0:00:00.000,0:00:03.204
(Toby Walsh) I wanna talk [br]about artificial intelligence: it's --

0:00:03.674,0:00:05.089
I'm a professor of artificial intelligence

0:00:05.089,0:00:09.271
and its a great time, 2015, [br]to be working in AI.

0:00:09.544,0:00:10.544
We're making real palpable progress

0:00:10.544,0:00:12.820
and there's loads of money [br]being thrown at us.

0:00:12.820,0:00:17.000
Google just spent [br]five hundred million dollars --

0:00:17.000,0:00:21.989
pounds buying an AI startup called Deep Mind a couple of weeks ago,

0:00:21.989,0:00:25.172
today they announced that they were [br]going to spend a billion dollars

0:00:25.671,0:00:28.469
setting up an AI lab in Silicon Valley.

0:00:29.001,0:00:31.591
IBM is betting about [br]a third of the company

0:00:31.830,0:00:34.858
on their cognitive AI computing effort.

0:00:35.084,0:00:38.216
So it's really interesting time to be working in AI.

0:00:38.631,0:00:41.782
But the first thing I wanted to [br]help inform you about

0:00:42.250,0:00:46.792
is what is the state of art, [br]what progress have we made in AI

0:00:46.792,0:00:51.187
because Hollywood paints [br]all these pictures,

0:00:51.187,0:00:53.488
these mostly dystopian pictures of AI.

0:00:53.495,0:00:58.672
Whenever the next science fiction movie[br]comes out, I put my head in my hands

0:00:58.672,0:01:01.755
and think Oh my God, what do people think[br]that we're doing?

0:01:01.755,0:01:08.729
So, I wanted to start by just giving you[br]a feel for what actually is really capable.

0:01:09.081,0:01:13.872
So a couple of years ago, IBM Watson[br]won the game show Jeopardy 1,

0:01:13.872,0:01:19.737
the million dollar prize in the game show,[br]Jeopardy P, the reigning human champions.

0:01:19.737,0:01:23.357
Now, you might think, well that's just[br]a party trick, isn't it?

0:01:23.357,0:01:27.625
It's a -- pour enough of Wikipedia [br]and the internet into a computer,

0:01:27.625,0:01:30.543
and it can answer general knowledge[br]questions.

0:01:30.543,0:01:33.204
Well, you guys are being a bit unfair to[br]IBM Watson,

0:01:33.204,0:01:35.881
there are the cryptic questions[br]they are answering.

0:01:35.881,0:01:40.878
But just to give you a real feel for [br]what is technically possible today,

0:01:40.878,0:01:43.586
something that was announced[br]just two days ago:

0:01:43.834,0:01:51.869
some colleagues of mine at NII in Japan[br]passed the University Entrance Exam

0:01:51.869,0:01:53.433
with an AI program.

0:01:53.433,0:01:56.326
Now, I thought long and hard about[br]putting up a page of maths.

0:01:56.326,0:01:58.007
I thought, well, I'm going to get --

0:01:58.007,0:02:00.007
half of the audience is going to [br]leave immediately

0:02:00.007,0:02:01.354
if I put up a page of math

0:02:01.354,0:02:03.133
But I wanted you to see, just to feel

0:02:03.962,0:02:07.403
the depth of questions [br]that they were answering.

0:02:07.403,0:02:11.507
So this is from the maths paper, you know,[br]a non trivial, sort of,

0:02:11.507,0:02:15.148
if you come from the UK, [br]A-level-like math question

0:02:15.148,0:02:16.468
that they were able to answer

0:02:17.272,0:02:21.889
Here is a physics question[br]about Newtonian dynamics

0:02:21.889,0:02:23.552
that they were able to answer.

0:02:23.552,0:02:28.545
Now they got 511 points, out of[br]a maximum of 950.

0:02:28.939,0:02:32.391
That's more than the average score[br]of Japanese students

0:02:32.391,0:02:34.007
sitting the entrance exam.

0:02:34.007,0:02:38.542
They would have got into most [br]Japanese universities with a score of 511.

0:02:38.908,0:02:40.621
That's what's possible today.

0:02:40.621,0:02:44.134
Their ambition in 10 years' time is to get[br]into Tokyo, University of Tokyo,

0:02:44.134,0:02:46.434
which is one of the best universities[br]in the world.

0:02:48.488,0:02:50.475
So, this is why I put up [br]a picture of Terminator,

0:02:50.475,0:02:53.468
because whenever I talk to the media[br]about what we do in AI,

0:02:53.476,0:02:55.769
they put up a picture of Terminator,[br]right?

0:02:55.769,0:02:58.098
So I don't want you to worry [br]about Terminator,

0:02:58.098,0:03:02.117
Terminator is at least [br]50 to 100 years away.

0:03:02.582,0:03:05.985
and there are lots of reasons why[br]we don't have to worry about it,

0:03:05.985,0:03:07.248
about Terminator.

0:03:07.248,0:03:08.979
I'm not going to go and spend[br]too much time

0:03:08.979,0:03:11.348
on why you don't have to worry about Terminator.

0:03:11.676,0:03:13.334
But there is actually things [br]that you should worry about,

0:03:13.334,0:03:15.352
much nearer than Terminator.

0:03:17.007,0:03:20.160
Many people have said, [br]Stephen Hawkins has said

0:03:20.160,0:03:22.713
that, you know, AI is going to spell[br]the end of the human race.

0:03:23.295,0:03:27.961
Elon Musk chimed in afterwards, said[br]"It's our biggest existential threat."

0:03:29.793,0:03:32.841
Bill Gates followed on by saying,[br]"Elon Musk was right."

0:03:33.498,0:03:39.519
Lots of people have said that's a --[br]AI is a real existential threat to us.

0:03:39.519,0:03:42.887
I don't want you to worry about[br]the existential threat that AI

0:03:42.887,0:03:44.900
or Terminator is going to bring.

0:03:45.953,0:03:47.711
There's actually a very common confusion,

0:03:48.121,0:03:51.139
which is that it's not the AI [br]that's going to be the existential threat,

0:03:51.139,0:03:52.139
it's autonomy, it's autonomous systems.

0:03:54.386,0:03:56.042
What's in the (check) --

0:03:56.042,0:03:57.972
it isn't it going to wake up [br]any time in the morning and say:

0:03:57.972,0:04:02.109
"You know what? I'm tired [br]of playing Jeopardy,

0:04:02.109,0:04:04.173
"I want to play Who Wants to Be a Millionaire!

0:04:05.299,0:04:07.468
"Or wait a second, I'm tired of playing [br]game shows,

0:04:07.468,0:04:09.393
I want to take over the universe."

0:04:09.821,0:04:12.195
It's just not in its code, there is no way,

0:04:12.195,0:04:15.065
it's not given any freedom to think [br]about anything other

0:04:15.065,0:04:17.000
than maximizing its Jeopardy score.

0:04:18.219,0:04:21.209
And it has no desires, no other desires than,

0:04:21.228,0:04:23.482
other to improve its maximum scores.

0:04:23.482,0:04:26.256
So I don't want you to worry about Terminator,

0:04:27.350,0:04:29.102
but I do want you to worry about jobs.

0:04:29.974,0:04:32.166
Because lots of people, [br]lots of very serious people,

0:04:32.166,0:04:35.076
have been saying [br]hat AI is going to end jobs,

0:04:35.076,0:04:38.157
and that is a very great consequence [br]for anyone working in education,

0:04:38.157,0:04:42.269
because, certainly, the jobs that are going [br]to exist in the future

0:04:42.269,0:04:44.925
are going to be different [br]than the jobs that exist today.

0:04:46.880,0:04:49.724
Now, who has an odd birthday?

0:04:51.251,0:04:52.938
Well, I haven't told you [br]what an odd birthday is yet,

0:04:52.938,0:04:54.990
so someone has an odd birthday, like me.

0:04:54.990,0:04:57.858
OK. Who was born on an odd-number [br]day of the month?

0:04:57.858,0:04:59.995
I was born on the 11th of April, right?

0:05:00.824,0:05:03.368
Come on, it's half the room, [br]I know it's half the room.

0:05:03.368,0:05:04.923
(Laughter)

0:05:04.923,0:05:08.690
OK.Well, you want to have [br]an odd birthday, by the way,

0:05:09.059,0:05:12.504
because that means, in 20 years' time, [br]you will be a person with a job.

0:05:13.426,0:05:16.277
As opposed to the even people, [br]who won't have jobs.

0:05:16.277,0:05:20.662
That's certainty -- if you believe [br]lots of serious people,

0:05:22.141,0:05:25.010
you might have missed this news [br]on Friday the 13th,

0:05:25.010,0:05:27.778
I thought this was a rather[br]depressing news story

0:05:27.778,0:05:29.748
....... (check) comparison otherwise ...... (check)

0:05:29.777,0:05:31.527
but the chief economist, Bank of England,

0:05:31.922,0:05:36.230
went on the record saying 50% of jobs [br]were under threat in the UK.

0:05:36.948,0:05:40.420
And he's not the first serious person[br]who should know what he's talking about

0:05:40.420,0:05:42.736
who said similar things.

0:05:43.029,0:05:46.255
There was a very influential Merrill Lynch[br]report that came out a month or to ago

0:05:46.270,0:05:49.430
saying very similar things about [br]the impact of AI,

0:05:49.430,0:05:51.375
robotics, automation on jobs.

0:05:53.439,0:05:56.278
And some of this goes back to, I think, [br]one of the first reports

0:05:56.278,0:05:59.288
that really hit the press, [br]that really got people's attention,

0:05:59.288,0:06:02.075
was a report that came out of [br]the Oxford Martin School.

0:06:02.075,0:06:05.702
They predicted that 47% of jobs [br]in the United States

0:06:06.110,0:06:08.550
were under threat of automation [br]in the next 20 years.

0:06:10.234,0:06:13.594
We followed that up [br]with a very similar study and analysis

0:06:13.880,0:06:15.961
for jobs in Australia, where I work.

0:06:16.865,0:06:19.323
And because it's a slightly different [br]profile of workers,

0:06:19.323,0:06:24.329
of the work force in Australia, we came up[br]with a number of around 40%.

0:06:24.329,0:06:27.129
These are non trivial numbers, right? [br]40-50%.

0:06:27.589,0:06:32.497
No, just an aside: 47%, I don't know [br]why they didn't say 47.2%, right?

0:06:32.497,0:06:35.551
You can't believe a number [br]that's far too precise

0:06:35.551,0:06:37.842
when you're predicting the future,[br]but nevertheless,

0:06:37.842,0:06:39.532
the fact that it's of this sort of scale,

0:06:39.532,0:06:44.601
you've got to take away: it wasn't 4%, [br]it was roughly about half the jobs.

0:06:47.784,0:06:50.495
Now, let's put some context to this.

0:06:50.495,0:06:52.413
I mean, is this really a credible claim?

0:06:52.413,0:06:55.549
The Number One job [br]in the United States today: truck driver.

0:06:55.962,0:06:57.526
Now you might have noticed,

0:06:57.536,0:07:00.079
the autonomous cars[br]are coming to us very soon.

0:07:00.506,0:07:03.579
We're going to be having -- tried[br]the first trial of autonomous cars

0:07:03.579,0:07:06.760
on the roads, public roads of Australia, [br]three weeks ago.

0:07:07.371,0:07:09.657
The Google Car has driven [br]over a million kilometers

0:07:09.657,0:07:12.542
-- or the Google cars, rather, [br]have driven over a million kilometers,

0:07:12.907,0:07:15.050
autonomously, on the roads of California.

0:07:15.893,0:07:18.553
In 20 years' time, we are going to have [br]autonomous cars.

0:07:18.553,0:07:20.255
We're also going to have [br]autonomous trucks.

0:07:21.033,0:07:24.769
So if you are in the Number One profession[br]in the United States,

0:07:24.769,0:07:28.480
you have to worry that your job [br]is not going to be automated away.

0:07:29.712,0:07:32.494
The Number Two job in the United States [br]is salesperson.

0:07:33.343,0:07:36.248
Again, since we use the internet,

0:07:36.248,0:07:39.618
we've actually mostly automated [br]that process ourselves,

0:07:39.618,0:07:43.382
but it's clear that a lot of those jobs [br]are going to be disappearing.

0:07:43.382,0:07:46.736
So I think these claims [br]have a lot of credibility.

0:07:48.896,0:07:53.104
There's actually a nice dinner party game [br]that my colleagues in AI play

0:07:53.104,0:07:55.736
at the end of our conferences,[br]where we sit around

0:07:55.736,0:07:58.250
and the game is, you have to name a job

0:07:59.115,0:08:02.671
and then, someone has to put up [br]some credible evidence

0:08:02.671,0:08:05.748
that we're actually well on the way [br]to actually automating that.

0:08:05.748,0:08:08.152
And this game is almost impossible to win.

0:08:08.737,0:08:10.423
If I had more time, [br]I'd play the game with you.

0:08:11.293,0:08:15.260
The only -- about the only winning answer[br]is politician.

0:08:15.650,0:08:17.931
(Laughter)

0:08:17.931,0:08:20.747
They will certainly regulate that[br]they'll be the last to be automated.

0:08:20.747,0:08:22.731
But that's about [br]the only winning answer we have.

0:08:24.049,0:08:28.836
So -- and it's not just technology [br]that is the cause of this.

0:08:28.836,0:08:33.035
There's many other, really, [br]sort of rather unhelpful trends.

0:08:33.035,0:08:35.558
If you were trying to set up [br]the world's economy,

0:08:35.558,0:08:38.842
you would not put these things [br]all down on the table at the same time:

0:08:38.842,0:08:40.825
the global, [br]ongoing global financial crisis,

0:08:40.825,0:08:43.840
which seems like [br]it will never disappear, I think;

0:08:44.755,0:08:48.013
the fact that we're all living longer: [br]this is great, great news for us

0:08:48.013,0:08:50.333
but bad news for employment;

0:08:50.961,0:08:54.425
the impact of globalization, the fact that [br]we can outsource our work

0:08:54.425,0:08:56.191
to cheaper economies.

0:08:56.191,0:09:00.191
All of these things [br]are compounding the impact

0:09:00.191,0:09:02.957
that technology is having [br]on the nature of work.

0:09:04.749,0:09:09.725
And this transformation is going to be [br]different than the last one,

0:09:09.725,0:09:11.357
the Industrial revolution.

0:09:12.021,0:09:15.591
There's no hard and fast [br]rule of economics that says:

0:09:16.911,0:09:20.039
"As many jobs need to be created [br]by a new technology as destroyed."

0:09:20.039,0:09:22.952
Every time we have a new technology, [br]of course, new jobs are created.

0:09:22.952,0:09:25.883
There's lots of, there's thousands, [br]hundreds of thousands of new jobs

0:09:26.204,0:09:27.848
enabled by technology today.

0:09:28.355,0:09:32.484
But there's no reason that they have to [br]balance exactly those that are destroyed.

0:09:33.092,0:09:36.199
In the last -- in the last revolution, [br]that did happen to be the case.

0:09:36.853,0:09:40.930
A third of the population was working [br]out in the fields, in agriculture.

0:09:40.930,0:09:44.552
Now, worldwide, [br]it's 3 or 4% of the world's population

0:09:44.552,0:09:45.657
working in agriculture.

0:09:45.657,0:09:48.029
Those people are working [br]in factories and offices now.

0:09:48.029,0:09:51.920
We employ far more people than we did [br]at the turn of the 19th century.

0:09:52.629,0:09:55.993
But this one looks different, this [br]information revolution looks different.

0:09:55.993,0:10:00.233
It looks like it has the potential [br]to take away more jobs, perhaps,

0:10:00.233,0:10:01.529
than it does.

0:10:01.529,0:10:03.684
And one of the other things is that [br]we used to think

0:10:03.684,0:10:05.293
it was the blue-collar jobs.

0:10:06.509,0:10:10.937
And that's true: if you go [br]to a car factory today, sure enough,

0:10:10.937,0:10:12.578
there are robots [br]that are doing the painting,

0:10:12.578,0:10:14.800
there are robots [br]that are doing the welding.

0:10:14.800,0:10:16.584
But nowadays, it's white-collar jobs:

0:10:16.584,0:10:20.006
it's journalists, it's lawyers, [br]it's accountants,

0:10:20.006,0:10:21.667
these jobs that are under threat.

0:10:21.667,0:10:25.444
These graphs here show [br]the percentage change in employment

0:10:25.444,0:10:31.244
and the change in employment rates.

0:10:31.660,0:10:34.672
And it's the middle, the middle class, [br]white-collar professions

0:10:34.672,0:10:37.923
that we thought that you would go [br]to university to make yourself safe,

0:10:37.926,0:10:40.624
but it seems to be the ones [br]that are most under threat.

9:59:59.000,9:59:59.000
If you are a ....... (check) it's probably -- you're too cheap to be replaced by something automated. But if you're a more expensive person, and this means (check) that the rich are getting richer and inequalities that we are seeing in society that are distressing our societies to day, seem to be magnified by these technological changes. And there is so many frightening graphs, if you're going to read Thomas Picketty (check), I encourage you. Go and look at one of his books, and you can see here that we're seeing a constant improvement in productivity. Technology is buying us those improvements in productivity, is increasing our wealth. But there's a leveling off of employment. And so, the challenge, then, is how -- it's a question for society, not for a technologist like myself -- how do we all benefit from this rising tide, not so that it is the rich get richer and the rest of us get further behind. So, many parts of many jobs looks likely (check) to be automated. One confusion is this: people say these jobs are going to disappear. Actually, it seems to be more likely that many parts of your job will be automated. But that still means that there is perhaps less employment around. So how can you make yourself more future-proof? Well, I have two pieces of advice as a technologist, in terms of what's going to be technically possible in AI. Either you've got to embrace the future, so become like me, become someone who's working on trying to invent that future. And if you're not technically minded, that's fine: I've got the other part of the equation, the other answer to your question which is completely at the other end of the spectrum, which is: focus on those things that find as AI, the hardest things, making a future that's more creative, making a future that can understand your emotional state, focusing on emotional terms and not intellectual intelligence. 12:23