Dwayne Bragonier, CPA (Canada), CITP, pays close attention to the technology trends affecting consumers and organizations. In a podcast episode recorded at AICPA ENGAGE in June, he explains how technology advances such as 5G will quickly change our world and why those trends matter for accountants.
What you’ll learn from this episode:
- The technology advances affecting the accounting profession.
- Why 5G technology will be “a game-changer.”
- Insight into a major language-related technology breakthrough.
- Bragonier’s opinion on the limitations of the internet of things.
- How professionals can manage the information overload that new technology often brings.
Play the episode below:
To comment on this podcast or to suggest an idea for another podcast, contact Neil Amato, a JofA senior editor, at Neil.Amato@aicpa-cima.com.
Neil Amato: Are there advances that are affecting finance professionals more than others on the technology front?
Dwayne Bragonier: You want me to speak to what's really happening in the fintech world with regards to automation, with regards to the AI. AI is a blanket statement for many, many different technologies and I'll be happy to go there, but we're not the only ones. The things that I'm speaking to you, with small variations, are affecting many of the other both professional sectors and industry and agricultural sectors as well.
People will talk about, no question, the deep learning concepts of AI — its capability through computer visioning to be able to look at a whole series of samples and decide from the samples really what's taking place. Many people use this analogy because — it's a story, not an analogy — because it's very effective.
Google fed a whole bunch of pictures to their AI, but they didn't identify that there was a cat in the picture. The reason they chose a cat is because it has no good shapes to it. The human face has an oblong sort of — it can search for it. A dog, because its hair is not normally as fluffy as a cat, they can see the bone structure and they can learn. But a cat isn't that way. They kept feeding it information until eventually it said, "I can see this is a common shape, and it must be the cat that you're talking about."
Those sorts of things now are coming to us so that when we apply that same sort of philosophy or that same sort of learning — the pattern recognition, the continual redundancy of going over and over millions of transactions — small little things kick out. Small little things like, "OK, this transaction took place on a Sunday after 6 p.m.," would fall out as a variable because the computer doesn't recognize it as anything but one piece of that puzzle. This puzzle is a little bit different than the other puzzles that I've investigated.
We're seeing lots of things coming from that deep learning, the recognition that we see within our applications — our software, finance, our transactional recording. We would have to have been sort of overseers of that before. A CFO was a guy that was always going, "OK, am I seeing any sort of ripples in the water, so I got to go see what's here. Is this just a little bit of wind or are we heading towards the shore or was there a rock?"
In today's world, they're looking at absolutely every variable all the time. That's just relatively new. So transactions that are recorded this morning can be spit out if there's an anomaly associated with it for quick review this afternoon. So before it's reported on, before it's relied on, before management takes a look at it, the data then is becoming purer.
There's an assurance level that now your data is the dataset that you want, and these anomalies are OK or they are corrected or they're handled appropriately. So we're getting that information on this moment decision-making taking place.
What happens is the technology handles so much volume, it's hard for us to comprehend and believe that it's true. The reason I'm saying that is that's how a lot of people will go. If I say GM's transactions today — and I'm making this up — is going to be 10 million transactions around the world that's going to affect both their production and their ERP systems, but every one of those transactions have been reviewed against all the things that GM has done over the last 10 years, it seems almost impossible, and yet it's nothing.
We don't even think about — you hit a key on your keyboard, you don't think about all the things that transpire with that one little action — what stopped, what started, how processing. And yet somehow or another when we start to talk about this deep learning, we're doubtful. Can it really be processing that much that quickly?
We're getting back to — we're getting into a stage where a continuous assurance can actually be given and spit out onto a website or to a very, very tight blockchain supply channel so that they can rely on what's really happening today, not just within the company, but with the third parties that are working with the company as well. All of this is really exciting because it's taking away all this rote activity, all the things that we used to try to find or test for now is tested all the time on every transaction.
Amato: This may not be the right example, but I think it hit home for me today because it made me think about how this would have been handled 10, 20 years ago. I got a text message from my credit card that said, "Fraud alert. Did you make a purchase of $58.12 at this vendor I never heard of?" It turned out I'm already on the path to getting a new card. Someone somehow tried to make a purchase with my card. Where I'm going with that is how would that have been done 10, 20 years ago? Yes, someone would have gotten it. It still would have been computerized, but it wouldn't have been so fast.
Bragonier: It would have taken a while. In fact, they probably would have had a card. They would have swiped it. You would have signed the piece of paper. That chip would have been entered at some time over the next couple of days because they may not enter it on the same day. It gets entered. Then it has to be processed at whichever credit card company you're using. Then they have to do whatever detection on top of that.
You're easily talking about a two-, three-, five-day timeline where now that transaction could have happened 30 seconds ago and that text to you could have been an automated response. "Here's the parameters on your behavior. This is what we've seen. Here's the anomaly. We don't understand it, or we don't understand this anomaly given exactly the last transaction that we saw and so we're going to automatically kick out a text to you saying, 'Could you verify?'"
Your response then will kick in whether it's automated or whether it is now picked up and handled on the human side. So if your response was everything was good, no human probably got involved until we're at that stage where every transaction is being checked all the time for its authenticity, for its recordkeeping. We're not always looking for fraud.
Sometimes it's just human error or the transaction is coded wrong or it should have been recorded at another time because of the delay — the whole accrual concept. Yes, those are things that are taking place already today. You can imagine what's going to take place in two or three years from now because all this technology is brand-new. We're just now building it into our infrastructures.
Amato: So you touched on what's coming in the future. That leads nicely into my next question. Have you watched the movie Back to the Future, and more specifically the sequel to Back to the Future?
Bragonier: Yeah, of course. Of course.
Amato: We'll go where I went in my practice. Michael J. Fox is Canadian. You're from Canada, so you had to watch it, right?
Bragonier: Not necessarily, but I did watch it because I like Michael J. Fox, and I definitely enjoyed that series at that stage of my life 30 years ago or whatever it was.
Amato: So in the sequel, when they went from the year — they went to 2015 from 1985. My opinion and others I've read, they did a pretty good job on some of the tech trends, guessing what was going to happen, even the Cubs winning the World Series. The Cubs were a little early, I think. No, a little bit late. Anyway, all of that's a long way of saying, is there any way to see what's coming in 2040 or do we just have to hire the Back to the Future writers to figure that out?
Bragonier: There's some wonderful stuff that's just around the corner. Twenty years from now, we are going to be in a completely changed environment. We're going to be that way for a couple of reasons. The first is, we have 5G technology that's just coming out. Everyone hears about a 5G, but we don't really understand what it's all about.
5G is a game-changer with regards to how Wi-Fi works. 5G is very short blasts. It's very quick blasts. In my world, we talk about latency is the time when I actually say it for when you hear it. In 5G, that's almost zero now. It handles 40% more — I said that wrong. It handles 40 times more information than 4G does.
This is just one of three components I'm going to mention. What's happening with 5G is 5G is coming out so that we're going to get a whole bunch of things taking place. It will no longer just be human-to-human communication. It will be device-to-device communication.
For instance, in an autonomous car, when it's driving itself now, it's relying on a whole bunch of one channel sort of communication. As soon as 5G kicks in, then each car will have its own transmitter so each car will actually communicate with one another so that I'll know if the car in front of me is stopping. I'll know just one or two milliseconds than I would have beforehand. If you're traveling 60, 70 miles an hour, one or two milliseconds is the difference between you hitting something and you not hitting something.
The first component is 5G. I'm going through your question about what's really taking place, but you have to understand that's information that the pipeline — that's what's really going to give us an advantage on the next couple things I'm going to mention to you. We've talked about now the capability of the channel taking more information, so how does information go? Now we'll talk about the human.
We now have, through natural language processing, unbelievably accurate almost immediate translation taking place now. There are devices already in the marketplace where you and I could be sitting across from one another, you could be talking United States-ese and I could be talking Canadian-ese. You could be talking English. I could be talking Spanish. You could be talking French.
It's automatically translated so that we're here together in a present. We're looking eyeball to eyeball. There's no latency in the translation, and so now we have a true way to collaborate, a true way to communicate without me hoping that I'm using the right language for my non-mother tongue. I don't have to worry about that anymore.
The second part of it is, this technology already exists, but what will make it very effective is when we can get it anywhere we want so I don't have to worry about, "Do I have a strong Wi-Fi right now so I can get it now?" The point is, I'll get it everywhere so that now I can be driving in my car or walking down the street or talking to the vending machine as we go forward. It's the more natural way to communicate in a fuller sense.
Language, even though it's being translated now and we can communicate across long distances, has one additional component to it. That's body language. If we get verbal communication and complete body language communication, we're going to have a wonderful world and that's already taking place today. In June of this year — this month, this year — the first holographic images were created, and I saw it demonstrated.
One person was in one location. Her images were being shown in another location, 3D, so that as she turned, she moved — it isn't an avatar. Her holographic image was doing that. So as soon as we get to that stage where we have fast transmission, we have immediate translation on the language side, and we have full body communication, now we're really talking about some wonderful collaboration.
Now we enter one last one and that's augmented reality, the AR, where we can take an object and put it in a space that it isn't native in. Now we're really talking about all kinds of design, all kinds of review taking place from people that don't have to be in the same location. When we talk about remote work, we can talk about remote work across the world or we can talk about remote work — I'm in Vegas right now and I don't want to travel all the way across the city to have a two-hour meeting with you. Now my hologram could be there, so you'll get my full body language. I get your hologram at my location.
It's as if we're meeting full language-wise. We have augmented reality in between and it could be anything from reviewing a workpaper, reviewing it on the screen, or it could be an object so that if you get into engineering or architectural design, now we have the object that we can collaborate on. All of this is taking place because it's all digitalized and we have the pipeline to be able to handle that.
The way humans interact in 20 years from now, it's going to be abundant. It's going to be wonderful. It's going to be rich. We're not losing things by digitalizing. We're gaining the capability of doing things that we normally couldn't because the distance separated us.
Amato: I may be alone in this, but I had never heard 5G and all those other things explained like that. Just on the 5G topic, I seriously just thought, "Oh, my cellphone will drop fewer calls."
Bragonier: It is amazing though that it's just around the corner, though there will have to be a whole bunch of new infrastructure put in for the first time because 5G is going to require that. But we'll definitely start to see it roll out by the end of this year. We'll start to see many applications start to kick in by 2020. But by 2025 in five years from now, the infrastructure will be across the world.
Now we're going to have manufacturers that know that they can build devices that can communicate with one another whether they're similar or not. It's just the transmission that's taking place. So what are we really going to see when it comes to — I don't know if you know this — taxi drones are already available. They're not commercial yet anywhere.
I'm thinking that Dubai, last year was supposed to do it in the summer of 2018. They held back at the last minute. We hope that they're actually going to release it in July, a month and change from now. We'll have our first passenger drones in Dubai. Uber's aerial is already looking for, three years from now, having as many as 20 — sorry — 200 drone taxis in San Francisco alone.
So when you're in the city now and you have to travel, you don't have to worry about taking the car and the road. So remove 200 trips at all times off the road and becoming more efficient at it. Add in that concept of having autonomous driving where in 20 years from now, our cars don't have to be built the way they're built now. We'll have devices that are comfortable. We'll have office space that we can work in or relaxing time that we can relax in.
Now you're going to have constant flow of traffic because it won't be the ebb and flow that really takes place because of the humans slowing and rubbernecking. You're going to remove some of that because they're actually going to be on drone taxis. Drone taxis, they anticipate will be at the same cost level as a regular taxi because the drones will be substantially cheaper to maintain. They'll be pilotless so that there will be no human involved in it.
With electrical power now, with the drones, because most drones will not be going hundreds and hundreds of miles, with the electrical component in it, the lightweight materials on it, it's going to be a fascinating world. That technology is here. They're already doing that now. Once we get the capability of controlling 200 drones flying people around the city, it's going to just change the way we live, the way we work. We're going to go to a taxi stand and choose whether we're on the road or we take the air.
Amato: Do you think, let's say, current 5-year-olds are going to have driver's licenses?
Bragonier: No, they won't. Current 5-year-olds won't have driver's licenses. My kids today delayed on their driver's license because there's so many options available to them because they do live in a metropolitan area. If you were in the country, that's different. The point is, my kids grew up in Toronto, the metropolitan Toronto area. They were delayed by several years. Where I remember I got my driver's license the day of my birthday. I was writing the thing because it meant the freedom. I had the chance. I was going to be experiencing something new.
They already don't see it that way. They see all the benefits that they have. They already see, "I only have to pay for Uber. I don't have to come up with these major insurance payments and everything else." They're already delaying getting their licensing now.
As soon as you go to the stage where they don't even have to drive, it will be a much safer environment for them. It will be a natural transition. So, yes, I believe that the 5-year-old now in 10 years from now and change, there will be a big chunk of them that just won't bother to get a driver's license.
Amato: Yeah. I can see it coming. Are there tech advances that maybe looking back on — we've looked ahead, now let's look back — that maybe you thought would happen faster than they have?
Bragonier: Yeah. I guess one of the things is I didn't recognize the limitations of communication when it came to the internet of things. So, yes, when it was first introduced directly and it was a very specific manufacturing world with one supplier controlling, without question.
But as soon as we see the internet of things, the light switch really does talk to my sound system. My sound system really does talk to my heating system. When that internet of things really becomes robust throughout all, it's going to change. I thought it would be quicker, but I didn't recognize I should have. In hindsight, always, but I didn't recognize the component that we were missing was what 5G is going to fill in.
We needed a faster communication channel, and we needed a channel that didn't have to travel much distance, that could actually connect in these mesh — they refer to them — networks so they can communicate with one another. Without question, what I saw — when I first started talking about the internet of things, I could see a whole bunch of things that was going to happen.
When they first talked about autonomous driving, I went, "Oh, my God, we're there," because they've got all the technology in the cars already. I was naive in the sense of not waiting for that last component. It's coming out. It's entering our world now. In retrospect, yeah, those things were — I didn't understand why the delay, but it's pretty evident now.
Amato: You talked about all the information that's available that can help us in decisions. How can those trying to learn more, whether it's getting more tech savvy or just — how to humans deal with that information overload?
Bragonier: I teach everyone that in the tech world, there's a concept that Chris Anderson, who was at that time the editor of Wired magazine — he wrote something that — he referred to it as the longtail. In the longtail, the concept is that as we migrate more and more into a complete digital world, products will differentiate themselves in just very, very small but important ways to the end user. I always speak about the beer industry. The beer industry has all these very small, unique flavors. You can actually get now pumpkin flavored beer. That may not appeal to you, necessarily —
Amato: Actually, it does.
Bragonier: Then you understand what I'm saying. What's happening is you're getting all these products that are not trying for the majority but are trying for a very good segment of the marketplace.
The reason I'm saying that is that stratification across all the product lines provides for a very smooth transition depending on wherever you start to wherever you want to go. Where if we only had two or three products in the marketplace or two or three major industries to deal with, there were always major jumps between those. There was major differences between them.
As we migrate forward with this longtail sort of concept, the marketplace will continue to grow exponentially. As the digital marketplace, our entire marketplace is going to double in less than five years. The projection is that by the year 2025, for all practical purposes, the world will be connected, and everyone will be connected to it. There will still be some anomalies, but that's only five years away.
If you look at it now, we're just beyond the 50% mark. What we're really saying is every product that's being sold today has a marketplace that's going to increase by 100% in five years if it can be digitalized, produced, and pushed out that way. In this digital world, the only cost is in creation. The cost of production, of course if it's digital, you copy it. If it's distribution, you just download it. Whether you're downloading one or you're downloading a million, there's no additional cost factor in that. It's all in creation.
The creative side is where people should really be looking. How do you spawn the creative side? I don't know, but you hear people playing with it now because we've never really had to try to figure out, how do we do this creative side? How do we encourage success through failure? We have a whole bunch of fancy ways of saying that's agile processing, that's lean processing. But we're just saying, "You're resilient because you tried and it didn't work and so you stood up and you tried again."
And so that in the creative side, a whole bunch of things is going to take place. When I talked earlier about people with the skillset that can communicate to others, that's all the creative side. There's no way that one individual can be as creative as a collaborative group can be. There's no way the collaborative group can communicate unless somebody quite often translates content for them. Not translate as in language but translate it in a way that somebody else would understand.
We've got this wonderful world emerging, huge marketplace emerging, huge opportunities because of the longtail stratification of the marketplace and then this whole concept of almost everything is in the creation side. The products are going to be pushed more and more to the true concept of being free in the marketplace.
Amato: Let's see if we can tie this all into kind of sum it up. How does all this matter? How does this play into the world of the accountant?
Bragonier: This analogy has been used by many, but at the turn of the century there were wheelwrights. They were carpenters in essence that had a specialty in creating wheels and there was a huge demand for them. Wheels constantly broke. Those that saw themselves as makers of wheels had a huge opportunity as technology started to kick in. Those of them that saw them as carpenters saw their world dying away.
In our world as CPAs, if you consider yourself an accountant — a transactional, recording, reporting person — then your world is going to fade away because a big chunk of it is already automated on the transactional side. But if you see yourself as someone who has the ability to look at data, give an unbiased opinion on the reliability of that data — the assurance so that it can then be reported on.
If you're that person that says, "Look — I now have hundreds and hundreds of points of data. How can I bring that up into proper dashboards so that people can oversee and allow them to drill down into proper levels that make sense?" then you see who we are because we provide two major services.
The first is the whole concept of assurance starting with the dataset. Can it because relied upon? And then all the reports that go above it. Our area of expertise is to say, "We need to include these things in a report for it to be relevant. We've always seen it as financial because we were constrained by time." Our training is really to say we have huge volumes of data and how do we get that at the proper level for someone to rely on it either internal or external?
So there's huge opportunities for us for exactly what you said. We have huge change taking place, huge datasets that are coming out of the blue, huge opportunities for somebody to say, "Yes, this appears to be reliable data and, yes, that appears to be in a proper report that summarizes that dataset."
That's what we've always done, and now our opportunities are growing exponentially. I don't know what the right language set is. I know that the AICPA and CPA Canada are wrestling around this. Are we business advisers? Are we trusted advisers? Society has always trusted us to say, "Yeah, we rely on your summation or accumulation of that data in that report that you've presented for us and the dataset is clean enough for us to rely on." We have huge opportunities here.