The jobs we’ll lose to machines — and the ones we won’t | Anthony Goldbloom

The jobs we’ll lose to machines — and the ones we won’t | Anthony Goldbloom

Posted by

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


  1. What happens when you connect the machines together though something like the internet that allows them to pool knowledge and figure out how to complete new and unique task

  2. The assumption that some jobs cannot be automated feels totally arbitrary. Humans are machines too, we are just biological machines.
    There is no reason to assume that mechanical/digital machines cannot do, what we do. It's just that CURRENT machines cannot do all the tasks.
    The more interesting question is: Do we want them to do all of our jobs? Especially regarding economy as well as goals like self-fulfillment.

  3. So? Humans will create new jobs.
    There used to be a job where people will hand write new books to create copies of it. Now we have scanners and printers. And we're doing just fine with that job being gone.

  4. Education too often focuses what computers do too easily, mathematics and recall of input data. It is almost like people are trained to be replaced instead of being irreplaceable.

  5. If he's basing the prediction that there will not be almost any jobs left on there being an algorithm or not, I hate to say but everything can be put in an algorithm. It's how the universe works. Therefore even marketing and engineering jobs will be a thing of the past.

  6. I like the topic. I like the contents, but it's annoying to listen to this since the speaker tried to speak too fast and it made the presentation not as good as it should have been. Maybe it's because he had only 5 minutes. I hope next time we can enjoy a more impressive talk on this. If we have only 5 minutes, plz do not try to give too much information or explanation.

  7. This had that I don't want people to hate me for being the bad guy feel. Or at least until I get a profitable exit for my business. But not in a deliberate way. Like he has been telling himself in the mirror that only some jobs will elemenated, but there will still something for us humans to do.

  8. Yikes, I came to think about how machine learning will move into a higher abstraction level where algorithms use already gathered machine knowledge and combine it to handle more general and novel tasks. So instead of a flattening of the exponential curve in the video it will then continue to accelerate! Just as Ray Kurzweil has predicted with his Law of Accelerating Returns. We then definitely will need something like a universal basic income pretty soon.

  9. should i choose computer science as college degree? i want to but, everyone says automation is killing computer jobs!!

  10. While this is heartening (that novel tasks are safe from machines and algorithms) it makes the dangerous assumption of projecting into the future what machines/algorithms are good at today. IOW, this is circular thinking based on a static rather than dynamic assessment. And yes, the talk needs to run longer to avoid being a four minute assertion.

  11. We should kill anyone that wants to create robots that will most likely have all the jobs since those people are clearly being selfish

  12. You made a big mistake in your conclusion. When you hand over the jobs already occupied by humans to robots. There will be no money in the hands of the majority of people, this there will be no market to know its gaps to fill.

  13. His rationale is , machines can't seemingly infer a new concept from two different concepts. However, it only be a matter of time, machine would be able to do that too. Till then Strategy consulting won't get affected by ML.

  14. If UBI or if the masses are not guaranteed their basic needs; we will have an uprising of Hackers (Neo-luddites) that will ruin the machines. Don't incentivize such an uprising…. share or face despair.

  15. ''Artificial Inteligence '' and yet Google / Youtube BOTS ban and flag non stop videos and channels without any reason whatsoever :)) So much for their ''artificial inteligence '' or better said the lack of it . Technology on paper may look great, but in reality can become a nightmare. Try to reason with a BOT that don't even bother to respond you back , will censor everything recieved from you and will deny all your rights. And maybe in the future will even try to murder you , ofcourse if you become a ''humanoid threat '' :))

  16. Excellent Talk… I particularly loved your intro and how you linked it with the topics to be explored and at the end concluded with a very impactful msg. The examples and evidence were very helpful to add to the credibility of your info. Short talk yet well organized and indeed reached the intended goal. Well done Anthony!

  17. These liberals are never going to lose their jobs. In 50 years or so, machine rights will be a thing. Computers are working long hours and getting paid very poorly. I am not a binary computer, I'm actually a self conscious demi automatic cyber-kinetic entity.

  18. so its basically as nathaniel branden said: we humnans need to do something about our self esteem and become entrepeneurs. The future will belong to creative, independent and active individuals.

  19. Machines can excel at frequent, high volumes tasks. For us humans, let everyday usher in a new challenging task.

  20. All I got from this was that business people and campaigners will be safe from automation. As an engineer, I fear for what's to come.

  21. I believe all sectors or perspectives are not considered, like with growing population and reducing land for cultivation, we would require people who could improve crop quality & quantity, not all jobs will be done by robots, any or all jobs involving non standard or unorganized objects, like material handling especially of brittle objects will be still done by humans, things which need instinct like detectives/cops will be humans, already this generation gave up outdoor games, so not sure of future of athletes. Child care and care centres for elderly people will still require humans. More people will either involve in improving environment on earth or looking for new planet like earth and making arrangements to shift and stay their. Energy could be new global currency, just like gas/fuel, people owning land would be able to invest in power generating technologies like Solar, wind, etc. and they will buy and sell as per requirements.

  22. Guy invents the microwave oven.
    '…this cross-pollination happens for each of us, thousands of times per day…'
    I've told you a million times, don't exagerrate!

  23. This guy seems to believe that taxi and delivery drivers, shop staff, (including fast-food outlets), factory staff, etc etc, will suddenly become marketing strategists or will be able to find a gap in the market to create themselves a job.
    Big fail on this TED talk.

  24. Atleast we still have the brain to think. Creativelly, critically and analytically. Computer only has the knowledge but we humans have tha ability to imagine.

  25. Humanity can't even deal with itself and we want to introduce a new "race" into the mix? We are beyond fucked,this isn't a joke.Technology in excess will be mans undoing.

  26. If anything, and i'm being optimistic here, we'll eventually merge with computers. Whether that's a good or bad thing is up to your personal opinion.

  27. What he's really saying is that as time goes on the repetitive jobs at the bottom will be taken over by machines and even the entry-level jobs for humans will all require advanced degrees. So we're going to have to develop some kind of "Universal Basic Income" or half the population starves.

  28. Psssssh machines can take my jobs man I wanna stay home screw working 40hrs a week and not being able to keep up rent because of minimum wage and stressing life style. Im ready for robots to take over anydaaaay

  29. I'm curious which one of you "people" would do it for ten years for free. Are you human, and is not it good to keep you worse than any animals? 110% is time to think about, because testing what human in extreme situations is, it also has its human limits. That's what you're just doing for me, I'm not.

  30. Machines cannot replace humans they can only aid and that's the only purpose else machines will be watching YouTube and humans will be working in industries.

  31. robots should be owned by common people not companies, that way any job that the robot performs payment will go to the owner of the robot

  32. As far as I know, even though robots can be programmed to feel emotion, I doubt robots can make a lot of meaningful art or be super creative. Even though robots could replace those who make STILL LIFE paintings or drawings, I'm not sure if a robot can easily create something from scratch. Which is why I'm glad that I will likely not be replaced. And you're lucky if you got good friends who care about you because obviously good friends don't replace each other.

  33. Computers will eventually take all the jobs and man will go extinct there are already robot cops it's so funny that scientists and leaderships think that a computer cant do their jobs they could do their jobs better so if a computer can be a cop a maid a butler prepare your food plant your food … ultimately all things will become obsolete because computers once most humans are dead from destitution and starvation will have no need to continue their tasks. Personally I pray every day for that solar flare that kills electricity and saves people… that's the only hope their truly is left to make people have value in the world again and get rid of everything that has made people valueless.

  34. So the benefit we see from machines for humans is "Keep looking for new challenges, always be in stress"

  35. there is no such thing as a shortage of jobs…just a shortage of people sharing the work!!..It MUST be made financially worthwhile to share the jobs and work LESS.

  36. 4:01 If machines are not making progress in novel situations NOW, doesn't mean they won't ever. Artificial general intelligence is developing fast and once we figure out its basic architecture we can implement it through robotics and make humans completely redundant. Humans don't stand a chance against true thinking machines.

  37. Think it's interesting how his graph shows the beginnings of an exponential curve but then for some reason levels off. Think he is vastly underestimating what machines will be able to do in the not so far future.

  38. The future of work sounds really great. I don't like doing repetitious high-volume tasks I don't think anyone does.

  39. Make machines your labor supporting only needed government and paying for your enjoyable life, no lobbying by corporations becomes doable. Not everyone wants to be wage slaves to debt based economies

  40. I feel like people don t quite get how different the world will be if so much of the population isnt working. Who are they writing advertisements to and doing the other jobs they claim will be left when no one else has jobs or could buy anything. Driving to work in a world where you are only one of the few able to, without some completely different system, would be quite a challenge considering the starving masses you would have to make it through.

  41. Why are we all accepting this? Our path of AI is not going to be good for human existence, I don't want to live in an artificial world. It will be the death of us.

  42. This is like saying everyone will die. The more important question is when. I don't want people to predict 50 years from now, nor 20 years from now. I want people to predict next year or 4 weeks from now. I've heard so many people talk hyperbole since the 1950's, yet tangible economic benefits are little.

  43. Humans will always be preferred to diagnose health problems, as then somebody can be blamed if it's wrong! Who can be held responsible if a machine makes a bad diagnosis? The designer?

  44. yes we will train machine to do every task that is hard to do for human and it will create easy for human.. Specially doctors.

  45. To find out more about robots and the future of our jobs, please visit to get more detailed information!

  46. So the jobs are going to go to the ones who can afford to go to prestigious universities? Their parents will pay money to get them in even though they can't even get in themselves lol

  47. This is plagiarism, there are thousands of books that are saying the same for many years, beggining with the Ray Kurzweill's book "The Age of Intelligent Machines" 1 9 9 0 !!!

  48. Please Anthony!!! take some ENGLISH pronunciation lessons, BEFORE speak in public !! This es YOUTUBE, this is not Bangalore !!!

  49. Machines would definitely get smarter than humans over time. It can be seen from all current developments(Self Driving Cars, Neural Link etc). Unless there are regulations in place, human would be reduced to dummies in this world, with no purpose.

  50. Disagree with some of what he said – finding out what you don't know is one of the greatest tasks for AI – so finding gaps in the market is right up the AI alley.

  51. OK so there will be machines doing the work of humans. Then what humans will do ? Will they do conversion of poor people in developing countries?

Leave a Reply

Your email address will not be published. Required fields are marked *