Latest / Giant Ideas / Waymo, Udacity, and Kitty Hawk Founder, Sebastian Thrun: Building Moonshots & His Next Company

Transcript

Waymo, Udacity, and Kitty Hawk Founder, Sebastian Thrun: Building Moonshots & His Next Company · Giant Ideas

0:01
SPEAKER_00: Today, I cannot even predict the next seven months because things are moving so insanely fast.
0:07
SPEAKER_00: That to me is a singularity.
0:08
SPEAKER_00: 90% of what a software engineer does today, in terms of writing computer code, they can now do this automatically.
0:15
SPEAKER_01: Hello, and welcome to Giant Ideas with me, Cameron McLean.
0:19
SPEAKER_01: And me, Tommy Statlin.
0:20
SPEAKER_01: We're co-founders of Giant Ventures, which builds and backs purpose-driven companies.
0:24
SPEAKER_02: At Giant, we're lucky to meet extraordinary people with giant ideas that are changing the world.
0:29
SPEAKER_02: This podcast brings you behind-the-scenes access to those ideas and the inspiring stories of the people behind them.
0:35
SPEAKER_01: We explore how one giant idea can kickstart a billion-dollar company, shape culture, and transform life as we know it.
0:42
SPEAKER_01: Today on the podcast, we've got someone who probably is going to have more impact on the way the world looks and feels than any of the guests we've had on giant ideas so far.
0:51
SPEAKER_01: He's Sebastian Thrun, the godfather of self-driving cars.
0:55
SPEAKER_01: Sebastian invented Google's autonomous vehicle, which is now known as Waymo, and Waymo is doing about a quarter of all rideshare journeys in San Francisco.
1:04
SPEAKER_02: But Sebastian's life has been full of moonshot ideas.
1:07
SPEAKER_02: He co-founded Google X, which is the moonshot factory at Google, as well as the flying taxi startup Kitty Hawk, and Udacity, which democratized elite education online.
1:16
SPEAKER_02: He's widely seen one of the greatest computer scientists and roboticists of our time.
1:21
SPEAKER_01: I first met Sebastian about 12 years ago when I was researching my book, Connect.
1:25
SPEAKER_01: It's really, really great to have you with us on Giant Ideas, Sebastian.
1:28
SPEAKER_01: Thanks for joining us.
1:29
SPEAKER_01: Hi, can we run ahead, Tommy?
1:31
SPEAKER_01: Well, lots to cover.
1:32
SPEAKER_01: Why don't we start with autonomous vehicles?
1:36
SPEAKER_01: Why do driverless cars matter, Sebastian?
1:39
SPEAKER_01: What's the impact going to be on society?
1:41
SPEAKER_00: Well, there's two fundamental things.
1:44
SPEAKER_00: One is safety and one is cost.
1:46
SPEAKER_00: On the safety side, we lose more than a million people every year in traffic accidents.
1:52
SPEAKER_00: It's the leading cause of death for young people worldwide.
1:56
SPEAKER_00: And that is just not acceptable to me.
1:59
SPEAKER_00: Self-driving cars have proven to be exponentially safer already at this point.
2:05
SPEAKER_00: But the second thing is cost.
2:08
SPEAKER_00: It turns out that a car is mostly not used.
2:12
SPEAKER_00: Whoever owns a car on this show, please check.
2:15
SPEAKER_00: Your car is probably parked right now.
2:17
SPEAKER_00: In fact, cars are parked 96% of the time, and maybe 3-4%.
2:22
SPEAKER_00: And that means that you're wasting a ton of resources, okay, by virtue of not using your car.
2:27
SPEAKER_00: So if the car is being shared among multiple people, you get much higher utilization, you don't get these inner cities full of parked cars, you don't have to have a garage at home and all that stuff, and never have to look for a parking spot again.
2:40
SPEAKER_01: Am I right in thinking that for you the safety issue and this goal of trying to eradicate almost all of those million lives lost every year was a very personal thing based on on the sort of tragedies that happened with you in the in Stanford with your colleagues?
2:56
SPEAKER_00: I've now um lost a number of friends and co-workers to traffic accidents, but the most stingy thing was when I was 18.
3:05
SPEAKER_00: And my neighbor Harold went on a drive with a friend, they found themselves on a sheet of ice, crashed into a truck, and both died instantaneously, and took like less than a second.
3:19
SPEAKER_00: I was thinking, wow, here's a promising life of an 18-year-old, my best friend at the time, gone for no reason.
3:28
SPEAKER_00: Like, why do we do this to ourselves?
3:30
SPEAKER_00: Uh why do we have teenagers drive on a sheet of ice in the first place?
3:35
SPEAKER_00: But then why do we use technology that's extremely dangerous as if it's just something completely casual?
3:42
SPEAKER_01: Has has this therefore been a a very, very mission-driven thing for you, all the way back to that incident that you felt that you simply had to solve this problem?
3:52
SPEAKER_00: I mean, that's basically the way I live my life, that I see things that are just not okay.
3:56
SPEAKER_00: I want to fix them.
3:58
SPEAKER_00: And it involves people's deaths, it's particularly painful.
4:01
SPEAKER_00: And I'm sure many of your esteemed listeners have similar experiences with themselves or loved ones.
4:07
SPEAKER_00: And yes, it's driven me to really think about as an AI researcher, as a roboticist, how can I fix this big problem of more than a million deaths every year?
4:17
SPEAKER_00: And the answer is make the car smart.
4:20
SPEAKER_00: It's artificial intelligence, just make the car smarter than people are.
4:23
SPEAKER_00: Like computer self-driving cars, they don't text, they don't drink, they're not distracted, they're not on the phone, they're not fatigued.
4:31
SPEAKER_00: They can look in all directions all the time.
4:34
SPEAKER_00: By virtue of doing this, they can be safer than human drivers can be.
4:38
SPEAKER_00: And now we know for a fact that they're safer than human drivers.
4:40
SPEAKER_02: Yeah, and we love we love incredibly mission-driven entrepreneurs, and sounds like you are exactly that.
4:45
SPEAKER_02: Um, you've been at the center of the autonomy story really since the very beginning.
4:49
SPEAKER_02: It'd be great if you could paint a picture for our audience of where it started, your role, uh, maybe give us some tidbits and then where we are today.
4:57
SPEAKER_00: Yeah, in 2001, 25 years ago, my God, the US government came up with this idea of creating a competition for self-driving cars.
5:07
SPEAKER_00: Their focus at the time was the war in Iraq.
5:10
SPEAKER_00: They lost lots of soldiers to IEDs, improvised explosive devices, and they wanted to be able to move about without putting people's lives at risk.
5:20
SPEAKER_00: Be it as it is, they came up with this thing called the DARPA Grand Challenge.
5:24
SPEAKER_00: DARPA is a part of the US government that's responsible for the internet and stealth bombers and many other great innovations.
5:31
SPEAKER_00: And the Grand Challenge was intended to be a challenge to the scientific community of professors around the world to build a car that could drive itself 140 miles, roughly 200 kilometers, through a desert environment.
5:45
SPEAKER_00: Similar to Iraq, they picked the Mojave Desert here in southwest United States.
5:50
SPEAKER_00: And they said, whoever can build a car that can drive itself without a person, it's like completely autonomous, 200 kilometers, it's gonna make a million bucks.
6:00
SPEAKER_00: And look, a million bucks is doesn't go far in terms of research budget.
6:03
SPEAKER_00: When you run a yeast at university, you have a you have billions of dollars uh that you use for research.
6:08
SPEAKER_00: But as an individual professor, I was at Stanford at the time, it felt like a lot of money.
6:13
SPEAKER_00: Like, my God, I can I can put a computer in a car, I can make it safe, I can live onto my mission and figure out how to make cars safe, and I can villain bucks.
6:22
SPEAKER_00: That's how it started.
6:24
SPEAKER_01: And then you you won that challenge, right?
6:26
SPEAKER_01: And am I right in thinking that Larry Page, the Google co-founder, was sort of incognito in dark glasses, watching who would win this?
6:33
SPEAKER_01: And then and then came up to you.
6:34
SPEAKER_00: It was the most surreal thing possible.
6:36
SPEAKER_00: Like because normally I do lots of racing, bicycles, and marathons, and the race is always the hard thing.
6:42
SPEAKER_00: You sweat, but when you send your self-driving car off to race, almost like sending your child off after training it for a year or so, and it races, but you're sitting there with you, zipping coffee and possibly champagne.
6:54
SPEAKER_00: Um, the race unfolded over seven hours.
6:57
SPEAKER_00: We had uh a total of almost 200, 296 contestants applying for it.
7:03
SPEAKER_00: 23 finalists were allowed to race.
7:06
SPEAKER_00: And of those uh 23 finalists, uh five made it.
7:10
SPEAKER_00: Um, four within the lotted time of 10 hours.
7:12
SPEAKER_00: And I was at Stanford, our little Volkswagen was the fastest.
7:15
SPEAKER_00: So we we kind of won.
7:16
SPEAKER_00: Yes, we did win.
7:17
SPEAKER_00: But I always told people look, it's not really important that we won at Stanford, it's important that the scientific community won because there was the birth moment, at least in the states for the self-diving car.
7:27
SPEAKER_02: So fast forward to the day and getting into a Waymo in San Francisco uh is one of the most amazing hero product experiences I've ever had.
7:35
SPEAKER_02: It just it just blows your mind.
7:37
SPEAKER_02: My son, uh, when I was there with him, was was uh very confused to see cars driving around without people in the front.
7:43
SPEAKER_02: And now he he's he's five years old, but he talks regularly about the Waymo.
7:46
SPEAKER_02: Um, but you you spend 25 years from you know from the time that Larry watched you in his dark glasses to today, 25 years has gone by.
7:54
SPEAKER_02: Give us a bit of a colour of how how we got to where we are today and where you think we are on the adoption curve.
7:59
SPEAKER_00: Well, thanks, Tommy and Cameron, for asking that question.
8:01
SPEAKER_00: And um, appreciate your son being enthusiastic.
8:04
SPEAKER_00: Uh, hopefully he's going to be a future computer science student game.
8:09
SPEAKER_02: Right now he's showing he's showing a lot of proclivity for art at the moment, but um okay.
8:14
SPEAKER_00: Um a little bit is art.
8:15
SPEAKER_00: Mostly it's just straight math, physics, computer science, and what we built.
8:19
SPEAKER_00: Um, we then uh founded a team, and the very first team, which actually a company that that Google acquired for me, uh, built Street View.
8:27
SPEAKER_00: And Street View showed the world a sliver of the self-driving kind that was able to record data, but it wasn't really robotic.
8:34
SPEAKER_00: It was mostly like human-driven to record the world from the street level.
8:37
SPEAKER_00: And then we moved into what we call project chauffeur.
8:40
SPEAKER_00: In 2009, we started a team at Google trying the impossible.
8:45
SPEAKER_00: And we we built in the beginning eight cars that we drove on public streets in California with a safety driver that could take over, but still, if you go like 100 kilometers per hour, it's pretty scary to be in a car that has been programmed overnight.
8:59
SPEAKER_00: And we started driving um pretty much all streets in California.
9:03
SPEAKER_00: Uh, there were city streets like San Francisco, including the Vicent Lombard Street, if you've ever been here, or like surface streets all the way from Highway 1, the ceiling highway to Los Angeles and around Lake Tahoe, and bridges, places where there was just no GPS because we had a tunnel, everything.
9:21
SPEAKER_00: And we had to get it to a point where the car was 100% correct.
9:26
SPEAKER_00: And just to lay out how the hard this is, when you um use Chat GPT today or any large language model, you have what's called hallucination.
9:34
SPEAKER_00: Hallucination is the moment when the model makes something up that's not correct.
9:39
SPEAKER_00: And it happens, okay?
9:40
SPEAKER_00: There's nothing wrong with it.
9:41
SPEAKER_00: The technology is amazing, but it's a side effect of the way these things are being trained.
9:46
SPEAKER_00: When you have a hallucination in a large language model, you shake your head and you move on, and that's okay.
9:52
SPEAKER_00: But if your self-driving car hallucinates, it'll run a red light, it'll hit a person.
9:59
SPEAKER_00: It's not acceptable.
10:00
SPEAKER_00: So we the bar in terms of reliability is so high that it took a much, much longer time to train our artificial intelligence to be really, really reliable.
10:12
SPEAKER_00: And it's you get a lot of appreciation for people because people are good at this stuff.
10:16
SPEAKER_00: They're really good at this stuff.
10:17
SPEAKER_00: When we focus and we don't drink, you're not fatigued, we're generally pretty safe.
10:22
SPEAKER_00: Okay.
10:23
SPEAKER_00: But to get the same capability in the computer took about 15 years.
10:27
SPEAKER_01: Maybe let's just talk about the landscape of competitors out there.
10:31
SPEAKER_01: Um, obviously, Waymo is is very dominant in the US.
10:35
SPEAKER_01: In in China, Baidu's offering, Apollo, seems to be doing very, very well.
10:39
SPEAKER_01: Uh, strangely uh undervalued, I think, because Baidu, I think, is valued essentially at its cash reserves, which seems odd given that Waymo some people think is worth almost$200 billion within Google, um, within Alphabet.
10:52
SPEAKER_01: I guess it'd be great to hear from you just sort of how do you see the landscape?
10:55
SPEAKER_01: There's also WAVE here in London, which was founded by a friend of ours, Amar Shah.
10:59
SPEAKER_01: They're doing great, very different approach technically.
11:02
SPEAKER_01: Maybe give us a sense of the competitive landscape.
11:04
SPEAKER_01: Is there going to be one dominant winner?
11:06
SPEAKER_01: Is it gonna be, you know, a whole bunch of players with different technical approaches?
11:10
SPEAKER_01: And also maybe unpick for us if you can, just the importance to to Google, to Alphabet of Waymo.
11:16
SPEAKER_01: How central to Alphabet's future do you think Waymo is?
11:19
SPEAKER_00: Look, um, in the US, the, and I'd say globally, the the number one player right now is obviously Waymo, in that Waymo is not driven more than 100 million miles and never harmed a person, which is something that people just can't do.
11:33
SPEAKER_00: Okay, in a hundred million miles of human driving, which takes many, many people, in expectation you kill more than one person.
11:41
SPEAKER_00: Um, on its heels in the US is certainly Tesla, which has a different approach.
11:46
SPEAKER_00: Tesla has been trying to do the same capability based on camera only.
11:51
SPEAKER_00: And while the hypothesis is correct that eyes are sufficient to drive, in practice, it has been a bit more challenging than using lighters and more advanced sensors.
12:01
SPEAKER_00: But yes, you're correct.
12:02
SPEAKER_00: Unbeknownst to most people in China, there's been massive progress and not just Baidu, but Pony and Diddy and other companies, who effectively started the Silicon Valley lab and hired a top dog from places like Waymo, and then eventually moved over development to China, and Chinese people just work twice as hard as Americans.
12:20
SPEAKER_00: 996 is the famous word from nine in the morning to nine in the evening, six days a week.
12:24
SPEAKER_00: If you work on the math, it's more than the German 37.5 hours.
12:29
SPEAKER_00: Okay.
12:30
SPEAKER_00: And they've been um now launching systems and the services in China that are effectively on par or getting very close to what VMO can do.
12:39
SPEAKER_00: Um and similar in we see this in large language models between, I don't know, Groc and Alibaba and Meta and OpenAI and Google, that there's a race going on, and um the the teams by and large are all like surpassing each other all the time.
12:54
SPEAKER_00: So there's nothing really secret about how you defend this.
12:58
SPEAKER_00: Um the basic rule is the more data you have, the more experience you have, the more reliable your system becomes.
13:04
SPEAKER_00: And anybody who puts the money into generating the data and trading the systems eventually can increase the liability to the point where you can drive in public.
13:13
SPEAKER_00: So big motion in China right now.
13:15
SPEAKER_00: We're all waiting for Germany to follow suit.
13:17
SPEAKER_00: German companies, of course, are German car companies are the best in the world, there's no question.
13:22
SPEAKER_00: They put a lot of emphasis on the driver assist system, where the driver is still the driver in charge.
13:27
SPEAKER_00: Um, and Germany has not yet made the move to a completely autonomous system, but that's just a matter of time.
13:33
SPEAKER_02: Given the compounding data mode you you alluded to there, do you think it's a winner-take-all dynamic where there will be one leading autonomous driving car or will it break up by region?
13:42
SPEAKER_02: How do you think about that?
13:44
SPEAKER_00: That it depends on first local regulations and how the big players are playing the game.
13:48
SPEAKER_00: Um, in the right-hailing or right-sharing world, we find historically it's a winner-take-all position.
13:55
SPEAKER_00: In the United States, it's effectively Uber who is the winner, uh, even though Lyft still exists.
14:00
SPEAKER_00: In China, we find Didi is the winner.
14:02
SPEAKER_00: In Singapore, it's Grab.
14:04
SPEAKER_00: And when you look historically, the smaller players who don't quite get the liquidity end up losing because there's a network effect.
14:12
SPEAKER_00: And the network effect says the more people participate, the more people participate, the better the service to everybody, right?
14:19
SPEAKER_00: So early network effects, for example, is the fax machine.
14:22
SPEAKER_00: The more people used to, I mean, this updates me, used to have fax machines, the better for everybody.
14:26
SPEAKER_00: And then ride sharing, the more people use your ride sharing services, the more you can afford putting cars all over the place.
14:33
SPEAKER_00: And the shorter the wait time for the next car, and the less the waste in terms of time wasted on the driver's side, so the cheaper you can make it.
14:40
SPEAKER_00: So there is a uh possibility that a single winner emerges in this place.
14:46
SPEAKER_00: But we are we are far away from this.
14:48
SPEAKER_00: We are far away from even making the capital investment, the ability to build, let's call it, a million self-driving cars.
14:54
SPEAKER_00: That is itself is already challenging.
14:56
SPEAKER_00: But then also you find that local places differ in the regulations and they favor certain companies, right?
15:02
SPEAKER_00: So I would say Uber worked really hard in China.
15:06
SPEAKER_00: Um I'm I wasn't part of the Uber or the China team at the time, but it reported the time it went into headwinds because China is China and America is America, and America likes American companies and China likes Chinese companies.
15:18
SPEAKER_00: Um, that might not be fair, but overall, um, local companies seem to have an advantage in these things.
15:24
SPEAKER_02: And if you were starting the Google self-driving car project today, how would your approach differ given there's been these huge advances in foundational models and multimodal AI?
15:33
SPEAKER_02: Or would you take the exact same approach?
15:35
SPEAKER_00: Well, when we started out first, uh we thought of this more as a like a 3D geometry game.
15:40
SPEAKER_00: Like think of it a video game where use a lighter to really figure out how far exactly is this car wave up to millimeter precision, how is the range changing, meaning what's its speed, and so on.
15:52
SPEAKER_00: And it became this big blocks world of like blocks moving around in the beginning.
15:57
SPEAKER_00: Uh, there was a good amount of machine learning involved ever since the Dapper Grand Challenge, but it wasn't the predominant uh solution.
16:04
SPEAKER_00: As time moved on, uh machine learning, AI has massively improved, especially when it comes to the understanding of images or video.
16:13
SPEAKER_00: So, for example, going from a pixel cloud of like a 3D point cloud to saying this is the person, this is a dog, this is a trash bin.
16:22
SPEAKER_00: That uh became much more feasible in the last call it 10 years or so.
16:28
SPEAKER_00: So there has been a shift on the team to use much, much more machine learning.
16:31
SPEAKER_00: But there's one important caveat, which is machine learning doesn't solve everything.
16:38
SPEAKER_00: There's always gonna be situations that are extremely rare, but you have to handle correctly.
16:44
SPEAKER_00: And if if you they're so rare that they're no training set, you can drive 100 million miles and you're never gonna see the situation.
16:50
SPEAKER_00: Maybe it's, I don't know.
16:52
SPEAKER_00: Uh we had a case where a lady lost a stroller with a crying baby inside, and the stroller zipped across the street.
17:00
SPEAKER_00: And you can't train for this, right?
17:02
SPEAKER_00: You still have to react correctly.
17:04
SPEAKER_00: The machine learning model cannot say, oh, I've never seen this, just keep driving.
17:07
SPEAKER_00: That's not okay.
17:08
SPEAKER_00: Uh, it has to be able to react correctly to even completely unforeseen situations.
17:13
SPEAKER_00: And that's where, to be quite frank, a hybrid method seems to win today, which is one that uses machine learning for understanding and prediction, but also uses this kind of 3D logic, the physics, the point cloud, to understand the physics of the space involved.
17:27
SPEAKER_01: So, Sebastian, are we envisaging a world where close to 100% of cars on the road are going to be driverless?
17:33
SPEAKER_01: Or are we not approaching anywhere near that?
17:36
SPEAKER_01: How long is it going to take?
17:38
SPEAKER_01: Try to paint a picture for our listeners, if you can, of when this world is going to arrive.
17:42
SPEAKER_01: Because it's arrived in San Francisco, but in cities around the world, you know, there are almost no driverless cars on the road today.
17:48
SPEAKER_00: I really recommend everybody come to San Francisco.
17:50
SPEAKER_00: Just like the cable car and the Golden Gate Bridge, the driverless car has become a tourist attraction.
17:56
SPEAKER_00: And it's surreal to sit in a car without a driver inside and see it drive you.
18:00
SPEAKER_00: It's something that takes a lot of courage in the beginning, and after a few minutes, a lot of appreciation, and you feel how safe this video has become.
18:09
SPEAKER_00: This technology can now be rolled out pretty much all across the world.
18:13
SPEAKER_00: Maybe not in crazy places.
18:16
SPEAKER_00: I recently visited India and India has different rules of traffic than the United States.
18:21
SPEAKER_00: Leave it at that.
18:23
SPEAKER_00: But by and large, in European cities could be rolled out if the regulatory framework and the governments play along.
18:30
SPEAKER_00: So it's not a um if question, it's more like a when question.
18:34
SPEAKER_00: Now, I want to say to this, um, I also worked on flying cars, and I believe those are basically where the self-driving car was like 10, 15 years ago.
18:44
SPEAKER_00: So there will be more innovation in traffic.
18:46
SPEAKER_00: It's not going to stop there.
18:48
SPEAKER_00: Like there's going to be, I mean, the car itself is no older than 150 years.
18:52
SPEAKER_00: The aircraft, the airplane is no older than 150 years.
18:55
SPEAKER_00: In the history of humanity, which is like hundreds of thousands of years, all these new things are relatively recent and don't expect innovation to stop.
19:04
SPEAKER_02: So is autonomous air travel the future then?
19:07
SPEAKER_00: Yeah, so if you look at the air-based stuff, the work we did at a company called Kitty Hawk that I founded with Larry Page and we recently sold to Boeing.
19:14
SPEAKER_00: Um, what happens there is um it's it's not really a flying car, it hasn't wheels, but it's something that takes off like a human drone uh electrically and then goes in a straight line.
19:24
SPEAKER_00: And we were able to prove that these systems are about three times as energy efficient as green as a Tesla is, which at the time was the benchmark for most energy efficient vehicles on the ground.
19:36
SPEAKER_00: And by virtue of being in the air, they're actually safer, turns out.
19:40
SPEAKER_00: Air travel is safer than ground travel, and that's because there's almost nothing to hit in the air, but there's lots of stuff to hit on the ground.
19:47
SPEAKER_00: Um, so you could uh conceivably have a future where you don't really need roads anymore, right?
19:53
SPEAKER_00: You can you can use the roads to turn them into parks or whatever, right?
19:56
SPEAKER_00: Into bicycle lanes, but then have these places where these things just take off like a drone and fly on a straight line.
20:02
SPEAKER_00: They're also um not just uh greener, which is actually really remarkable.
20:06
SPEAKER_00: People don't think of aviation as greener than ground transport, but but by and large, it actually is greener.
20:11
SPEAKER_00: Um, this is something we can debate at length.
20:14
SPEAKER_00: Um when you look at the actual energy consumed per traveler passenger mile, a plane wins or a train, turns out, even though people hate when I say this, especially in Europe.
20:23
SPEAKER_00: Um, but it's also, I mean, it's also quieter.
20:25
SPEAKER_00: It's I mean these things are so incredibly quiet that you can't hear them anymore.
20:29
SPEAKER_02: We had Reed Hoffman on the podcast uh a couple months ago, and we he came up with an interesting analogy that you know intelligence is to the uh AI era, what connectivity was to the internet era, and everything is gonna have intelligence, everything is gonna have autonomy.
20:42
SPEAKER_02: You are a godfather of autonomy.
20:44
SPEAKER_02: Um, you know, we see we see aircraft, we see cars.
20:48
SPEAKER_02: What are other big ideas or big sectors that will become autonomous and really impact the world?
20:53
SPEAKER_00: If you look at um companies, let's say telcos, complete different example, right?
20:59
SPEAKER_00: So telcos employ tens of thousands of people.
21:02
SPEAKER_00: Um and what do they do?
21:04
SPEAKER_00: I mean, lots of stuff from sales to customer service to fixing the network and cell phone towers and so on.
21:10
SPEAKER_00: Um, you can imagine that with intelligence, you could build a telco with like 10 people.
21:17
SPEAKER_00: Um how would that work?
21:20
SPEAKER_00: Well, all these things that break all the time, you have to fix, you would automate.
21:24
SPEAKER_00: Every time you interact with a customer, you automate this.
21:27
SPEAKER_00: And that's underway.
21:28
SPEAKER_00: This is I'm not making this up.
21:29
SPEAKER_00: This is definitely underway.
21:30
SPEAKER_00: So you can take industries that are very labor-intense and highly automate them.
21:35
SPEAKER_00: Um, that's happening in manufacturing.
21:37
SPEAKER_00: So, SoftCons Vision, the the company that employs 1.5 million people to manufacture things like your iPhone, massive numbers of people.
21:46
SPEAKER_00: Teregu, the CEO, recently said he would like to guide one employee.
21:50
SPEAKER_00: Okay, then the entire company is run by one person, and the rest is robotic.
21:54
SPEAKER_00: Uh, that sounds very much like science fiction, but bear with me in the following sense.
22:00
SPEAKER_00: Everything that you take for granted, listener, okay, like your cell phone, your car, even like very basic things, your food, your, your, your, your clothing, is a result of massive automation in the last 100 or so years and didn't exist before.
22:16
SPEAKER_00: And then even like your light switch at home, right?
22:18
SPEAKER_00: Didn't exist 150 years ago.
22:20
SPEAKER_00: You almost certainly didn't have a flashing toilet at home, like basic stuff, like a warm shower.
22:24
SPEAKER_00: Um, so if you if you look a little bit, zoom a little bit out and ask yourself what's happening in society, you find that um things are moving really fast.
22:32
SPEAKER_00: And there's really no reason why human labor is being used the way it's being used today.
22:37
SPEAKER_00: The same way there's no reason why 150 years ago we all worked in farming and should have stayed this way.
22:43
SPEAKER_00: Right now, less than 2% of us work in farming because of automation.
22:47
SPEAKER_00: And in the future, less of 2% of us would have worked on to any company that we know today.
22:52
SPEAKER_01: Yeah, I think that's really interesting when you think about the debate of reshoring manufacturing jobs in the US and Europe.
22:57
SPEAKER_01: You know, if you look at BYD, one of the big, the biggest uh EV producers, I think they make one car per day per two employees of the factory.
23:06
SPEAKER_01: And I think the US is closer to six, Europe probably even more.
23:10
SPEAKER_01: And so you look at that and wonder where those jobs are going to be as we reshore them into the US and Europe.
23:16
SPEAKER_00: Yeah, look, when I talk to my Silicon Valley friends about the mandate to rebuild manufacturing in the US, we talk about things like 3D printing, which would allow everybody to have factory factory in their basement.
23:28
SPEAKER_00: Uh, the best 3D printers today cost less than 2,000 bucks.
23:33
SPEAKER_00: Best in the sense that there's a lot of um innovation happening at the grassroots, the same way in the 70s and 80s, there was a lot of innovation for PCs in the grassroots for computing, um, with where all the innovation was back then, and we all benefit from it today.
23:48
SPEAKER_00: Um, so there's there's thinking about how we can rechange manufacturing, and maybe we can make it on demand as opposed to, I don't know, like you take the clothing industry.
23:58
SPEAKER_00: Um, it the clothing industry is responsible for 10% of global emissions.
24:02
SPEAKER_00: People don't know this, and only two in three garments ever get sold.
24:06
SPEAKER_00: Okay, so a third of the stuff just gets disposed, destroyed, sent to Africa, what have you, or donated.
24:12
SPEAKER_00: Um, if you could make clothing on demand, right, all of a sudden you could cut global emissions by, I don't know, 3%.
24:18
SPEAKER_00: It'd be amazing.
24:19
SPEAKER_01: Let's talk a little bit about what you describe and others describe as the singularity.
24:23
SPEAKER_01: You said we are living through the singularity now in terms of super intelligence.
24:28
SPEAKER_01: What does that mean for you in practice?
24:31
SPEAKER_01: So, what is a singularity?
24:32
SPEAKER_00: A singularity is a moment where things accelerate so vastly that you won't be able to say what happens at the other end.
24:40
SPEAKER_00: Okay, that's a singularity.
24:42
SPEAKER_00: Uh, and we can we can scribble about the definition.
24:45
SPEAKER_00: But you look at um look at society today.
24:48
SPEAKER_00: Um, if you if you think about itself back, say 100,000 years, you could have easily predicted the next 10,000 years because not much really changed, to be honest.
24:56
SPEAKER_00: Maybe a few little innovations, like, I don't know, steel, what have you, but that's pretty much it.
25:02
SPEAKER_00: Stone Ages were boring, okay?
25:04
SPEAKER_00: If you lived 500 years ago, same thing.
25:06
SPEAKER_00: Europe would be engulfed in war after war.
25:08
SPEAKER_00: If you lived 100 years ago, things a little bit faster.
25:11
SPEAKER_00: We have the steam engine, you have uh the airplane innovation and stuff like this.
25:15
SPEAKER_00: Um, you have the 50 years ago, 20 years ago.
25:18
SPEAKER_00: When I was a student, I was able to predict the next seven years.
25:22
SPEAKER_00: Okay, so in computer science, my own field, every seven years there was a major new programming language and a major new programming framework, and things changed.
25:30
SPEAKER_00: And past that point, I couldn't predict.
25:32
SPEAKER_00: Today, I cannot even predict the next seven months because things are moving so insanely fast.
25:38
SPEAKER_00: And that to me, if even the experts in the world can't predict what's happening, that to me is a singularity.
25:44
SPEAKER_00: We now see for the first time that for the vast majority of people, the machines are as good or better in big aspects of their daily work.
25:54
SPEAKER_00: And I'm not scared of it, I actually love it myself because I think we're wasting a ton of time ourselves doing stupid stuff.
26:00
SPEAKER_00: But it's amazing to see how big machines now are so smart that they be the world's best chess player and the world's best go players, but also are as good as the world's best, or maybe close as good as the world's best lawyers and doctors and all these things that we inspire our kids to do to be safe in life and have a great job.
26:17
SPEAKER_02: And why is the pace, the exponential pace of improvement and change, what's going on there?
26:23
SPEAKER_02: Why is it happening now?
26:24
SPEAKER_00: Well, so the big innovation in the last few years, which I'm sure all your listeners have been talking about every night for many, many years now, is data.
26:33
SPEAKER_00: Okay, so um OpenAI was probably the first to see this.
26:37
SPEAKER_00: They got the world's text data together, which is hundreds of billions of documents and emails and pieces of software, and they're put in a big table, a big machine to predict more data.
26:50
SPEAKER_00: So, how does this work?
26:51
SPEAKER_00: Like um, you can always predict the next word in a simple way.
26:55
SPEAKER_00: Let's say like the dog ate the dog food is a good word, or maybe homework, but not the propeller, right?
27:03
SPEAKER_00: That's not a good word.
27:04
SPEAKER_00: So that idea of statistically figuring out what's a good next word um based on three, four, five words, has been around for several decades in uh statistics, machine translation linguistics.
27:17
SPEAKER_00: But now with the latest large language technology, you can predict the next word based on hundreds of thousands of words in context.
27:26
SPEAKER_00: And it turns out, strangely enough, uh predicting the next word, if you take enough context and you're smart enough about it, looks really intelligent.
27:34
SPEAKER_00: So intelligent that large language models can write a rap song, they can write a haiku, they can translate languages.
27:42
SPEAKER_00: In fact, they can do 90% of what a software engineer does today in terms of writing computer code, software engineering software, they can now do this automatically just by looking at past data and extrapolating from it.
27:57
SPEAKER_01: Sebastian, I think I've read an interview where you talked about in the next period of time, because of those advancements, the pace of those advancements, we're gonna enter a world where the basic needs of humans are gonna be met in a very significantly different way.
28:11
SPEAKER_01: We're gonna have essentially close to free food, energy, even clothing, potentially housing.
28:17
SPEAKER_01: And it's something that almost every one of the senior leaders in AI that I've spoken to and that we've had on the show, uh, on John Ideas, people like Mustafa Sullivan, who founded DeepMind, that seems to be a view that people right at the center of AI's progress share.
28:33
SPEAKER_01: And I don't, it's never actually been particularly clear to me the leap between the pace of change we're talking about and all the amazing things that AI can do and it can be a great lawyer and so on.
28:43
SPEAKER_01: But why does that mean we're going to get almost free energy, almost free food?
28:47
SPEAKER_01: Um, these very physical basic needs that human beings need.
28:50
SPEAKER_00: Yeah, I mean, if you look at even pre-AI and local automation and what we're doing in society, food is now so massively available that obesity has become a bigger problem than malnutrition.
29:04
SPEAKER_00: There's still pockets in the world where food is very short right now, usually for political reasons.
29:10
SPEAKER_00: But if you were to distribute our food better, we could feed every human being, to be honest.
29:16
SPEAKER_00: Um, and that's new in history.
29:18
SPEAKER_00: Like if you go back two, three, four hundred years ago, all Europeans affected, we lived in extreme poverty.
29:26
SPEAKER_00: Uh take clothing.
29:28
SPEAKER_00: Uh, three, four hundred years ago, you would rip the clothing off dead people's body as your supply for clothing for many people.
29:36
SPEAKER_00: Now, I mean, often the teacher is cheaper than the laundry to wash it.
29:41
SPEAKER_00: It's amazing how disposable it has become.
29:43
SPEAKER_00: Or look at education.
29:45
SPEAKER_00: Um, education is effectively available worldwide uh for pretty much everybody, maybe not the same quality, but online education.
29:53
SPEAKER_00: I built a company on that topic that really democratized access to high-quality online education.
29:58
SPEAKER_00: Um, K through 12 education for kids is available, vaccinations are available.
30:03
SPEAKER_00: I think 95% of kids under the age of five are vaccinated, some very large number.
30:09
SPEAKER_00: Um, so all these things have been improving even pre-AI.
30:12
SPEAKER_00: And they're all amazing um accomplishments because what they did to us people is they allowed us to escape the daily churn of farm working and learn, for example, how to read or write, and how to become scientists, and how to really advance the world knowledge.
30:28
SPEAKER_00: We've done amazing work in making agriculture more effective and all these things.
30:33
SPEAKER_00: It's all pre-AI.
30:35
SPEAKER_00: Now, AI will superpower all this even faster, right?
30:39
SPEAKER_00: So we still live in a world where all of us effectively work, most of us do, and we work mostly in offices now.
30:46
SPEAKER_00: And the Western world is like 70% or so.
30:49
SPEAKER_00: And what we do in offices is communicate with other people, one manada, right?
30:52
SPEAKER_00: So we write memos, we we write orders, we listen, we go to meetings, what have you, we write email.
30:58
SPEAKER_00: Um, all that stuff now can be automated very so much.
31:01
SPEAKER_00: Free up enormous creativity in society, which can put us into an incredible position, historically, never happened before, that we have a lot of free time to be creative.
31:11
SPEAKER_00: How amazing would that be?
31:13
SPEAKER_02: One last question on the topic of autonomy.
31:16
SPEAKER_02: Uh, you have designed some of these systems from the ground up.
31:19
SPEAKER_02: You've been really at ground zero.
31:21
SPEAKER_02: What are some of the potential failure modes that can be designed into these autonomous systems that aren't maybe being talked about enough?
31:29
SPEAKER_00: Especially in self-driving cars, um, what do you find is that there is um a long tail of situations that even the smartest engineers might never anticipate that might come to bite you.
31:44
SPEAKER_00: Um, so for example, um, let's take Waymo.
31:47
SPEAKER_00: Waymo is just about now to go to autonomous highway driving.
31:51
SPEAKER_00: Um, and highway driving in many ways is the simplest or the easiest way of driving because if if it goes well in a good day, nothing really moves.
31:58
SPEAKER_00: In the United States, all the different lanes have the same speed limit.
32:01
SPEAKER_00: So we all basically run rolling around the same speed, and relative to you, all the other traffic is effectively standing still.
32:09
SPEAKER_00: But why did it take way more a long time to do autonomous driving, driverless driving on highways?
32:14
SPEAKER_00: It's because there's this very rare incident where like you have an accident in front of you or like a pedestrian or a cow or what have you, something that you haven't really thought about.
32:24
SPEAKER_00: Like a cow on a highway, okay?
32:25
SPEAKER_00: You think that it's impossible unless you're in India.
32:28
SPEAKER_00: But in reality, yeah, there'll be a cow on a highway occasionally, or a sheep, or whatever, a dog, a coyote, a deer.
32:34
SPEAKER_00: Deers are pretty common in highways.
32:36
SPEAKER_00: So, how do you um deal with these very rare instances that could be conceivably fatal?
32:42
SPEAKER_00: Like deer can actually kill people.
32:44
SPEAKER_00: They're very hard to avoid.
32:45
SPEAKER_00: Moose kill people regularly when cars in highways hit a moose.
32:50
SPEAKER_00: So, how do you deal with this incredibly rare instance and still make your car safe?
32:54
SPEAKER_00: And that's challenging because, as I said before, your tolerance for error, your tolerance for hallucinations in self-driving is effectively zero.
33:03
SPEAKER_01: Maybe let's talk a little about education.
33:05
SPEAKER_01: You obviously founded Udacity with this massive online learning startup.
33:10
SPEAKER_01: Um, maybe before we go into some of the questions about the future of education, just tell us, if you can, the story of how Udacity came about.
33:17
SPEAKER_00: The company was started uh 2011 when I was teaching as a professor at Stanford University.
33:22
SPEAKER_00: And I taught a graduate level class on AI.
33:26
SPEAKER_00: Uh, today a very hot topic, but trust me, in 2011 it was an esoteric topic.
33:30
SPEAKER_00: And I decided to um to put the class online and not just the lectures as videos, but I created a classroom environment where people had to take homework exams and quizzes and pass the midterm exam and the final exam.
33:45
SPEAKER_00: And I made it so that the online students would effectively have to pass the same exams as like the Stanford PhD doctoral students.
33:53
SPEAKER_00: Okay.
33:54
SPEAKER_00: Uh, we we put it online, we sent out one email to a couple of friends, and I expected maybe 500 students would sign up.
34:02
SPEAKER_00: I mean, Stanford's campus was about 200 every year, so maybe a thousand.
34:06
SPEAKER_00: That would be optimistic.
34:07
SPEAKER_00: Uh, the next morning, this was a Friday evening, we had 5,000 sign up.
34:12
SPEAKER_00: On Sunday morning, 10,000.
34:15
SPEAKER_00: On Monday morning, 14,000 people had already signed up.
34:20
SPEAKER_00: Which is when my dean at Stanford found out because I was being plotted on the blogosphere for finally doing away with Stanford tuition, which a message he didn't quite appreciate.
34:34
SPEAKER_00: And then I went on to reach about 160,000.
34:38
SPEAKER_00: And I realized, my God, my impact teaching 160,000 students is more than I can do in 10 years of life, 10 consecutive lifetimes at Stanford University.
34:49
SPEAKER_00: So I really focused on how I can teach these 160,000 students effectively.
34:54
SPEAKER_00: So we really built this new modality, online, interactive, quiz-driven learning system overnight that would teach 160,000 students.
35:03
SPEAKER_00: And these were not just your typical Stanford students.
35:05
SPEAKER_00: Some of them were like soldiers in Afghanistan.
35:07
SPEAKER_00: Some were like a single mother trying to raise their infant.
35:10
SPEAKER_00: There was a person on the deathbed whose last wish really was to finish this class before he died.
35:15
SPEAKER_00: People you would, Stanford would normally not admit, let's put it this way, okay?
35:18
SPEAKER_00: But anywhere as smart as the best Stanford student.
35:21
SPEAKER_00: In the end, 23,000 of these students finished.
35:24
SPEAKER_00: And I got a chance to stack ranked the Stanford students, well, these online students.
35:30
SPEAKER_00: And guess what?
35:32
SPEAKER_00: The top 412 finishers were not at Stanford.
35:36
SPEAKER_00: And the best Stanford doctoral student ranked number 413.
35:42
SPEAKER_00: And that opened my eyes to I mean, Stanford is an amazing institution.
35:47
SPEAKER_00: I love Stanford, and it does a fantastic job, but it's very confined.
35:51
SPEAKER_00: It only accepts a few thousand students every year, a few hundred students.
35:56
SPEAKER_00: And the world has eight billion people.
35:59
SPEAKER_00: So if we just opened the floodgates and let everybody study at Stanford, we would have so much more impact than having a small elite university here in California.
36:08
SPEAKER_01: And is that the way that the world is going to go, particularly now with AI?
36:12
SPEAKER_01: And how do you think AI is going to supercharge the efforts that you began really with Udacity to democratize education?
36:18
SPEAKER_00: Yeah, first let's, I mean, Udacity became a global force.
36:21
SPEAKER_00: We do a lot of Middle Eastern education in places like Egypt and Saudi Arabia, where people in Nigeria and Northern Africa, where people really have no chance to ever get a great technical education.
36:39
SPEAKER_00: Now, AI is the next chapter.
36:41
SPEAKER_00: And there's a very famous uh 1982 paper by a gentleman named Bloom who proved that if you give a normal kid a tutor to the entire like youth and lifetime when they grow up, they will perform at the level of a highly exceptional gifted kid.
36:59
SPEAKER_00: So you can turn a normal kid into a world-class gifted child, it's all relative, of course, just by tutoring it one-on-one.
37:07
SPEAKER_00: And what does a tutor do?
37:08
SPEAKER_00: A tutor is much better to adapt to kids' needs, right?
37:12
SPEAKER_00: So whereas a teacher might teach 30 kids at a time, a tutor teaches one kid at a time.
37:17
SPEAKER_00: And it turns out in terms of math, English, and so on, a tutor delivers better results.
37:22
SPEAKER_00: Um, now what AI can do is AI can now understand the learner deeper than ever before.
37:27
SPEAKER_00: And rather than putting a single curriculum out for every kid identical, like today, where we teach kids, we have an identical curriculum for all the kids effectively, it could adapt to the child and say, this child is more of a visual learner, or this needs more exercise, or this has kids more about history and less about math.
37:43
SPEAKER_00: And then really evoke a conversation with the child and help the child to become better and better.
37:49
SPEAKER_00: It's unproven, it's a hypothesis.
37:51
SPEAKER_00: Um, in child education, it hasn't been shown.
37:54
SPEAKER_00: In adult in uh professional training where I work, it has been shown to work better.
38:00
SPEAKER_00: Um, and I think they're gonna go into a world where all of a sudden, even learning gets superpowered, and all the kids learn so much more than in the past.
38:08
SPEAKER_02: So, to maybe summarize, moving from sort of democratization of education to mass personalization, where everyone gets an individual tutor tailored to them.
38:17
SPEAKER_02: Amazing.
38:18
SPEAKER_02: You've had some huge successes like Waymo, but you've also had some challenges.
38:21
SPEAKER_02: You had to wind down Kitty Hawk.
38:23
SPEAKER_02: What heuristics do you use when you're committing years of your life to an idea?
38:26
SPEAKER_02: We know building a company is a challenge, it's an adventure, uh, but always has ups and downs.
38:31
SPEAKER_02: What are the heuristics you use?
38:33
SPEAKER_00: The majority of Kitty Hawk was sold to Boeings.
38:35
SPEAKER_00: We had it okay.
38:36
SPEAKER_00: It's still available as a company right now called whisk.ero.
38:41
SPEAKER_00: It wasn't exactly the path I anticipated in some cases.
38:44
SPEAKER_00: Um, I always think that there's some these amazing things where we can change the world.
38:48
SPEAKER_00: And my goal is to make the world better for other people.
38:50
SPEAKER_00: So, for example, education, what if you can truly democratize education, right?
38:55
SPEAKER_00: That's the mountain you want to climb.
38:56
SPEAKER_00: So you pick a mountain in life that will motivate you 10 years later, right?
39:00
SPEAKER_00: So self-driving cars motivates me 10 years later.
39:04
SPEAKER_00: Transportation, uh, health motivates me 10 years later.
39:07
SPEAKER_00: And that's the mountain you're ought to climb.
39:09
SPEAKER_00: But then you, if you actually, this is a mountain never been climbed before, right?
39:12
SPEAKER_00: So you pick something that hasn't been done before.
39:14
SPEAKER_00: There's no book you can buy on Amazon that says how to do it because it hasn't been done before.
39:18
SPEAKER_00: And then you go and um and start climbing, right?
39:20
SPEAKER_00: And as anyone here knows, that's a completely different story.
39:23
SPEAKER_00: You're gonna put your first foot forward the second, and then you, I know, you you run at the false summit and you have to retract, which happens all the time in innovation.
39:31
SPEAKER_00: That you build something, it doesn't work.
39:33
SPEAKER_00: And then you have to undo it, possibly even fire people in this process.
39:37
SPEAKER_00: Um, that is painful for most people, but it has to keep motivating you because rather than you didn't make any progress, maybe in terms of getting your goal done, but you learned something interesting that other people don't know yet and you didn't know, so you're not going to make the same mistake again.
39:50
SPEAKER_00: So you're constantly making mistakes.
39:52
SPEAKER_00: You go into, as a climber, let's say, take the mountain climbing allergy, you go into bad weather, right?
39:56
SPEAKER_00: So all of a sudden there's a thunderstorm, right?
39:59
SPEAKER_00: You can't see the summit anymore.
40:00
SPEAKER_00: That doesn't mean the summit is gone.
40:02
SPEAKER_00: You still have to believe it exists, but maybe your colleagues can't see it and they get frustrated.
40:06
SPEAKER_00: And you have to keep climbing and you have to be very flexible.
40:10
SPEAKER_00: Um, my estimate is if you know what you're doing, if someone can be a little bird in your ear and tell you exactly every every moment exactly what to do, you're typically five times faster than if you're a true entrepreneur.
40:22
SPEAKER_00: It takes five times longer to figure things out than if you could pick up a book at Amazon and would tell you how to do it.
40:29
SPEAKER_00: So if you do something new like building democracy and education or building self-driving cars, four out of five days are effectively spent learning something interesting without making real progress towards the goal.
40:39
SPEAKER_00: And then pick something that really is meaningful.
40:41
SPEAKER_00: And to me, the most meaningful is stuff.
40:43
SPEAKER_00: I always have this thing called grandmother test.
40:45
SPEAKER_00: Like, can I go to my fictitious my grandmother's or passed away, obviously, my fictitious grandmother and tell her I'm working with something like this.
40:52
SPEAKER_00: Okay.
40:53
SPEAKER_00: And if I tell her, look, I'm building you a car that can get you anywhere that's self-driving, she probably gets it.
40:59
SPEAKER_00: But as say I'm building a lock likelihood discriminator that goes into blah, blah, blah and does blah, blah, blah, she's um, I don't care about that stuff.
41:07
SPEAKER_00: So if you're, I think, what are the basic things that people really deeply care about?
41:11
SPEAKER_00: It's communication, it's health, it's education, it's safety.
41:16
SPEAKER_00: Um, these are the things we truly care, maybe travel experiences we care about.
41:20
SPEAKER_00: Can you relate your innovation to those things?
41:22
SPEAKER_00: And I would say when you do things like transportation or education, or most recently I've been working on shopping and e-commerce, these are things people do care about.
41:32
SPEAKER_00: And you can explain to somebody who's non-technical.
41:36
SPEAKER_01: Tell us about the new startup.
41:37
SPEAKER_00: Yes, I haven't really talked much about it, but we have a startup that's called Shop on Gold, which has launched just last week.
41:45
SPEAKER_00: Uh, it's a personalization app that goes out on the internet and in the space of fashion, finds everyday stuff for you that has been recently discounted and you might actually like, and uses AI to match those millions of things we find every night that are showing up new in the fashion industry, and finds those 50 pieces it believes you will love the most.
42:09
SPEAKER_00: And then it also has an agentic AI where once you like something, uh, with a single click, it does all the shopping for you.
42:15
SPEAKER_00: So you don't have to go and create a password on some obscure website and sign up for some newsletter you didn't want to sign up for and all that stuff.
42:22
SPEAKER_00: So it really changes the interface uh to the World Wide Web instead of having 30, 40, 50, 100 different shops you might actually attend to in your lifetime.
42:32
SPEAKER_00: There's just one that now it does all the shopping for you.
42:35
SPEAKER_01: Fantastic.
42:36
SPEAKER_01: Well, we should be watching the future of that uh very, very closely.
42:40
SPEAKER_01: Maybe if we can just wrap things up, Sebastian, by asking something we ask everyone who comes on giant ideas.
42:45
SPEAKER_01: Uh, and modesty is not allowed in this answer.
42:47
SPEAKER_01: So, what is it about you, do you think, that has allowed you to have the kind of extraordinary success that we've just talked about for the past half an hour?
42:55
SPEAKER_00: I'm both extremely arrogant and extremely humble.
42:57
SPEAKER_00: Okay, I'm arrogant when I say, let's democratize the world's education.
43:01
SPEAKER_00: I go to Stanford dealing and say, screw this, Stanford should admit 10 million people, not just 7,700 people.
43:08
SPEAKER_00: Um and that's certainly unappreciated by the people who care about the status quo.
43:13
SPEAKER_00: Um and I do this for everything, it does for government, it does this for health.
43:16
SPEAKER_00: Um, these mountains that have to be climbed, you have to pick them.
43:19
SPEAKER_00: And when you when you proclaim you're going to climb the mountain of making cars drive themselves, you're surrounded by people who basically think you're an idiot.
43:27
SPEAKER_00: Okay, let's be plain in the beginning.
43:29
SPEAKER_00: That's the way it feels.
43:30
SPEAKER_00: But then you can't just pretend you know everything.
43:33
SPEAKER_00: Um, the actual process of climbing this mountain is extremely humbling because you're constantly proven wrong.
43:40
SPEAKER_00: And you have to be able to motivate yourself when you constantly do the wrong thing.
43:45
SPEAKER_00: So give an example for my most recently startup, which is now doing extremely well, uh, Shop on Gold.
43:50
SPEAKER_00: Uh, we did five iterations of building something that just didn't work.
43:54
SPEAKER_00: We tested it with users and they basically hated it.
43:57
SPEAKER_00: And it's not because I'm not smart enough.
44:00
SPEAKER_00: Um, maybe it is.
44:02
SPEAKER_00: Um, but it's very hard to understand what makes people really tick and what makes things work.
44:06
SPEAKER_00: The same is true for Cadacity, the same was true for self-driving cars.
44:09
SPEAKER_00: Many, many, many iterations didn't work.
44:12
SPEAKER_00: And to me, uh, those give me energy.
44:14
SPEAKER_00: I love uncertainty and I love learning.
44:16
SPEAKER_00: I love learning new things, I love being proven wrong.
44:19
SPEAKER_00: I have a rule in my life.
44:20
SPEAKER_00: I'm allowed to make every mistake, but only once.
44:24
SPEAKER_00: If I make it the second time, I haven't learned anything about it.
44:26
SPEAKER_00: So I mean, then I'm actually an idiot.
44:28
SPEAKER_00: This is a word.
44:30
SPEAKER_00: But I I drive enormous energy out of this idea of like you do something, it doesn't work, and you make a mistake, and you look in the mirror and say, Shh I screw this up.
44:38
SPEAKER_00: And it's okay.
44:39
SPEAKER_00: That's great.
44:40
SPEAKER_00: I allow myself to make mistakes.
44:42
SPEAKER_00: And that's I think where many people fail.
44:44
SPEAKER_00: They they often move in and say, I know everything.
44:46
SPEAKER_00: And if you take the position you know everything, you gotta be extremely lucky to be correct because you're dealing with something it's never been done before and it's hard.
44:55
SPEAKER_00: It's more likely you believe you know everything, but you don't, and then you're being stuck with your mistakes and you can't understand that you actually have to learn, right?
45:04
SPEAKER_00: So I feel like an like a first grader in what I do every single day.
45:08
SPEAKER_00: I feel like a complete novice and have no clue what I'm doing, and I'm learning from the firehouse every single time I build a new company.
45:16
SPEAKER_02: One final question, Sebastian.
45:17
SPEAKER_02: What advice would you give to your 10-year-old self?
45:20
SPEAKER_00: Be curious.
45:21
SPEAKER_00: Um, I think the world is changing so rapidly.
45:23
SPEAKER_00: Don't get married to anything that exists today.
45:26
SPEAKER_00: And let's say uh develop grit.
45:28
SPEAKER_00: Um, grit to me is the one thing that is really driving success.
45:33
SPEAKER_00: Grit means you stick with it, right?
45:35
SPEAKER_00: When you if you're the person who jumps off quickly when the first failures come along, then you're not made for Silicon Valley, you're not made for this world of innovation.
45:43
SPEAKER_00: Pick a job, like work for the German government or the UK government where things are more predictable.
45:48
SPEAKER_00: But if you want to be in a world where where you invent something, you innovate, you push something forward.
45:53
SPEAKER_00: In fact, pretty much any job, if you have grit, you have a much higher chance for success.
45:59
SPEAKER_01: Superb place to end.
46:00
SPEAKER_01: Couldn't agree more.
46:01
SPEAKER_01: Sebastian, thank you so much for sharing not just one giant idea, but about three or four in this episode.
46:06
SPEAKER_01: Really, really appreciate you taking the time to join us.
46:08
SPEAKER_00: Thanks so much.
46:09
SPEAKER_00: Coming everyone, it was a real pleasure meeting you.
46:11
SPEAKER_02: Likewise.
46:16
SPEAKER_02: Fascinating conversation.
46:18
SPEAKER_02: Uh, one one insight I take away from that was just how advanced the technology is today.
46:23
SPEAKER_02: It's essentially pretty close to perfect.
46:25
SPEAKER_02: And uh by Sebastian's estimation, we have close to a million deaths a year from human drivers.
46:31
SPEAKER_02: Yeah, given that's the situation, we still, as a society, as a people, aren't quite comfortable with the role of that technology, even though the cost of doing that is is fairly high.
46:40
SPEAKER_02: That's pretty fascinating, both philosophically and and uh and socially.
46:43
SPEAKER_01: Yeah, I agree.
46:44
SPEAKER_01: And I think it's so striking when the technology is so far rolled out in San Francisco.
46:48
SPEAKER_01: So we heard there that a quarter of ride shares, right?
46:51
SPEAKER_01: A quarter of the kind of Uber equivalents are now Waymo driverless cars.
46:55
SPEAKER_01: It's working.
46:56
SPEAKER_01: It's just working, it's there, and you know, go to San Francisco and it's part of daily life.
46:59
SPEAKER_01: And it's so completely foreign to people in Rome or London where you just don't see that at all.
47:04
SPEAKER_01: Long way away.
47:04
SPEAKER_01: And I think it's unusual because mostly the technology that we get out of Silicon Valley is just widely adopted everywhere, all at the same time.
47:10
SPEAKER_01: And it's it's it's an example of that not being true at all, basically because of regulatory differences.
47:14
SPEAKER_01: And I think that will probably increase over time because of this kind of geopolitical differences that are now being accentuated.
47:19
SPEAKER_01: Um it's a very tangible example of the future is here, but it's not yet equally distributed.
47:24
SPEAKER_01: Yes, totally.
47:24
SPEAKER_01: And I feel like we need in Europe to get on it pretty quickly.
47:27
SPEAKER_01: Absolutely.
47:27
SPEAKER_01: We're gonna miss out.
47:28
SPEAKER_01: I think one thing I would have liked to push Sebastian on more was this question of why exactly AI is going to suddenly deliver free food, free energy, free clothes, which, as I said in the interview, like every single one of these top AI thinkers are all saying.
47:42
SPEAKER_01: But no one, I think to me at least, is actually nailing down exactly what is the jump between AI and free energy, which obviously would solve climate change and all this, and is to me just seems a bit of a fantasy.
47:53
SPEAKER_01: Did for you, do you see that?
47:55
SPEAKER_02: Yeah, I think it's an assumption of this uh extrapolating from the progress that's happening that it will continue, which I think is is an assumption.
48:02
SPEAKER_02: Yeah, and that there will be this these emergent incredible benefits, and we have to use our imagination.
48:07
SPEAKER_02: But I also feel yeah, it depends on the day that you ask me whether I think that's gonna happen or not.
48:11
SPEAKER_02: Yeah.
48:11
SPEAKER_02: Um, one thing I wish he would have gone a little bit deeper on was the failure modes of building these autonomous systems.
48:15
SPEAKER_02: He's built them from the ground up.
48:17
SPEAKER_02: Uh, you know, uh it would have been great to go maybe a level deeper on how it could go wrong and how we should design these systems for those edge case scenarios to protect us and protect the planet and and protect uh yeah, our lives.
48:29
SPEAKER_01: Agreed.
48:30
SPEAKER_01: But overall, I thought it was great.
48:31
SPEAKER_01: I mean, it was really nice and incredible.
48:32
SPEAKER_01: I saw him 12 years ago in Silicon Valley in his office, and he told me driverless cars are gonna be a thing, it is gonna work.
48:37
SPEAKER_01: And at that point, no, it was a kind of the pit of despair for for autonomous vehicles.
48:41
SPEAKER_01: Yeah.
48:41
SPEAKER_01: And it's now happening, and it's kind of down to him.
48:43
SPEAKER_01: Lovely fellow and a clear visionary.
48:45
SPEAKER_01: Yeah.
48:49
SPEAKER_01: If you like the episode, don't forget to like, subscribe, and follow the show.
48:52
SPEAKER_01: And you can find many more giant ideas wherever you get your podcasts.