Duration 25:49
16+
Play
Video

AI and the Post Work World | SXSW 2021

Kai-Fu Lee
Chairman and CEO at Sinovation Ventures
  • Video
  • Table of contents
  • Video
SXSW 2021
March 18, 2021, Online, Austin, USA
SXSW 2021
Request Q&A
Video
AI and the Post Work World | SXSW 2021
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Add to favorites
188
I like 0
I dislike 0
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
  • Description
  • Transcript
  • Discussion

About speaker

Kai-Fu Lee
Chairman and CEO at Sinovation Ventures

About the talk

The world renowned AI expert Dr. Kai-Fu Lee describes how AI is moving from making headlines to implementation into our daily lives. AI applications built in the past few years can now replace human eyes, hands/feet and sometimes our brain to make decisions. In the age of AI, our traditional definition of work ethics must be reinvented and people need to be better prepared for job displacement and transition. Dr. Lee proposes a 3Rs strategy - relearn, recalibrate, and renaissance - for us to contemplate and take actions in order to embrace the real benefits of AI to humanity.

About SXSW:

SXSW dedicates itself to helping creative people achieve their goals. Founded in 1987 in Austin, Texas, SXSW is best known for its conference and festivals that celebrate the convergence of the interactive, film, and music industries. An essential destination for global professionals, this year’s online event features sessions, showcases, screenings, exhibitions, professional development and a variety of networking opportunities. For more information, please visit sxsw.com.

Subscribe: http://www.youtube.com/user/sxsw?sub_...

Connect with SXSW:

Website: https://www.sxsw.com

Facebook: https://www.facebook.com/SXSWFestival/

Twitter: https://twitter.com/sxsw

Instagram: https://www.instagram.com/sxsw/

YouTube: https://www.youtube.com/sxsw

Share

Hi, I'm carefully. A sign of Ace Adventures. It's my pleasure today to talk to you about artificial intelligence. Its growth opportunities and challenges, given how fast is grown in the past were seeing that AI is making all kinds of headlines. We're seeing that AI has beaten the best human game. Players, AI has passed all kinds of exams, and AI has even being able to generate natural language in surprising intelligence. And finally, AI has conquered a fifty-year-old biology Grin Challenge,

bro, protein folding. So how long will this continue? How many of human capabilities will? I be able to take over or will AI in human work with each other symbiotically? Well, let's first. Look at how a I can essentially, replicate human hands and feet, and that's really robotic. Working. And we've seen traditional robotics that are have really robots on wheels, that might say, welcome as a, as a product and Zen products, like that are kind of cute. But in the more recent, five years, robotics have become real. Here are some examples you see, on the upper left is a soft robot

head that you can pick up something like an egg yolk even better than people and then on the upper, right is a giant robot that doesn't look like a robot because it's not humanoid, but it's really more. Like the laboratory is a fully automated self running. Laboratory that can be applied to something like covid. NM. Cannot only be safe because you don't have to humans in the middle of Laboratories playing with potentially toxic types of materials. And it's predictable humans, make mistakes robots like this kind

of dumb, and also it works 24 by 7. So when you come back that with Scientists or to a terrific Spirit, then you can do all kinds of experiments from a research and or just passed a bunch of patients on whether they have covid-19 thing else. So no longer will we need a lab technician, will have a much more effective accurate and lower-cost mechanism to do that. With revivex the third example, lower left is autonomous farming vehicles. These are harvesting cars that will drive in the farm and basically Harvest apples

oranges or any fruit. And of course, drone can do the seedings and then the fertilizing, and then spreading the inspector side. So farming is very rapidly, becoming autonomous the lower right-hand side. We see Economist loading vehicle or a forklift. We all know that in big Warehouse is where there's Amazon warehouse or a warehouse with a large company. There are people who need to move boxes around for things that that are needed to as packages or to ship within the company. But actually that is very routine and

repetitive work within a constrained environment much easier than autonomous driving. So, we have a number of companies that have already built very powerful, inexpensive predictable again. 24 by 7 forklifts that work within the factory or the warehouse. Let's take a look at the second area where a, I using computer vision is replacing human eyes. And Humans can see things through our eyes, a lot of our input comes from our eyes. So with a computer with AI or we can have computer vision as an

algorithm that takes pixels as an input and recognizes objects. So the input would be potentially through a camera but it could also be camera and sensors. So here again, are four examples of a I'm working with cameras and sensors able to do things equal or better than human eyes. On the upper left is an apparel Factory that uses AI to check all the shirts that it makes to make sure the colors. Correct. And the labels are correct. The sizes are correct. So it saves about 80% to 90% of the human

cost. At this point, you still need a human to straighten out the shirt to before it inspected. But that can't be all the men that later as well on the upper right. Are we seeing an Electronics Factory where a eyes being used to check on blemishes or imperfections or mistakes on a phone on the lower left is a similar case but we're examining something. Much bigger, namely motor assembly again, checking for errors there or problems or things that might cause a potential safety issue in the future.

And then on the lower, right? We see a i reading lung scan weather is for covid Organics. ER machines can recognize objects including tumors much more accurately than people. So with these tasks, they're also somewhat a natural for people to do. They're not things that where we learn to do as we grow up all so they can wear out people's cell site and vision. So it's wonderful that way I can do it, more accurately and more cheaply and more reliably The third example, using a high to replace human brains. So before you get

all scared about what that's going to do to all of us. This is actually only a very small part of the human. Brain is only repetitive routine tasks. A lot of our brain is still a mystery to us. So, here are four examples on the upper left is a, I being used in our automated trading. Of course, you know, that a lot of the human trading is already replaced by automatic trading and Polly Taylor trading, but that's really not funny. It's still largely the human brains that drives the most of the trading Behavior. Example, you

or your fund manager. But we are seeing now and is able to absorb so much more data then people so that it can make decisions that people probably don't have enough data to make. So I think, increasingly secondary Market trades, will be taking over more and more by Artificial Intelligence on the upper, right. You see RPA or robotic process Automation and what it is is a smart computer that's installed on someone's computer, who's doing a lot of routine work. So what is the worker on whether workers doing free sample? If you're in

the expense report department or the human resources department doing recruiting, you're probably not replying routinely to people about their expense reports are about their job application, you're probably setting up payments for the two for the Processing or setting up interview scheduling. All of that is very routine. So AI algorithm, the RPA can sit on your computer and watch it work for a while and start to figure out parts of your work that it can do on the lower left. You see basically a chatbot that's able to

help people who have covid and instead of having people help, you can now have a chatbot, do 24 x 7, and I can also be trained from Human transactions on and on the lower, right. You'll see something that combines our brains and as well as eyes and our hands, which is autonomous driving. But autonomous driving, that is driving in a way that's as good as people aren't any wrote. Well, that is really complex, soapy a lot of people view autonomous-driving, as one of the grand challenges in the next 5 to 10 years because it is a

complex area. Do you have to see? You have to ask that, you have to think, and you have to react and you have to deal with complex issues such as how to get me from play state to place. Be safely, unsafely for myself, as well as all the pedestrians and other drivers, I encounter, and also don't hurt the car. So these priorities are a lot more complex. And also you have to understand, not only the rules of the road, but The Pedestrian out there, whether he, or she is likely to Cross or actually headed home in the other direction. And then, observe, and predicting human

intention. That's against something more difficult. So, as I just found with price, a lot of data in order to work well. So naturally the first wave is an internet wave, that is the Amazon or Alibaba Google and sense of the world. Are you able to collect a lot of data and use your internet application, using your data and your own clicks as a way to learn to do things. So Amazon will show you more things that you click on it and some it will get better because of it. The second way it is really over

businesses that happen to have a lot of data. For example, thanks can use it for credit card, fraud detection. Insurance company can use it to rate, insurance applications and all of that is trained based on the outcome. So a, if you can, if you know which are the Frog fraudulent transactions, you can detect fraud, if, you know, which users have defaulted, you can detect, likely defaulters. So if you have a lot of data weather is in public policy or apple in public sector, or education or health care or retail are all of that can be aggregated

and used to make smart predictions. The third wave is caused by Give examples of computer vision and computer speech. So understanding your environment, they have the ability to hear and see and recognize and understand. So that's something a lot of a, I can do them all week. We see that even our phone a has cameras and it has sensors, it has microphone, you can talk to it, it can see. So we're seeing that deployed in more and more situations so that our homes will become smart, our cities will become smart as sensors are

spread everywhere. You probably heard the term iot internet of things and these are effectively sensors spread everywhere in the past iot didn't quite take off. But now you can collect all the data and you say hi to process the data. Make sense out of it, make smart decisions, and I think that will cause all flowers to bloom in the area of perception. Economist Ai and we've also seen some examples robots autonomous vehicles, smart forklifts and the cars and arms so that will continue to expand further

and and able to really take over a lot of the blue color work. So you can see that these four ways, I really impacting the industry's under the four ways and they will mature. The more data the better they get and all four waves will improve over the next decade or two and able to do a lot of things for us. So that brings us to a key question, which is, as AI becomes more and more powerful. Buy more data. What type of jobs might they hide this place. So as we've described a, I spent the first displays the repetitive jobs and that's probably within the next 5

years then it will replace the more routine jobs and then it will go after the optimizing jobs. Don't worry party thinking but just requires checking out Radiology picture and MRI or CT and determining whether someone has cancer or not. So that would be an optimizing job. What remains safe are jobs that are really complex that require strategy and the mythical thinking reasoning across domains common sense and then jobs that are creative jobs that requires you to come up with new ideas and no one has ever had before.

And when we think about The AI taking over some of the jobs at tasks. It's really not just the blue collar. Many people think all the factory worker, the assembly line, those will go first or yes, those jobs are endangered, but the white collar jobs are also potentially endangered. I gave the example of robotic process automation, you can imagine if an HR department starts using that and it's being used for collecting resumes and once that works, well, some people are replaced. It can go on to do, email, communication with the

candidates. It can set up the interviews, it can coordinate the feedback, it can help facilitate decision-making and even do an offer negotiation once Adventure stand two candidates. So once HR departments uses a half a recruiting than it can think about using a AI for, for training for orientation, for performance evaluation. And when HR department, Amore. I said apartment, like Finance legal, Marketing sales, customer service, they will follow. So I will not be immediate, it will take some time for it to find an

entry into a company for once it goes in technologies will spread. So AI displacement is inevitable. Its gradual and eventually he is also nearly total for the purpose of taking over routine work. So as we think about us and our next Generations, many of our jobs will be potentially in danger. And and also, when we think about our children also, what are the hopes of their careers? How would they get educated? And if you have a one-year-old, how do you envision

their future, and what should they study and a house, you the future of Education, the evolved. So the answer these questions I think we should Ask an important question, which is one day I cannot do. And what I can tell you is that the AI there are several things that say I cannot create conceptualize or plan strategically objective is unable to pick out what objective should be or set its own goals or think creatively. So nor can I have any common sense or think across complex domains, but the other two

things for us to realize is that a, I also cannot feel or interact with empathy and compassion. That's something we believed in mate with human beings. So a robot trying to take care of an elderly person or making someone feel understood in hair for, that's not likely to be easy to do that. Even if a I can kind of Humans are unlikely to accept. It also has an additional point and I cannot yet accomplished complex physical work, that requires a high degree of dexterity or very precise, hand-eye coordination,

and deal with unknown and unstructured spaces, especially ones that it hasn't seen before. So putting their sweet Dimensions here. Let's put the first two dimensions on the picture we've seen the x-axis before, going from repetitive tasks, all the way to create a task and more to the right of the human more to the left are for the AI to do. But we also talked about the 2nd Dimension which is about empathy and compassion. So let's add that as a second axis, the y-axis at the lower part with me, there would need to be less

of empathy and compassion higher would be more empathy and compassion. So we now have to buy that all the job. Tossed into four quadrants on the lower left, one's the low creative and the low compassion. Of course, we would expect these to be done gradually by those are the jobs that we talked about replaced in the first parts of this talk, those are really the only guard jobs. Seriously endangered, we move to the lower, right? These are maybe lower compassion required but weep are high degree of productivity like a scientist. Well, that's actually an area where

there's a great opportunity for human AI, symbiosis for AI to assist the scientists example by proposing possible, compound's, chemical compounds, and proteins and generating them in a Tannerite chemistry using AI in terms of what are most likely to pass clinical trials. And what are most likely to be effective for certain diseases. And then the human scientists will then select from those and do experiments. So it's really human AI working together on the upper

are the most interesting examples because they are not necessarily the most creative jobs, but they require a lot of human touch on compassion and trust between humans and humans. So that's where one could potentially take a i as an analytical engine and really wrap the human warm around it, in these kinds of people in these kinds of jobs example, the doctor's job will no longer be the doctor having to memorize and have an encyclopedic knowledge of everything about medicine

but rather doctor can rely on and analytical AI engine that can give very accurate diagnosis and prescription and treatment. But the doctor will use his or her warmth and Care in touch and win the trust of the patient in order to come up with the best diagnosis to tease out the patience of all, the presents and conditions family issues, and them other issues, and get the patient to talk and to trust and visit the patient at home. Give the patient confidence to that, he or she will recover. And all these things

will actually be very helpful to the patient. Is actually quite lacking in our Healthcare systems but if the doctors can move from attempting to be a i and memorize all the new new treatments and drugs going into more of a compassionate care, giver letting it do what it does best with numbers. I think it's a transformative impact on the physician profession and the same can be strapped elated to teachers and many other jobs. So again, it's a great. Symbiosis symbiosis of a, I doing what it does, fast and letting people do, what only people can do. And of

course on the upper right with high creativity and compassion. That's what humans will excel. So there you have. It is a blueprint of how humans and a I can and we'll work together. So while we're seeing increasingly a I can do more and more tasks. There are still many things that only humans can do. So let's move to the last topic. In light of the still significant challenges of AI, displacement of jobs, or some would call Technologic technologically driven. And then play, what are the solutions for us? And for our children,

eye outline these in three categories, in order to survive and thrive in the AI economy, I would say we have to relearn recalibrate and build a new Renaissance. So, what do I do? I mean by we learned, we learned is the process for the people who are this place from routine, work to learn work. That is not this placeable. So I think vocational schools need to redesign their curricula to consider the fact that auto mechanics are going to be needed less and less plumbers may be a little

safer but nurses will be a job that will increase a lot. So this Reality. She needs to be made. The second area is recalibrate. I think a lot of jobs, we really need to creatively. Think about how to recalibrate them. We need to make sure all the professionals are not just doing the same job they've been doing but are able to learn AI based tools so that as we discussed a, I buy a scientist. Should USA itools, AI based medical scientist should you I use AI based generation programs are there will be a, I.

Being used all kinds of places for people who are artists or writers and the future, doctors will need great. AI diagnostic tools, the future of teachers will need to rely on AI for helping each student two To get a customized education and an entertaining education by having one-on-one AI to student kind of customized teaching. But then the teacher will change the job into one where the teacher is more of a mentor to help the child to grow into someone

that fits into the AI, me. So, we also think it will be many new jobs that will be created and we don't quite know what they are, but we should definitely be on the watch for what they can't vote for the opportunities that will arise the last opportunities. It's the most fun, I called that Renaissance because it's really celebrating. The human creativity compassion and Humanity. Just like in the Italian cities, that's really the merchants funded the Renaissance on a. I will create a phenomenal amount of wealth that can fund a new Renaissance way. I will

inject, System flexibility people can perhaps work a few hours, follow their passion, develop their talents, really go after and become the people. They really wanted them. And and with more time, I think there will be people who could be part-time artist, sculptors writers, photographers. And also Educators, I think freed from the drudgery of having the grade homework, would you say I will do. The teachers will be able to unleash the energy to design lesson that encourage curiosity, creative thinking and the principal thinking. So this Renaissance, I think it's the most

exciting and the fun part. So, lastly, just to conclude, I think we had, we're definitely seeing AI making a huge impact in the society and I will do routine jobs for us. And that means we have to figure out ways to transition people along and also make sure that it people Are where they need to be looking at a new AI are up where the most important skills are different than before. So, to conclude on this talk, we anticipate a, I will dramatically change the world at first. It looks

like a I will take away a lot of jobs and tasks, but when we analyze that we actually see what day I can and will do is take away the routine jobs. And while even that place has significant challenges that for us to really have to think about the three ours until or how to overcome the next 15 to 20 years, where routine jobs are being taken away. But if we look a bit beyond that, be on the 15, and 20 years, of what we see is that way, I will have truly liberated us from having to do routine work and really be

able to focus on the exciting. The fun that passionately And these after all are the reasons why we came to this earth. Thank you.

Cackle comments for the website

Buy this talk

Access to the talk “AI and the Post Work World | SXSW 2021”
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free

Ticket

Get access to all videos “SXSW 2021”
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Ticket

Similar talks

Joost van Dreunen
Investor, author of One Up, board advisor, lecturer at New Breukelen
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Sebastian Tomich
SVP, Global Head Of Advertising, Marketing Solutions at The New York Times
+ 3 speakers
Monica Drake
Assistant Managing Editor at The New York Times
+ 3 speakers
Danielle Betras
Executive Strategy Director at The New York Times
+ 3 speakers
Marc Lavallee
Head of R&D at The New York Times
+ 3 speakers
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free
Alexander Vindman
Doctoral Candidate at Johns Hopkins School of Advanced International Studies
+ 3 speakers
Igor Jablokov
Founder and CEO at Pryon Inc
+ 3 speakers
Michael Kanaan
Author at T-Minus AI
+ 3 speakers
Nate Yohannes
Chief Technology Officer at Microsoft
+ 3 speakers
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free

Buy this video

Video

Access to the talk “AI and the Post Work World | SXSW 2021”
Available
In cart
Free
Free
Free
Free
Free
Free
Free
Free

Conference Cast

With ConferenceCast.tv, you get access to our library of the world's best conference talks.

Conference Cast
647 conferences
26477 speakers
9839 hours of content