Science Corner特别节目！David Friedberg、Cleo Abram、Alex Filippenko与Keller Rinaudo Cliffton

Good morning. Where are my besties? They’re not here. What does that mean? Science Corner. I have a guest host because my besties abandoned me for Science Corner. Let’s see who it is.

Cleo, you’re one of the fastest-growing channels on YouTube right now.

Former Vox journalist who left to go independent on YouTube. She went from 0 to 5 million subscribers in just 3 years.

I don’t know many other YouTube creators who are going to go to those lengths.

There’s a lot of very lucrative fearmongering going on. That’s why I want to bring a more optimistic point of view into the conversation, to help people imagine what could go right. That’s why I went independent.

Ladies and gentlemen, please welcome Cleo Abram.

Welcome. Thank you. Thanks for being here.

Thanks for having me.

Grab a seat. You were here all day yesterday.

Yeah.

How was it?

Having a great time. This is my first All-In Summit. I’m so excited to be here.

Welcome. So, Cleo, you have 6 million subscribers on your YouTube channel. We have under 1 million. Thank you for having us on your show.

How did you do it? What happened? You were at Vox before.

I was. Yeah.

And you were an independent director. I mean, you were doing other projects. Tell us how you set up this channel on YouTube, why you did it, and how it got so big so fast.

Huge If True seems to me to be a bit of a microcosm of this big shift that we’re in with media generally right now. I was at a media company making what we call explainer journalism—taking complicated issues and making them understandable both to me and to millions of people.

I went independent to start this show because there was something that I felt like I was missing when I looked out into my media diet. I really wanted to find a show that was optimistic, that helped me see where the people working on hard problems were, and how they were making them better every day in a way that I could understand and participate in.

So, I left the media company where I was, started this show, and had the opportunity, because of what YouTube offers, to reach a global audience very quickly and find out, “Oh my God, I’m not alone. Oh my God, there are millions of people who also want this kind of show.”

YouTube made the bet that if you allow anyone to create their best creative work, the widest audience will watch. YouTube has become, in the last 18 months, I think, the most-watched streaming platform on televisions. So, we’re in the middle of this big moment of change in media and how media gets made. I don’t think most people know that it’s really happening. They know that YouTube shows can get big, but they don’t really understand this shift that we’re in.

By the way, the shift is also very exciting for streamers, because they’re looking at this and saying, “You know, the Netflixes of the world—I used to make a Netflix show as well—can look at this incredible new wealth of creativity and IP and say, ‘Oh my God, who do we want to work with to give upfront capital to make something that’s even bigger?’”

That’s something you and I have talked about, because if you’re on Netflix today—and we’re going to talk with Neal and Ari today—you can, if you’re an independent director, go to Netflix and they’re like, “Okay, we’ll pay for your production costs plus 10%.”

That’s quite different from what it used to be like. when you made friends, you could make the show and then you could eventually make like a VC. You can make hundreds of millions of dollars if it worked out and became a massive show, but you’re basically capped at Netflix.

YouTube is quite different. So, there seem to be two incentives to go to YouTube: creative freedom and economic incentive. But how does financing happen? Where can creators drive the engine to fund and create new content?

Well, most YouTubers have ad-funded businesses. What that requires is that you go out on your own, as we did with Huge If True. We went independent, started this show, and the show grows. Then you’re able to get sponsors who, in turn, fund better and better work, and it continues to scale.

The traditional model of paying upfront for a show that the streamer then owns offers something very different. I think we’re in an interesting, flexible moment of change right now, where Netflix might say, “Wow, we see a really exciting show on YouTube. We want to allow that creator to make a version of that IP that is bigger, and we’ll invest in that upfront.”

What I think we’re seeing is that, for the same creator and the same kind of IP, you can have a really wonderful relationship between the kind of show that you can make when you can reach global audiences immediately, grow, and see how far you can take it with an advertising model.

At the same time, you might be able to take that gem of an idea and say, “What would I do if I had upfront capital?” I think there’s a really interesting way in which these things all work together.

Some of the headlines make it seem as though this is an antagonistic moment in media. I think it’s really great for everybody. I think it’s really, really exciting. I’m also the optimist, so of course I’m going to say that.

Yeah. But I’m actually curious to hear what Neal thinks at YouTube. I know you know him, because it opens up a window right now for YouTube to suck up some of the best content creators in the world from the more traditional platforms—broadcast and streaming.

Yeah, it goes both ways.

Yeah. But just talking about your show, your show is so great because it really meets what I always say is missing in media today, which is that we’ve got this deep sort of techno-pessimism. Everyone thinks that technology always has a catch. There’s always something bad emerging. Robots are going to kill us all. AI is going to wipe out human civilization. Nuclear power is going to melt down and destroy neighborhoods.

Every technology has some negative angle, but then that becomes the cycle. You watch all the shows on Netflix, you watch all the movies—Erin Brockovich, for example. The ones that work, the ones that seem to resonate, which means that’s what people truly want, are the ones that talk about things gone wrong.

But your show is quite different. You talk about, “What if things go right?” Why do you think that’s resonating? Are we changing, or are you capturing a small audience while the bigger one is still sort of technopessimistic?

The best thing by far about making this show is realizing that there are millions of people out there who also want that same kind of work. You see it with Science Corner in so many ways; those are very similar in tone.

When I started this show, I was really looking for a part of my media diet that I wasn’t getting anywhere else. That’s what makes making something yourself on YouTube so special. You’re creating something and asking, “Are there millions of people out there like me?” The answer turns out to be yes.

With respect to optimistic science and tech content specifically, the reason why I make it in the first place is that we spend months on these episodes. We travel all around the world. We invest a huge amount in the animations and the technical explainers so that you can understand, without any background at all, quantum computing, the impacts of supersonic planes, and how we’re trying to bring them back.

I was in a zero-gravity plane the other day, trying to explain the cutting edge of gravity research and theoretical physics. These are things that millions of people can understand if you explain them in the right kinds of ways. That’s what we try to do every day.

The reason why we do that is because we genuinely believe that when people see those better futures, they’ll help build them. That’s what I want to do. I’m not an engineer. I’m not a scientist. I look out at the world and think, “Wow, there are so many people working on hard problems. I want to know how I can participate.”

My hope is that’s what we’re doing every day.

We used to have that after World War II.

I always tell people that Disneyland opened in 1955. There’s a YouTube channel called the Disney History Institute, and it shows what Tomorrowland was like when Disney opened in 1955. It was all about building a better world with all of these crazy technologies: rockets to the Moon, plastic so we could all have cheap furniture.

There was a crazy device that they had in the kitchen called the microwave, where you could cook in 30 seconds so you wouldn’t have to sit around and cook for hours. But we’ve lost that. I really hope that your content resonates with more people and that we get there again.

Cleo’s going to join me this morning for 2 really fun panels that we’re going to have, and we’re going to kick it off now.

We used to look up in the sky and wonder at our place in the stars. How thrilling must it be to truly discover something or understand something that no human on Earth has ever seen or understood?

He was a member of both the Supernova Cosmology Project and the High-Z Supernova Search Team, which discovered that the universe is accelerating—a leader in all of these undertakings.

That’s one of the big questions of cosmology. Ladies and gentlemen, please welcome Alex Filippenko.

Wow. Wow. This is so fantastic to see you all here. Good morning, David. Thank you for inviting me to Science Corner. It’s such a pleasure.

Most of you probably don’t know that, in fact, David was a student of mine at UC Berkeley 28 years ago and became an astrophysics major.

In fact, I feel like I had some influence on him. I’ll take some credit. As Joe Tsai said yesterday, teachers want their students to become more successful, to become better than they are. I always knew that David would be very successful in his career, but I didn’t know that he’d be quite this successful. So, good job, David.

I’d also like to officially acknowledge California’s 175th birthday today—California Admission Day. We were told that yesterday, and I looked it up, and it’s true. California is beginning its 176th orbit around the Sun. May it be revolutionary, so to speak. Right? Get it? Okay.

It’s my pleasure to be speaking today about the James Webb Space Telescope as just one example of an amazing mission where humans are pursuing science and exploring the universe. It’s an amazing device, and it’s already brought us so many interesting results. It was launched on Christmas Day 2021 aboard an Ariane 5 rocket, and it’s a wonderful example of how international collaboration and cooperation—in this case, between the United States, Europe, and Canada—can lead to incredible achievements in very complex projects.

There are many comparisons with NASA and ESA’s Hubble Space Telescope, which has been serving us well for over 3 decades. The primary one is that Webb has a much bigger mirror. A mirror can be thought of as a gigantic eyeball, a collecting area that brings together faint starlight from distant parts of the universe. The bigger the collecting area, the fainter the object you can see. Webb has 6 times the collecting area of Hubble, so it’s a more powerful telescope.

Fundamentally, Webb was designed to explore our origins. Where did we come from? How are we evolving? What’s going to happen far, far in the future? How do galaxies like the Milky Way form, and how do they evolve with time?

We now know that many galaxies merge together, like the group that you’re seeing here in a beautiful Webb image. By the way, to the lower right of the word “time,” there’s a star with a bunch of spikes. Ignore the spikes. They’re not beautiful; they’re ugly, okay? They’re just a consequence of the interaction of light with the telescope. So, ignore the spikes. But here are a bunch of merging galaxies.

The first image NASA released publicly a little over 3 years ago was of a tiny part of the sky. Imagine a grain of sand held at arm’s length. Imagine how small that looks. Yet in that tiny patch of the sky, there are thousands of galaxies. These fuzzy things you see out there—you can count them if you’re interested. Over the whole sky, we can see about 1 trillion galaxies, a million million galaxies. Some of them we see forming just a few hundred million years after the explosive birth of the universe, the so-called Big Bang.

One of the interesting aspects of this image is that galaxies started forming and evolving earlier than expected. We’re working on that interesting puzzle right now.

How do stars like our Sun form? They form in stellar nurseries—giant clouds of gas and dust, fine little particles that collect as a result of gravity. The central, densest regions collapse and form these stars. But they’re hidden from view when looked at with most telescopes because we can’t peer through the dust. Webb, looking at infrared wavelengths—heat wavelengths—is able to peer inside and see newly formed stars and stars that are still forming.

We can also look at discs of gas and dust around newly forming stars. This is essentially the mechanism by which our solar system formed about 4.5 billion years ago: debris around the newly formed Sun that gradually collected to form bigger and bigger objects—planets.

How about the death of stars? This is a snapshot, a preview of the Sun’s future in about 7 billion years, when the outer atmosphere will start getting gently ejected off, leaving a hot, dying star in the middle that makes the gases glow. The star, the fainter of the 2 that you see there, looks faint because there’s dust—fine little particles that have formed in the ejected gases.

These particles consist of elements heavier than hydrogen and helium that were cooked up in the nuclear furnace of the star during its life. These dust particles can later form new stars, planets, and ultimately life. To get most of the heavy elements, you need the explosions, the cataclysmic disruptions of certain varieties of stars at the end of their lives. Our Sun won’t explode in this titanic way, but some do.

Here’s one that we started studying about 40 years ago. Analysis of the Webb data shows the kinds of elements of which we are made: the calcium in our bones, the phosphorus in our DNA, the oxygen that we breathe, the carbon in our cells, and the iron in our red blood cells. These elements were created through nuclear reactions in stars billions of years ago.

Humans understand that. Is that cosmic or what? As Carl Sagan used to say, we are made of star stuff.

We can move closer to home and image planets in our own solar system, like Neptune here with its moons and rings. Those bright spots on Neptune are a storm that has been developing. You can monitor planetary storms and come to a better understanding of climate on Earth.

We can move to other stars and search for planets orbiting them—so-called exoplanets. It turns out that nearly every star you see in the sky has a collection of planets around it. They’re just really hard to see. To see this one, the Webb telescope had to place a disc in front of the star, where that little 5-pointed thing is in the circle, revealing the exoplanet orbiting it.

The hope is that through studies of the atmospheres of these exoplanets, we will find places where life could have arisen and maybe even did arise independently of life on Earth. We don’t have such evidence yet. But once we do have compelling evidence for life elsewhere, it’ll be one of the most monumental discoveries in all of humanity.

You could say this is all very interesting, intellectually titillating, but so what? Why spend national funds on pure research of this type, rather than applied research that will lead in the short term to new gizmos, pacemakers, iPhones, and things like that? Why should we pursue this kind of research with taxpayer money? It’s a legitimate question.

Let me give you 3 reasons. The first is that, of all known animals, humans are the only ones with the curiosity to ask complex questions, abstract questions, questions about their very origins. We have the intellectual capability to pursue answers to those questions and the hands with the opposable thumbs with which to build machines like telescopes and particle colliders to help us answer those questions. If some subset of humanity were to not do this, we would be selling ourselves short as Homo sapiens. You don’t need many of us, but it’s good to have some.

The second point is that astronomy is a gateway science. It’s like the bug that bites kids and gets them interested in STEM fields. Most won’t go on to become astrophysicists. Again, that’s an okay thing. But they’ll be more motivated to pursue fields of science and technology that will lead them to careers that are more immediately beneficial to society: computer science, engineering, medical physics, applied physics, and those sorts of things.

I see this all the time as a board member of the Chabot Space & Science Center and also at Lick Observatory, in the hills east of Silicon Valley, where I conduct much of my research and public outreach. Kids love this stuff, just like I and some of my friends were inspired in our youth by the Apollo lunar landings. What an amazing accomplishment that was. We are on the Moon. Wouldn’t it be great to contribute to this grand enterprise and go boldly where no one has gone before? It’s an incredibly inspiring moment, and the Hubble, Webb, and things like that are providing that moment for kids now.

Then there are the technological spin-offs and unanticipated applications, like quantum physics. Over a century ago, there were 2 outstanding questions in physics: What is the nature of light, and why are atoms stable? You could say, “As long as we know how to make light bulbs and as long as the floor doesn’t collapse underneath me, who cares what light really is and why atoms are stable? You don’t need to know, do you?”

Physicists over a century ago, like Einstein, Schrödinger, Heisenberg, Bohr, and Planck, cared about the workings of nature simply to satisfy their curiosity. There were no practical applications immediately in sight. Fast-forward a century: you couldn’t imagine today’s world without an understanding of quantum physics.

One example is lasers, a $13.5 billion industry in the United States with innumerable applications. Computer chips—Moore’s law, with 3 and even 2 nanometers per pixel. Now we have the equivalent of half a billion transistors on the head of a pin. That is amazing. That’s quantum mechanics, folks—quantum electronics.

Specifically from something like Webb, there are lots of spin-offs. Infrared detectors—similar ones are now used in medical imaging, night-vision systems, and environmental monitoring. Cryogenic engineering—you had to cool down the telescope. This led to advances in cooling systems now used in quantum computing, superconducting electronics, medical imaging, and so on.

As just one other example, among many, precision optics and materials: segmented, gold-coated mirrors and deployment mechanisms for Webb led to innovations in robotics, metrology, and high-precision manufacturing. Those are just some of the spin-offs from Webb itself.

I hope I’ve convinced you that spending some small amount of money on research of this type is exciting and important. It extends our grand vision as pioneers of the universe, exploring our origins.

To give you a sense of scale, over 10 years, the $10 billion cost of the Webb was one $6 hamburger per U.S. taxpayer per year. That's what you contributed to the Webb. Thank you very much. I hope that you feel it was worth it to give up this one hamburger.

Now, listen. If you're interested in this sort of stuff, I give much longer talks with more details to corporate groups and others. Just contact me if you're interested. Thank you so much for being here.

Thanks, Alex. All right. Good to see you. Grab a seat. So, Alex, you are one of the world's greatest scientists and science communicators. David and I have prepared a set of rapid-fire questions for you.

I'll give rapid-fire answers.

Based on what our audience might have seen in headlines or might be understanding and want to know more about—not just James Webb, but generally.

Yes. I just gave one example of Webb time.

Mhm. Yeah. So, one of the places I want to start is searching for life on exoplanets. I think many people might understand that James Webb is doing that but might not fully understand how, and what the implications might be. So, as a way to understand this, if we were looking at Earth from 100 light-years away, what would we see and how would we understand that as life?

Yeah. What you want to find is some sort of chemical disequilibrium. That sounds fancy, but what do I mean? In the case of the atmosphere of Earth, the simultaneous presence of oxygen and methane is very curious because methane oxidizes; that is, it reacts with oxygen very quickly. So, you wouldn't expect any methane in the atmosphere unless there were some more or less continuous source of that methane.

Although methane can be produced through chemical means having nothing to do with biology, it's also produced by biology. Carl Sagan called it bovine flatulence. It's the decay of biological organisms. If we were to find that in another exoplanet atmosphere, that wouldn't be absolutely definitive, but it would be sort of a flash point: “Wow, we better study that planet more,” because that's one that could have life.

And we're seeing that?

Yeah, we're beginning to see that. We've not seen methane and oxygen in any other planetary atmosphere yet, but certainly there are interesting signs of elements that are reported by the Webb through these kinds of atmospheric studies.

One of the other big discoveries with the Webb was these early massive galaxies.

Yeah, I mentioned the early massive galaxies.

And there was a paper that followed. You and I talked about this, and there's been a lot of social media and nerdy YouTube videos about this paper and the theory that these early massive galaxies may actually disprove the Big Bang theory. The idea is that these early massive galaxies may be responsible for, or account for, what we see as the cosmic microwave background radiation. That may mean that what we assumed was coming from the early universe—from the Big Bang—may actually come from these galaxies. It's like, do we have it all wrong? We may, and these papers are getting a lot of attention. Is the Big Bang theory disproven now with this discovery?

The Big Bang theory is on very solid ground. There are many details we don't understand. The basic tenets, however, of the theory are threefold: The universe long ago was hot, dense, and expanding. Nothing in those studies contradicts any of that.

As I mentioned, the early formation of galaxies is an interesting puzzle. It means that our understanding of how galaxies formed and evolved is still incomplete. But that's part of the fun of doing science. There are new things. The cosmic microwave background radiation is the afterglow of the Big Bang, and it turns out that it agrees to very high precision with a very nearly single-temperature blackbody thermal spectrum. That means that the universe everywhere was the same temperature, then expanded by the same amount, and we see the same temperature everywhere.

There's no way you can do that with galaxies forming at a range of times and distances. They would each contribute light that would not give this so-called blackbody thermal spectrum. There are many other details of the microwave background—the spots and stuff—that are not at all addressed.

As scientists, we can dream up things and put them out there to be explored more. But, of course, the media likes to highlight the really snazzy-sounding things, so sometimes the very speculative ideas get way too much attention. We're exploring them, but they're probably wrong.

Yeah.

Okay.

Yeah. Okay. So, I don't need to throw out what I learned in high school.

No, no. The Big Bang is on very solid ground.

What about the theories coming out? We've been talking a little bit about this, on whether or not we are inside a black hole.

Oh, yeah. Are we inside a black hole? A black hole is a region of space where matter is compressed so much that nothing, not even light, can escape.

It turns out that, in a sense, our universe—if you look at the total amount of matter, dark matter, dark energy, and all that stuff, the visible matter in the volume out to which we can see—has about the right value to make the universe as a whole resemble a black hole. That would mean that our universe is finite. We don't actually know whether it's finite or infinite. We only know that it's much bigger than what we can see.

Mathematically, there is some correspondence between the equations governing a black hole and those governing the universe. But there are some important differences. A black hole is a physical structure within our 4 spacetime dimensions, like right here, whereas applying that to the whole universe is a qualitatively different idea. There are some mathematical correspondences that are useful and interesting.

I personally doubt that we are a giant black hole. Some people say it's actually a black hole that was born from another universe. For that, there's really no evidence.

But Alex, one of the many things that blows my mind about astrophysics and cosmology is that the further out we look, the faster objects are moving away from us—to a point that, at a certain distance from us, the objects in whatever direction we look are moving away from us at nearly the speed of light or even faster. And so that becomes the observational limit of our ability to see or ultimately experience our universe. There's this boundary that, without crossing the speed of light, we will never get to, and we will never see what's beyond it. Right?

That's our observable universe.

That feels pretty up.

Well, you know, space can become really big. In fact, good student, you asked me the right question.

You need some help?

Ah, thank you very much, my assistant. Okay, so I've got these galaxies here. They don't expand, by the way. They're held together by gravity in the case of real galaxies, but the hose between them expands. So, let's expand it here. Try not to aim at your eyes or David's eyes. That would be very bad—lawsuits and stuff.

From the perspective of our galaxy here, the more distant ones, with each bit of space expanding, can and do go away faster than the speed of light. Einstein wouldn't rap me on the knuckles for that. Einstein simply said that no material object or no information can travel through pre-existing space faster than light.

But space itself expanding, especially if it expands exponentially—which we think it did early on in its existence—grows faster than the speed of light, and you get a truly humongous universe, maybe even an infinite universe. Most of it we can't see, but there are other independent volumes out there where we could be having this conversation right now. Or you could not like what I said and punch me in the face, but then I would respond by punching you in the face. In other words, all these possibilities could occur in these parallel observable universes beyond the observable part that we can see.

It's freaky, but this is the kind of stuff we get to think about. And I'll ask you the question I asked you on the phone the other day, which is: There's mathematics that shows that the geometry may be inverse inside of a black hole, or that some things are reversed or inverse. What's the right term?

Therefore, is the expanding universe that we see our version of being inside a black hole, which is effectively an accelerating contraction toward the singularity?

Yeah. So, what David is referring to is that if you look at the mathematics of a black hole, from our perspective, what we call space and time outside reverse their meaning. Time becomes space and space becomes time in terms of directionality.

For example, if you're in a black hole, there's no way you can avoid the so-called singularity, where you get squished into nothing, because it's in your future no matter what you do. Now, applying that, as you wanted to do, to the whole universe, I don't think that the correspondence is such that the expansion that we see is the reversal effect of going toward the singularity, because of some technical issues.

Again, if you look at the mathematics, there are some interesting correspondences, but they shouldn't be taken too literally in most cases.

Okay.

Yeah. So, the question I would be wondering, if I were in the audience listening, is: We have an expanding universe. It is potentially infinite.

My question would be: So where is everybody?

Yeah. So, yeah, where are they all? The Fermi paradox.

Is the Great Filter in front of us?

Yeah. I actually think the Great Filter is in front of us. That's an idea where civilizations such as ours rarely get past this point, where they can achieve interstellar travel easily and stuff. Something happens, either intentionally or unintentionally or through neglect, and they get destroyed.

I actually think that, first, life at our level is very rare. I'm not saying we're alone, okay? But very rare. And the second punch of the one-two punch is that there's almost always a Great Filter. And so rarely do civilizations reach interstellar capability to the extent where they colonize a galaxy.

If it had happened even once in our Milky Way, we would easily see the aliens here. Not just the sketchy UFO evidence that's been presented—that doesn't reach the bar of credibility in science, by the way—but we would be the aliens, more likely, right? Because they would already have colonized Earth, and we would have been the aliens.

It makes sense to maybe not travel and just transmit information back and forth. Maybe we just don't know how to see or understand the information that's being sent our way, and we don't know how to transmit it.

Yeah, certainly communication techniques could be different. So, I'm not saying we know at all. There could even be this dark forest where they're intentionally not transmitting toward us because they don't want us to know about them. They're sort of maybe even pursuing us and going to kill us before we kill them.

These are all possibilities, but I think the most likely, in my view, is what I said. Also, the vastness of space means that we wouldn't be able to communicate with or hear from aliens that were much farther away than 100 or 1,000 light-years, and the galaxy is 100,000 light-years in extent. So unless they colonized the galaxy, if they're very rare, we won't see them because the signals are too faint and they haven't had a chance to get here.

I want to give you an opportunity to share with us what's going on with respect to hiring graduate students and funding research right now.

Yeah.

I've heard from lots of scientists that NIH grants have been cut.

Yeah.

And it's affecting their ability to hire and build out their labs and do some of their research. Are you seeing the same today? Maybe just give us a sense of what's going on on the ground with respect to what you're seeing in funding.

Yeah. The issue is a very serious one. In a sense, science is under attack to some degree, intentionally or unintentionally, maybe part of a broader thing. But it's having an enormous effect. The number of National Science Foundation graduate fellowships, for example, was cut in half this year.

NASA funding has been cut in half, and I'm all for going to the Moon and Mars. But if all of the remaining NASA funding goes toward those ideals, then nothing will be left for professors and their students and postdocs to analyze the great data that Hubble and Webb and all that are giving us. Various space telescopes are now in jeopardy of not being launched, including the Nancy Grace Roman Space Telescope.

So graduate schools are now reluctant to accept new graduate students and to hire new postdocs because, frankly, we don't have the funding with which to do so. I'm personally very worried about my own research group. I'm not taking on any new researchers until I personally can fund my existing group. That's got to be my primary concern right now, and I don't know how I'm going to do it.

Others throughout my field, and even, in a sense, you could say more immediately useful fields like NIH—you said, right?—are facing cuts in funding there. These are researchers who are going to be doing things that are going to be good for humanity soon, not these unanticipated spin-offs. But the kind of stuff I do should be supported as well.

Yeah. Well, I was a physics and math major. I don't know if I would have gotten the math degree, I'll be honest. But I took Alex's Astro 10 class because I was partying a little bit too much that year, and I'm like, "I got to take an easier class." There were 800 people in the class. It was the most inspirational class I've ever taken, and every student who's taken it says the same.

Alex became the favorite professor at UC Berkeley 9 times, 10 times—I don't know how many times. I think you can all understand why. His contributions to students and to science are profound. So please join me in thanking Alex.

Thank you. Thanks so much.

Regulators are now approving drone deliveries.

There is one company that is huge in this space. They’re called Zipline.

Keller Rinaudo Cliffton is the co-founder and CEO of Zipline, the world’s largest autonomous logistics and delivery system.

We should get back to building real things in the real world.

What they’ve been showing is way more advanced than anything from Google or Amazon. What nerds are working on during the weekends in their garages today are what will be the giant companies of 5 or 10 years from now.

Ladies and gentlemen, please welcome Keller Rinaudo Cliffton.

Well, good morning, everybody. David was talking a little bit about techno-pessimism. I hadn't heard that before, but by a quick show of hands, how many of you have read an article in the last year about robots trying to kill you or take your jobs? Okay, so basically everybody. The cool thing is, today we get to talk about robots that save lives.

I thought it'd be cool to just take you back to 2016. In 2016, our backgrounds were in automation and robotics. We had this simple, naive idea that it should be possible to build a new kind of logistics system—a fully automated logistics system—that would be 10 times as fast, half the cost, and zero-emission.

The first contract we signed was with the government of Rwanda to deliver blood transfusions, primarily to moms with postpartum hemorrhage, at about 21 different hospitals across the country. I thought it'd be cool to just show you this video. It's actually a video I took on my iPhone, so nothing fancy, but you can actually see what we call Zips.

This is the very first version of this autonomous aircraft that we had built. We were delivering, using a really simple paper parachute, to a hospital called Kabgayi, which is in a rural part of Rwanda. Here we were delivering, I think, 3 units of packed red blood cells and platelets. We could deliver to a couple of parking spaces in a way that was about 10 times as fast.

You can see the women in this video are like, "What the hell did we just see?" Which is funny. I was taking the video; I looked up at them, and they were looking at me very suspiciously. We often try to describe what we're going to do either to doctors, nurses, or hospital administrators, and they look at us like we're completely crazy or on drugs. So we have to do the first delivery.

Once we do that first delivery, a doctor looked at me and said, "It's as though Jesus Christ is delivering blood from the sky." But what's hilarious is that you get about 7 days of science-fiction amazement, and then people are completely bored of it. It's totally normal. In fact, I had one nurse look at her watch and then look at me and say, "It's 30 seconds late."

That made me realize humans go from science fiction to entitlement in approximately 7 days, which is great. That's what technology should do. It should fade into the background. Let doctors and nurses do the work that they were trained to do, which is save lives, and logistics should just work. That was always the vision.

So, a quick tour of the distribution center. Zipline builds, designs, manufactures, and operates these vehicles completely from scratch. This is one of our flight operators launching a Zip. It accelerates from 0 to about 100 km/h in a third of a second. From the moment the vehicle leaves the end of that launcher, it's fully autonomous.

It will fly out up to 100 miles to make a delivery to a hospital and then fly all the way back. Why do we have to have a launcher like that? Because we don't have runways, obviously, and the vehicle has no landing gear. Taking off is one thing; landing is even a little more complicated. We were inspired by aircraft carriers.

As this vehicle is flying back, we're aiming for a 1-centimeter tail hook on the back of that aircraft. This is really only possible with autonomy and robotic solutions that can be far, far more precise in controlling these kinds of vehicles than humans. The system at this point can recover an aircraft about every 60 seconds, and we operate about 20 distribution centers across 8 countries.

People always think, "Drone delivery isn't really real." So I thought it'd be cool to actually just show you a time-lapse. This is one of our distribution centers. You can see it's 1:00 a.m. The system operates 24/7, 365. It never takes a day off.

It's 3:00 a.m. here. You're seeing fulfillment operations where we're packing and loading packages, getting them packed into vehicles. Here you can see the launcher and the recovery system, with sunrise just happening in the back at 5 or 6 a.m. This is a second distribution center, another fulfillment center.

And so this is basically both fulfillment centers across the country of Rwanda, which is the smallest country we operate in today. But the cool thing is, you can see that at 8:00 a.m., every single one of these little triangles on the map—this is what we call the sky map—is an autonomous aircraft going out and making a lifesaving delivery of blood, vaccines, transfusions, infusions, cancer products, and almost the entire public healthcare supply chain.

By 10:00 a.m., there are 50 autonomous aircraft out making deliveries simultaneously to all of the 500 hospitals and health facilities that we serve in the country. I actually used to show this video to investors, and we would get to the end of the presentation and they would say, “Oh, I think my favorite slide was that simulation of what this could look like one day.” And I got so pissed off because it’s not a simulation—that happened yesterday.

So we put the CCTV on the right-hand side so you can actually see the teams doing this work, so people understand this is not the far future. This is happening day in and day out in a way that is saving lives.

And on that point, it’s not just about making logistics more efficient. It turns out that if you can deploy AI and robotics infrastructure for healthcare, you can save a lot of lives. The system has been able to reduce maternal mortality, as measured by the University of Pennsylvania, by 51% across the hospitals we serve.

Had you told us when we were starting the company that we were going to reduce maternal mortality by 5%, we would have said, “Hell yes, we have to do this.” A new study came out a couple of months ago actually showing a 60% reduction in under-5 childhood mortality due to malnutrition, from one of the new products we’ve begun delivering in the last few years.

And when this was studied by a major global health institution for the cost-effectiveness of delivering vaccines, it was found to be the most cost-effective way of delivering vaccines to zero-dose children ever studied. So it turns out that, yeah, it’s exciting. People think about robotics as being expensive or fancy, or maybe solving problems for rich people. It’s not just that. We can solve some of the most important problems that we face as a world. We can make this technology work for everybody.

So, stepping back, Zipline has now surpassed 115 million commercial autonomous miles. We serve about 5,000 hospitals and health facilities globally, with over 1.6 million deliveries like the one you saw in that video, and zero safety incidents—which is important, not just for saving lives, but for being safe for the communities that we serve.

It’s actually become the largest commercial autonomous system on Earth of any kind, ground or air, based on those flight miles. I thought it would be kind of cool to show you a bit about how this technology is evolving—how what we started doing in 2016 is evolving into the next-generation technology and launching in the US.

We play this.

Wow, that was so cool.

We love the fun.

So if you’re like, “Okay, that’s cool, but when can I use it?” The good news is very soon. As we started doing this, focusing on healthcare and operating outside the US, a lot of the biggest brands in the US started to get pretty excited and say, “Hey, we want teleportation from our hospitals, our primary care facilities, our stores, or our restaurants directly to customer homes.”

Not only did a lot of the biggest healthcare systems in the US sign up to start using Zipline, but we’ve also seen these additional major verticals in food and retail. We’ve been scaling incredibly fast with Walmart over the last 6 to 9 months. We just launched Chipotle, along with a lot of other amazing food partners, over the last month. I’ll show you a little bit more about what that looks like.

One of the amazing things is that, over the last 3 months, the service has been growing between 20% and 30% week over week. So it’s more than doubling flight volume every month. This is a little bit startling: We only launched Dallas, which is the major metro we’re scaling in in the US right now, in April.

And by July—just to give you a sense of the customer behavior we’re seeing—customers were ordering 3 to 4 times per week from Zipline. The service has a net promoter score of 94. I was talking to a grandma a couple of weeks ago. She’s 78 years old, and she’s ordered from Zipline 350 times in the last 9 months.

We were doing a little customer research, and she was showing me on her phone, clicking around and ordering everything she needed for the day. She double-clicks, uses Face ID and Apple Pay, and she’s like, “It’s on its way. It’ll be here in 8 minutes.” This woman’s living in the future.

By July, we were sufficiently nervous about the capacity of the system. We ended up turning off all the demand-generation marketing because we were trying to slow down growth. You can see the impact that turning off our marketing had on the growth of the system, which is approximately zero.

We were trying to figure out why that is, and basically, it just turns out that having a robot deliver whatever you need to your home in less than 10 minutes is really good content for TikTok. A lot of our different customers have been making tons of TikToks of receiving these deliveries, and a lot of these videos have gone viral. They get seen 8, 10, 12 million times.

We’re delivering to universities, offices, hotels, townhomes, and apartment buildings. So every time one person is getting a delivery, there are 10 other people who are like, “What the hell is that, and how do I get it?”

Even cooler than that, as we’re launching new sites in Dallas, the first site that we launched in April took us about 2 and a half months to get to 100 deliveries a day, which was the break-even point for the site. The site that we launched 2 weeks ago hit 100 deliveries a day in 5 days. So we’re seeing the sites themselves ramp way, way faster.

A big part of that is that it’s getting simpler and simpler for us to build this infrastructure. To give you a sense of what the infrastructure looks like, we integrate right into the side of hospitals, primary care facilities, stores, and restaurants. You can basically think of it like a magical portal. Zipline is just building a magical portal in the wall, and now any healthcare worker, Walmart employee, or Chipotle employee can pass whatever they want through this magical portal, and it’s teleported directly to the home that it needs to go to.

We do this for a lot of different kinds of buildings. We also have what we call Zip Points, which you can see there on the bottom right. Zip Points can be installed in 1 hour. So if you’re a business and you want to access Zipline, we show up, boop, drop a Zip Point, and now that business is enabled with Zipline. It can deliver in this way.

To give you a quick sense of what this infrastructure looks like, we’re now building these sites and launching about 1 a week. By Q1 of next year, we expect to accelerate to about 1 a day. This infrastructure is relatively quick to build and enables up to 500 deliveries a day from a site like this.

One of the cool things is that customers are all just using the Zipline app to order these things. When they’re ordering for the first time, they type in their address. We actually show them a satellite image of their home, and they tell us exactly where they want us to deliver. You can pick the dinner-plate-level area, whether it’s in your backyard, on the side of your house, in your parking lot, or at an apartment building. We can even deliver onto roofs.

You can scroll and see all the different brands that are available on the app, and order whatever you want. The average delivery time right now is 18 minutes. A lot of deliveries happen in under 10 minutes. In fact, we just launched Chipotle 2 weeks ago, and the first delivery happened in under 7 minutes, from the customer ordering to it being delivered to their house.

I think it’s going to redefine what is possible in terms of instant delivery in people’s minds. And just to hint at something cool that we can’t announce just yet, we’ll be adding a lot of people’s favorite brands to the service very soon, over the coming weeks.

And, just one last thought: I joked before about this sense of going from science fiction to entitlement in about 7 days. We do enjoy that period of sci-fi amazement. The similar version of Jesus Christ delivering blood from the sky is pretty cute to see in families and kids.

Kids are telling their parents what they want to do for the weekend: They want to go and watch the Zipline aircraft. We take these pictures when we’re at the sites of people hanging out on the hoods of their cars, or a mom with her kids sitting in her lap, or kids looking through the window of the car just watching the system operate.

And that brings me to my last provocative point. Our parents had this incredibly inspiring mission, right? The United States was in this geopolitical race to get to the moon—the space race—and it united all the best engineers. It inspired us. It made us dream with optimism about what the future could represent. And we did something impossible: We put men on the moon in 9 years.

Obviously, the US is in a similar technological race today.

It’s a race for AI and robotics. But what does winning that race for the US really mean? I want to leave you all with just a slightly provocative answer to that question.

But first, who knows what city this is? Shout it out if you know.

Yes.

Okay, good. There are nerds in the audience. This is Wakanda. Wakanda is a fictional, radically advanced African city hiding in plain sight from one of my favorite movies, Black Panther.

The provocative idea is that we can go build this in the real world. I think that winning the AI and robotics race for America isn’t just us building exquisite AI technology to serve the richest people on the coast of this country. It’s about extending the reach and influence of the United States. It’s using AI and robotics infrastructure to lift the rest of the world up with us.

These countries want to leapfrog into the future. They want access to the best technology that America has to offer. If we extend it, we want these countries building on US AI and robotics infrastructure, not that of our geopolitical adversaries. If we can do that, we can make the world a safer place and a wealthier place.

We could potentially eliminate maternal mortality and childhood mortality in a lot of these countries. In doing so, we can secure US technological and manufacturing leadership for the decade to come. So, thank you all.

We want to do a little bit of time travel with you today. We want to go back to your origin story, and then we want to play it out into the Wakanda future that you’re imagining.

Cool.

Taking it back to where you began, why start in Rwanda?

It’s funny. Everybody makes this assumption that the most advanced technology in the world is going to start in the United States and then trickle its way out. It’ll start in the rich cities, and then maybe it’ll trickle its way to rural areas in the US. After years, it might trickle its way out to developing countries.

I think that paradigm is largely wrong, and it has a lot to do with which countries are hungry and entrepreneurial and willing to move super fast to build new kinds of regulatory paradigms. Rwanda is kind of like the Singapore of Africa. It moves incredibly fast. It’s very entrepreneurial. It’s kind of a startup country, and it was perfect for us to work with them.

They wanted to take this risk on us when we were 20 people. We were totally naive nerds who had no idea what we were talking about. In fact, I remember this conversation with the Minister of Health in 2016 where I was saying, “Oh, you know, we’re going to use autonomous aircraft to deliver all the different medical products in your health system.”

She looked at me and was like, “Keller, shut up. Just do blood.” She explained to me that 50% of blood transfusions go to moms with postpartum hemorrhaging, and 30% go toward kids with severe anemia due to malaria. She was like, “Just show us that you can do that.”

That was the best advice the company ever received. We’ve really just been following their lead for the last 8 years as we’ve developed the technology from there.

The reduction in maternal mortality when you showed that stat—I got goosebumps. It’s just incredible.

Yeah. By the way, I think a lot of times people in the US think, “Oh, those poor Africans. It’s unbelievable that they have those kinds of healthcare problems.” We have these exact same problems in the US.

People in the audience may not know, but the US has the highest rate of maternal mortality of any developed country. Rates for African-American women are 3 times the average. We have a lot of challenges with rural healthcare in this country.

I think, honestly, people probably think that these countries are more different than they are. Almost every health system is dealing with the same kinds of challenges.

So, you come to the United States and launch it here. Tell me about the first period of launching in the US.

We originally launched the first version of the technology—the fixed-wing technology that you could see—in 2020. Honestly, it’s shocking. We were delivering birthday cakes and rotisserie chickens via those paper parachutes. It’s pretty unfancy, but customers loved it.

We were rapidly iterating to build something that we thought would be the future version of logistics, which is ultimately Platform 2. It’s the video I showed today. We only launched Platform 2 on January 15, and we really only started scaling it in April or May. This is all happening in real time. A lot of those videos we showed were just from yesterday or the day before.

What has it felt like?

It’s stressful. Hardware is incredibly hard. We have been scaling a hardware product while the tariff craziness has been going on through March, April, and May.

Building a global supply chain is a huge undertaking. To put it into perspective, Zipline designed the flight computer and all of the avionics on the aircraft. We designed the aircraft itself, all the mechanical components, and the primary structure.

From a software perspective, it’s flight-control algorithms, multi-vehicle deconfliction, and communications architecture. We design and build an unmanned traffic management system that we provide to regulators like the FAA. We also design the app that you saw, which is our customer ordering platform.

All of that has to work. Then you also have to figure out supply chain, maintenance, manufacturing, operations, and logistics. All of it has to work for the end customer to have this magical experience of teleportation.

There’s no part of it that doesn’t feel desperate and stressful as you’re scaling a system at that level of exponential growth and launching in the US.

Do you have a sense, Keller? Are you going to beat the unit cost to deliver with delivery drivers today? And by how much? Can you give us a sense of, if I want Chipotle, why would I go to the Zipline ordering system or use Chipotle’s app and have Zipline fulfill for me? What’s the cost difference going to be, percentage-wise, over traditional food delivery?

Interestingly, people may not realize that instant delivery has grown incredibly fast, especially through COVID. Even before that, there are now 5–12 billion instant deliveries being done every year just in the US. That’s not Amazon or UPS; that’s just instant deliveries.

We’re using a 4,000-pound gas-combustion vehicle driven by a human to deliver something to your home that weighs, on average, 4–5 pounds. If aliens were to land on the planet and look at the way we’re solving that problem, they would conclude there’s no intelligent life on Earth. This is a bizarre solution.

The reality is that we have this new demand, and the demand is vast. People want things delivered quickly, and they want to have more time with their family rather than spending time in traffic or in a store. But we’re using technology that’s 100 years old to solve that problem.

All you have to realize is that instead of using a 4,000-pound gas-combustion vehicle driven by a human, you should use a 50-pound vehicle that is autonomous and electric. That’s just reasoning from physics first principles. You don’t have to be a genius. As soon as you’ve realized that, I think you know something really fundamental about the future that few people actually understand.

We think it’s very inevitable that these systems will be less expensive than using a 4,000-pound gas-combustion vehicle. By the way, if you were to just extend the customer ordering behavior that we see with our customers today, there would be 50 billion instant deliveries happening in the US.

Wow.

Based on the order data, yeah. Basically, if you make the deliveries less expensive, 10 times as fast, and just a way better experience, people order a lot more. It’s not that surprising.

I think the reality is that these kinds of systems will definitely be less expensive than using a 4,000-pound gas-combustion vehicle. More importantly, the reason customers are using them so much is that it’s just a way better experience when you can have something delivered in 7 minutes, 8 minutes, or 12 minutes.

Does it need to be much cheaper?

I think it will be naturally, but I don’t believe it needs to be. It’s the reason that Waymo right now is more expensive, and people prefer Waymo.

What’s the weight limit, and then how much of the market does that address?

Right now, the system is designed to deliver up to 8 pounds, and 8 pounds gets you about 95% of all packages delivered by Amazon. I think it’s about 95% of food-delivery orders.

Suffice it to say, you’re not going to deliver flat-screen TVs in this way anytime soon, but the vast majority of stuff actually fits and can be delivered like this.

While all this is happening, while Zipline is exploding, I think many Americans came to believe that the era of drone delivery had somehow passed—that this wasn’t a near future that they were going to experience. Why do you think that misconception happened? And what should all of the people in this audience go out and say to the people who might ask them what they’ve seen here?

Yeah.

Well, it definitely didn't help that the CEO of one of the largest companies in the world went on 60 Minutes in 2013 and promised everybody drone delivery in the next 1 or 2 years. Maybe some of you saw that interview.

Who was it?

I don't even know what that interview was.

Jeff Bezos from Amazon. Yeah, that was 2013. They announced Amazon Prime Air and said that by 2015, it would be serving everybody in the U.S. I think people probably believed it, and then they were really disappointed when it didn't happen.

Maybe you saw a similar trend with autonomous vehicles and autonomous cars. In 2015, so many companies were raising billions of dollars, and it seemed like it was right around the corner. People could see it working for the first time. But obviously, it's a whole decade later that we actually see Waymo and robotaxis scaling commercially.

The reality with these kinds of technologies is that you always have the bubble and the peak of hype, then the trough of disillusionment, and then 8 to n years of the actual hard work of making the technology work. Zipline launched in 2016, and we've spent 10 years driving the economics down and driving the reliability up.

You saw that statistic of 115 million miles with zero safety incidents. That's hard. It requires time to get the manufacturing technology, operations, and maintenance right in a way that achieves that.

The good news is that, with both autonomous cars and this technology, we've now done the 10 years of hard work. We now see it scaling in a way that's fundamentally changing the way people live their lives. When I talk to that grandma, or when you talk to a mom who's using Zipline every single day, they're getting hours back each week to spend with their family or loved ones. They don't have to stress about buckling their kids into a car and driving to a store. And obviously, that's the retail use case, let alone the lifesaving implications this has for health care logistics.

What are the competitive barriers? Google has, I think, Wing, a drone-delivery company. I don't know what the status is. Amazon has obviously invested. By the way, I thought it was either Elon, someone from Google, or Jeff.

There have been companies like Meituan, I think, out of China, that have famously shown videos of delivering food to the Great Wall with a drone. How much of an advantage does Zipline have versus others, and how quickly can they catch up? Help us understand how hard the technology is. What did you have to engineer to get the unit-cost advantages you're having today, and how persistent will that be?

There are a lot of people out there. Over the last 10 years, we've seen so many companies or teams buy a quadcopter off the shelf, duct-tape a Snickers bar to the bottom of it, and then manually fly it a mile. They'd get TechCrunch to write an article about it and say, “It's a Kitty Hawk moment. Drone delivery is here.”

We've seen that 50 times at this point. I think people know that it's not real. The trick is designing a system that can operate 24/7, 365 days a year in a way that people can depend on with their lives, that works in all weather, is reliable and safe, and can achieve hundreds of millions or billions of autonomous miles. That's hard to do. It takes time.

Zipline has now spent a decade scaling these systems. I think the realization is that there's no off-the-shelf hardware you can buy for this. You can look at these cheap plastic quadcopters that DJI makes in China, or you can look at Predator drones, but something in the middle that's more automotive-grade—something that can do, for example, 1 million miles with a single aircraft—that's hard, and it has to be built from scratch.

We honestly don't worry. Our competition is motorcycles and cars. If we are better than motorcycles and cars, I'm very confident someone is going to build a multihundred-billion-dollar company in automated logistics over the next 5 to 10 years. It's so obvious that this needs to exist. The demand is unbelievably vast. It's going to be one of the biggest markets on Earth.

A lot of people are excited about a lot of different kinds of robotics, but this is the area of robotics that, in my opinion, is going to scale the fastest and is most ready for prime time.

Want to talk about Rwanda?

I do. So, just to jump ahead—yeah.

If we were interviewing you here in 10 years, and you're back, what do you hope you're saying about the impact of drone delivery, both on the golden billion and on everybody else? To your point earlier, the relationship between those 2 things and those 2 groups might be closer than we think.

The thing that has always really inspired us—you talk about logistics. Logistics is boring, right? Who wants to work in logistics? It's incredibly boring. You just do the same thing day after day, just doing the same deliveries. But that's also what makes it great for robotics and automation.

The key thing to realize is that the golden billion Cleo is talking about—the richest billion people on Earth—my assumption is that we're all in the golden billion. Our access to logistics is really good. There are 7 billion people on Earth who are not in the golden billion, whose access either sucks or is nonexistent.

As a result, 5.5 million kids lose their lives every year due to a lack of access to basic medical products. This isn't, “Oh, we need some advanced therapy.” It's, “No, no, we couldn't get them the basic, almost-free drug that they needed to save their life.” We've been making excuses for decades about why we can't solve these problems.

Logistics is boring, but it's only boring when it's working well for you. The thing that gets me so excited about why AI and robotics matter, and why we should be applying them to this industry, is that it's not just about making people's lives better, giving them new kinds of economic opportunity, saving them time, and letting them spend more time with their kids.

It's also because reducing the cost of logistics, automating it, expanding it, and improving the performance of these kinds of systems is going to extend access to logistics to 7 billion people on Earth who don't have it today. That is going to save lives, increase economic opportunity, and make the world a more stable place.

That's really our vision. It's time to stop making excuses. We should eliminate these problems. The thing that gets me excited is that I know we're both kind of solarpunk techno-optimists, right? That's the future that I want to build, and that I want to tell my kids about. If we can play a small part in it, that would be a good life.

That's the future I think we all want to be part of.