> Exactly. >> Yeah. >> The the capability of the human mind to
[17:34] do this is very hard. There's something I've been observing over the last few years, and I'd love for you to comment on this. Peter wrote about this in one of his original books, the demonetization of the world, right? So, throughout human history, it's always been true that advanced technologies cost a lot. And today, for the first time in human history, advanced technologies are cheap. See solar energy, cheap sensors are cheap, blockchain's open source and free, etc., etc. This should enable, you know, for a CFO or for a big company,
[18:04] when you want to do very something very disruptive, you had to make a big bet, put a lot of resources. CFO hated you, they would defund it as fast as they could. Today, you can take a portfolio of very disruptive experiments at the edge. So, now it just comes down to mindset, and are you willing to go for it? Do you agree with that, and how much are you seeing companies do that? >> Uh I I do agree. I think it is a I hear a lot, "Oh, well, Alphabet has lots of money, and so that's why this is happening." I guarantee you that is not
[18:34] why. We have tiny teams. The teams that graduate from us, Google Brain when they graduated, they were like 18 people, and that's pretty normal. We keep our teams tiny because we want I we I I say to them all the time is, "I want you to find a cheat in the video game of life, a Gordian knot-chopping moment. If you can't find that, I don't want you solving things the normal way. I'd rather you stop and we go find a different moonshot, cuz that's what moonshots mean."
[19:04] So, I believe it is mostly a mindset issue, and for those of you out there who are interested in learning more, we did a masterclass certificate course. It's about 20 hours if you do all the work. It's about 3 and 1/2 hours if you just watch the videos. I encourage people to go watch. It's on breakthrough innovation and it will step people through. It's our secret sauce is not secret. We're telling people exactly how to do it. >> And tonight, Astro, there's an evening program by your team doing a 50-minute condensed version of that, you know,
[19:36] that 10-hour course. >> Exactly. 7:00 tonight, Joe Sergeant. It's going to be a lot of fun. It's on rapid prototyping. >> Amazing. >> I'd also love to have maybe better understand and develop the record on the economics of moonshots. I I could make on the one hand an argument that the real cost of moonshots increases over time because expectations go up and we see arduous problems solved left and right by people sitting on the toilet casually solving hard problems. I could make the argument in the opposite direction that the real cost of
[20:06] achieving a moonshot is hyper-deflating because we have super intelligence now and we can solve lots of hard problems. I could probably make some sort of mediocre middle-of-the-road argument as well. What would be your position on the long-term economics of the real costs of moonshots hyper-inflating, hyper-deflating, staying the same. >> Uh I mean, we track because we're obsessed with the efficiency of getting at moonshots. We track what it costs for us to get to our graduates very carefully and I can tell you that it's
[20:37] down by about a factor three over the last 16 years. There's a lot of waste and complexity and it's a bit of a lagging indicator. So, some of what >> Wait, wait, factor of three over the that that's inflation. >> No, no, >> [laughter] >> it is a factor of three cheaper, just to be clear. It's a factor >> Okay. So, but it that factor of three over 16 years sounds rhymes to me like 10% per year, which or at least market returns. It may be market returns. >> Well, we're talking about two different things. Yeah, I'm talking about the cost to get to a Waymo. I guarantee you the
[21:09] value of the Waymo is so huge relative to not only the cost of developing that team but developing all of the other things that didn't work, it is a very good return on investment. So, I'm not worried about that. I'm looking at what does it cost for each individual thing that makes it all the way through our process and that is getting cheaper at I don't know, 10 20% year-over-year, something like that. Some of that is cuz we're getting better at our job. Some of that is we're graduating things a little bit earlier. A lot of it is because of
[21:39] artificial intelligence. So, it's a mix of all of those things. But, it's cheaper to get to moonshots now than it was in the past. >> But, there are some maybe just to develop this a a little bit more. Some would argue that to solve some of the grandest challenges in science and engineering, it's not going to be like a sort of a one-shot spiky capability launch. It's going to be the result of our economy macro growth that we we grow our capacity of our economy holistically and that enables particular moonshots to
[22:09] become easier. And when I hear you say 3x cost reduction, presumably in real terms, over 16 years, that to me, maybe you take a different position, that to me sounds basically like just the result of markets doubling, say, every 10 years. Stock market doubles every 10 years. So, the result of the reduced cost of moonshots sounds to me like basically just our economy is capable of more things and is basically the same cost. >> I think you're confusing the the cost to paint a house with the value of the house. They're like apples and oranges.
[22:40] They're not the same thing. I'm talking about how cheap it is for us to paint our house, not what the world will pay for our house. >> two moonshots a year, Astro? I mean, why aren't you pushing out 100 baby moonshot Alphabet companies per year? I mean, this is about 10x growth, right? >> True. I mean, again, if all I cared about was moonshots and I didn't care about efficiency, I guess we would do that. Because I'm obsessed about efficiency, there is a cultural rhythm and a set of habits that are really easy to describe,
[23:12] but fiendishly hard to get people to practice. Things like playing the long game, being intellectually honest, showing up with humility, being a team player, and not being an egomaniac. Like, everyone's like, "Yeah, yeah, duh, everyone wants that." But, just take a moment and ask yourself super honestly, is that how your organization really functions today? Right, I can hear the laughter, cuz it isn't. It's not your fault, but someone would have to be maniacally determined
[23:42] to create a microcosm in which those things were actually the adaptive behaviors in order to do that. That's what I'm maniacally focused on is create is the culture engineering necessary to get people to practice the habits that drive innovation efficiently. Fortunately, I know how to do that with a few hundred people. I don't know how to do that with a few thousand people. It's a culture problem. I just don't know how to solve it. >> So, Astro, isn't aren't you making the argument then that the ultimate moonshot is the meta moonshot of solving the
[24:14] organizational problem of launching more moonshots? >> Um, yes, I I'm going to connect your question with the way I think about it, which is I would say the meta moonshot we have been talking about is the meta moonshot is how to systematize radical innovation. So, to your question then, there is a meta meta moonshot, which is how to systematize systematizing radical innovation. How to copy and paste moonshot factories. Which, back to what I was saying before, the world is coming and asking us that. We would love to help. I
[24:45] haven't yet figured out how to do the meta meta moonshot, but we're excited to help. >> Organizational singularity, Salim. >> Yeah, I think we might be we might have that uh methodology for you, because we've been building for 10 years of of us societal layer. We're now 50,000 people globally on how do you actually create an environment where you can spin off these things very, very quickly. So, I think there it's almost like you've solved the inner few layers of Stuart Brand's pace layer thing and we've solved the outer
[25:15] through love. So, >> We should talk about that. >> Dave, you were going to say >> I'd like to take your exact quote and say, "Look, you know, if you Hey, dirty little secret, everybody. Doing a moonshot is really, really easy. You take fanatically focused people who care about a mission, who are really, really smart, and you pour money on them. Great. Now, there will be a world where you swap out the people and you swap in AGIs, and you take out the money and you pour compute on them. That world isn't here yet. There's got to be some transition world we're in
[25:45] right now where there are three AIs and three people, and I poured compute and money on them. And then there's another ratio, and there's Have you Have you started down that journey yet? >> Totally. I mean, All right. I mean, I connect your question to like a side rant. On the one hand, if someone came to you and said, "I want to pitch you my company. Are you ready for my pitch? We use electricity." You would think I was nuts. Like, on the one hand, you'd be frightened if I didn't use electricity
[26:15] in my business, but you'd be like, "So what? That's an implementation detail. What are you doing?" >> Yeah. >> I feel the same way about AI. I actually try to say AI as little as possible, cuz it's it's driving me crazy the way I >> We should say SI now, super intelligence. >> [laughter] >> Sorry, I I I got a PhD 30 years ago in artificial intelligence, so a little stuck on that one. But, >> Me too. >> Uh on the one hand, like what you just described, every one of our moonshots, that's it. It's a bunch of agents and a bunch of people working together to try
[26:46] to solve a problem. That's an implementation detail. What we're obsessed with is the problem, whether it's clean water or it's a way to make grid-scale energy storage or it's education, whether it's how to fix the electric grid, how to make our world serve actually circular so we're not sending trillions of dollars a year to landfill. Like those are the things that we're obsessed about. And then in the background, yeah, it's humans and agents like making it happen. But most of our obsession is the problem and like how to couple that into a business so that it
[27:16] gets bigger over time, so it doesn't cost too much. Let me try >> that into a budget question. So if you said here, clean water is a great one, great you know, moonshot. I need What was it like a a penny a liter or about a penny a liter to really change the world? Okay, so that's a great moonshot. If if I said, let's turn Have you already got a team on that one? Is that >> We have had teams on them in the past. I Yes, we do have a team on it right now. >> Okay, okay. So is there What's bigger, their AI bill or their salary bill? >> It's still their salary in that particular case, but yeah, I don't know by how much, honestly. So
[27:47] I I don't know what fraction it is for sure, but I would say they're absolutely teammates in the process of innovation. And in fact, you know, we are finding ways all the time to automate parts of the moonshot factory so we can go faster and be more efficient. >> Your best guess then, you know, if you're you're hyper-focused on the efficiency of that innovation machine, there's a case study. Right now their AI bill is smaller than their head count bill. When's the day when it hits 50/50, 75% AI and then 100%?
[28:19] >> Well, first of all, I'm not sure anytime in the next decade it's going to be 100% because I think figuring out which problems to solve and making sure they're being solved in societally acceptable ways is an important part of the problem. If you need to have distributors helping to sell this, if you have to land it in communities in ways that that community feels are acceptable to them, like you need humans as part of that process for at least the next decade or two. >> Okay. >> So, but look, it'll go as as fast as is
[28:51] efficient. Like, we're not going to use compute just to use compute. And there were people who are spinning up agents because it feels powerful to spin up agents. But, there is a difference between spinning up an agent actually getting benefit. >> Mhm. >> Astro, you go ahead, Salim. >> Yeah, there's I think of moonshots as two classes. Okay. >> Okay. >> uh pain points. Solve hunger, solve education, solve whatever. And then there's the aspirational moonshots. Elon, get us to Mars, right? Uh um um
[29:21] solve aging, etc. How much of your work in moonshots focuses on one versus the other? Is it an equal balance? Or do you focus only on the problem side? >> I mean, if you worked at X, what I would say is, Jason, can't I have both? I want you to find the most audacious things you can work on that we can move from crazy idea to not crazy idea in less than a decade. And then from not crazy idea to an enduring business in less than another decade. So, if you want to
[29:52] go to Mars, I would start with, "Okay, I'm listening. How is that in less than two decades going to be an enduring business?" I'm not saying it can't, but that would be my task. >> So, it's time your time you're constraining yourself just so you can give a sense of if we can reliably work on this problem over this next decade and spin out something that becomes the main criteria. >> Because >> That makes sense. >> because I'm getting money. Our investor is Alphabet. I have to tell Alphabet, "Here's what I'm doing with your money." And they have to feel good about the
[30:22] sort of flywheel that I'm creating for them. That's >> Astro, one of the things I find fascinating is the the fact that you kill the vast majority of the ideas that are started. Just to give us some numbers here, total number of projects started versus graduated, it's under 1%? >> Uh I think it's close to 2% but call it like 2,000 things over the last 16 years that got code names of which we've had, a little bit depending on how you count, sort of 35 to 50 graduates.
[30:52] >> So, talk about this idea of the importance of killing your babies early. >> Well, um >> No, not not his actual babies. >> Yes, I I I'm going to use I'm going to use the monkey and pedestal analogy >> Please, don't. >> you know well, but imagine we were working on a moonshot together, all of us, and we were trying to get a monkey to stand on the top of a 10-ft pedestal and recite Shakespeare. Which should we do first? Train the monkey or build the pedestal? It is so weirdly tempting, if we're
[31:24] being honest with each other, in something where it's not that extreme a situation, to build the pedestal, cuz then you're half done, and you can get another round of funding, live to fight another day. You can go to your boss and say, "Can I have a promotion? I'm half done, man. Can I Can I have a bonus?" This is what drives the world, and it drives me crazy. So, what we say at X is forget the pedestal, only focus on the monkey. If you can train the monkey, we can always build a pedestal afterwards,
[31:54] and if you can't train the monkey, thank god we didn't waste Alphabet's money and our time on that. And so, it's not that I want to lose. I am like the crown prince of failure, but it's not because I like to fail, it's because you learn nothing when you're right. You only learn things when you're wrong. >> [applause] >> I love it. Crown prince of failure, love it. >> And so, if you accept, which it sounds like you do, that that's true, and
[32:24] moonshots is a process almost entirely taken up with learning, then we're going to go on an adventure, which is defined entirely by how fast we learn, which means if we in any way feel bad in the moment where we learn because we had a model of the world that turned out to be wrong in some way, then we will unconsciously or maybe even consciously avoid that moment. So, we're going to de-stigmatize failure. If you work at X, I need you to be on the mission with me to optimize X's portfolio.
[32:56] It's not that I don't want the teleporter to work, Peter, who's working on this for me, but I want to find out as efficiently as possible if it's going to work. I don't want you in denial about the fact that you just really want the the teleporter to work and we go for like 5 years and we spend tens of millions of dollars and it's still kind of going sideways and it's a zombie. How many people have you here have spent at least 5 years on something that was a zombie where you knew for at least 2 or 3 years that it was a zombie. >> What's the What's the worst the worst failure? Like the the failure where you
[33:26] were like that's really crazy and turns out, you know what? That really was really dumb. >> Oh, we've had >> [laughter] >> lots of those. We We have one right now which is almost certainly wrong. Um it would be a a problem for physics if it turned out to be right. So, um but it's it's working. We do not know why it's working in the lab. We're working ferociously hard to find why
[33:56] we're kidding ourselves so that we can kill it off. But it's about 5 months into trying to discover this sort of like cold fusion moment, which I hate. Cold fusion, it's not cold and it's not fusion. But [laughter] >> It's okay, good. It didn't work anyway. It's all right. >> we don't want to be that group, right? We are trying to be the opposite of that group. >> going to tell us what this thing is? >> Nope. >> Oh, come on. >> Okay, if it works out, we'll tell you about it. >> All right. So, uh >> We have one right now that's like makes me deeply uncomfortable and deeply excited.
[34:26] >> Okay, a couple audience questions here. Rich asked, are there any moonshots that you regret denying in the past as our world's problems have changed, that you would pursue today. >> I mean, we were wrong a bunch. So, I'm about to say no to that question, but I do not want that to be confused with we've never been wrong. We're wrong all the time. >> the question though the question >> Do I regret it though? >> But do you it was something that you reject you never even try. Okay. >> So, I want to be really clear. The cost for a false positive where we believe
[34:57] it's a moonshot, we run it for many years, and then it turns out not to be is very high, many tens of millions of dollars potentially. >> Yeah. >> The cost of a false negative where it actually is a moonshot, but I rejected it is zero. As long as you believe that there's an infinite number of problems in the world and an infinite number of creative solutions, we can always just go draw again from the distribution of moonshots. It costs us nothing to say no. So, I say no and have no regrets even
[35:28] about the stuff we were wrong about. >> We probably don't do it fast enough. >> Paul asks, "What about moonshots and material sciences? Uh do you see any particular promising fields to inquire in that area?" >> Well, I mean, side rant, I'm super excited about the new material space. >> Yeah. >> And I think people haven't played out in their minds cuz there's like such a fervor about the technology and everyone's like, "And then we'll make a business out of it." Yeah, that's harder than it sounds. Look
[35:58] how many centuries it took for us to really refine steel. Steel completely changed the world, but not until it was industrialized in a bunch of ways. >> Mhm. >> There are a lot of new materials that if you could make one flake of room temperature superconductor, that would be pretty cool, Nobel Prize, yay. That is not a business. That's like the first 5% of a business. How do you make, you know, thousands of tons a day of
[36:28] that material. How do you make it with very high ductility so that you could put it into like winding wires and things like that? >> Yeah. >> All of the stuff that it takes, the other 95% of the work for new materials or otherwise, that's what it really takes to make a moonshot. So, I'm super excited about lots of things like new materials, but I'm kind of focused on the hill after the one that people are focused on, which is how would we scale these up as they show up in the world? Not just how do we find them in the first place? >> All right, last quick question from the
[36:58] audience and a quick answer, hopefully. So, other companies keep trying to replicate X and fail. What's the single hardest piece for them to replicate that you think most outside observers still underestimate? >> I really believe it's what I said, which is you have to have a leadership team that is maniacally focused on engineering a culture in which people can show up in the ways that tend to drive radical innovation most efficiently. And, if I can make a plug,
[37:30] just like you have a moonshot podcast, we have a moonshot podcast. There are two seasons out now where you can actually hear about people who went through the process. So, this is not me blah blah blah-ing about moonshots, it's actually the moonshot takers talking about their successes in >> find it? On YouTube? >> If you just search for moonshot factory moonshot podcast, you can find it anywhere you get podcasts. And I would encourage people to listen to this because I think what you'll hear is a reflection of exactly what I said, which
[38:01] is the culture, the microcosm which encourages people to do the things that actually drive innovation. It's it's obvious, you all know what it is, it's making that space, that protected space for those explorers, that's the really hard part. >> I I would like to say something to answer that question. Um you have created a mechanism where you've structured it such that the mothership has left you alone for a long period of time. I've never seen another company do
[38:31] it. Every other company I've ever seen, over time the immune system cannot resist, reaches out its sticky little hands, and snatches it back, and then messes it up invariably. So, congratulations to you for having lasted this long, and congratulations to the Google execs who have left it and had the discipline to leave it on the edge, because that's where all the magic happens. >> Hey everybody, give it up for Astro Teller. >> [music]
[0:35] childhood, yes. >> Yeah, I I guess so. So, um let me let me kick off the first question. So, you've been in the moonshot business for over 20 years. Uh walk us through your advice of what it actually takes to come up with a moonshot and launch a moonshot. You know, how do you make this real? We've got a lot of individuals here who are here because they believe in optimism, they believe in moonshots. You know, in 10 seconds. No, just
[1:05] kidding. Uh what's your advice? I mean, you've learned so much. >> Let Let me Let me give you two quick things. The first one is, let's just define moonshot, at least how we think of a moonshot. >> Which is how I think of it as well. >> Excellent. >> from you. >> So, three basic components. First, there has to be a huge problem with the world that you can name and you want to solve. If you can't name the huge problem, then it's arguably an academic exercise. Second, there has to be some kind of science-fiction sounding product or service that, no matter how unlikely it
[1:36] is you could make it, we could pre-agree that if you made that science-fiction sounding product or service, it would resolve that huge problem with the world. And then, three, there has to be some kind of breakthrough technology that gives us a prayer, at least a tiny chance, at minimum, of being able to make that science-fiction sounding product or service and resolving that huge problem with the world. When you have that, we would call that a moonshot story hypothesis. That does not mean you're going to win, but at least means it's testable.
[2:08] The second thing I would say is like one of the fundamental issues that we found at X is that you have to have two things in equal amounts in order to be a moonshot explorer, to be a moonshot team. The first one is you have to have very high audacity. You have to be willing to suspend your disbelief for non-stupid reasons and go on unlikely journeys that might actually profoundly help the world. But
[2:40] the second thing you have to have in equal measure is humility. You have to know from the first moment you set out on that unlikely journey that it's unlikely because if you don't, you will go ruinously far down that path before you find out that that is one of the 99% that just isn't going to work out. So, being able to say right at the beginning, "I'm proud that I'm trying this, but it probably won't work. Let me learn as fast and as cheaply as I can whether this is even possible." allows
[3:11] you to get onto the next moonshot when that moonshot's not the right one to do. >> So, your advice for folks, I mean, at with the challenge of picking a moonshot, I mean, right now the speed at which you can test things, create things, right? But what about the beginning, choosing the idea? Any advice there? Well, I mean, let me put it this way. What do you do to shut down an idea quickly? >> So, sometimes the idea is just not big enough. Sometimes the idea is too
[3:41] reasonable or likely to succeed. Sometimes it's just not good for the world or the potential for too many sort of secondary problems associated with it is just too high. Uh often it's techno-economics. I cannot tell you the number of moonshots that feel so good to people in the early days. Often these are very technical people and it's not until you say, "Really? Like let's go step by step. What is the best case scenario for the
[4:12] build of materials that you would make this from? What is like just the raw cost in weight of the stuff that would go into that? What are people really going to pay? Could this even plausibly zero >> uh first principle thinking in that regard. >> Right. First principles thinking about the techno-economics. You can often kill moonshots in the very early days because it's just not going to turn into an enduring business and at least what I believe is that purpose and profit can
[4:43] support each other and then if you're doing something that's going to lose money, it's probably not going to change the world. >> Yeah. >> Dave >> Yeah, I'm really curious about you know Alex and Peter have this solve everything book. It's phenomenal but the rate at which physics and material science chemistry is going to change just astronomically accelerating. So when you joined Google what 25 23 years ago? How long ago? >> Uh 16 and a half years ago is I co-founded X 16 and a half years
[5:13] >> 16 and a half years ago. Okay. And at the time it was about a 30 billion dollar company? >> Ish. Yes. It was It was very sweet. It was very small back then. >> Yeah. A little 30 billion dollar company. Now just about half a trillion dollar company. So the scale of R&D in a corporate environment like that has grown miraculously. Just absolutely tremendously. And I think you were saying backstage that the ideal target for a new idea 100 ideas a year two of them succeed. >> That's right. We we start one to 200 ideas a year that make it far enough
[5:43] that they end up with a code name. I who knows how many we actually look at but one to 200 ideas a year that are make it far enough we get a code name. Of those, about 5 to 6 years later, we graduate to moonshots out of X. >> Wow. So, 2% hit rate. >> And so then if you know if if AGI is imminent and the idea rate >> No, it's here. >> is astronomically >> Right. >> Does does that grow on that same exponential curve? Just that the number of things at the front of the funnel go from 100 to 1,000 to 10,000 to
[6:13] >> Oh, for sure. I mean, there's no shortage of problems with the world. There's no shortage of creative ideas for running at those problems. I think AI is going to somewhat shrink that time from a crazy idea to us having de-risked it to the point where we can say definitively here's evidence it's not a crazy idea anymore. So, that's getting shorter. I hope that AI is going to let us raise our audacity even higher, so we'll be shooting higher, and so we'll be able to do more in a shorter amount of time. But I'm But fundamentally, I
[6:44] think the structure will be >> May I think I think uh you know, if you asked me 3 years ago, you know, what are the big things that came out of out of X, uh Waymo is at the top of the list. Um Google Google Brain >> So, that's what I was going to ask about. So, Google Brain was probably like, "Ah, yeah, this is cool." Up until some point a couple years ago when it was like, "No, this is the entire future of Google now." >> guys just spent 80% of the last 3 hours talking about the stuff that arguably like Google Brain sort of birthed into the >> What What is Google Brain? Define it
[7:15] because we You know DeepMind, but we don't know Google Brain. >> So, Google Brain, uh which started at X 15 and a half years ago or so, was one of the earliest things that we started. At the time, there were a set of academics who were kind of out of vogue. This was like Andrew Ng, Yann LeCun, a few others who were saying, "I think scale might matter." And neural nets were absolutely dead in the field at the time. And they're saying, "I think scale might matter. It it actually bring neural networks back. And everyone else in the field said, Nope, you're crazy.
[7:46] Andrew Ng came to Google X at the time, and we set up what it was what was called Google Brain with Jeff Dean. So, the two of them paired together to say, What if we industrialized neural networks? What if we made them tens of thousands of times bigger than anyone had ever done before? Could that actually be the breakthrough that everyone's been waiting for? So, we were the first ones who really made what Fast Forward >> called deep learning. >> TPUs ultimately came from Google Brain.
[8:20] The Transformer, which is sort of underpins everything, it's literally the T in ChatGPT, that came from Google Brain. So, you know, our job is to make the seed crystals that ultimately go on to change the world, and I think Google Brain counts as one of them. >> But there must have been a long period of time there where that was kind of quaint and promising, and then a period of time just a recently where it's like, Holy crap, that was a lot more than quaint and promising. >> Of course, but I think the reality of any innovation is it takes a super long time. Every single innovation is like
[8:52] a overnight success that was 15 to 20 years in the making. And just because the general public only wakes up to it in year 18 of that slog, it feels like it happened overnight. >> Every overnight success after 11 years of hard work. >> For sure. And >> So, Astro, you you think I watched hundreds of moonshots, evaluated hundreds. What are a couple of examples of moonshots that you would like to have seen that didn't quite make it, didn't cut the mustard, whatever you would have loved to have had them come to fruition?
[9:24] >> I I mean, I have a very long list. I'll give you two examples, but >> Sure. >> we've been working on clean water over and over again. If you can make clean water, if you could pull it from the atmosphere, if you could desal for a tenth the price, you have to be able to get to like a penny a liter all-in cost for it to really change the world. But we have almost 3 billion people in the world who are water stressed, who don't have clean water to drink. That's going to get worse as climate change picks up. We're
[9:54] going to have hundreds of millions of climate refugees around the world, and it's often going to start because they don't have enough clean water. This is one of the biggest problems to solve. >> Mhm. >> But techno-economics we've done some great work in the space, but that it's going to be 10 cents a liter. >> Nope. >> Well, we'll stop doing that. Try it again. So, that's an example of something we keep coming back to. Um >> Two more. >> Two more, okay. Um being able to uh time
[10:26] shift and location shift energy. And I say it that way on purpose because if you say transmission lines and batteries, people tend to narrow their focus in. But being able to time shift and location shift energy is absolutely going to change the world. We've taken a bunch of runs at those. So far, you know, we haven't been super excited with what we've found. Education is another one. Education is not working. It's It's not working in the developed world. It's not working in the developing world.
[10:56] >> [applause] >> We're going to keep coming back to that one till we find something cuz that has got to be one of the sort of most fundamental moonshots that you could do. >> Got it. >> Love it. >> Alex, please. >> Yeah, I'd love to talk a bit about the history of moonshots. So, one of your grandfathers invented the hydrogen bomb. And then if if you look at history, and I I should add maybe humanity was in the process once upon a time of landing humans on the moon in actual moonshots. And then that stopped. Sometime one can quibble over the chronology, but
[11:26] arguably depending on which version of the great stagnation thesis one subscribes to, there was maybe up to a 50-year period, call it 1970 to 2020-ish. Again, one can quibble at the margins, when it would seem that there was, call it a reduced launch velocity of moonshots from humanity as all. We stopped landing humans on the moon. We stopped inventing, arguably, applied physical innovations, regardless of their warfare potential or not, like hydrogen bombs. And then, again,
[11:57] arguably, sometime around 2020 or so, plus or minus a few years, moonshots started again. We got superintelligence that actually works. We're getting robots that actually work. Physical intelligence, so-called, is seeming to work. We got, arguably, medical nanotechnology in the form of lipid nanoparticles that helped to treat a pandemic. So, a few related subquestions. One, do you believe, do you subscribe to any variant of a great stagnation hypothesis, that there was a half-century pause in moonshots? And to
[12:28] the extent you do, another question, how do we prevent that from happening again? >> Well, um the first thing is I know you didn't maybe intend it that way, but I'm going to take that as an enormous compliment and 11-year in the making joke that we just made. Since we started in 2010 and you said 2020 several times, exactly my point is we started literally a moonshot factory in 2010. >> you're actually you brought moonshots back. >> Well, yes, to some extent, I would like to
[12:59] think that I helped to lead we did. >> [applause] >> So, that's the first half of the answer. And and yes, I do think that I don't know I don't think anyone was planning on it being a stagnation, but I think functionally we had lost our way a little bit somewhere between, you know, you had your Star Trek stuff up here a little bit. >> Yes. Somewhere between like the 1960s and NASA and the inspiration that that created, there was a period where we sort of lost the explorer spirit.
[13:29] >> you think? What do do you have an ideology, a diagnosis? What do you think went wrong? >> I don't know, but but here is a partial answer to that, and then I want to get to the future. Here's Can I tell you all a a dirty little secret? It'll be our little secret. Taking moonshots is really, really easy. >> Huh. >> It's it's it's laughably easy if you don't care about efficiency. You just find some super energetic people who are sort of delusionally optimistic, pour a bunch of money on
[14:01] them, you will absolutely get some moonshots. It's just not a very good return on investment. And so, I think we had a period, mostly because of World War II and the Cold War, where people didn't care about the return on investment. >> Yeah. >> Then there was a period where people probably still wanted moonshots, but they cared about the return on investment, and so there was less of it. What X has been trying to practice is making a moonshot factory, being able to
[14:32] get at that audacity efficiently enough to systematize the process, not just to do it a lot, but to do it efficiently so it is rational for people to put a lot more money in. And to your point about the future, good news, I think, we get approached now on a weekly basis from some large company or some country saying they want to set up their own moonshot factory. I'm not sure exactly how we can help them with that, but there is a hunger to
[15:02] do more of this, and so good news, I expect you'll see a lot more of it in the future, not just for that. >> But Astro, take a second, and there are a number of people here in the audience that are running companies that do want to set up a moonshot factory, and and we have discussed a little bit about this, right? Which Let me paraphrase what I've learned from you. You're my moonshot mentor here, uh which is, you know, your core organization is responsible for delivering 10% profit year after year.
[15:32] And And Salim and I have written about this that your moonshot factory, if you would, needs to be on the edge of the organization where your crazy thinkers exist, right? And you Again, quoting you, you know, if you work on anything at 10%, you're fired. Focus on 10x, the 1,000%. And then have that side moonshot organization report directly to the CEO. Is that still the formulation? >> Absolutely. You know, I've run the following experiment. I apologize to Peter who's uh heard this a bunch.
[16:02] Salim's probably heard it a few times, too, but just bear with me for 45-50 seconds. Choice A, choice B. Choice A, you can give a a million dollars of value to your business this year, but it's And it's guaranteed. Or B, choice B, billion, you can give a billion dollars of value to your business this year, but it's not guaranteed. It's one chance in a hundred. So, A, million guaranteed, B, billion, one chance in a hundred. Who's choosing choice A?
[16:32] All right, almost nobody here, almost nobody in the world raises their hand. WHO'S CHOOSING CHOICE B? WOO! All right, so you all passed the math test. It cuz it has 10 times the expected utility. I've done this with CXOs the world over. And I've said, "Okay, now keep your hand up if on their best days, in your wildest dreams, your manager, your CEO, your board of directors actually supports you doing choice B." >> Yeah. >> And every hand in the room goes down.
[17:02] Yeah. And I say, "You don't need a lecture on innovation, you need a new manager." >> Yeah. >> This So, the formulation The formulation you just described is exactly right, and it's because of this thing. You have to sequester the choice B's to a place where they're actually wanted and and the mess that comes with them is tolerated. >> This is Lockheed Skunk Works off on the edge. This is Steve Jobs taking the Mac team off campus to build it. >> Exactly. >> Yeah. >> The the capability of the human mind to
[17:34] do this is very hard. There's something I've been observing over the last few years, and I'd love for you to comment on this. Peter wrote about this in one of his original books, the demonetization of the world, right? So, throughout human history, it's always been true that advanced technologies cost a lot. And today, for the first time in human history, advanced technologies are cheap. See solar energy, cheap sensors are cheap, blockchain's open source and free, etc., etc. This should enable, you know, for a CFO or for a big company,
[18:04] when you want to do very something very disruptive, you had to make a big bet, put a lot of resources. CFO hated you, they would defund it as fast as they could. Today, you can take a portfolio of very disruptive experiments at the edge. So, now it just comes down to mindset, and are you willing to go for it? Do you agree with that, and how much are you seeing companies do that? >> Uh I I do agree. I think it is a I hear a lot, "Oh, well, Alphabet has lots of money, and so that's why this is happening." I guarantee you that is not
[18:34] why. We have tiny teams. The teams that graduate from us, Google Brain when they graduated, they were like 18 people, and that's pretty normal. We keep our teams tiny because we want I we I I say to them all the time is, "I want you to find a cheat in the video game of life, a Gordian knot-chopping moment. If you can't find that, I don't want you solving things the normal way. I'd rather you stop and we go find a different moonshot, cuz that's what moonshots mean."
[19:04] So, I believe it is mostly a mindset issue, and for those of you out there who are interested in learning more, we did a masterclass certificate course. It's about 20 hours if you do all the work. It's about 3 and 1/2 hours if you just watch the videos. I encourage people to go watch. It's on breakthrough innovation and it will step people through. It's our secret sauce is not secret. We're telling people exactly how to do it. >> And tonight, Astro, there's an evening program by your team doing a 50-minute condensed version of that, you know,
[19:36] that 10-hour course. >> Exactly. 7:00 tonight, Joe Sergeant. It's going to be a lot of fun. It's on rapid prototyping. >> Amazing. >> I'd also love to have maybe better understand and develop the record on the economics of moonshots. I I could make on the one hand an argument that the real cost of moonshots increases over time because expectations go up and we see arduous problems solved left and right by people sitting on the toilet casually solving hard problems. I could make the argument in the opposite direction that the real cost of
[20:06] achieving a moonshot is hyper-deflating because we have super intelligence now and we can solve lots of hard problems. I could probably make some sort of mediocre middle-of-the-road argument as well. What would be your position on the long-term economics of the real costs of moonshots hyper-inflating, hyper-deflating, staying the same. >> Uh I mean, we track because we're obsessed with the efficiency of getting at moonshots. We track what it costs for us to get to our graduates very carefully and I can tell you that it's
[20:37] down by about a factor three over the last 16 years. There's a lot of waste and complexity and it's a bit of a lagging indicator. So, some of what >> Wait, wait, factor of three over the that that's inflation. >> No, no, >> [laughter] >> it is a factor of three cheaper, just to be clear. It's a factor >> Okay. So, but it that factor of three over 16 years sounds rhymes to me like 10% per year, which or at least market returns. It may be market returns. >> Well, we're talking about two different things. Yeah, I'm talking about the cost to get to a Waymo. I guarantee you the
[21:09] value of the Waymo is so huge relative to not only the cost of developing that team but developing all of the other things that didn't work, it is a very good return on investment. So, I'm not worried about that. I'm looking at what does it cost for each individual thing that makes it all the way through our process and that is getting cheaper at I don't know, 10 20% year-over-year, something like that. Some of that is cuz we're getting better at our job. Some of that is we're graduating things a little bit earlier. A lot of it is because of
[21:39] artificial intelligence. So, it's a mix of all of those things. But, it's cheaper to get to moonshots now than it was in the past. >> But, there are some maybe just to develop this a a little bit more. Some would argue that to solve some of the grandest challenges in science and engineering, it's not going to be like a sort of a one-shot spiky capability launch. It's going to be the result of our economy macro growth that we we grow our capacity of our economy holistically and that enables particular moonshots to
[22:09] become easier. And when I hear you say 3x cost reduction, presumably in real terms, over 16 years, that to me, maybe you take a different position, that to me sounds basically like just the result of markets doubling, say, every 10 years. Stock market doubles every 10 years. So, the result of the reduced cost of moonshots sounds to me like basically just our economy is capable of more things and is basically the same cost. >> I think you're confusing the the cost to paint a house with the value of the house. They're like apples and oranges.
[22:40] They're not the same thing. I'm talking about how cheap it is for us to paint our house, not what the world will pay for our house. >> two moonshots a year, Astro? I mean, why aren't you pushing out 100 baby moonshot Alphabet companies per year? I mean, this is about 10x growth, right? >> True. I mean, again, if all I cared about was moonshots and I didn't care about efficiency, I guess we would do that. Because I'm obsessed about efficiency, there is a cultural rhythm and a set of habits that are really easy to describe,
[23:12] but fiendishly hard to get people to practice. Things like playing the long game, being intellectually honest, showing up with humility, being a team player, and not being an egomaniac. Like, everyone's like, "Yeah, yeah, duh, everyone wants that." But, just take a moment and ask yourself super honestly, is that how your organization really functions today? Right, I can hear the laughter, cuz it isn't. It's not your fault, but someone would have to be maniacally determined
[23:42] to create a microcosm in which those things were actually the adaptive behaviors in order to do that. That's what I'm maniacally focused on is create is the culture engineering necessary to get people to practice the habits that drive innovation efficiently. Fortunately, I know how to do that with a few hundred people. I don't know how to do that with a few thousand people. It's a culture problem. I just don't know how to solve it. >> So, Astro, isn't aren't you making the argument then that the ultimate moonshot is the meta moonshot of solving the
[24:14] organizational problem of launching more moonshots? >> Um, yes, I I'm going to connect your question with the way I think about it, which is I would say the meta moonshot we have been talking about is the meta moonshot is how to systematize radical innovation. So, to your question then, there is a meta meta moonshot, which is how to systematize systematizing radical innovation. How to copy and paste moonshot factories. Which, back to what I was saying before, the world is coming and asking us that. We would love to help. I
[24:45] haven't yet figured out how to do the meta meta moonshot, but we're excited to help. >> Organizational singularity, Salim. >> Yeah, I think we might be we might have that uh methodology for you, because we've been building for 10 years of of us societal layer. We're now 50,000 people globally on how do you actually create an environment where you can spin off these things very, very quickly. So, I think there it's almost like you've solved the inner few layers of Stuart Brand's pace layer thing and we've solved the outer
[25:15] through love. So, >> We should talk about that. >> Dave, you were going to say >> I'd like to take your exact quote and say, "Look, you know, if you Hey, dirty little secret, everybody. Doing a moonshot is really, really easy. You take fanatically focused people who care about a mission, who are really, really smart, and you pour money on them. Great. Now, there will be a world where you swap out the people and you swap in AGIs, and you take out the money and you pour compute on them. That world isn't here yet. There's got to be some transition world we're in
[25:45] right now where there are three AIs and three people, and I poured compute and money on them. And then there's another ratio, and there's Have you Have you started down that journey yet? >> Totally. I mean, All right. I mean, I connect your question to like a side rant. On the one hand, if someone came to you and said, "I want to pitch you my company. Are you ready for my pitch? We use electricity." You would think I was nuts. Like, on the one hand, you'd be frightened if I didn't use electricity
[26:15] in my business, but you'd be like, "So what? That's an implementation detail. What are you doing?" >> Yeah. >> I feel the same way about AI. I actually try to say AI as little as possible, cuz it's it's driving me crazy the way I >> We should say SI now, super intelligence. >> [laughter] >> Sorry, I I I got a PhD 30 years ago in artificial intelligence, so a little stuck on that one. But, >> Me too. >> Uh on the one hand, like what you just described, every one of our moonshots, that's it. It's a bunch of agents and a bunch of people working together to try
[26:46] to solve a problem. That's an implementation detail. What we're obsessed with is the problem, whether it's clean water or it's a way to make grid-scale energy storage or it's education, whether it's how to fix the electric grid, how to make our world serve actually circular so we're not sending trillions of dollars a year to landfill. Like those are the things that we're obsessed about. And then in the background, yeah, it's humans and agents like making it happen. But most of our obsession is the problem and like how to couple that into a business so that it
[27:16] gets bigger over time, so it doesn't cost too much. Let me try >> that into a budget question. So if you said here, clean water is a great one, great you know, moonshot. I need What was it like a a penny a liter or about a penny a liter to really change the world? Okay, so that's a great moonshot. If if I said, let's turn Have you already got a team on that one? Is that >> We have had teams on them in the past. I Yes, we do have a team on it right now. >> Okay, okay. So is there What's bigger, their AI bill or their salary bill? >> It's still their salary in that particular case, but yeah, I don't know by how much, honestly. So
[27:47] I I don't know what fraction it is for sure, but I would say they're absolutely teammates in the process of innovation. And in fact, you know, we are finding ways all the time to automate parts of the moonshot factory so we can go faster and be more efficient. >> Your best guess then, you know, if you're you're hyper-focused on the efficiency of that innovation machine, there's a case study. Right now their AI bill is smaller than their head count bill. When's the day when it hits 50/50, 75% AI and then 100%?
[28:19] >> Well, first of all, I'm not sure anytime in the next decade it's going to be 100% because I think figuring out which problems to solve and making sure they're being solved in societally acceptable ways is an important part of the problem. If you need to have distributors helping to sell this, if you have to land it in communities in ways that that community feels are acceptable to them, like you need humans as part of that process for at least the next decade or two. >> Okay. >> So, but look, it'll go as as fast as is
[28:51] efficient. Like, we're not going to use compute just to use compute. And there were people who are spinning up agents because it feels powerful to spin up agents. But, there is a difference between spinning up an agent actually getting benefit. >> Mhm. >> Astro, you go ahead, Salim. >> Yeah, there's I think of moonshots as two classes. Okay. >> Okay. >> uh pain points. Solve hunger, solve education, solve whatever. And then there's the aspirational moonshots. Elon, get us to Mars, right? Uh um um
[29:21] solve aging, etc. How much of your work in moonshots focuses on one versus the other? Is it an equal balance? Or do you focus only on the problem side? >> I mean, if you worked at X, what I would say is, Jason, can't I have both? I want you to find the most audacious things you can work on that we can move from crazy idea to not crazy idea in less than a decade. And then from not crazy idea to an enduring business in less than another decade. So, if you want to
[29:52] go to Mars, I would start with, "Okay, I'm listening. How is that in less than two decades going to be an enduring business?" I'm not saying it can't, but that would be my task. >> So, it's time your time you're constraining yourself just so you can give a sense of if we can reliably work on this problem over this next decade and spin out something that becomes the main criteria. >> Because >> That makes sense. >> because I'm getting money. Our investor is Alphabet. I have to tell Alphabet, "Here's what I'm doing with your money." And they have to feel good about the
[30:22] sort of flywheel that I'm creating for them. That's >> Astro, one of the things I find fascinating is the the fact that you kill the vast majority of the ideas that are started. Just to give us some numbers here, total number of projects started versus graduated, it's under 1%? >> Uh I think it's close to 2% but call it like 2,000 things over the last 16 years that got code names of which we've had, a little bit depending on how you count, sort of 35 to 50 graduates.
[30:52] >> So, talk about this idea of the importance of killing your babies early. >> Well, um >> No, not not his actual babies. >> Yes, I I I'm going to use I'm going to use the monkey and pedestal analogy >> Please, don't. >> you know well, but imagine we were working on a moonshot together, all of us, and we were trying to get a monkey to stand on the top of a 10-ft pedestal and recite Shakespeare. Which should we do first? Train the monkey or build the pedestal? It is so weirdly tempting, if we're
[31:24] being honest with each other, in something where it's not that extreme a situation, to build the pedestal, cuz then you're half done, and you can get another round of funding, live to fight another day. You can go to your boss and say, "Can I have a promotion? I'm half done, man. Can I Can I have a bonus?" This is what drives the world, and it drives me crazy. So, what we say at X is forget the pedestal, only focus on the monkey. If you can train the monkey, we can always build a pedestal afterwards,
[31:54] and if you can't train the monkey, thank god we didn't waste Alphabet's money and our time on that. And so, it's not that I want to lose. I am like the crown prince of failure, but it's not because I like to fail, it's because you learn nothing when you're right. You only learn things when you're wrong. >> [applause] >> I love it. Crown prince of failure, love it. >> And so, if you accept, which it sounds like you do, that that's true, and
[32:24] moonshots is a process almost entirely taken up with learning, then we're going to go on an adventure, which is defined entirely by how fast we learn, which means if we in any way feel bad in the moment where we learn because we had a model of the world that turned out to be wrong in some way, then we will unconsciously or maybe even consciously avoid that moment. So, we're going to de-stigmatize failure. If you work at X, I need you to be on the mission with me to optimize X's portfolio.
[32:56] It's not that I don't want the teleporter to work, Peter, who's working on this for me, but I want to find out as efficiently as possible if it's going to work. I don't want you in denial about the fact that you just really want the the teleporter to work and we go for like 5 years and we spend tens of millions of dollars and it's still kind of going sideways and it's a zombie. How many people have you here have spent at least 5 years on something that was a zombie where you knew for at least 2 or 3 years that it was a zombie. >> What's the What's the worst the worst failure? Like the the failure where you
[33:26] were like that's really crazy and turns out, you know what? That really was really dumb. >> Oh, we've had >> [laughter] >> lots of those. We We have one right now which is almost certainly wrong. Um it would be a a problem for physics if it turned out to be right. So, um but it's it's working. We do not know why it's working in the lab. We're working ferociously hard to find why
[33:56] we're kidding ourselves so that we can kill it off. But it's about 5 months into trying to discover this sort of like cold fusion moment, which I hate. Cold fusion, it's not cold and it's not fusion. But [laughter] >> It's okay, good. It didn't work anyway. It's all right. >> we don't want to be that group, right? We are trying to be the opposite of that group. >> going to tell us what this thing is? >> Nope. >> Oh, come on. >> Okay, if it works out, we'll tell you about it. >> All right. So, uh >> We have one right now that's like makes me deeply uncomfortable and deeply excited.
[34:26] >> Okay, a couple audience questions here. Rich asked, are there any moonshots that you regret denying in the past as our world's problems have changed, that you would pursue today. >> I mean, we were wrong a bunch. So, I'm about to say no to that question, but I do not want that to be confused with we've never been wrong. We're wrong all the time. >> the question though the question >> Do I regret it though? >> But do you it was something that you reject you never even try. Okay. >> So, I want to be really clear. The cost for a false positive where we believe
[34:57] it's a moonshot, we run it for many years, and then it turns out not to be is very high, many tens of millions of dollars potentially. >> Yeah. >> The cost of a false negative where it actually is a moonshot, but I rejected it is zero. As long as you believe that there's an infinite number of problems in the world and an infinite number of creative solutions, we can always just go draw again from the distribution of moonshots. It costs us nothing to say no. So, I say no and have no regrets even
[35:28] about the stuff we were wrong about. >> We probably don't do it fast enough. >> Paul asks, "What about moonshots and material sciences? Uh do you see any particular promising fields to inquire in that area?" >> Well, I mean, side rant, I'm super excited about the new material space. >> Yeah. >> And I think people haven't played out in their minds cuz there's like such a fervor about the technology and everyone's like, "And then we'll make a business out of it." Yeah, that's harder than it sounds. Look
[35:58] how many centuries it took for us to really refine steel. Steel completely changed the world, but not until it was industrialized in a bunch of ways. >> Mhm. >> There are a lot of new materials that if you could make one flake of room temperature superconductor, that would be pretty cool, Nobel Prize, yay. That is not a business. That's like the first 5% of a business. How do you make, you know, thousands of tons a day of
[36:28] that material. How do you make it with very high ductility so that you could put it into like winding wires and things like that? >> Yeah. >> All of the stuff that it takes, the other 95% of the work for new materials or otherwise, that's what it really takes to make a moonshot. So, I'm super excited about lots of things like new materials, but I'm kind of focused on the hill after the one that people are focused on, which is how would we scale these up as they show up in the world? Not just how do we find them in the first place? >> All right, last quick question from the
[36:58] audience and a quick answer, hopefully. So, other companies keep trying to replicate X and fail. What's the single hardest piece for them to replicate that you think most outside observers still underestimate? >> I really believe it's what I said, which is you have to have a leadership team that is maniacally focused on engineering a culture in which people can show up in the ways that tend to drive radical innovation most efficiently. And, if I can make a plug,
[37:30] just like you have a moonshot podcast, we have a moonshot podcast. There are two seasons out now where you can actually hear about people who went through the process. So, this is not me blah blah blah-ing about moonshots, it's actually the moonshot takers talking about their successes in >> find it? On YouTube? >> If you just search for moonshot factory moonshot podcast, you can find it anywhere you get podcasts. And I would encourage people to listen to this because I think what you'll hear is a reflection of exactly what I said, which
[38:01] is the culture, the microcosm which encourages people to do the things that actually drive innovation. It's it's obvious, you all know what it is, it's making that space, that protected space for those explorers, that's the really hard part. >> I I would like to say something to answer that question. Um you have created a mechanism where you've structured it such that the mothership has left you alone for a long period of time. I've never seen another company do
[38:31] it. Every other company I've ever seen, over time the immune system cannot resist, reaches out its sticky little hands, and snatches it back, and then messes it up invariably. So, congratulations to you for having lasted this long, and congratulations to the Google execs who have left it and had the discipline to leave it on the edge, because that's where all the magic happens. >> Hey everybody, give it up for Astro Teller. >> [music]