Russ: Welcome back to The BusinessMakers Show coming to you today from the campus of Rice University where my guest is Dr. Moshe Vardi, the George Distinguished Service Professor in Computational Engineering and Director of the Ken Kennedy Institute for Information Technology here at Rice University. He’s the recipient of numerous technology and scientific awards and author and co-author of numerous books and over 500 papers; Moshe welcome to The BusinessMakers Show.
Dr. Vardi: It’s a pleasure; it’s good to be here.
Russ: You bet. So you’re pretty well known for this category called the future of work and the future of jobs based upon the impact that the digital economy, artificial intelligence, robotics is having on our workforce; how did you get in to that path?
Dr. Vardi: Well for the biggest portion of my career I was just a technologist. I viewed my job as I do a lot of technology, other people deploy this, and thinking about social consequences for this way of sociologists and humanists, that’s not the job of a technologist. And also I thought that the real artificial intelligence seemed to be very far in the future; we were making good progress and I was happy with the progress I was making but the things that people talk about, all of those things are very, very far off. And then the first inkling that something might be different happened about 20 years ago… IBM invited me to watch Deep Blue playing against Kasparov.
Russ: The chess game.
Dr. Vardi: And I watched – I saw the first game and Kasparov won the first game and I remember telling myself well one day computers will win but the time has not come yet. And I left the tournament and I missed watching a very historical game, Kasparov lost the second game and wow, a big milestone that people thought was far in the future, you know, that milestone has been reached.
Russ: That’s impressive. I remember very well when it happened, I was with IBM at the time too and it’s funny your comment about artificial intelligence, thinking it was far into the future; I was even kind of dissing it for a while, calling it imaginary intelligence.
Dr. Vardi: Yeah, people always talk that this is a field that over-promised and under-delivered. In fact when I did my PhD it had a bit of an unsavory reputation. I mean people considered it as a used car salesman kind of thing.
Russ: But now I also know that what also impacted you was – was really 3 games that the computer beat the human, go on to the second one.
Dr. Vardi: So the second took a while so it’s about 14 years later so now we are in 2011. And now again, IBM – it’s not the same IBM of the time, this is a company that’s not the same as last time but still put a very impressive engineering accomplishment into developing Watson and it goes on to beat two great players on Jeopardy and Jeopardy is different than chess. Chess is kind of almost people thought a mechanical game; if you just compute hard enough you can do that. But Jeopardy was about human culture and history and that language, you need to understand it and integrate many different sources with contradictory information, so that was an amazing accomplishment.
Russ: Well I totally was impressed too, blown away, and even I – think I mentioned this to you earlier – had Sridhar Sudarsan, an IBM – a PhD in computer science with IBM that’s with IBM Watson on the ecosystem – talk about it and a guest on the show. It just blew me away and then just the other day on 60 Minutes I was watching the way that cancer research is working and it made me think how different it is. In the old days you could do some significant computation but you had to standardize and scrub the data that was garbage in/garbage out and it’s not like that at all now. I mean they input hundreds of thousands of documents on curing cancer and now use that as opposed to sitting around a table with 10 experts.
Dr. Vardi: So in fact shortly after the winning – of Watson winning Jeopardy we contacted our friends at IBM and we said we would like to host a technical event here at Rice.
Dr. Vardi: And we invited people from Rice, people from the Medical Center and we had a nice half day symposium. And IBM told us make sure the people from MD Anderson show up, make sure people from MD Anderson show up. And we did and, you know, it was an open event and they follow it up with signing an agreement with MD Anderson.
Dr. Vardi: Just give it to us. We will scan, we will do the category cognition, we will do the analytics and now do you see in the news that many examples were looking at the result of mammography and which one is better than doctors in identifying cancer cases. They also – I just recently read another article about very, very difficult cancer cases; somebody has a cancer, it metastasized, they’ve tried different therapies and nothing works – you know, thousands of articles published on a regular basis and who knows, every day.
So Watson came up with some therapy that somebody investigated and so some people are gaining months of life because you can come up with – you know, sometimes in cancer it’s just a battle between the therapy and the cancer. And so in fact, just a couple of weeks ago I was in a medical conference and the title of the session was “Must the Physician be Human?” So people now in the medical field are thinking wow, must the physician be human? What is the role of a human, what is the role of a machine in healing people?
So this was 2011 so it was kind of the third big kind of milestone. And this was the point at which I started thinking very, very hard about wow, this is – the future is now so to speak; this is not something in the future. Now, what are going to be the consequences? And in 2012 I wrote the first article I wrote on the topic, it was titled “The Consequences of Machine Intelligence” where I speculated what the consequences for labor may be and that article kind of went viral. And in 3 months I had more references to that article than I had to all of my papers published over the previous 35 years.
Russ: So that’s when you decided wow, this must be my specialty from here on in?
Dr. Vardi: And so I kind of started dedicating more and more part of my time in just digging and understanding what’s going to happen.
Russ: Wow, wow.
Dr. Vardi: And then the third kind of big turning point happened just this year earlier in the spring.
Russ: A game again.
Dr. Vardi: And again it’s a game and now it’s not IBM anymore, it’s Google, and now they have a program and it beats the Korean master of Go.
Russ: Which is a very complicated game, right?
Dr. Vardi: And Go – actually to see them play the game is something else. There only 2 pieces; I mean there are black and white, you don’t have different pieces, but the number of configurations of the game is much, much, much larger. So it was well known that Deep Blue kind of techniques by themselves cannot – cannot be used to – to – to win computer Go. And some people thought this is it, I mean we have finally found a game that computers cannot win. And Google came up with a new approach called Deep Learning where the program developed intuition how to play the game.
Russ: Intuition how to play?
Dr. Vardi: Intuition. What is an intuition? You have an intuition when you know it’s the right thing to do but if I ask you why all you say is well I…
Russ: I just have a feeling about it.
Dr. Vardi: I have a feeling about it all right? I cannot give you 1, 2, 3 rationale why I’m doing it. I have a feeling this is the right thing to do. So the computer developed this what we call Neural Network model, what we call Deep Learning models, which I would say are similar to what we would call intuition; when a computer knew what’s the right move to do in what circumstances.
Russ: Do you consider that an upgrade over Watson?
Dr. Vardi: No, it’s different from Watson, it’s different. It’s more sophisticated than Deep Blue, it’s complimentary to Watson.
Russ: So, back to the whole societal impact of it, I mean even – even kind of the medical thing that you talked about, I mean I’ve heard people say wow, the medical industry is going to be totally disrupted. Is that possible?
Dr. Vardi: So you know this is actually very interesting because – and the answer is nobody knows. So there are aspects of medicine which are very technical; think of radiology, the person who just had to read the images, and the computers are going to become better than people. Now medicine as far as the human aspect you need to be able to talk to someone in the eyes and say look it’s cancer but you have a chance, you can fight it. And it’s a very, very important part of medicine right? The bedside manner may emerge to be the main thing that doctors do because for the rest they’ll be able to use computers for the rest. So this is the 21st century, we see what’s happening.
Russ: Okay, so kind of moving in a different direction now, artificial intelligence and robotics though seems to really be having an impact on the whole manufacturing world and with significant job losses, right?
Dr. Vardi: So when we start to pay attention to manufacturing we see this has been going on since 1980, we started to accelerate in the 90s when China opened up, when we started globalizing. What did globalization do? It did not kill American industry, American industry is thriving. We’re a huge manufacturing country. We manufacture more than many, you know, then U.K. and Brazil France and Germany together.
Dr. Vardi: Combined. It’s a huge sector of the economy. But employment has been going down. It went roughly down from about roughly 20 million to 12 million today. We have lost millions of jobs in manufacturing.
Russ: And you’re attributing that to technology and artificial intelligence and robotics.
Dr. Vardi: Well go look at these factories. They have been under tremendous competitive pressure because of globalization so that is probably the way they are supposed to respond. They become more competitive, they’ve increased productivity and how do you do that? You employ fewer people and in place buy new machines. So today look at the modern car factory floor you’ll see what they call dark manufacturing floor; you don’t need lights, it’s all done by machines. I mean they turn on the lights so they can take photographs.
Russ: So blaming this on immigrants coming in is probably all wrong or blaming it on outsourcing is probably all wrong.
Dr. Vardi: So these are two different things. So one of the things that globalization did it puts more competitive pressure on American industry but it would have happened anyway. The pace might have different because the – what we know about competition is that either the Chinese compete with us or someone in Tennessee builds a better factory, thus creating pressure on everybody else. So globalization accelerated it but it’s not only globalization. Now what went on with immigrants, they have an impact; they have an impact only on the very, very low end, very low skill workers.
Russ: Well it’s interesting because when you talk about manufacturing I remember something else I recently heard about automation, robotics taking over a lot of the Chinese manufacturing as well and they’re already having layoffs.
Dr. Vardi: So we started this 15 years before them, but they have to be more competitive so they are automating, they are buying robots. In fact, the most recent thing I read is automation is getting cheaper and cheaper so you can start pushing it down to a lower and lower margin of manufacturing. They are now expected to lose millions of jobs to automation.
Russ: Man, so the numbers are mounting up here but another one that I heard you talk about is that – and we haven’t even touched on the autonomous vehicle thing yet – but I just think it’s a delight to think that that’s what’s going to be moving us around. I look at it not having to drive, being able to read the paper on the way to work and travel long distances without worrying but there’s an impact on actual jobs from autonomous vehicles as well, right?
Dr. Vardi: So if you go outside and just look at an urban setting and compare it today to the way it was let’s say 50 years ago it’s not going to be radically different. The count is going to be slightly different – you see more minivans, you see more SUVs. The next 50 years it will be completely different because the automation of driving will change the very fabric of our urban structure; it will change everything.
Russ: And how is that? Because we still have to go places, we’ll still go to the same places.
Dr. Vardi: For example, right now we have – cars are grossly underutilized; 90% of the time they do absolutely nothing. People estimate that 25% of our urban area is just for cars that are sitting and doing nothing. We’ll be able to use and utilize our cars in a much more efficient way. I mean it’s almost an expense to us just to provide parking to all the employees. Now the car will go and park itself somewhere and it will come back to pick me up. In fact it will take other people who need to drive during the day.
So we don’t know will people be willing to commute longer if they don’t have to drive; if they can sit in the car and read and work in the car? Maybe we’ll see more support because people won’t mind commuting longer. So it’s going to be a huge, huge dramatic change I think in our whole urban setting. But overall I mean the main thing to me why this is a wonderful invention is because we are lousy drivers and we keep killing each other and maiming each other and injuring each other and causing property damage – and supporting a whole bunch of industries that are only in existence because we are such terrible drivers, right – the insurance companies, the lawyers, the hospitals.
Russ: The repair – yeah, the hospital, yes.
Dr. Vardi: Yeah, a huge part of it is just the maim that we cause by driving. If you can reduce this by let’s say 90% it’s going to be enormous. So think about it from this point of view, it’s the most amazing invention that I will have seen in my lifetime until you start thinking about what it’s going to do to the job market.
Russ: To the job market.
Dr. Vardi: To the job market.
Russ: What’s it going to do to the job market?
Dr. Vardi: Well in more than half the states the job category with the most workers are drivers.
Dr. Vardi: Drivers. In 29 states this is the most common job is a driver. Millions of people are driving for a living; just people who are professional drivers – taxi and truck drivers – 4 million people. If you look at people who – where driving is an important key component of their job like postal delivery workers, we’re talking about 50 million workers. Now think of the whole – think of the whole industry that’s around to support them, all the motels and the restaurants along the way, what happens if you automate all of this?
And it’s not just this, if you automate driving you automate the ports, you automate the warehouses; Amazon is also working very hard on automating their warehouses. By now we are talking about many, many millions of jobs and a whole industry that just supports this part of the job market is going to be disrupted in a very, very significant way.
Russ: So and you kind of implied that you’ve become a sort of student in sociology in thinking about this, is it just right to conclude that it’s very unlikely that what is going to accompany this progress is, you know, new ideas, new businesses, new jobs. It almost seems like it’s impossible for anything to give that much support to the economy like we have today.
Dr. Vardi: Well what we are seeing where we’re creating new jobs is very often in the high end where somebody needs to develop the software for all of this automation, right? But it’s unlikely that drivers who lost their jobs are going to become the people who write the software; this is different people. And so we are seeing the people – economy is called job polarization – we’re creating jobs at the high end and the low end. The people in the middle that lost their jobs they can move up or they can move down; they can so call upscale or downscale. Most likely they are going to downscale. So that’s why we see a tremendous amount of economic frustration in this country.
It used to be that 97% of men between 25 and 54 used to work for a living. This is what men did. Now the umber is down to 88% which means about 1 in 8 men are not working. You extend this trend for another 35 years and 1 in 4 men are not working. The Industrial Revolution forced us to rethink how we arrange our socioeconomic life. The Great Depression forced us again to think of social setting at home. So far for the last 30 years we have not done something with that. In fact people talked about cutting away entitlement programs. Now, those candidates, nobody’s talking about cutting entitlement anymore. People realize that we have to create a safety network. We will have vigorous discussion and huge acrimonious debates on what is the right solution.
Russ: Yeah, when you really think about it too, the way that the economy – we all still think that having a real good, strong economy is good and we’ve also put this importance in consumer spending for driving the economy more than ever before; if they quit spending the economy craters. It seems like in a worst case scenario like you just pictured there will be a lot of people that won’t be spending because they won’t have money. I heard you speak recently too you mentioned these brain storming sessions on this topic and there was this – there was this idea brought up, I think it was called – the acronym was UBI.
Dr. Vardi: Universal Basic Income. Universal Basic Income
Russ: Okay, and where did – where did that surface?
Dr. Vardi: This actually turns out it is an old idea. You go back and Thomas Paine was thinking about it. I mean this is an old idea and then it came up by some economist were arguing for it and the idea was let’s get away from all of these entitlement programs and just – no more Welfare, no more many programs – just make sure that everyone gets like a basic stipend for a living.
Russ: Which is a check every month?
Dr. Vardi: A check; everybody gets a check a month; Alaska, that’s what happens in Alaska.
Russ: Yeah, well Alaska does. Well doesn’t Saudi Arabia kind of do that these days too?
Dr. Vardi: Some places they do it. And so now there was a vote in Switzerland do we do it. It lost but when you look at the vote you see the Millennials were much more in favor of it than the older people. Finland is discussing it, Canada is discussing it; it’s very controversial. It has supporters on the left and the right. It has people who object to it from the left and from the right. So this is one of the things I think we will see happening; we will have vigorous discussion about how do we arrange our socioeconomic life. And the economic aspect and what will happen to people who are not working – they have the stipend and are not working – how do they give meaning to their life?
Russ: It reminds me, I made a trip to Cuba probably 4 years ago now and I mainly went there – I just wanted to see what the people were like. I mean I knew the cars, the architecture is all spectacular.
Dr. Vardi: It’s like a time machine, a time machine.
Russ: Yeah, but the people – it kind of bothered me. There was no drive. I mean they live off of the government, I mean they still have to kind of fight to survive because the government can’t feed them completely. And then I have friends like from Norway and stuff and they love it, it’s a great country, but when they talk about entrepreneurship they say it’s not – well it’s not as risk-taking as it is here in America. And so there’s all these things about our country that I like that are probably going to be diluted somehow.
Dr. Vardi: They are going – they are going to be changing. There is going to be big change, yeah, absolutely.
Russ: Okay. Well I really appreciate you spending some time with me today. I’m going to start following you more, you’ve got a Twitter account too and you’ve got a conference coming up too that the university does, right? December – when is it?
Dr. Vardi: December 5th and 6th this year.
Russ: 5th and 6th
Dr. Vardi: December 5th and 6th here on the campus of Rice University. So I’ve been trying to promote, over the last few years I’ve been trying to convince people that this is a big issue and we got a grant to run a conference here at Rice. And we’re bringing together technologists, economists, humanists, management experts, social scientists to look at this issue from all aspects and if people are interested they can go to – the title of the conference after the name of the sponsor is DeLange.rice.edu and it’s going to be very interesting, aimed at a broad audience; anybody that cares about this issue.
Russ: December 5th and 6th?
Dr. Vardi: December 5th and 6th.
Russ: Okay, Moshe I really appreciate it, thanks a lot
Dr. Vardi: Pleasure talking to you.
Russ: You bet. And that wraps up my discussion with Dr. Moshe Vardi and this is The BusinessMakers Show.
brought to you by