0:00:00.735,0:00:03.024
How many of you are creatives,

0:00:03.048,0:00:06.672
designers, engineers,[br]entrepreneurs, artists,

0:00:06.696,0:00:09.083
or maybe you just have[br]a really big imagination?

0:00:09.107,0:00:10.955
Show of hands? (Cheers)

0:00:10.979,0:00:12.160
That's most of you.

0:00:13.334,0:00:15.628
I have some news for us creatives.

0:00:16.714,0:00:19.287
Over the course of the next 20 years,

0:00:21.471,0:00:24.444
more will change around[br]the way we do our work

0:00:25.382,0:00:27.539
than has happened in the last 2,000.

0:00:28.511,0:00:33.139
In fact, I think we're at the dawn[br]of a new age in human history.

0:00:33.645,0:00:38.406
Now, there have been four major historical[br]eras defined by the way we work.

0:00:39.404,0:00:42.679
The Hunter-Gatherer Age[br]lasted several million years.

0:00:43.163,0:00:46.739
And then the Agricultural Age[br]lasted several thousand years.

0:00:47.195,0:00:50.685
The Industrial Age lasted[br]a couple of centuries.

0:00:50.709,0:00:54.996
And now the Information Age[br]has lasted just a few decades.

0:00:55.020,0:01:00.240
And now today, we're on the cusp[br]of our next great era as a species.

0:01:01.296,0:01:03.976
Welcome to the Augmented Age.

0:01:04.000,0:01:07.693
In this new era, your natural human[br]capabilities are going to be augmented

0:01:07.717,0:01:10.785
by computational systems[br]that help you think,

0:01:10.809,0:01:12.995
robotic systems that help you make,

0:01:13.019,0:01:14.667
and a digital nervous system

0:01:14.691,0:01:18.381
that connects you to the world[br]far beyond your natural senses.

0:01:19.437,0:01:21.379
Let's start with cognitive augmentation.

0:01:21.403,0:01:23.603
How many of you are augmented cyborgs?

0:01:24.133,0:01:26.783
(Laughter)

0:01:26.807,0:01:29.628
I would actually argue[br]that we're already augmented.

0:01:30.288,0:01:31.792
Imagine you're at a party,

0:01:31.816,0:01:35.336
and somebody asks you a question[br]that you don't know the answer to.

0:01:35.360,0:01:39.120
If you have one of these,[br]in a few seconds, you can know the answer.

0:01:39.869,0:01:42.168
But this is just a primitive beginning.

0:01:42.863,0:01:46.194
Even Siri is just a passive tool.

0:01:46.660,0:01:50.041
In fact, for the last[br]three-and-a-half million years,

0:01:50.065,0:01:53.174
the tools that we've had[br]have been completely passive.

0:01:54.203,0:01:57.858
They do exactly what we tell them[br]and nothing more.

0:01:57.882,0:02:00.983
Our very first tool only cut[br]where we struck it.

0:02:01.822,0:02:04.862
The chisel only carves[br]where the artist points it.

0:02:05.343,0:02:10.984
And even our most advanced tools[br]do nothing without our explicit direction.

0:02:11.008,0:02:14.189
In fact, to date, and this[br]is something that frustrates me,

0:02:14.213,0:02:15.661
we've always been limited

0:02:15.685,0:02:19.186
by this need to manually[br]push our wills into our tools --

0:02:19.210,0:02:21.507
like, manual,[br]literally using our hands,

0:02:21.531,0:02:22.959
even with computers.

0:02:24.072,0:02:26.535
But I'm more like Scotty in "Star Trek."

0:02:26.559,0:02:28.409
(Laughter)

0:02:28.433,0:02:30.579
I want to have a conversation[br]with a computer.

0:02:30.603,0:02:33.573
I want to say, "Computer,[br]let's design a car,"

0:02:33.597,0:02:35.136
and the computer shows me a car.

0:02:35.160,0:02:37.768
And I say, "No, more fast-looking,[br]and less German,"

0:02:37.792,0:02:39.955
and bang, the computer shows me an option.

0:02:39.979,0:02:41.844
(Laughter)

0:02:42.208,0:02:44.514
That conversation might be[br]a little ways off,

0:02:44.538,0:02:47.203
probably less than many of us think,

0:02:47.227,0:02:48.990
but right now,

0:02:49.014,0:02:50.165
we're working on it.

0:02:50.189,0:02:54.222
Tools are making this leap[br]from being passive to being generative.

0:02:54.831,0:02:58.139
Generative design tools[br]use a computer and algorithms

0:02:58.163,0:03:00.771
to synthesize geometry

0:03:00.795,0:03:03.549
to come up with new designs[br]all by themselves.

0:03:03.996,0:03:06.744
All it needs are your goals[br]and your constraints.

0:03:06.768,0:03:08.176
I'll give you an example.

0:03:08.200,0:03:10.988
In the case of this aerial drone chassis,

0:03:11.012,0:03:13.638
all you would need to do[br]is tell it something like,

0:03:13.662,0:03:14.935
it has four propellers,

0:03:14.959,0:03:17.090
you want it to be[br]as lightweight as possible,

0:03:17.114,0:03:19.384
and you need it to be[br]aerodynamically efficient.

0:03:19.408,0:03:24.322
Then what the computer does[br]is it explores the entire solution space:

0:03:24.346,0:03:28.273
every single possibility that solves[br]and meets your criteria --

0:03:28.297,0:03:29.739
millions of them.

0:03:29.763,0:03:31.738
It takes big computers to do this.

0:03:31.762,0:03:33.717
But it comes back to us with designs

0:03:33.741,0:03:36.884
that we, by ourselves,[br]never could've imagined.

0:03:37.326,0:03:40.238
And the computer's coming up[br]with this stuff all by itself --

0:03:40.262,0:03:41.940
no one ever drew anything,

0:03:41.964,0:03:44.050
and it started completely from scratch.

0:03:45.038,0:03:47.425
And by the way, it's no accident

0:03:47.449,0:03:50.930
that the drone body looks just like[br]the pelvis of a flying squirrel.

0:03:51.287,0:03:53.294
(Laughter)

0:03:54.040,0:03:56.342
It's because the algorithms[br]are designed to work

0:03:56.366,0:03:58.003
the same way evolution does.

0:03:58.715,0:04:01.375
What's exciting is we're starting[br]to see this technology

0:04:01.399,0:04:02.558
out in the real world.

0:04:02.582,0:04:05.034
We've been working with Airbus[br]for a couple of years

0:04:05.058,0:04:06.967
on this concept plane for the future.

0:04:06.991,0:04:09.061
It's a ways out still.

0:04:09.085,0:04:12.865
But just recently we used[br]a generative-design AI

0:04:12.889,0:04:14.696
to come up with this.

0:04:15.609,0:04:20.762
This is a 3D-printed cabin partition[br]that's been designed by a computer.

0:04:20.786,0:04:23.610
It's stronger than the original[br]yet half the weight,

0:04:23.634,0:04:26.780
and it will be flying[br]in the Airbus A320 later this year.

0:04:27.405,0:04:28.964
So computers can now generate;

0:04:28.988,0:04:33.583
they can come up with their own solutions[br]to our well-defined problems.

0:04:34.677,0:04:35.987
But they're not intuitive.

0:04:36.011,0:04:39.097
They still have to start from scratch[br]every single time,

0:04:39.121,0:04:41.686
and that's because they never learn.

0:04:42.368,0:04:44.134
Unlike Maggie.

0:04:44.158,0:04:45.739
(Laughter)

0:04:45.763,0:04:49.060
Maggie's actually smarter[br]than our most advanced design tools.

0:04:49.467,0:04:50.907
What do I mean by that?

0:04:50.931,0:04:52.521
If her owner picks up that leash,

0:04:52.545,0:04:54.613
Maggie knows with a fair[br]degree of certainty

0:04:54.637,0:04:56.041
it's time to go for a walk.

0:04:56.065,0:04:57.250
And how did she learn?

0:04:57.274,0:05:00.598
Well, every time the owner picked up[br]the leash, they went for a walk.

0:05:00.622,0:05:02.500
And Maggie did three things:

0:05:02.524,0:05:04.393
she had to pay attention,

0:05:04.417,0:05:06.499
she had to remember what happened

0:05:06.523,0:05:10.540
and she had to retain and create[br]a pattern in her mind.

0:05:11.429,0:05:13.524
Interestingly, that's exactly what

0:05:13.548,0:05:16.071
computer scientists[br]have been trying to get AIs to do

0:05:16.095,0:05:17.954
for the last 60 or so years.

0:05:18.683,0:05:20.032
Back in 1952,

0:05:20.056,0:05:23.857
they built this computer[br]that could play Tic-Tac-Toe.

0:05:25.081,0:05:26.241
Big deal.

0:05:27.029,0:05:30.029
Then 45 years later, in 1997,

0:05:30.053,0:05:32.525
Deep Blue beats Kasparov at chess.

0:05:34.046,0:05:39.014
2011, Watson beats these two[br]humans at Jeopardy,

0:05:39.038,0:05:41.966
which is much harder for a computer[br]to play than chess is.

0:05:41.990,0:05:45.802
In fact, rather than working[br]from predefined recipes,

0:05:45.826,0:05:49.149
Watson had to use reasoning[br]to overcome his human opponents.

0:05:50.393,0:05:52.832
And then a couple of weeks ago,

0:05:52.856,0:05:57.118
DeepMind's AlphaGo beats[br]the world's best human at Go,

0:05:57.142,0:05:59.354
which is the most difficult[br]game that we have.

0:05:59.378,0:06:02.274
In fact, in Go, there are more[br]possible moves

0:06:02.298,0:06:04.322
than there are atoms in the universe.

0:06:06.210,0:06:08.036
So in order to win,

0:06:08.060,0:06:10.678
what AlphaGo had to do[br]was develop intuition.

0:06:11.098,0:06:15.208
And in fact, at some points,[br]AlphaGo's programmers didn't understand

0:06:15.232,0:06:17.518
why AlphaGo was doing what it was doing.

0:06:19.451,0:06:21.111
And things are moving really fast.

0:06:21.135,0:06:24.362
I mean, consider --[br]in the space of a human lifetime,

0:06:24.386,0:06:26.619
computers have gone from a child's game

0:06:27.920,0:06:30.968
to what's recognized as the pinnacle[br]of strategic thought.

0:06:31.999,0:06:34.416
What's basically happening

0:06:34.440,0:06:37.750
is computers are going[br]from being like Spock

0:06:37.774,0:06:39.723
to being a lot more like Kirk.

0:06:39.747,0:06:43.365
(Laughter)

0:06:43.389,0:06:46.813
Right? From pure logic to intuition.

0:06:48.184,0:06:49.927
Would you cross this bridge?

0:06:50.609,0:06:52.932
Most of you are saying, "Oh, hell no!"

0:06:52.956,0:06:54.264
(Laughter)

0:06:54.288,0:06:56.945
And you arrived at that decision[br]in a split second.

0:06:56.969,0:06:59.397
You just sort of knew[br]that bridge was unsafe.

0:06:59.421,0:07:01.410
And that's exactly the kind of intuition

0:07:01.434,0:07:05.002
that our deep-learning systems[br]are starting to develop right now.

0:07:05.722,0:07:07.429
Very soon, you'll literally be able

0:07:07.453,0:07:09.659
to show something you've made,[br]you've designed,

0:07:09.683,0:07:10.836
to a computer,

0:07:10.860,0:07:12.349
and it will look at it and say,

0:07:12.373,0:07:15.196
"Sorry, homie, that'll never work.[br]You have to try again."

0:07:15.854,0:07:18.924
Or you could ask it if people[br]are going to like your next song,

0:07:19.773,0:07:21.836
or your next flavor of ice cream.

0:07:23.549,0:07:26.128
Or, much more importantly,

0:07:26.152,0:07:28.516
you could work with a computer[br]to solve a problem

0:07:28.540,0:07:30.177
that we've never faced before.

0:07:30.201,0:07:31.602
For instance, climate change.

0:07:31.626,0:07:33.646
We're not doing a very[br]good job on our own,

0:07:33.670,0:07:35.915
we could certainly use[br]all the help we can get.

0:07:35.939,0:07:37.397
That's what I'm talking about,

0:07:37.421,0:07:39.976
technology amplifying[br]our cognitive abilities

0:07:40.000,0:07:43.552
so we can imagine and design things[br]that were simply out of our reach

0:07:43.576,0:07:46.135
as plain old un-augmented humans.

0:07:47.984,0:07:50.925
So what about making[br]all of this crazy new stuff

0:07:50.949,0:07:53.390
that we're going to invent and design?

0:07:53.952,0:07:58.045
I think the era of human augmentation[br]is as much about the physical world

0:07:58.069,0:08:01.134
as it is about the virtual,[br]intellectual realm.

0:08:01.833,0:08:03.754
How will technology augment us?

0:08:04.261,0:08:06.734
In the physical world, robotic systems.

0:08:07.620,0:08:09.356
OK, there's certainly a fear

0:08:09.380,0:08:11.868
that robots are going to take[br]jobs away from humans,

0:08:11.892,0:08:13.722
and that is true in certain sectors.

0:08:14.174,0:08:17.052
But I'm much more interested in this idea

0:08:17.076,0:08:22.086
that humans and robots working together[br]are going to augment each other,

0:08:22.110,0:08:24.168
and start to inhabit a new space.

0:08:24.192,0:08:26.554
This is our applied research lab[br]in San Francisco,

0:08:26.578,0:08:29.720
where one of our areas of focus[br]is advanced robotics,

0:08:29.744,0:08:32.255
specifically, human-robot collaboration.

0:08:33.034,0:08:35.793
And this is Bishop, one of our robots.

0:08:35.817,0:08:37.606
As an experiment, we set it up

0:08:37.630,0:08:41.090
to help a person working in construction[br]doing repetitive tasks --

0:08:41.984,0:08:46.178
tasks like cutting out holes for outlets[br]or light switches in drywall.

0:08:46.202,0:08:48.668
(Laughter)

0:08:49.877,0:08:52.988
So, Bishop's human partner[br]can tell what to do in plain English

0:08:53.012,0:08:54.317
and with simple gestures,

0:08:54.341,0:08:55.788
kind of like talking to a dog,

0:08:55.812,0:08:57.955
and then Bishop executes[br]on those instructions

0:08:57.979,0:08:59.871
with perfect precision.

0:08:59.895,0:09:02.884
We're using the human[br]for what the human is good at:

0:09:02.908,0:09:05.241
awareness, perception and decision making.

0:09:05.265,0:09:07.505
And we're using the robot[br]for what it's good at:

0:09:07.529,0:09:09.277
precision and repetitiveness.

0:09:10.252,0:09:12.619
Here's another cool project[br]that Bishop worked on.

0:09:12.643,0:09:15.718
The goal of this project,[br]which we called the HIVE,

0:09:15.742,0:09:19.593
was to prototype the experience[br]of humans, computers and robots

0:09:19.617,0:09:22.837
all working together to solve[br]a highly complex design problem.

0:09:23.793,0:09:25.244
The humans acted as labor.

0:09:25.268,0:09:28.741
They cruised around the construction site,[br]they manipulated the bamboo --

0:09:28.765,0:09:31.521
which, by the way,[br]because it's a non-isomorphic material,

0:09:31.545,0:09:33.419
is super hard for robots to deal with.

0:09:33.443,0:09:35.465
But then the robots[br]did this fiber winding,

0:09:35.489,0:09:37.940
which was almost impossible[br]for a human to do.

0:09:37.964,0:09:41.585
And then we had an AI[br]that was controlling everything.

0:09:41.609,0:09:44.899
It was telling the humans what to do,[br]telling the robots what to do

0:09:44.923,0:09:47.838
and keeping track of thousands[br]of individual components.

0:09:47.862,0:09:49.042
What's interesting is,

0:09:49.066,0:09:52.207
building this pavilion[br]was simply not possible

0:09:52.231,0:09:56.755
without human, robot and AI[br]augmenting each other.

0:09:57.890,0:10:01.210
OK, I'll share one more project.[br]This one's a little bit crazy.

0:10:01.234,0:10:05.702
We're working with Amsterdam-based artist[br]Joris Laarman and his team at MX3D

0:10:05.726,0:10:08.604
to generatively design[br]and robotically print

0:10:08.628,0:10:11.623
the world's first autonomously[br]manufactured bridge.

0:10:12.315,0:10:16.000
So, Joris and an AI are designing[br]this thing right now, as we speak,

0:10:16.024,0:10:17.196
in Amsterdam.

0:10:17.220,0:10:19.541
And when they're done,[br]we're going to hit "Go,"

0:10:19.565,0:10:22.876
and robots will start 3D printing[br]in stainless steel,

0:10:22.900,0:10:26.183
and then they're going to keep printing,[br]without human intervention,

0:10:26.207,0:10:27.765
until the bridge is finished.

0:10:29.099,0:10:32.027
So, as computers are going[br]to augment our ability

0:10:32.051,0:10:34.201
to imagine and design new stuff,

0:10:34.225,0:10:37.120
robotic systems are going to help us[br]build and make things

0:10:37.144,0:10:39.228
that we've never been able to make before.

0:10:40.347,0:10:44.507
But what about our ability[br]to sense and control these things?

0:10:44.531,0:10:48.562
What about a nervous system[br]for the things that we make?

0:10:48.586,0:10:51.098
Our nervous system,[br]the human nervous system,

0:10:51.122,0:10:53.433
tells us everything[br]that's going on around us.

0:10:54.186,0:10:57.870
But the nervous system of the things[br]we make is rudimentary at best.

0:10:57.894,0:11:01.457
For instance, a car doesn't tell[br]the city's public works department

0:11:01.481,0:11:04.611
that it just hit a pothole at the corner[br]of Broadway and Morrison.

0:11:04.635,0:11:06.667
A building doesn't tell its designers

0:11:06.691,0:11:09.375
whether or not the people inside[br]like being there,

0:11:09.399,0:11:12.409
and the toy manufacturer doesn't know

0:11:12.433,0:11:14.440
if a toy is actually being played with --

0:11:14.464,0:11:17.003
how and where and whether[br]or not it's any fun.

0:11:17.620,0:11:21.434
Look, I'm sure that the designers[br]imagined this lifestyle for Barbie

0:11:21.458,0:11:22.682
when they designed her.

0:11:22.706,0:11:24.153
(Laughter)

0:11:24.177,0:11:27.083
But what if it turns out that Barbie's[br]actually really lonely?

0:11:27.107,0:11:30.254
(Laughter)

0:11:31.266,0:11:32.554
If the designers had known

0:11:32.578,0:11:34.685
what was really happening[br]in the real world

0:11:34.709,0:11:37.292
with their designs -- the road,[br]the building, Barbie --

0:11:37.316,0:11:40.010
they could've used that knowledge[br]to create an experience

0:11:40.034,0:11:41.434
that was better for the user.

0:11:41.458,0:11:43.249
What's missing is a nervous system

0:11:43.273,0:11:46.982
connecting us to all of the things[br]that we design, make and use.

0:11:47.915,0:11:51.470
What if all of you had that kind[br]of information flowing to you

0:11:51.494,0:11:53.677
from the things you create[br]in the real world?

0:11:55.432,0:11:56.883
With all of the stuff we make,

0:11:56.907,0:11:59.342
we spend a tremendous amount[br]of money and energy --

0:11:59.366,0:12:01.742
in fact, last year,[br]about two trillion dollars --

0:12:01.766,0:12:04.620
convincing people to buy[br]the things we've made.

0:12:04.644,0:12:08.032
But if you had this connection[br]to the things that you design and create

0:12:08.056,0:12:09.783
after they're out in the real world,

0:12:09.807,0:12:13.421
after they've been sold[br]or launched or whatever,

0:12:13.445,0:12:15.065
we could actually change that,

0:12:15.089,0:12:18.136
and go from making people want our stuff,

0:12:18.160,0:12:21.594
to just making stuff that people[br]want in the first place.

0:12:21.618,0:12:24.405
The good news is, we're working[br]on digital nervous systems

0:12:24.429,0:12:27.230
that connect us to the things we design.

0:12:28.365,0:12:29.992
We're working on one project

0:12:30.016,0:12:33.728
with a couple of guys down in Los Angeles[br]called the Bandito Brothers

0:12:33.752,0:12:35.159
and their team.

0:12:35.183,0:12:38.616
And one of the things these guys do[br]is build insane cars

0:12:38.640,0:12:41.513
that do absolutely insane things.

0:12:42.905,0:12:44.355
These guys are crazy --

0:12:44.379,0:12:45.415
(Laughter)

0:12:45.439,0:12:46.842
in the best way.

0:12:48.993,0:12:50.756
And what we're doing with them

0:12:50.780,0:12:53.220
is taking a traditional race-car chassis

0:12:53.244,0:12:54.829
and giving it a nervous system.

0:12:54.853,0:12:57.911
So we instrumented it[br]with dozens of sensors,

0:12:57.935,0:13:00.570
put a world-class driver behind the wheel,

0:13:00.594,0:13:03.951
took it out to the desert[br]and drove the hell out of it for a week.

0:13:03.975,0:13:06.466
And the car's nervous system[br]captured everything

0:13:06.490,0:13:07.972
that was happening to the car.

0:13:07.996,0:13:10.617
We captured four billion data points;

0:13:10.641,0:13:12.951
all of the forces[br]that it was subjected to.

0:13:12.975,0:13:14.634
And then we did something crazy.

0:13:15.268,0:13:16.768
We took all of that data,

0:13:16.792,0:13:20.528
and plugged it into a generative-design AI[br]we call "Dreamcatcher."

0:13:21.270,0:13:25.234
So what do get when you give[br]a design tool a nervous system,

0:13:25.258,0:13:28.140
and you ask it to build you[br]the ultimate car chassis?

0:13:28.723,0:13:30.696
You get this.

0:13:32.293,0:13:36.006
This is something that a human[br]could never have designed.

0:13:36.707,0:13:38.595
Except a human did design this,

0:13:38.619,0:13:42.928
but it was a human that was augmented[br]by a generative-design AI,

0:13:42.952,0:13:44.183
a digital nervous system

0:13:44.207,0:13:47.212
and robots that can actually[br]fabricate something like this.

0:13:47.680,0:13:51.275
So if this is the future,[br]the Augmented Age,

0:13:51.299,0:13:55.560
and we're going to be augmented[br]cognitively, physically and perceptually,

0:13:55.584,0:13:56.992
what will that look like?

0:13:57.576,0:14:00.897
What is this wonderland going to be like?

0:14:00.921,0:14:02.630
I think we're going to see a world

0:14:02.654,0:14:05.722
where we're moving[br]from things that are fabricated

0:14:05.746,0:14:07.191
to things that are farmed.

0:14:08.159,0:14:11.612
Where we're moving from things[br]that are constructed

0:14:11.636,0:14:13.340
to that which is grown.

0:14:14.134,0:14:16.322
We're going to move from being isolated

0:14:16.346,0:14:17.956
to being connected.

0:14:18.634,0:14:21.045
And we'll move away from extraction

0:14:21.069,0:14:22.942
to embrace aggregation.

0:14:23.967,0:14:27.734
I also think we'll shift[br]from craving obedience from our things

0:14:27.758,0:14:29.399
to valuing autonomy.

0:14:30.510,0:14:32.415
Thanks to our augmented capabilities,

0:14:32.439,0:14:34.816
our world is going to change dramatically.

0:14:35.576,0:14:38.822
We're going to have a world[br]with more variety, more connectedness,

0:14:38.846,0:14:41.133
more dynamism, more complexity,

0:14:41.157,0:14:43.475
more adaptability and, of course,

0:14:43.499,0:14:44.716
more beauty.

0:14:45.231,0:14:46.795
The shape of things to come

0:14:46.819,0:14:49.109
will be unlike anything[br]we've ever seen before.

0:14:49.133,0:14:50.292
Why?

0:14:50.316,0:14:54.071
Because what will be shaping those things[br]is this new partnership

0:14:54.095,0:14:57.765
between technology, nature and humanity.

0:14:59.279,0:15:03.083
That, to me, is a future[br]well worth looking forward to.

0:15:03.107,0:15:04.378
Thank you all so much.

0:15:04.402,0:15:10.071
(Applause)