United States, Watson, IBM discussed on WSJ The Future of Everything
Displacement. Yeah. New jobs that come into being that we don't know. And even for those of us who who don't have jobs, gone, the jobs change a lot. So really you. Can you talk a little bit about the what you because you've said, it's a very dramatic chain. Yeah. And how we were coming that will have been ramification. This has been my my biggest topic to want to keep coming. Actually brought this up at Davos is my third year on this topic first off because the first couple years. It was everybody was what percentage is going to change. And I said, it doesn't really matter to be honest. I mean technologies change jobs every area of looked. But this time is one hundred percent in this faster. And so if it's one hundred percent in this is what I believe leads to one of the biggest issues, I've said all different ways skills as the issue of our time because now you've got generations of people working in their job is changing in. So I may be fifty years old six year in my job. What I did. Or what I did is now going to be paid less, and I've got to add another skill to it. That's to me this big issue. So if you if you believe which I do really believe that we are going to leave people behind like in the US, Matt you cannot have a country where two sides have all the benefits of this technology history shows when there's great new technology inventions, the structure of physical infrastructure in societal infrastructure has to change it wasn't maybe so Vert. It's always like when you can look back. It's always clear and. It took longer the relaunch periods. If they were do adjust right here. I really think for us one of the biggest biggest things is. So you can't leave everyone a big swath of society feeling disenfranchise and left behind in. That's potentially. What's in front of us are right in front of us, in some cases. So because for one look all of us. We get seven thousand resumes a day, we can hire the people we need an IBM. However, the whole world doesn't have university degrees. The whole world does not have PHD's. And so you look at what caused Brexit. What caused some of this unrest is issues of people saying, hey, look my future. I'm not sure this looks much better out here. So what's in this technology, any of the they're not gonna make it easier for me? So what I'm a really strong proponent of is that this won't be sowed by government, but business in government together can make a huge dent. But you've got to believe a couple of things you have to believe some different paradigms that you'll hire people preach. Jobs on skill versus diplomas or degrees gate. I've I have a whole the whole company of university impeach dis. But there are jobs that can be done different ways. You have to believe that there are different pathways to get a good paying job other than a college degree, we've made some inroads than I think other companies have in there's great hope for this. And I'm not deceptive mystic Optimus based on fact. And so, and then I so you have to have different pathways. And I think you have to remove a stigma that it is a bad thing to be people say blue collar, white collar, we've come up with this title of new collar, meaning no problem. You don't have to have university degree, you can have these contemporary skills, and you could be a cyber expert not necessarily with a university degree. So couple ideas. I would share one of them would be all of us do reskilling. I do half a billion years. So that that goes without saying any any company, I should prepare you to be employable. I don't guarantee employment, but I should prepare you to be employable that. It should be that should be a commitment. How do you? I mean, do you have a lot more mandatory training and reskilling either within the company that yes? And I talk a lot with our clients on this one of the things is and I believe everyone should do this first off we used to do skill planning. Once a year, we do it every quarter. Now, we also go through what skills are in surplus in the market in what skills are leading edge skills or lagging in everyone knows where their skills fall on that matrix in you know, that if you're in surplus in a in a skill that is lagging you gotta get yourself to another area. This idea that people don't need to be left behind in. By the way, age doesn't matter. Our data proves it. We've retrained the whole workforce it is irrespective of age their ability to move forward into these new areas just goes to the point a bit a bit. How many companies and customers you work with all kinds of companies all sizes get a I get what it is. What's the gap between the perception of what it is the reality of what a community groom? But they okay. Well, they've formed or. Or are we just sort of filled with size Seifi images that aren't accurate? Well, I think it's a shame that some of the rhetoric is about how bad it could be like any technology, it it can do great things and it can be used for challenging things in. So I this is why I think it's really important that we all bring it in safely into the world right in that's true with any kind of technology. So when you think of what you hear on one end is what people call generally, I like can imitate a human actually imitating human most of the I know today people would call narrow it learns to do one thing in one domain, that's kind of the two spectrums. That's where we're at today, but we're moving into something called broad AI. And this is where we're doing a lot of work, which says, you know, I knew how to do one thing in a domain. I can learn the next Jason thing. And and that would be as that example, like we're just talking about the roofs in the hail, right? I know roofs, I've seen I've seen hail. Okay. Now, I can see wind damage. I can learn that my. Self in. So it's an adjacent task. I learn in that that is going to make it far more valuable as well. So, but you're still that idea that equal to human is decades, if it all away here still so it's a shame because it causes fear mongering right about amongst people at that other end so US where are people today? We've done I think last count we try to count them twenty five thousand engagements. And so I kind of feel we've we've gone through a lot of EROs and learned a lot of lessons learned about where clients are, and I would say that a lot of clients are to stage where they have random acts of digital in a I or all over, and if those clients in the room mo- many identify with that thought in its this belief that gotta kick it live behind bitter try, and I got a million flowers blooming in their number one issue is they can't scale, and they can't necessarily prove that the same data was used in the same places into this model. You know, I'm a Bank dry. No one I made that. Decision what the model looked like in the algorithms in? So they're hitting that Rohbock. So on the other hand, I've seen clients now go from process to process, and again, they start with customer service and customer service HR than any kind of knowledge worker that needs expertise would be the next thing that I see rolling through. But here would be the lessons learned I one. I mentioned people spend eighty five percent of their time on trying to get the data. Just find data that they want us to train. So most people I there's kind of funny saying you can't have AI without a you can't have a without an information architecture. Okay. Boring as it sounds. But if you don't get this put together, this is a long road second one some of the early mistakes if you just try to put in and you don't rethink how you're gonna do your work. We learned that early on with doctors. If you try to give them a tool and say, no, you keep doing everything to add this to what you're doing. They like realize I see a patient, eight minutes. Okay. This is not going to be able to be. Avid here. So when we rethink how they do their work. That's the second thing. People had get onto was rethink how they did their work. The third thing is you gotta be able to explain the eye and can't be a blackbox bias. So we just released a tool that you could take anybody's AI and put it through that looks for biased doesn't have to be Watson can be anything. Google tensor flow, put it through in the init- can show you where it has potential bias meeting it 'cause without telling it with biases. It's looking for. Why is this always happening with these set of attributes like, so why did you always give this group and increase in this group? So it's it's and you could also tell it will for gender look for this, look, but you can also let it say, what looks like potential bias to you. You need these tools. Right. So you've got that's the third element. I've seen this is what I think is interesting particularly with your experience there, which is we've think of watt. I mean, I mean, you know, we have these moments it bursts of the consciousness. No Google wins. The game of go or Watson wins. Chest. Part of your experience Watson has been sort of accumulation of knowledge experience learning adding functionality. It's kind of a constant it is that's a CS. We when you look at what's next on these things cannot Allen. What's next? Maybe if it's okay, it's a little bit related. The idea of what is what you would have seen. We've been working on something Watson called debater. And we let it interact with everybody. I know that you might not like what it's going to tell you can it's able to do is debate and argument. Well, I think if you talked to any home device and talk to it for four minutes. What does it do back if for minutes? Okay. So what this next levels being able to teach that it can actually comprehend listen comprehend in then go search and make an argument in. So we the first day were there. There was should gambling illegal. I, you know, not legalized or not legal in in Las Vegas. We pick some interesting topics every day to in the pro and the con argument, that's one of the things next. The next one of the next other things we're working on right now, we had just announced a cooperation with Michael J, FOX foundation on Parkinson's. But we've already done the work. It's an idea that there's something called deep data coming out there. And this is one estimates that less than one percent of all the data in the world's actually collected and used less than one percents collected in us. And so what we've been working on with sensors on your fingernails that that they really can get the slightest motions in its early. The ability to predict Parkinson's because of it in the I we built to do it as well as schizophrenia. So I think these are the kinds of this idea of deep data meaning will think about it. Nobody's collecting the vibrations off at your fingernails right now. So that it's massive when you think of that amount of data, even what we're doing with the weather company. Now, we just announced new weather prediction that an hour in advance. Because we are now able to say we can get I won't teach about whether forecasting. That's a really interesting topic to be honest, but every three kilometers we can now accurately get the weather, and we're spoiled in the US of the ability the models that are in the US. There do not exist around the rest of the world they do. Now what we just did. But you could predict turbulence in our in advance. I mean, just think how many times are you on airplane? When someone says I'm gonna look for smoother air, you know, in this idea, but that has repercussions weather's number one thing that impacts every every number one Xtra factor that impacts the economy is whether and so the ability of what? That has to do with hospitals with disease with forecasting for retailers. All of what you get better. That's that's what I call deep data that when we're collecting off of an airplane every five seconds or the barometric reading off your cell phone if they can do it. We're now ingesting those with your permission. We're ingesting those that's deep data. So this idea of deep data, and we're we're heading with the, you know, things like the debater is this idea of broad data or the ability to hey, I've learned one task here. I'm going to jump over. And learn another without much more new data is where the future goes on. Jay. Thank you forget. My pleasure. Thank you. IBM's? Ginny remedy and the Wall Street Journal's editor in chief Matt Murray. The future of everything is a production of the Wall Street Journal. This episode was produced by Anthony green with help from Rebecca dough wineman and George downs. Stan perish is the editor. In chief of the future of everything. Our technical director is Jacob Gorski. Thanks for listening. I'm Jennifer strong.