>> From Times Square, in the heart of New York City, it's theCUBE. Covering Imagine 2018. Brought to you by, Automation Anywhere. >> Welcome back everybody, Jeff Frick here with theCUBE. We're in Manhattan, New York City, at Automation Anywhere's Imagine Conference 2018. About 1,100 professionals really talking about the future of work bots, and really how automation is gonna help people do the mundane a little bit easier, and hopefully free us all up to do stuff that's a little bit more important, a little higher value. We're excited to have our next guest, he's Craig Le Clair, the VP and Principal Analyst from Forrester, and he's been covering this space for a long time. Craig, great to see ya. >> Yeah, nice to see you, thanks for having me on. >> So, first off, just kind of general impressions of the event? Have you been to this before? It's our first time. >> Yes, I did a talk here last year, so it was a little bit smaller then. There's obviously more people here today, but it's pretty much, I think it was in Brooklyn last year. >> It was in Brooklyn, okay. >> So, this is an upgrade. >> So, RP Robotic Process Automation, more affectionately, probably termed as bots. >> Yeah. >> They're growing, we're seeing more and more time and our own interactions with companies, kind of on the customer service side. How are they changing the face of work? How are they evolving as really a way for companies to get more leverage? >> Yeah, so I'll make one clarification of your sentence, and that's, you know, bots do things on behalf of people. What we're talking to in a call center environment is a chat bot. So, they have the ability to communicate or really, I would say, attempt to communicate with people. They're not doing a very good job of it in my view. But, bots work more in the background, and they'll do things for you, right? So, you know, they're having a tremendous effect. I mean, one of the statistics I was looking at the other day, per one billion dollars of revenue, the average company had about 150 employees in finance and accounting ten years ago. Now, instead of having 120 or 130, it's already down to 70 or 80, and that's because the bots that we're talking about here can mimic that human activity for posting to a general ledger, for switching between applications, and really, move those folks on to different occupations, shall we say. >> Right, right. >> Yeah. >> Well it's funny, Jeff Immelt just gave his little keynote address, and he said, "This is the easiest money you'll find in digital transformation is implementing these types of technology." >> Yeah, it's a good point, and it was a great talk, by the way, by Jeff. But, you know, companies have been under a lot of pressure to digitally transform. >> Right. >> You know, due to really the mobile, you know, mobile peaked around 2012, and that pushed everyone into this gap that companies couldn't really deal with the consumer technology that was out there, right? So then you had the Ubers of the world and digital transformation. So, there's been a tremendous focus on digital transformation, but very little progress. >> Right. >> When we do surveys, only 11% are showing any progress at all. So, along comes this technology, Robotic Process Automation that allows you to build bots without changing any of the back end systems. There's no data integration. You know, there's no APIs involved. There's no big transformation consultants flying in. There's not even a Requirements Document because you're gonna start with recording the actual human activity at a work station. >> Right. >> So, it's been an elixir, you know, frankly for CIOs to go into their boss and say, "You know what, we're doing great, you know, I've just made this invoice process exist in a lot better way." You know, we're on our path to digital transformation. >> And it's really a different strategy, because, like you said, it's not kind of rip and replace the old infrastructure, you're not rewriting a lot of applications, you're really overlaying it, right? >> Which is one of the potential downfalls is that, you know, sometimes you need to move to that new cloud platform. You don't want, to some extent, the technology institutionalizes what could be a very bad process, one that needs to be modernized, one that needs to be blown up. You know, we're still using the airline reservation systems from 1950s, and layers, and layers, and layers and layers built upon them. At some point, you're gonna have to design a new experience with new technology, so there's some dangers with the seduction of building bots against core systems. >> Right, so the other thing that's happening is the ongoing, I love Moore's Law, it's much more about an attitude then the physics of a microprocessor, but you know, compute, and store, and networking, 5Gs just around the corner, cloud-based systems now really make that available in a much different way, and as you said, mobile experience delivers it to us. So as those continue to march on and asymptomatically approach zero and infinite scale, we're not there yet, but we're everyday getting a little bit closer. Now we're seeing AI, we're seeing machine-learning, >> Yes. >> We're seeing a new kind of class of horsepower, if you will, that just wasn't available before at the scale it's at today. So, now you throw that into the mix, these guys have been around 14 years, how does AI start to really impact things? >> It's a fascinating subject and question. I mean, we're, at Forrester, talking about the forces of automation. And, by the way, RPA is just a subset of a whole set of technologies: AI, you mentioned, and AI is a subset of automation, and there's Deep Learning, is a subset of AI and you go on and on, there are 30, 40 different automation technologies. And these will have tremendous force, both on jobs in the future, and on shifting control really to machines. So, right now, you can look at this little bubble we had of consumer technology and mobile, shifting a lot of power to the consumer, and that's been great for our convenience, but now with algorithms being developed that are gonna make more and more decisions, you could argue that the power is going to shift back to those who own the machines, and those who own the algorithms. So, there's a power shift, a control shift that we're really concerned about. There's a convergence of the physical and digital world, which is IOT and so forth, and that's going to drive new scale in companies, which are gonna further dehumanize some of our life, right? So that affects, it squeezes humans out of the process. Blockchain gets rid of intermediaries that are there to really transfer ideas and money and so forth. So, all of these forces of automation, which we think is gonna be the next big conversation in the industry, are gonna have tremendous effect societally and in business. >> Right. Well, there's certainly, you know, there's the case where you just you can't necessarily rescale a whole class of an occupation, right? The one that we're all watching for, obviously, is truck drivers, right? Employs a ton of people, autonomous vehicles are right around the corner. >> Right. >> On the other hand, there's going to be new jobs that we don't even know what they're gonna be yet, to quote all the graduating seniors, it's graduation season, most of them are going to work in jobs that don't even exist 10 years from now. >> Correct, correct, very true. >> And the other thing is every company we talk to has got tons of open reqs, and they can't get enough people to fulfill what they need, and then Mihir, I think touched on an interesting point in the keynote, where, ya know, now we're starting to see literal population growth slow down in developed countries, >> Yes. >> Like in Japan is at the leading edge, and you mentioned Europe, and I'm not sure where the US is, so it's kind of this interesting dichotomy: On one side, machines are going to take more and more of our jobs, or more and more portions of our job. On the other hand, we don't have people to do those jobs necessarily anyway, not necessarily today, but down the road, and you know, will we get to more of this nirvana-state where people are being used to do higher-value types of activities, and we can push off some of this, the crap and mundane that still, unfortunately, takes such a huge portion of our day to day world? >> Yeah, yeah. So, one thought that some of us believe at Forrester, I being one of them, is that we're at a, kind of, neutral right point now where a lot of the AI, which is really the most disruptive element we're talking about here, our PA is no autonomous learning capability, there's no AI component to our PA. But, when AI kicks in, and we've seen evidence of it as we always do first in the consumer world where it's a light version of AI in Netflix. There's no unlimited spreadsheets sitting there figuring out which one to watch, right? They're taking in data about your behavior, putting you in clusters, mapping them to correlating them, and so forth. We think that business hasn't really gotten going with AI yet, so in other words, this period that you just described, where there seems to be 200,000 people hired every month in the ADP reports, you know, and there's actually 50,000 truck driver jobs open right now. And you see help-wanted signs everywhere. >> Right, right. >> We think that's really just because business hasn't really figured out what to do with technology yet. If you project three or four years, our projections are that there will be a significant number of, particular in the cubicles that our PA attacks, a significant number of dislocation of current employment. And that's going to create this job transformation, we think, is going to be more the issue then replacement. And if you go back in history, automations have always led to transformation. >> Right. >> And I won't go through the examples because we don't have time, but there are many. And we think that's going to be the case here in that automation dividends, we call them, are going to be, are being way underestimated, that they're going to be new opportunities, and so forth. The skills mis-match is the issue that, you know, you have what RPA attacks are the 60 million that are in cubicles today in the US. And the average education there is high school. So, they're not gonna be thrown out of the cubicles and become data scientists overnight, right? So, there's going to be a massive growth in the gig economy, and there's an informal and a formal segment of that, that's going to result in people having to patch together their lives in ways they they hadn't had before, so there's gonna be some pain there. But there are also going to be some strong dividends that will result from this level of productivity that we're gonna see, again, in a few years, cause I think we're at a neutral point right now. >> Well, Amara's Law doesn't get enough credit, right? We overestimate in the short-term, and then underestimate the long-term needs affect. >> Absolutely. >> And one of the big things on AI is really moving from this, in real time, right? And all these fast databases and fast analytics, is we move from a world where we are looking in the rear view mirror and making decisions on what happened in the past to you know, getting more predictive, and then even more prescriptive. >> Yes. >> So, you know, the value unlock there is very very real, I'm never fascinated to be amazed by how much inefficiency there still is every time we go to these conferences. (Craig laughs) You know we thought we solved it all at SAP and ERP, that was clearly-- >> Clearly not the case. Funny work to do. >> But, it's even interesting, even from last year, you mentioned that there the significant delta just from year to year is pretty amazing. >> Yes, I've been amazed at the level of innovation in the core digital worker platforms, the RPA platforms, in the last year has been pretty amazing work. What we were talking about a year ago when I spoke at this conference, and what we're talking about now, the areas are different. You know, we're not talking about basic control of the applications of the desktop. We're talking about integration with text analytics. We're talking about comp combining process mining information with desktop analytics to create new visions of the process. You know, we weren't talking about any of that a year ago. We're talking about bot stores. They're out there, and downloadable robots. Again, not talking about last year at all. So, just a lot of good progress, good solid progress, and I'm very happy to be a part of it. >> And really this kind of the front end scene of so much of the development is manifested on the front end, where we used to always talk about citizen developers back in the day. You know, Fred Luddy, who was just highlighted Service Now, most innovative company. That was his, you know, vision of Citizen Developer. And then we've talked about citizen integrators, which is really an interesting concept, and now we're talking about really citizens, or analysts, having the ability via these tools to do integrations and to deliver new kind of work flows that really weren't possible before unless you were a hardcore programmer. >> Yeah, although I think that conversation is a little bit premature in this space, right? I think that most of the bot development requires programming skills today, and they're going to get more complicated in that most of the bot activities today are doing, you know, three decisions or less. Or they're looking at four or five apps that are involved, or they're doing a series of four or five hundred clicks that they're emulating. And the progression is to get the digital workers to get smarter and incorporating various AI components, so you're going to have to build, be able to deal statistically with algorithm developments, and data, and learning, and all of that. So, it's not.... The core of this, the part of it that's going to be more disruptive to business is going to be done by pretty skilled developers, and programmers, and data scientists, and statistical, you know, folks that are going through. But, having said that, you're going to have a digital workforce that's got to be managed, and you know, has to be viewed as an employee at some level to get the proper governance. So you have to know when that digital worker was born, when they were hired, who do they report to, when were they terminated, and what their performance review is. You gotta be doing performance reviews on the digital workers with the kind of dashboard analytics that we have. And that's the only way to really govern, because the distinction in this category is that you're giving these bots human credentials, and you're letting them access the most trusted application boundaries, areas, in a company. So, you better treat them like employees if you want proper governance. >> Which becomes tricky as Mihir said when you go from one bot to ten bots to ten thousand. Then the management of this becomes not insignificant. >> Right. >> So Craig, I want to give you the last word. You said, you know, big changes since last year. If we sit down a year from now, 2019, _ Oh. >> Lord knows where we'll be. What are we gonna talk about? What do you see as kind of the next, you know, 12-month progression? >> You know, I hope we don't go to Jersey after Brooklyn, New York, and-- >> Keep moving. >> I see Jersey over there, but it's where it belongs, you know, across the river. I'm from Jersey, so I can say that. You know, I think next year we're gonna see more integration of AI modules into the digital worker. I think with a lot of these explosive markets, like RPA is, there's always a bit of cooling off period, and I think you're going to see some tapering off of the growth of some of the platform companies, AA, but also their peers and compatriots. That's natural. I think that the area has been a little bit, you know, analysis and tech-industry loves change. If there's no change, there's nothing for us to write about. So, we usually over-project. Now, in this case, the 2.8 billion-dollar market project five years out that I did is being exceeded, which is rare. But I expect some tapering off in a year where there's not a ceiling hit, but that, you know, you end up with going through these more simple applications that can be robotized easily. And now you're looking at slightly more complicated scenarios that take a little more, you know, AI and analytics embedded-ness, and require a little more care, they have a little more opaque, and a little more thought, and that'll slow things down a bit. But, I still think we're on our way to a supermarket and a lot of productivity here. >> So just a little less low-hanging fruit, and you gotta step up the game a little bit. >> I guess you could, you said it much simpler then I did. >> I'm a simple guy, Craig. >> But that's why you're the expert on this panelist. >> Alright, Craig, well thanks for sharing your insight, >> Alright. >> Really appreciate it, and do look forward to talking to you next year, and we'll see if that comes true. >> Alright, appreciate it, take care now. >> He's Craig Le Clair and I'm Jeff Frick. You're watching theCUBE from Automation Anywhere Imagine 2018.