Home
登錄
註冊
學習內容
Loading...
聽力練習
聽力練習
/
Video
/
TED Talk
/
Waymo’s Case for a Driverless Future | Tekedra Mawakana, Sal Khan | TED
Waymo’s Case for a Driverless Future | Tekedra Mawakana, Sal Khan | TED
選擇學習模式:
查看字幕
選詞
重寫單字
Highlight:
3000 Oxford Words
4000 IELTS Words
5000 Oxford Words
3000 Common Words
1000 TOEIC Words
5000 TOEFL Words
字幕 (416)
0:03
Sal Khan: Well great to have you here, Tekedra.
0:06
Tekedra Mawakana: It's great to be here.
0:08
Thank you so much.
0:10
SK: Well I thought before we start, because there was,
0:13
it seemed like, about 40 percent of the audience
0:15
who had never had that experience,
0:17
I thought I would share some of our family videos
0:20
of our first Waymo ride with my two sons.
0:23
I also have a daughter, but we kind of split up into two cars.
0:26
But if we could play the video.
0:28
(Video) Yeah, yeah. Start the ride.
0:30
Yeah yeah.
0:32
Son 1: Alright, here we go.
0:33
Waymo: Hello from Waymo.
0:35
Son 2: "Hello from ..." oh, my ...
0:37
(Laughter)
0:38
Son 1: This is cool.
0:40
Son 2: Oh, my gosh.
0:42
Waymo: If you're traveling with little ones or expecting,
0:45
please see the --
0:46
SK: Are you expecting?
0:49
We're at the beach,
0:51
being driven by a ghost.
0:54
Waymo: Almost there.
0:55
Don't forget your phone, keys or wallet.
0:58
SK: And sons, don't forget your sons.
1:02
(Laughter)
1:04
TM: That's great.
1:05
SK: So Tekedra, what’s next?
1:07
I mean, every time I look at the news,
1:09
I see a new city, I see new models.
1:11
What's new?
1:12
TM: So we are currently operating in 11 cities.
1:16
In the US,
1:19
we're providing 500,000 trips per week.
1:23
And we have a goal, by the end of the year,
1:25
to be at a million trips per week.
1:28
And just to sort of help people wrap their minds around what that means,
1:33
or how to think about it,
1:35
our fleet of vehicles is about 3,000 cars ...
1:39
The single Waymo driver
1:43
drives over four million miles per week.
1:47
And so the way to think about that
1:49
is a human drives around 700,000 miles in a lifetime --
1:54
a human who lives sort of the full lifetime --
1:58
and so this is like six lifetimes of human driving per week.
2:03
And so we're, you know, really excited.
2:06
We're also in four airports right now.
2:08
You can fly in to San Antonio or Phoenix,
2:11
San Francisco or San Jose
2:13
and jump in a Waymo, use the Waymo app.
2:16
It'll take you from point A to point B.
2:19
And we're at this point
2:21
where the technology has moved from, “Will it work?”
2:26
and “Can we produce the kind of safety outcomes that we care about?”
2:31
to scaling.
2:32
And so we're really excited.
2:34
We are also in London, in Tokyo, [and]
2:37
we've moved from manual driving to fully autonomous driving.
2:41
We still have people behind the wheel,
2:43
so we haven’t launched a service yet,
2:45
but those will be our first two international markets.
2:48
SK: What’s the rate-limiting factor:
2:50
is it some notion of safety, or is it regulation?
2:53
What is it?
2:55
TM: It really depends.
2:56
In certain markets, it’s absolutely regulation.
2:59
I think when you think about the evolution of this technology as a whole,
3:03
it's been making sure that we could have the safety outcomes that we want.
3:07
And all of that is the technology.
3:09
Can the driver produce these kinds of outcomes?
3:13
We’ve driven over 200 million miles in our time
3:17
and based on 170 million miles of data ...
3:21
we are now seeing [a] 13x reduction
3:25
in serious-injury-causing crashes over a human,
3:28
and the same amount of reduction
3:31
as it relates to injuries with pedestrians.
3:34
So we can now say that we are just over 10 times safer than a human
3:38
at 170 million miles.
3:41
That’s the kind of superhuman performance
3:45
that we were seeking
3:46
in order to then be able to scale this technology.
3:49
Now we have markets that are closed, markets that are open,
3:52
and we have to do all of the work with the cities,
3:54
with the regulators, first responders, everyone,
3:57
in order to advance our technology.
3:59
So we're laser-focused on doing that.
4:01
And the reality is, you know, [for] citizens, consumers, riders --
4:06
those who’ve been cut out of the mobility systems and cities --
4:09
they're demanding and requesting and advocating for us to come.
4:13
And so that’s a really exciting inflection point.
4:17
SK: It’s worth triply underlining what you just said:
4:19
over 10 times safer, at least based on the current data.
4:22
We were talking earlier,
4:23
y'all have already driven the equivalent of 300 lifetimes --
4:29
TM: Two hundred and forty lifetimes.
4:31
SK: And Waymo has not caused a death,
4:36
knock on wood.
4:38
And there’s an argument:
4:39
roughly 40,000 people in America, every year,
4:42
who die of a car accident.
4:44
I think in the world it’s on the order of a million every year.
4:47
So in theory, if this became everywhere,
4:51
you could be saving 900,000 lives globally.
4:53
You could be saving 30, 35,000 to 36,000 lives
4:57
just in the United States.
4:58
Why isn't there more --
5:00
I know you all are growing maybe as fast,
5:02
but why not roll out as fast as possible?
5:04
What's stopping that from happening?
5:06
TM: I think we are now focused on rolling out safely
5:11
as quickly as possible.
5:13
There is an adoption curve that has to happen, though, right?
5:16
I think what you’re saying is what undergirds our entire company,
5:20
which is: safety is urgent, saving lives.
5:24
If this technology can perform at this level,
5:27
then we have a responsibility to figure out how to do it.
5:30
Doesn't mean, though, that we're not being met with opposition.
5:33
And so we also have the burden of making those policymakers,
5:38
who are not welcoming us with open arms
5:41
... to demonstrate our safety record.
5:43
So we have a safety hub.
5:45
We make all of this data transparent, [and]
5:47
we hand it to researchers so they can make their arguments.
5:50
And so it's been really fun lately
5:52
to watch different sectors of the economy
5:55
start to talk about how powerful this technology is.
5:59
You know, we have Dr. John Slotkin, a neurosurgeon,
6:02
who has just calculated that $900 billion to $1.25 trillion
6:08
could be saved
6:09
if our technology was rolled out,
6:12
if every car on the road was a Waymo.
6:15
But this is the problem:
6:17
people are really comfortable
6:18
with the number of people who die on the roads.
6:21
40,000 people dead,
6:24
1.2 million globally
6:26
and avoidable in many cases,
6:29
isn't something upon which we share collective outrage
6:34
or even moderate discomfort.
6:37
And so we have to introduce sort of that reality
6:41
in order for the technology to be viewed
6:44
as solving a problem that society currently faces,
6:48
because we've just accepted it.
6:50
Why?
6:51
Because we don't experience these as 737s falling out of the sky every day.
6:56
But that's what it would be
6:58
if we actually experienced it as mass casualties.
7:01
We experience it, you know, [as] this person’s mom,
7:04
my uncle.
7:06
And so because of that,
7:07
we have to remind people
7:09
that the status quo is totally unacceptable.
7:12
And it's even more unacceptable
7:14
when you have technology that could help address it.
7:16
SK: Yes, definitely.
7:17
(Applause)
7:19
I think it was a couple of months ago -- you would know it much better --
7:22
I saw a headline, it was a very clickbaity headline:
7:24
“Waymo hits an eight-year-old outside of a school.”
7:28
And I was like, oh my God.
7:30
And I click on it.
7:31
And then they described what happened,
7:33
and it was outside of a school and there was a car parked.
7:38
And if I remember correctly, there was a kid,
7:40
he just darted into the street.
7:42
And the Waymo,
7:43
everyone who was observing that assumed the kid was going to die
7:48
because he just darted into a street where traffic was coming.
7:51
And the Waymo, I don't know what the reflexes are,
7:53
you probably know the numbers,
7:55
it just immediately came to a stop,
7:57
and it hit him at like, four miles an hour or something.
7:59
Like he literally just fell on his knees,
8:01
brushed it off and looked at the Waymo kind of weird.
8:04
Why isn't --
8:06
You know, it's convenient.
8:08
I mean, you feel like you're in the future.
8:11
Why is it the safety --
8:12
I mean, it feels like that message,
8:14
Is that something that y'all might want to lead with?
8:17
Because that seems like such a clear value proposition.
8:19
(Laughter)
8:20
TM: Yeah, I mean, you know,
8:22
I think what you're saying is how comfortable are humans
8:27
with this technology being better drivers than humans?
8:31
You know, it goes back to the status quo.
8:34
People don't generally view themselves or their neighbors or their friends
8:37
or everyone they know as unsafe drivers,
8:40
even when they're distracted, drunk, tired and angry.
8:43
We just kind of accept that that's the human condition.
8:46
And so I think, you know, in the case of the child in Santa Monica ...
8:51
our car was traveling 17 miles per hour.
8:56
And we were able to get to six miles per hour
9:00
before making contact.
9:02
Obviously, we want to make no contact,
9:04
but we know that was superhuman performance
9:08
and as so many of the bystanders said after observing it --
9:11
“The child would not have survived” -- is what we were told.
9:15
We weren't obviously there.
9:17
Those moments actually cause entire communities
9:21
to think differently about this technology.
9:24
And so, you know, to your point: it’s clickbaity.
9:27
And on the other hand,
9:29
it actually demonstrates what superhuman performance is.
9:32
SK: Yeah, I couldn't agree more.
9:34
I actually do assume most of the people I know
9:36
are not good drivers.
9:37
I'm the biggest backseat driver.
9:40
The economics of the industry,
9:41
y'all are the first mover.
9:43
Obviously there's a Tesla robotaxi.
9:45
They've taken some different approaches.
9:47
You have many other competitors.
9:50
Amazon's coming into it.
9:51
How are you thinking about the competitive space?
9:53
What do you think is going to happen to the economics of it?
9:56
It's already cost competitive with a typical rideshare drive.
9:59
TM: Yeah, I think, you know, right now for us,
10:02
we're the only company operating at this scale.
10:05
We think it’s really important to just stay laser-focused on that.
10:09
You know the way to think about it,
10:11
because everyone’s calling everything either autonomous or robotaxi,
10:18
is if you do not need a driver’s license to be in the car by yourself,
10:23
which is for many, many people who do not have a driver's license,
10:27
it’s their first time ever in life --
10:29
and so those are very special moments --
10:32
but if you do not need a driver’s license
10:34
to be in a Waymo to get from point A to point B,
10:36
that is fully autonomous,
10:37
[or] what is referred to, in the geeky sense,
10:40
[as] level four and level five.
10:42
If it is anything less than that --
10:44
if you have to be behind the wheel,
10:46
if you have to take over when there's a beep,
10:49
when there’s anything --
10:51
that's level three or level two or level one.
10:55
And it's really important for people to internalize this distinction,
11:00
because what we found back in 2012,
11:02
when we had sort of a supercruise capability,
11:06
the Waymo driver was at that ability,
11:08
[was that] we could go from ramp to ramp, no problem.
11:11
But when we got to surface streets,
11:12
you needed to pay attention.
11:15
And we told employees this is early days:
11:17
“When we go from ramp to ramp, we have a camera in the car” --
11:20
this was just for employees --
11:22
"we need you to pay attention."
11:24
And people immediately unplugged from the driving task,
11:28
and they started shaving and curling their eyelashes,
11:31
and they were like, picking up things from the backseat and plugging things in.
11:35
And so that's when we decided, you know,
11:38
cars have gotten safer over the years, right?
11:41
They've gotten airbags and seat belts
11:43
and all of these things as a result of a lot of deaths.
11:46
But what hasn't gotten safer is the human.
11:49
And so we decided we're patching around the problem.
11:54
Actually, the way to make sure that this multi-ton vehicle
11:58
is not as dangerous is to actually let humans be humans.
12:03
Be distracted, make your phone call,
12:05
do the thing you want to be doing.
12:07
Let the Waymo driver get you from point A to point B.
12:10
And so when you ask me about competition,
12:12
I think there are a host of companies
12:15
that are in the level-two-plus and level three.
12:19
They need an attentive human
12:20
after inviting the human to be inattentive,
12:24
to be the most attentive at the moment
12:27
when being most attentive is least convenient.
12:31
And so to me, that is not --
12:33
that is still patching around the problem.
12:36
Doesn't mean it won't result in safer driving when you are attentive,
12:39
but it isn't actually level four autonomy.
12:42
SK: Yeah, yeah.
12:44
I mean, if you extrapolate all of this,
12:46
how much do you all think about what the world is going to look
12:49
like just because of this?
12:51
Maybe just because of you all and some of the other players,
12:54
10 [to] 15 years, what will it look like?
12:55
You mentioned, if you don't have a driver's license,
12:58
could this be a replacement in some ways for traditional public transportation,
13:02
how are our roads different,
13:03
are parking lots going to turn into parks -- what’s going to happen?
13:06
TM: Yeah, we’ve partnered, from the early days, with transit.
13:09
We think it's a really interesting way
13:11
to take some of the congestion out of the city,
13:14
have sort of the kiss-and-ride from the suburbs in,
13:17
have people take Waymo rides.
13:19
We've gotten really favorable sort of reviews in the reports
13:22
that have been done there.
13:24
We've also partnered with people
13:26
who take public transit in LA.
13:29
You know, we've offered discounts if you do public transit
13:32
to a Waymo or vice versa.
13:33
You know, we want to think about the ways
13:35
that this could become part of that fabric
13:39
because we think it's important
13:40
that this is also very accessible to a lot of people.
13:44
So that's one.
13:45
But when we think about 10 to 15 years from now,
13:48
it's really fun to think about parts of cities that would get recaptured,
13:53
you know, space right now where cars are just sitting all day,
13:56
people are at work, the cars are just sitting.
13:58
Most of those parking lots
14:00
in major cities around the world
14:03
are central.
14:05
And so the idea that that could get recaptured --
14:07
there are city planners who've reached out to us to try to understand,
14:11
you know, do we have to make these investments
14:14
for the next 10 or 15 years,
14:15
or can we start imagining a world where parking structures could move out,
14:19
parks could come back [and]
14:21
community life could sort of come back to the center of a city.
14:24
So that's a vision.
14:25
I mean, who knows, right?
14:27
I think some of this is as the technology rolls out more and more,
14:30
I think people's imagination actually starts to get sparked
14:34
in ways that are very exciting.
14:36
I said to you before, it's like, in some ways people go from, you know,
14:40
is it safe, kind of to your poll, right?
14:43
Is it safe, should I get in?
14:45
OK, I got in.
14:46
It's kind of boring and amazing.
14:48
I really like it.
14:49
Oh, this is kind of not a big deal.
14:51
Now I'm going to be the coolest person I know in my friend group.
14:54
And then all of a sudden, it's like, oh,
14:57
what could I do in this thing if it got really exciting?
15:00
So I get pitched on like, you should make it a yoga studio
15:04
you know, on wheels.
15:05
SK: I'm glad that was what was pitched.
15:07
I thought this was going in a whole other direction.
15:09
TM: This is my point, is everyone kind of starts to imagine,
15:12
what do I want to do instead of commuting?
15:14
So that's the fun part.
15:15
SK: I had a very G-rated thought in a Waymo on the highway on 101,
15:19
and then I remembered.
15:21
But it would have been embarrassing.
15:22
And then I remembered, you do have a camera without sound,
15:25
so I didn't even act on my G-rated thought.
15:27
(Laughter)
15:29
TM: Just to be clear, I do not have a camera.
15:32
SK: No, no, I was just going to do some grooming type of thing.
15:36
Jobs.
15:37
I don't see this as something that's your responsibility in totality,
15:42
but my uncle drives an Uber in New Orleans
15:45
and I believe that one of the largest jobs for men globally
15:49
and not just men -- obviously women also drive vehicles --
15:52
but definitely for men is driving some form of a vehicle,
15:54
taxis, rideshares, long-haul trucking, etc.
15:58
What happens?
16:00
TM: Yeah, I think, you know, I'll answer the question in a few ways.
16:04
My uncle was actually a truck driver too,
16:06
and that job was really important to him.
16:09
And for better or for worse,
16:12
it was also a really hard job.
16:14
And so I think there’s an opportunity,
16:18
and it’s going to take a really thoughtful approach, right?
16:22
Because, I mean, I'm sitting here talking about building
16:25
the world's most trusted driver.
16:28
And so it's a really tough place for me to sit in
16:31
and [to] facilitate the dialogue around it.
16:35
However, over time,
16:36
what I’m starting to see just from us is [that] there are new jobs being created.
16:42
So we have running our fleet,
16:45
we have fleet technicians who we need,
16:48
fleet operators as well as smart depot professionals.
16:54
All of that didn't exist ten years ago.
16:56
These are like AV-focused jobs.
16:59
One.
17:00
Two, jobs that exist today, everyday drivers,
17:03
we have to hire them too,
17:05
in order to validate our software.
17:07
We have to accumulate miles in order to validate our software.
17:10
So we hire drivers constantly,
17:13
or our partners hire drivers in different parts of the world.
17:17
We also need people who can, you know, take care of our depots,
17:20
build our depots, EV charging infrastructure.
17:23
And so what we’ve started to see,
17:25
as we have these partnerships of fleet operators,
17:28
is they are driving a lot of job growth as we scale.
17:33
We’re just too early to be able to really quantify it.
17:37
But in addition to that, earlier this year,
17:40
we did a partnership with TechForce Foundation
17:43
where we’re helping
17:45
people who are in school to be mechanics,
17:47
to actually become mechanics of autonomous vehicles,
17:51
because we think that transition doesn't have to pass anyone by.
17:55
We just need to be mindful about it.
17:57
So that's one thing we're doing.
17:58
And then in addition, we are doing apprenticeship programs.
18:01
We’re doing one in LA,
18:04
and we’re doing curriculum development with [a] community college:
18:07
Bronx Community College, City University of New York,
18:10
because again, there's an opportunity for people
18:13
to learn the skills that are needed.
18:15
And then I think while there's a lot of focus on sort of the jobs,
18:19
there are also small businesses in these places where we’re launching
18:24
where because Waymo has launched,
18:26
like in San Francisco, our first year,
18:29
we actually drove $40 million of economic development
18:32
for local businesses from out-of-towners
18:35
because we were the number-one destination for tourists to try out a Waymo
18:39
and then to go to a coffee shop and go to a restaurant.
18:42
And so I think this is a long conversation.
18:45
I appreciate you for saying, you know, we're one part of it.
18:49
We take this seriously.
18:51
We think it's important for us to be focused on it.
18:53
And we're also really excited when we get to go
18:56
and talk to folks as they move from logistics
18:59
into now AV-adjacent roles
19:01
with their current skill set.
19:03
So it doesn't all require upskilling.
19:06
SK: Well, we could talk for hours about this, but thank you so much.
19:09
TM: Thank you.
19:11
(Applause)
