(Toby Walsh) I want to talk
about artificial intelligence: it's --
I'm a professor of artificial intelligence
and its a great time, 2015,
to be working in AI.
We're making real palpable progress
and there's loads of money
being thrown at us.
Google just spent
five hundred million dollars --
pounds buying an AI startup called Deep Mind a couple of weeks ago,
today they announced that they were
going to spend a billion dollars
setting up an AI lab in Silicon Valley.
IBM is betting about
a third of the company
on their cognitive AI computing effort.
So it's really interesting time to be working in AI.
But the first thing I wanted to
help inform you about
is what is the state of art,
what progress have we made in AI
because Hollywood paints
all these pictures,
these mostly dystopian pictures of AI.
Whenever the next science fiction movie
comes out, I put my head in my hands
and think Oh my God, what do people think
that we're doing?
So, I wanted to start by just giving you
a feel for what actually is really capable.
So a couple of years ago, IBM Watson
won the game show Jeopardy 1,
the million dollar prize in the game show,
Jeopardy P, the reigning human champions.
Now, you might think, well that's just
a party trick, isn't it?
It's a -- pour enough of Wikipedia
and the internet into a computer,
and it can answer general knowledge
questions.
Well, you guys are being a bit unfair to
IBM Watson,
there are the cryptic questions
they are answering.
But just to give you a real feel for
what is technically possible today,
something that was announced
just two days ago:
some colleagues of mine at NII in Japan
passed the University Entrance Exam
with an AI program.
Now, I thought long and hard about
putting up a page of maths.
I thought, well, I'm going to get --
half of the audience is going to
leave immediately
if I put up a page of math
But I wanted you to see, just to feel
the depth of questions
that they were answering.
So this is from the maths paper, you know,
a non trivial, sort of,
if you come from the UK,
A-level-like math question
that they were able to answer
Here is a physics question
about Newtonian dynamics
that they were able to answer.
Now they got 511 points, out of
a maximum of 950.
That's more than the average score
of Japanese students
sitting the entrance exam.
They would have got into most
Japanese universities with a score of 511.
That's what's possible today.
Their ambition in 10 years' time is to get
into Tokyo, University of Tokyo,
which is one of the best universities
in the world.
So, this is why I put up
a picture of Terminator,
because whenever I talk to the media
about what we do in AI,
they put up a picture of Terminator,
right?
So I don't want you to worry
about Terminator,
Terminator is at least
50 to 100 years away.
and there are lots of reasons why
we don't have to worry about it,
about Terminator.
I'm not going to go and spend
too much time
on why you don't have to worry about Terminator.
But there is actually things
that you should worry about,
much nearer than Terminator.
Many people have said,
Stephen Hawkins has said
that, you know, AI is going to spell
the end of the human race.
Elon Musk chimed in afterwards, said
"It's our biggest existential threat."
Bill Gates followed on by saying,
"Elon Musk was right."
Lots of people have said that's a --
AI is a real existential threat to us.
I don't want you to worry about
the existential threat that AI
or Terminator is going to bring.
There's actually a very common confusion,
which is that it's not the AI
that's going to be the existential threat,
it's autonomy, it's autonomous systems.
What's in the (check) --
it isn't it going to wake up
any time in the morning and say:
"You know what? I'm tired
of playing Jeopardy,
"I want to play Who Wants to Be a Millionaire!
"Or wait a second, I'm tired of playing
game shows,
I want to take over the universe."
It's just not in its code, there is no way,
it's not given any freedom to think
about anything other
than maximizing its Jeopardy score.
And it has no desires, no other desires than,
other to improve its maximum scores.
So I don't want you to worry about Terminator,
but I do want you to worry about jobs.
Because lots of people,
lots of very serious people,
have been saying
hat AI is going to end jobs,
and that is a very great consequence
for anyone working in education,
because, certainly, the jobs that are going
to exist in the future
are going to be different
than the jobs that exist today.
Now, who has an odd birthday?
Well, I haven't told you
what an odd birthday is yet,
so someone has an odd birthday, like me.
OK. Who was born on an odd-number
day of the month?
I was born on the 11th of April, right?
Come on, it's half the room,
I know it's half the room.
(Laughter)
OK.Well, you want to have
an odd birthday, by the way,
because that means, in 20 years' time,
you will be a person with a job.
As opposed to the even people,
who won't have jobs.
That's certainty -- if you believe
lots of serious people,
you might have missed this news
on Friday the 13th,
I thought this was a rather
depressing news story
....... (check) comparison otherwise ...... (check)
but the chief economist, Bank of England,
went on the record saying 50% of jobs
were under threat in the UK.
And he's not the first serious person
who should know what he's talking about
who said similar things.
There was a very influential Merrill Lynch
report that came out a month or to ago
saying very similar things about
the impact of AI,
robotics, automation on jobs.
And some of this goes back to, I think,
one of the first reports
that really hit the press,
that really got people's attention,
was a report that came out of
the Oxford Martin School.
They predicted that 47% of jobs
in the United States
were under threat of automation
in the next 20 years.
We followed that up
with a very similar study and analysis
for jobs in Australia, where I work.
And because it's a slightly different
profile of workers,
of the work force in Australia, we came up
with a number of around 40%.
These are non trivial numbers, right?
40-50%.
No, just an aside: 47%, I don't know
why they didn't say 47.2%, right?
You can't believe a number
that's far too precise
when you're predicting the future,
but nevertheless,
the fact that it's of this sort of scale,
you've got to take away: it wasn't 4%,
it was roughly about half the jobs.
Now, let's put some context to this.
I mean, is this really a credible claim?
The Number One job
in the United States today: truck driver.
Now you might have noticed,
the autonomous cars
are coming to us very soon.
We're going to be having -- tried
the first trial of autonomous cars
on the roads, public roads of Australia,
three weeks ago.
The Google Car has driven
over a million kilometers
-- or the Google cars, rather,
have driven over a million kilometers,
autonomously, on the roads of California.
In 20 years' time, we are going to have
autonomous cars.
We're also going to have
autonomous trucks.
So if you are in the Number One profession
in the United States,
you have to worry that your job
is not going to be automated away.
The Number Two job in the United States
is salesperson.
Again, since we use the internet,
we've actually mostly automated
that process ourselves,
but it's clear that a lot of those jobs
are going to be disappearing.
So I think these claims
have a lot of credibility.
There's actually a nice dinner party game
that my colleagues in AI play
at the end of our conferences,
where we sit around
and the game is, you have to name a job
and then, someone has to put up
some credible evidence
that we're actually well on the way
to actually automating that.
And this game is almost impossible to win.
If I had more time,
I'd play the game with you.
The only -- about the only winning answer
is politician.
(Laughter)
They will certainly regulate that
they'll be the last to be automated.
But that's about
the only winning answer we have.
So -- and it's not just technology
that is the cause of this.
There's many other, really,
sort of rather unhelpful trends.
If you were trying to set up
the world's economy,
you would not put these things
all down on the table at the same time:
the global,
ongoing global financial crisis,
which seems like
it will never disappear, I think;
the fact that we're all living longer:
this is great, great news for us
but bad news for employment;
the impact of globalization, the fact that
we can outsource our work
to cheaper economies.
All of these things
are compounding the impact
that technology is having
on the nature of work.
And this transformation is going to be
different than the last one,
the Industrial revolution.
There's no hard and fast
rule of economics that says:
"As many jobs need to be created
by a new technology as destroyed."
Every time we have a new technology,
of course, new jobs are created.
There's lots of, there's thousands,
hundreds of thousands of new jobs
enabled by technology today.
But there's no reason that they have to
balance exactly those that are destroyed.
In the last -- in the last revolution,
that did happen to be the case.
A third of the population was working
out in the fields, in agriculture.
Now, worldwide,
it's 3 or 4% of the world's population
working in agriculture.
Those people are working
in factories and offices now.
We employ far more people than we did
at the turn of the 19th century.
But this one looks different, this
information revolution looks different.
It looks like it has the potential
to take away more jobs, perhaps,
than it does.
And one of the other things is that
we used to think
it was the blue-collar jobs.
And that's true: if you go
to a car factory today, sure enough,
there are robots
that are doing the painting,
there are robots
that are doing the welding.
But nowadays, it's white-collar jobs:
it's journalists, it's lawyers,
it's accountants,
these jobs that are under threat.
These graphs here show
the percentage change in employment
and the change in employment rates.
And it's the middle, the middle class,
white-collar professions
that we thought that you would go
to university to make yourself safe,
but it seems to be the ones
that are most under threat.
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,
Go and 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, (check)
the hardest things,
making computers more creative,
making computers that can understand
your emotional state,
focusing on emotional terms
and not intellectual intelligence.
So, how safe is education?
The room is here, full of people
working in education.
How safe are your jobs?
Well these are the numbers
from that Oxford Martin report I wrote.
So if you're a telemarketeer,
99% chance that
you're going to be automated.
Not surprising, right?
Easy to automate, it's down the phone.
Some of the numbers
I just don't want you to take away:
they used machine only, they used AI
to actually generate the report
and I don't believe some of the numbers,
I don't believe:
"Bicycle repairmen: 94%."
There's no chance in hell
that the bicycle repair person
is going to be automated:
far too cheap and intricate a job.
"Parking lot attendant: 87%."
I don't know why it's not 100%:
well, you're not going to have
parking lot attendants, for sure.
But look: luckily,
you and me are
right at the bottom of the list,
down at the 0's and 1%, right?
I think those numbers
probably underestimate
how replaceable or
how irreplaceable we are,
but nevertheless,
you can take some heart away
from the sort of numbers you see there.
And the reason being?
The first reason is, because
we're dealing with people,
we're trying
to understand people,
understand their motivations,
what are their mental blocks.
These are things that are
really hard to get computers,
I can tell you, really hard to get
computers, program computers to do.
So, one of the things, I think,
we have to realize,
is that with this impact that automation
and AI in particular
is going to have on jobs
-- and as was mentioned in several
of the earlier talks --
is that of course, we try to educate
people for a future that does not exist,
for technologies
that have yet to be invented.
Education, therefore, inherently,
is going to have to be a lifelong process.
We can't teach you
the programming language of the future,
because we haven't invented it.
It's going to be --
we have to teach you fundamental ideas
so that you can then go off and learn
as you go on in life.
So, to have a job,
you're going to have to
keep yourself abreast
of the latest technologies
and AI and technologies should be there
to help us do that.
AI will help that lifelong journey, and
I think the big thing that AI can help is,
is to personalize
that learning experience,
to construct a model
of your understanding of the topic.
So, to generate you
infinite numbers of test problems
that can be tailored to exactly
where your state of knowledge is,
and then mark those test problems
instantly.
So, the fact that we can answer,
that we can answer
Japanese college entrance exams means
that we can also mark
Japanese college entrance exams.
We can do all those things, right?
with technology, and we can do it for you.
And MOOCs will turn into
-- I've invented a little acronym there --
POOCs, Personal Open Online Courses,
right?
I don't think they should be massive,
they should be for you.
You can choose your own trajectory.
I always find it very strange
that we're there still
just like the classroom experience.
We should be able to follow
whatever interesting trajectory we want
through the material.
And then, of course, use AI techniques
like data mining analytics
across the massive part of the MOOC
to really improve
your learning experience.
So, I just wanted to conclude by saying
AI is not quite
what you see in the movies
but we are making impressive progress
and it's certainly going to be
an interesting part of the equation
for this interesting future
that we all face.
Thank you very much.
(Applause)