Sebastian Thrun: Flying Cars Has Always Been the Dream | AI Podcast Clips
00:00:00.000 | - Many, many years ago, 1903,
00:00:05.000 | the Wright brothers flew in Kitty Hawk for the first time.
00:00:10.560 | And you've launched a company of the same name, Kitty Hawk,
00:00:14.900 | with the dream of building flying cars, EVtols.
00:00:19.900 | So at the big picture,
00:00:22.540 | what are the big challenges of making this thing
00:00:24.580 | that actually have inspired generations of people
00:00:27.960 | about what the future looks like?
00:00:29.700 | What does it take?
00:00:30.540 | What are the biggest challenges?
00:00:31.640 | - So flying cars has always been a dream.
00:00:35.200 | Every boy, every girl wants to fly.
00:00:37.680 | Let's be honest.
00:00:38.520 | - Yes.
00:00:39.340 | - And let's go back in our history
00:00:40.320 | of your dreaming of flying.
00:00:41.740 | I think my, honestly,
00:00:42.640 | my single most remembered childhood dream has been a dream
00:00:46.280 | where I was sitting on a pillow and I could fly.
00:00:48.700 | I was like five years old.
00:00:50.000 | I remember like maybe three dreams of my childhood,
00:00:52.120 | but that's the one I remember most vividly.
00:00:54.360 | And then Peter Thiel famously said,
00:00:57.360 | "They promised us flying cars and they gave us 140 characters,"
00:01:00.760 | pointing at Twitter at the time,
00:01:03.200 | limited message size to 140 characters.
00:01:06.320 | So we're coming back now to really go
00:01:08.160 | for this super impactful stuff like flying cars.
00:01:11.160 | And to be precise, they're not really cars.
00:01:13.840 | They don't have wheels.
00:01:15.080 | They're actually much closer to a helicopter
00:01:16.520 | than anything else.
00:01:17.560 | They take off vertically and they fly horizontally,
00:01:20.000 | but they have important differences.
00:01:22.320 | One difference is that they are much quieter.
00:01:25.680 | We just released a vehicle called Project Heaviside
00:01:29.560 | that can fly over you as low as a helicopter.
00:01:31.480 | And you basically can't hear it.
00:01:33.240 | It's like 38 decibels.
00:01:34.680 | It's like, if you were inside the library,
00:01:37.220 | you might be able to hear it,
00:01:38.200 | but anywhere outdoors, your ambient noise is higher.
00:01:40.960 | Secondly, they're much more affordable.
00:01:45.000 | They're much more affordable than helicopters.
00:01:46.940 | And the reason is helicopters are expensive
00:01:49.880 | for many reasons.
00:01:50.980 | There's lots of single point of figures in a helicopter.
00:01:54.960 | There's a bolt between the blades
00:01:57.080 | that's called Jesus bolt.
00:01:58.720 | And the reason why it's called Jesus bolt
00:02:00.360 | is that if this bolt breaks, you will die.
00:02:04.320 | There is no second solution in helicopter flight.
00:02:07.440 | Whereas we have these distributed mechanism.
00:02:09.440 | When you go from gasoline to electric,
00:02:11.680 | you can now have many, many, many small motors
00:02:13.760 | as opposed to one big motor.
00:02:15.200 | And that means if you lose one of those motors,
00:02:16.720 | not a big deal.
00:02:17.560 | Heaviside, if it loses a motor, has eight of those,
00:02:20.760 | if it loses one of those eight motors,
00:02:21.960 | so it's seven left,
00:02:23.120 | it can take off just like before
00:02:25.200 | and land just like before.
00:02:26.760 | We are now also moving into a technology
00:02:29.960 | that doesn't require a commercial pilot
00:02:32.080 | because in some level,
00:02:33.440 | flight is actually easier than ground transportation.
00:02:36.920 | Like in self-driving cars,
00:02:38.660 | the world is full of like children and bicycles
00:02:42.440 | and other cars and mailboxes and curbs
00:02:44.560 | and shrubs and what have you.
00:02:46.360 | All these things you have to avoid.
00:02:48.440 | When you go above the buildings and tree lines,
00:02:51.680 | there's nothing there.
00:02:52.520 | I mean, you can do the test right now,
00:02:54.040 | look outside and count the number of things you see flying.
00:02:57.360 | I'd be shocked if you could see more than two things.
00:02:59.440 | It's probably just zero.
00:03:00.800 | In the Bay Area, the most I've ever seen was six
00:03:04.880 | and maybe it's 15 or 20, but not 10,000.
00:03:08.320 | So the sky is very ample and very empty and very free.
00:03:11.920 | So the vision is, can we build a socially acceptable
00:03:15.760 | mass transit solution for daily transportation
00:03:20.280 | that is affordable?
00:03:22.200 | And we have an existence proof.
00:03:24.280 | Heaviside can fly a hundred miles in range
00:03:27.720 | with still 30% electric reserves.
00:03:31.200 | It can fly up to like 180 miles an hour.
00:03:34.000 | We know that that solution at scale
00:03:36.800 | would make your ground transportation 10 times as fast
00:03:40.560 | as a car based on use census or statistics data,
00:03:45.320 | which means you would take your 300 hours
00:03:48.240 | of yearly commute down to 30 hours
00:03:50.960 | and give you 270 hours back.
00:03:53.120 | Who wouldn't want, I mean, who doesn't hate traffic?
00:03:55.640 | Like I hate, give me the person who doesn't hate traffic.
00:03:58.760 | I hate traffic.
00:03:59.600 | Every time I'm in traffic, I hate it.
00:04:01.600 | And if we could free the world from traffic,
00:04:05.480 | we have technology, we can free the world from traffic.
00:04:08.000 | We have the technology.
00:04:09.280 | It's there, we have an existence proof.
00:04:11.000 | It's not a technological problem anymore.
00:04:13.360 | - Do you think there is a future where tens of thousands,
00:04:17.280 | maybe hundreds of thousands of both delivery drones
00:04:22.280 | and flying cars of this kind, EV towers fill the sky?
00:04:27.320 | - I absolutely believe this.
00:04:28.880 | And there's obviously the societal acceptance
00:04:31.520 | is a major question.
00:04:33.400 | And of course, safety is.
00:04:34.880 | I believe in safety,
00:04:36.000 | we're gonna exceed ground transportation safety
00:04:38.280 | as has happened for aviation already, commercial aviation.
00:04:42.440 | And in terms of acceptance,
00:04:44.560 | I think one of the key things is noise.
00:04:46.240 | That's why we are focusing relentlessly on noise
00:04:48.880 | and we built perhaps the quietest electric VTOL vehicle
00:04:53.560 | ever built.
00:04:54.400 | The nice thing about the sky is it's three dimensional.
00:04:57.680 | So any mathematician will immediately recognize
00:05:00.440 | the difference between 1D of like a regular highway
00:05:02.880 | to 3D of a sky.
00:05:04.160 | But to make it clear for the layman,
00:05:07.280 | say you wanna make 100 vertical lanes
00:05:10.680 | of highway 101 in San Francisco,
00:05:12.960 | because you believe building 100 vertical lanes
00:05:15.160 | is the right solution.
00:05:16.800 | Imagine how much it would cost
00:05:18.120 | to stack 100 vertical lanes physically onto 101.
00:05:21.320 | That would be prohibitive.
00:05:22.240 | That would be consuming the world's GDP for an entire year,
00:05:25.720 | just for one highway.
00:05:27.160 | It's amazingly expensive.
00:05:29.200 | In the sky, it would just be a recompilation
00:05:31.640 | of a piece of software
00:05:32.480 | because all these lanes are virtual.
00:05:34.480 | That means any vehicle that is in conflict
00:05:37.160 | with another vehicle would just go to different altitudes
00:05:39.760 | and then the conflict is gone.
00:05:41.240 | And if you don't believe this,
00:05:43.280 | that's exactly how commercial aviation works.
00:05:46.480 | When you fly from New York to San Francisco,
00:05:49.360 | another plane flies from San Francisco to New York,
00:05:52.120 | they're at different altitudes,
00:05:53.200 | so they don't hit each other.
00:05:54.640 | It's a solved problem for the jet space
00:05:58.320 | and it will be a solved problem for the urban space.
00:06:00.640 | There's companies like Google, Bing and Amazon
00:06:03.280 | working on very innovative solutions
00:06:04.960 | how do we have space management.
00:06:06.480 | They use exactly the same principles as we use today
00:06:09.560 | to route today's jets.
00:06:11.200 | There's nothing hard about this.
00:06:12.880 | - Do you envision autonomy being a key part of it
00:06:16.920 | so that the flying vehicles
00:06:20.360 | are either semi-autonomous or fully autonomous?
00:06:24.840 | - 100% autonomous, you don't want idiots like me
00:06:27.520 | flying in the sky, I promise you.
00:06:29.880 | And if you have 10,000,
00:06:31.160 | watch the movie "The Fifth Element"
00:06:33.960 | to get a view for what would happen
00:06:36.160 | if it's not autonomous.
00:06:37.400 | - And a centralized, that's a really interesting idea
00:06:39.680 | of a centralized sort of management system
00:06:43.200 | for lanes and so on.
00:06:44.280 | So actually just being able to have
00:06:46.760 | similar as we have in the current commercial aviation,
00:06:50.960 | but scale it up to much more vehicles.
00:06:53.520 | That's a really interesting optimization problem.
00:06:55.640 | - It is mathematically very, very straightforward.
00:06:59.080 | Like the gap we leave between jets is gargantuous
00:07:01.480 | and part of the reason is there isn't that many jets.
00:07:04.360 | So it just feels like a good solution.
00:07:06.760 | Today, when you get vectored by air traffic control,
00:07:10.320 | someone talks to you, right?
00:07:11.840 | So an ATC controller might have up to maybe 20 planes
00:07:14.880 | on the same frequency and then they talk to you,
00:07:17.080 | you have to talk back.
00:07:18.280 | And that feels right because there isn't more than 20 planes
00:07:20.640 | around anyhow, so you can talk to everybody.
00:07:22.880 | But if there's 20,000 things around,
00:07:24.680 | you can't talk to everybody anymore.
00:07:25.920 | So we have to do something that's called digital,
00:07:28.200 | like text messaging.
00:07:29.440 | Like we do have solutions,
00:07:30.960 | like we have what, four or five billion smartphones
00:07:33.480 | in the world now, right?
00:07:34.400 | And they're all connected.
00:07:35.680 | And somehow we solve the scale problem for smartphones.
00:07:38.680 | We know where they all are, they can talk to somebody
00:07:41.520 | and they're very reliable, they're amazingly reliable.
00:07:44.400 | We could use the same system,
00:07:46.560 | the same scale for air traffic control.
00:07:49.000 | So instead of me as a pilot talking to a human being
00:07:52.000 | in the middle of the conversation,
00:07:54.200 | receiving a new frequency, like how ancient is that?
00:07:57.600 | We could digitize this stuff
00:07:59.200 | and digitally transmit the right flight coordinates
00:08:03.200 | and that solution will automatically scale
00:08:06.120 | to 10,000 vehicles.
00:08:07.480 | (upbeat music)
00:08:10.080 | (upbeat music)
00:08:12.680 | (upbeat music)
00:08:15.280 | (upbeat music)
00:08:17.880 | (upbeat music)
00:08:20.480 | (upbeat music)
00:08:23.080 | [BLANK_AUDIO]