0:00:00.760,0:00:04.600
When I was a kid,[br]I was the quintessential nerd.

0:00:05.320,0:00:07.496
I think some of you were, too.

0:00:07.520,0:00:08.736
(Laughter)

0:00:08.760,0:00:11.976
And you, sir, who laughed the loudest,[br]you probably still are.

0:00:12.000,0:00:14.256
(Laughter)

0:00:14.280,0:00:17.776
I grew up in a small town[br]in the dusty plains of north Texas,

0:00:17.800,0:00:21.136
the son of a sheriff[br]who was the son of a pastor.

0:00:21.160,0:00:23.080
Getting into trouble was not an option.

0:00:24.040,0:00:27.296
And so I started reading[br]calculus books for fun.

0:00:27.320,0:00:28.856
(Laughter)

0:00:28.880,0:00:30.576
You did, too.

0:00:30.600,0:00:34.336
That led me to building a laser[br]and a computer and model rockets,

0:00:34.360,0:00:37.360
and that led me to making[br]rocket fuel in my bedroom.

0:00:37.960,0:00:41.616
Now, in scientific terms,

0:00:41.640,0:00:44.896
we call this a very bad idea.

0:00:44.920,0:00:46.136
(Laughter)

0:00:46.160,0:00:48.336
Around that same time,

0:00:48.360,0:00:51.576
Stanley Kubrick's "2001: A Space Odyssey"[br]came to the theaters,

0:00:51.600,0:00:53.800
and my life was forever changed.

0:00:54.280,0:00:56.336
I loved everything about that movie,

0:00:56.360,0:00:58.896
especially the HAL 9000.

0:00:58.920,0:01:00.976
Now, HAL was a sentient computer

0:01:01.000,0:01:03.456
designed to guide the Discovery spacecraft

0:01:03.480,0:01:06.016
from the Earth to Jupiter.

0:01:06.040,0:01:08.096
HAL was also a flawed character,

0:01:08.120,0:01:12.400
for in the end he chose[br]to value the mission over human life.

0:01:12.840,0:01:14.936
Now, HAL was a fictional character,

0:01:14.960,0:01:17.616
but nonetheless he speaks to our fears,

0:01:17.640,0:01:19.736
our fears of being subjugated

0:01:19.760,0:01:22.776
by some unfeeling, artificial intelligence

0:01:22.800,0:01:24.760
who is indifferent to our humanity.

0:01:25.880,0:01:28.456
I believe that such fears are unfounded.

0:01:28.480,0:01:31.176
Indeed, we stand at a remarkable time

0:01:31.200,0:01:32.736
in human history,

0:01:32.760,0:01:37.736
where, driven by refusal to accept[br]the limits of our bodies and our minds,

0:01:37.760,0:01:39.456
we are building machines

0:01:39.480,0:01:43.096
of exquisite, beautiful[br]complexity and grace

0:01:43.120,0:01:45.176
that will extend the human experience

0:01:45.200,0:01:46.880
in ways beyond our imagining.

0:01:47.720,0:01:50.296
After a career that led me[br]from the Air Force Academy

0:01:50.320,0:01:52.256
to Space Command to now,

0:01:52.280,0:01:53.976
I became a systems engineer,

0:01:54.000,0:01:56.736
and recently I was drawn[br]into an engineering problem

0:01:56.760,0:01:59.336
associated with NASA's mission to Mars.

0:01:59.360,0:02:01.856
Now, in space flights to the Moon,

0:02:01.880,0:02:05.016
we can rely upon[br]mission control in Houston

0:02:05.040,0:02:07.016
to watch over all aspects of a flight.

0:02:07.040,0:02:10.576
However, Mars is 200 times further away,

0:02:10.600,0:02:13.816
and as a result it takes[br]on average 13 minutes

0:02:13.840,0:02:16.976
for a signal to travel[br]from the Earth to Mars.

0:02:17.000,0:02:20.400
If there's trouble,[br]there's not enough time.

0:02:20.840,0:02:23.336
And so a reasonable engineering solution

0:02:23.360,0:02:25.936
calls for us to put mission control

0:02:25.960,0:02:28.976
inside the walls of the Orion spacecraft.

0:02:29.000,0:02:31.896
Another fascinating idea[br]in the mission profile

0:02:31.920,0:02:34.816
places humanoid robots[br]on the surface of Mars

0:02:34.840,0:02:36.696
before the humans themselves arrive,

0:02:36.720,0:02:38.376
first to build facilities

0:02:38.400,0:02:41.760
and later to serve as collaborative[br]members of the science team.

0:02:43.400,0:02:46.136
Now, as I looked at this[br]from an engineering perspective,

0:02:46.160,0:02:49.336
it became very clear to me[br]that what I needed to architect

0:02:49.360,0:02:51.536
was a smart, collaborative,

0:02:51.560,0:02:53.936
socially intelligent[br]artificial intelligence.

0:02:53.960,0:02:58.256
In other words, I needed to build[br]something very much like a HAL

0:02:58.280,0:03:00.696
but without the homicidal tendencies.

0:03:00.720,0:03:02.080
(Laughter)

0:03:02.920,0:03:04.736
Let's pause for a moment.

0:03:04.760,0:03:08.656
Is it really possible to build[br]an artificial intelligence like that?

0:03:08.680,0:03:10.136
Actually, it is.

0:03:10.160,0:03:11.416
In many ways,

0:03:11.440,0:03:13.416
this is a hard engineering problem

0:03:13.440,0:03:14.896
with elements of AI,

0:03:14.920,0:03:19.616
not some wet hair ball of an AI problem[br]that needs to be engineered.

0:03:19.640,0:03:22.296
To paraphrase Alan Turing,

0:03:22.320,0:03:24.696
I'm not interested[br]in building a sentient machine.

0:03:24.720,0:03:26.296
I'm not building a HAL.

0:03:26.320,0:03:28.736
All I'm after is a simple brain,

0:03:28.760,0:03:31.880
something that offers[br]the illusion of intelligence.

0:03:33.000,0:03:36.136
The art and the science of computing[br]have come a long way

0:03:36.160,0:03:37.656
since HAL was onscreen,

0:03:37.680,0:03:40.896
and I'd imagine if his inventor[br]Dr. Chandra were here today,

0:03:40.920,0:03:43.256
he'd have a whole lot of questions for us.

0:03:43.280,0:03:45.376
Is it really possible for us

0:03:45.400,0:03:49.416
to take a system of millions[br]upon millions of devices,

0:03:49.440,0:03:50.896
to read in their data streams,

0:03:50.920,0:03:53.176
to predict their failures[br]and act in advance?

0:03:53.200,0:03:54.416
Yes.

0:03:54.440,0:03:57.616
Can we build systems that converse[br]with humans in natural language?

0:03:57.640,0:03:58.856
Yes.

0:03:58.880,0:04:01.856
Can we build systems[br]that recognize objects, identify emotions,

0:04:01.880,0:04:05.256
emote themselves,[br]play games and even read lips?

0:04:05.280,0:04:06.496
Yes.

0:04:06.520,0:04:08.656
Can we build a system that sets goals,

0:04:08.680,0:04:12.296
that carries out plans against those goals[br]and learns along the way?

0:04:12.320,0:04:13.536
Yes.

0:04:13.560,0:04:16.896
Can we build systems[br]that have a theory of mind?

0:04:16.920,0:04:18.416
This we are learning to do.

0:04:18.440,0:04:21.920
Can we build systems that have[br]an ethical and moral foundation?

0:04:22.480,0:04:24.520
This we must learn how to do.

0:04:25.360,0:04:26.736
So let's accept for a moment

0:04:26.760,0:04:29.656
that it's possible to build[br]such an artificial intelligence

0:04:29.680,0:04:31.816
for this kind of mission and others.

0:04:31.840,0:04:34.376
The next question[br]you must ask yourself is,

0:04:34.400,0:04:35.856
should we fear it?

0:04:35.880,0:04:37.856
Now, every new technology

0:04:37.880,0:04:40.776
brings with it[br]some measure of trepidation.

0:04:40.800,0:04:42.496
When we first saw cars,

0:04:42.520,0:04:46.536
people lamented that we would see[br]the destruction of the family.

0:04:46.560,0:04:49.256
When we first saw telephones come in,

0:04:49.280,0:04:52.176
people were worried it would destroy[br]all civil conversation.

0:04:52.200,0:04:56.136
At a point in time we saw[br]the written word become pervasive,

0:04:56.160,0:04:58.656
people thought we would lose[br]our ability to memorize.

0:04:58.680,0:05:00.736
These things are all true to a degree,

0:05:00.760,0:05:03.176
but it's also the case[br]that these technologies

0:05:03.200,0:05:06.576
brought to us things[br]that extended the human experience

0:05:06.600,0:05:08.480
in some profound ways.

0:05:09.840,0:05:12.120
So let's take this a little further.

0:05:13.120,0:05:17.856
I do not fear the creation[br]of an AI like this,

0:05:17.880,0:05:21.696
because it will eventually[br]embody some of our values.

0:05:21.720,0:05:25.216
Consider this: building a cognitive system[br]is fundamentally different

0:05:25.240,0:05:28.536
than building a traditional[br]software-intensive system of the past.

0:05:28.560,0:05:31.016
We don't program them. We teach them.

0:05:31.040,0:05:33.696
In order to teach a system[br]how to recognize flowers,

0:05:33.720,0:05:36.736
I show it thousands of flowers[br]of the kinds I like.

0:05:36.760,0:05:39.016
In order to teach a system[br]how to play a game --

0:05:39.040,0:05:41.000
Well, I would. You would, too.

0:05:42.600,0:05:44.640
I like flowers. Come on.

0:05:45.440,0:05:48.296
To teach a system[br]how to play a game like Go,

0:05:48.320,0:05:50.376
I'd have it play thousands of games of Go,

0:05:50.400,0:05:52.056
but in the process I also teach it

0:05:52.080,0:05:54.496
how to discern[br]a good game from a bad game.

0:05:54.520,0:05:58.216
If I want to create an artificially[br]intelligent legal assistant,

0:05:58.240,0:06:00.016
I will teach it some corpus of law

0:06:00.040,0:06:02.896
but at the same time I am fusing with it

0:06:02.920,0:06:05.800
the sense of mercy and justice[br]that is part of that law.

0:06:06.560,0:06:09.536
In scientific terms,[br]this is what we call ground truth,

0:06:09.560,0:06:11.576
and here's the important point:

0:06:11.600,0:06:13.056
in producing these machines,

0:06:13.080,0:06:16.496
we are therefore teaching them[br]a sense of our values.

0:06:16.520,0:06:19.656
To that end, I trust[br]an artificial intelligence

0:06:19.680,0:06:23.320
the same, if not more,[br]as a human who is well-trained.

0:06:24.080,0:06:25.296
But, you may ask,

0:06:25.320,0:06:27.936
what about rogue agents,

0:06:27.960,0:06:31.296
some well-funded[br]nongovernment organization?

0:06:31.320,0:06:35.136
I do not fear an artificial intelligence[br]in the hand of a lone wolf.

0:06:35.160,0:06:39.696
Clearly, we cannot protect ourselves[br]against all random acts of violence,

0:06:39.720,0:06:41.856
but the reality is such a system

0:06:41.880,0:06:44.976
requires substantial training[br]and subtle training

0:06:45.000,0:06:47.296
far beyond the resources of an individual.

0:06:47.320,0:06:48.536
And furthermore,

0:06:48.560,0:06:51.816
it's far more than just injecting[br]an internet virus to the world,

0:06:51.840,0:06:54.936
where you push a button,[br]all of a sudden it's in a million places

0:06:54.960,0:06:57.416
and laptops start blowing up[br]all over the place.

0:06:57.440,0:07:00.256
Now, these kinds of substances[br]are much larger,

0:07:00.280,0:07:01.995
and we'll certainly see them coming.

0:07:02.520,0:07:05.576
Do I fear that such[br]an artificial intelligence

0:07:05.600,0:07:07.560
might threaten all of humanity?

0:07:08.280,0:07:12.656
If you look at movies[br]such as "The Matrix," "Metropolis,"

0:07:12.680,0:07:15.856
"The Terminator,"[br]shows such as "Westworld,"

0:07:15.880,0:07:18.016
they all speak of this kind of fear.

0:07:18.040,0:07:22.336
Indeed, in the book "Superintelligence"[br]by the philosopher Nick Bostrom,

0:07:22.360,0:07:23.896
he picks up on this theme

0:07:23.920,0:07:27.936
and observes that a superintelligence[br]might not only be dangerous,

0:07:27.960,0:07:31.816
it could represent an existential threat[br]to all of humanity.

0:07:31.840,0:07:34.056
Dr. Bostrom's basic argument

0:07:34.080,0:07:36.816
is that such systems will eventually

0:07:36.840,0:07:40.096
have such an insatiable[br]thirst for information

0:07:40.120,0:07:43.016
that they will perhaps learn how to learn

0:07:43.040,0:07:45.656
and eventually discover[br]that they may have goals

0:07:45.680,0:07:47.976
that are contrary to human needs.

0:07:48.000,0:07:49.856
Dr. Bostrom has a number of followers.

0:07:49.880,0:07:54.200
He is supported by people[br]such as Elon Musk and Stephen Hawking.

0:07:54.880,0:07:57.280
With all due respect

0:07:58.160,0:08:00.176
to these brilliant minds,

0:08:00.200,0:08:02.456
I believe that they[br]are fundamentally wrong.

0:08:02.480,0:08:05.656
Now, there are a lot of pieces[br]of Dr. Bostrom's argument to unpack,

0:08:05.680,0:08:07.816
and I don't have time to unpack them all,

0:08:07.840,0:08:10.536
but very briefly, consider this:

0:08:10.560,0:08:14.296
super knowing is very different[br]than super doing.

0:08:14.320,0:08:16.216
HAL was a threat to the Discovery crew

0:08:16.240,0:08:20.656
only insofar as HAL commanded[br]all aspects of the Discovery.

0:08:20.680,0:08:23.176
So it would have to be[br]with a superintelligence.

0:08:23.200,0:08:25.696
It would have to have dominion[br]over all of our world.

0:08:25.720,0:08:28.536
This is the stuff of Skynet[br]from the movie "The Terminator"

0:08:28.560,0:08:30.416
in which we had a superintelligence

0:08:30.440,0:08:31.816
that commanded human will,

0:08:31.840,0:08:35.696
that directed every device[br]that was in every corner of the world.

0:08:35.720,0:08:37.176
Practically speaking,

0:08:37.200,0:08:39.296
it ain't gonna happen.

0:08:39.320,0:08:42.376
We are not building AIs[br]that control the weather,

0:08:42.400,0:08:43.736
that direct the tides,

0:08:43.760,0:08:47.136
that command us[br]capricious, chaotic humans.

0:08:47.160,0:08:51.056
And furthermore, if such[br]an artificial intelligence existed,

0:08:51.080,0:08:54.016
it would have to compete[br]with human economies,

0:08:54.040,0:08:56.560
and thereby compete for resources with us.

0:08:57.200,0:08:58.416
And in the end --

0:08:58.440,0:08:59.680
don't tell Siri this --

0:09:00.440,0:09:01.816
we can always unplug them.

0:09:01.840,0:09:03.960
(Laughter)

0:09:05.360,0:09:07.816
We are on an incredible journey

0:09:07.840,0:09:10.336
of coevolution with our machines.

0:09:10.360,0:09:12.856
The humans we are today

0:09:12.880,0:09:15.416
are not the humans we will be then.

0:09:15.440,0:09:18.576
To worry now about the rise[br]of a superintelligence

0:09:18.600,0:09:21.656
is in many ways a dangerous distraction

0:09:21.680,0:09:24.016
because the rise of computing itself

0:09:24.040,0:09:27.056
brings to us a number[br]of human and societal issues

0:09:27.080,0:09:28.720
to which we must now attend.

0:09:29.360,0:09:32.176
How shall I best organize society

0:09:32.200,0:09:34.536
when the need for human labor diminishes?

0:09:34.560,0:09:38.376
How can I bring understanding[br]and education throughout the globe

0:09:38.400,0:09:40.176
and still respect our differences?

0:09:40.200,0:09:44.456
How might I extend and enhance human life[br]through cognitive healthcare?

0:09:44.480,0:09:47.336
How might I use computing

0:09:47.360,0:09:49.120
to help take us to the stars?

0:09:49.760,0:09:51.800
And that's the exciting thing.

0:09:52.400,0:09:54.736
The opportunities to use computing

0:09:54.760,0:09:56.296
to advance the human experience

0:09:56.320,0:09:57.736
are within our reach,

0:09:57.760,0:09:59.616
here and now,

0:09:59.640,0:10:01.320
and we are just beginning.

0:10:02.280,0:10:03.496
Thank you very much.

0:10:03.520,0:10:07.806
(Applause)