04:37
TED2016

Anthony Goldbloom: The jobs we'll lose to machines -- and the ones we won't

Filmed:

Machine learning isn't just for simple tasks like assessing credit risk and sorting mail anymore -- today, it's capable of far more complex applications, like grading essays and diagnosing diseases. With these advances comes an uneasy question: Will a robot do your job in the future?

- Machine learning expert
Anthony Goldbloom crowdsources solutions to difficult problems using machine learning. Full bio

So this is my niece.
00:12
Her name is Yahli.
00:14
She is nine months old.
00:16
Her mum is a doctor,
and her dad is a lawyer.
00:18
By the time Yahli goes to college,
00:21
the jobs her parents do
are going to look dramatically different.
00:23
In 2013, researchers at Oxford University
did a study on the future of work.
00:27
They concluded that almost one
in every two jobs have a high risk
00:32
of being automated by machines.
00:36
Machine learning is the technology
00:40
that's responsible for most
of this disruption.
00:42
It's the most powerful branch
of artificial intelligence.
00:44
It allows machines to learn from data
00:47
and mimic some of the things
that humans can do.
00:49
My company, Kaggle, operates
on the cutting edge of machine learning.
00:51
We bring together
hundreds of thousands of experts
00:55
to solve important problems
for industry and academia.
00:57
This gives us a unique perspective
on what machines can do,
01:01
what they can't do
01:04
and what jobs they might
automate or threaten.
01:05
Machine learning started making its way
into industry in the early '90s.
01:09
It started with relatively simple tasks.
01:12
It started with things like assessing
credit risk from loan applications,
01:15
sorting the mail by reading
handwritten characters from zip codes.
01:19
Over the past few years, we have made
dramatic breakthroughs.
01:24
Machine learning is now capable
of far, far more complex tasks.
01:27
In 2012, Kaggle challenged its community
01:31
to build an algorithm
that could grade high-school essays.
01:35
The winning algorithms
were able to match the grades
01:38
given by human teachers.
01:40
Last year, we issued
an even more difficult challenge.
01:43
Can you take images of the eye
and diagnose an eye disease
01:46
called diabetic retinopathy?
01:49
Again, the winning algorithms
were able to match the diagnoses
01:51
given by human ophthalmologists.
01:55
Now, given the right data,
machines are going to outperform humans
01:57
at tasks like this.
02:00
A teacher might read 10,000 essays
over a 40-year career.
02:01
An ophthalmologist might see 50,000 eyes.
02:06
A machine can read millions of essays
or see millions of eyes
02:08
within minutes.
02:12
We have no chance of competing
against machines
02:14
on frequent, high-volume tasks.
02:17
But there are things we can do
that machines can't do.
02:20
Where machines have made
very little progress
02:24
is in tackling novel situations.
02:27
They can't handle things
they haven't seen many times before.
02:28
The fundamental limitations
of machine learning
02:33
is that it needs to learn
from large volumes of past data.
02:35
Now, humans don't.
02:39
We have the ability to connect
seemingly disparate threads
02:41
to solve problems we've never seen before.
02:44
Percy Spencer was a physicist
working on radar during World War II,
02:46
when he noticed the magnetron
was melting his chocolate bar.
02:51
He was able to connect his understanding
of electromagnetic radiation
02:54
with his knowledge of cooking
02:58
in order to invent -- any guesses? --
the microwave oven.
02:59
Now, this is a particularly remarkable
example of creativity.
03:03
But this sort of cross-pollination
happens for each of us in small ways
03:06
thousands of times per day.
03:10
Machines cannot compete with us
03:12
when it comes to tackling
novel situations,
03:14
and this puts a fundamental limit
on the human tasks
03:16
that machines will automate.
03:19
So what does this mean
for the future of work?
03:22
The future state of any single job lies
in the answer to a single question:
03:24
To what extent is that job reducible
to frequent, high-volume tasks,
03:29
and to what extent does it involve
tackling novel situations?
03:34
On frequent, high-volume tasks,
machines are getting smarter and smarter.
03:37
Today they grade essays.
They diagnose certain diseases.
03:42
Over coming years,
they're going to conduct our audits,
03:44
and they're going to read boilerplate
from legal contracts.
03:47
Accountants and lawyers are still needed.
03:50
They're going to be needed
for complex tax structuring,
03:52
for pathbreaking litigation.
03:55
But machines will shrink their ranks
03:57
and make these jobs harder to come by.
03:58
Now, as mentioned,
04:00
machines are not making progress
on novel situations.
04:01
The copy behind a marketing campaign
needs to grab consumers' attention.
04:04
It has to stand out from the crowd.
04:08
Business strategy means
finding gaps in the market,
04:10
things that nobody else is doing.
04:12
It will be humans that are creating
the copy behind our marketing campaigns,
04:14
and it will be humans that are developing
our business strategy.
04:18
So Yahli, whatever you decide to do,
04:21
let every day bring you a new challenge.
04:24
If it does, then you will stay
ahead of the machines.
04:27
Thank you.
04:31
(Applause)
04:32

▲Back to top

About the Speaker:

Anthony Goldbloom - Machine learning expert
Anthony Goldbloom crowdsources solutions to difficult problems using machine learning.

Why you should listen

Anthony Goldbloom is the co-founder and CEO of Kaggle. Kaggle hosts machine learning competitions, where data scientists download data and upload solutions to difficult problems. Kaggle has a community of over 600,000 data scientists and has worked with companies ranging Facebook to GE on problems ranging from predicting friendships to flight arrival times.

Before Kaggle, Anthony worked as an econometrician at the Reserve Bank of Australia, and before that the Australian Treasury. In 2011 and 2012, Forbes named Anthony one of the 30 under 30 in technology; in 2013 the MIT Tech Review named him one of top 35 innovators under the age of 35, and the University of Melbourne awarded him an Alumni of Distinction Award. He holds a first call honors degree in Econometrics from the University of Melbourne.  

More profile about the speaker
Anthony Goldbloom | Speaker | TED.com