AI Wisdom Ep. 29: The Risks and Opportunities of Autonomous Vehicles

Insurance Industry News & Views - February 24 2021

On this episode of the “AI Wisdom – Talking Innovation in Insurance” podcast, host Ron Glozman speaks with Guy Fraker, Chief Innovation Officer, IE Advisory, and one of the world’s leading authorities on the risks and opportunities associated with autonomous vehicles. Ron and Guy discuss how autonomy-enabling technologies like AI, computer vision, machine learning and intelligent machines are transforming commercial auto insurance. What does the autonomous future hold for insurers, automakers, and drivers? Buckle up and click the play button to listen or read the full transcript below.

Full Transcript

Ron Glozman: Hello and welcome to “AI Wisdom Talking Innovation in Insurance." On this podcast, we talk to business and InsurTech leaders about how artificial intelligence is transforming the way we buy and sell insurance. I’m your host Ron Glozman, Founder and CEO of Chisel AI and a strong believer in the power of AI to help people work smart and enrich their lives. So, let’s get into it.

Ron: I’m very pleased to have Guy Fraker, founding principal at IE Advisory, join me today, as we discuss the future of auto insurance in the age of autonomous vehicles, and the impacts of COVID-19 on the commercial insurance industry and insurance innovation. Guy, so glad you could join us today. Tell us a little bit about yourself?

Guy: Absolutely and I really appreciate the invitation and the opportunity to share some thoughts given everything that’s going on right now. I started in the insurance industry literally as a staff person for an agent in 1985, was an insurance agent and most of my career was in the Bay Area. I was out there during the various bubbles in Silicon Valley. Worked my way up to marketing executive ranks, and then largely through experience with Global Business Network, or GBN the folks who predominantly invented scenario planning started to get into innovation work by attending their events and learning the practice of scenario planning.

And it occurred to me that the insurance...that’s my first realization that the insurance industry is, by and large, a car driven with the rearview mirror. All of the decisions are based on mountains of past data that’s how the industry has evolved. I became really interested in this concept of peeking around corners and kind of being able to plan in advance for what’s coming. Then left California and assumed a couple of other roles in redesigning the company...the structures that I was employed by.

Then co-founded an internal innovation team with some in-house partners and this was at State Farm. Then took my first ride from Sunnyvale, up the Bayshore Freeway in rush hour traffic through the financial district to San Francisco and back down to Sunnyvale, from the backseat of a car. There was somebody in the driver’s seat, but there was no intervention at any point, in 2009.

I walked away from that experience, went to Singularity University the very first executive program, and then had the opportunity to work with IDEO and Innosight, and some of these world-class innovators.

And I realized that the world was about to change in relative terms dramatically for the insurance industry.

I left State Farm in 2012, and then, really, at that point, began a coaching practice, an advisory practice for insurance companies on solving intractable problems, facing things that they thought were...that’ll never happen here and all of a sudden, they’re faced with it. And creating new markets and new revenue streams through the best practices that I’d learned over all these years. So, for about 15 years, I’ve been really at the heart of insurance innovation.

Ron: I love that. I’m curious though, like what car in 2009 was self-driving?

Guy: Google’s. I was in awe of the perspective and the goal they had set out in the form of Dr. Chris Urmson, and Sebastian Thrun, and others on that small team and realized that this was as fundamental of a shift beginning at the earliest tip of the spear stages than the industry had ever faced before.

Ron: That’s perfect because I’ve heard you argue that innovation is the DNA of insurance and at the same time, that resource constraint is what drives innovation in our industry. I would love for you to share your thoughts on that, especially in light of COVD-19 how it impacted insurers and insurtechs and really has put a resource constraint. Do you think that’s driving transformation and innovation?

Guy: I think it’s driving transformation innovation around the industry that will ultimately impact the industry. I think the industry itself, by and large, the culture of innovation has evolved to a point where something isn’t real unless it’s in writing and a group of people sit around a table and agree on that document. So, the whole concept of working virtually was culturally I think, a huge shift that has ended in a new normal, that’s resulted in a new normal. I’ve spoken with several insurance CEOs about reflecting upon their own experiences of COVID-19 and helping them come to grips with what they’ve learned themselves the changes they’ve had to make.

For companies who want to innovate in the space of insurance who are not in the space now, this has been a wonderful opportunity to raise capital, hire talent, and double down to accelerate their plans. And by a relative few number of companies that I’ve spoken with, that is a consistent theme, now is the time for us to make the leapfrog.

I don’t think the industry is necessarily aware of that. I think the insurance industry right now is patting itself on its back rightfully so, in accepting these new ways of doing business, and keeping the wheels on the bus, so to speak for them, that’s success. I don’t think many companies are taking the time and opportunity to sit back and actually document and capture what have we learned from this. Because it kind of all happened... the lockdowns and things happened so quickly. But the industry is going to face a transformation that is inevitable. So, let me speak to the insurtech part of this for a second. I don’t know that insurtechs have yet successfully been a catalyst for transformation. I think they have been a catalyst for a needed wake up call.

I tracked early-stage companies for the past 4 years across 176 countries, so that’s roughly give or take on any given week 48,000 to 50,000 startups and early-stage entities. It’s a lot of money. Insurtechs are only about 15% of that. Eighty-five percent of those startups are fundamentally...without necessarily realizing it, are fundamentally reinventing the very nature of risk. And pushing the insurance industry away from its history of simplified rating and recovery as core business to prediction and prevention.

Autonomous car is a perfect example. Now the reason I say innovation is in the DNA of insurance goes all the way back to the founding of the United States. In 1752, Ben Franklin launched the first homeowners’ insurance company, which is really a relevant case study for now, on the core mission of insurance. The core mission of insurance is to enable economies by freeing up consumer cash flow that they would otherwise stockpile, over fear of occurring a loss that they couldn’t recover from. With that blanket of security that the insurance industry provides people spend money that they wouldn’t otherwise spend. They’d have to stick it in a mattress or bank account.

Well, Franklin and others knew this, knew that the colonials would be stockpiling their newfound freedom to earn and be entrepreneurs, and launched this homeowners’ company. And by the way, the people who funded it with secret shipments of gold were committing high treason against the crown, but they saw the inevitability of the British laws and this was 17 years before the ratification of the constitution. So, everybody involved was taking enormous risk and being very innovative to launch this first company.

Now we saw it again, turn of the century coming out of World War I, coming out of the last pandemic the U.S. and big cities, Canada as well faced a real health crisis which was social distancing. Had impacted public transportation which was horse-drawn. And major cities were finding themselves overloaded with dead horses riding in the streets.

Well, along comes Henry Ford, and the ability to mass-produce. And the car was really kind of struggling on adoption and then a bunch of Mavericks stepped in and invented this new industry. And for the insurance audience, let me remind you, they had no data and launched the auto insurance industry. By 1925, with the help of banks inventing consumer credit, by 1925, the streets of most American cities and Canadian cities were clogged with cars, the car had gone viral.

So, two of the most significant societal shifts that have occurred already when insurance embraced the change by taking risk, not protecting insurance investment portfolios, or relying on mountains of historic data. But when they looked ahead and took risks, and enabled these economies, the result was arguably a couple of the most powerful positive societal shifts forward we’ve experienced globally. Now, you see this in countries where insurance is fighting for relevancy, where 2% of the population has any kind of policy at all and there’s tremendous economic instability.

And so, innovation, which democratizes access to capabilities for everybody, or for most people, insurance democratizes access to the purchase of goods and services, and they share the same mission. So, it is all of us who benefit from success in both those areas of economic activity. As the insurance industry has matured and become very much about managing huge portfolios and mountains of data, then the risk-taking has slowly evolved away.

So now what we’re seeing is new entrants approaching the insurance industry with a totally different business model, willing to take risks, finding capital, and really getting ready to propel the next generation of insurance, which is to support prediction and prevention. And move people and the mindset away from just this rear looking rating and recovery.

Ron: I love that. The story of Henry Ford, I think is so powerful. And many people who’ve gone to business school will have probably heard of... if you ask people what they wanted; they would have said faster horses. But I’ve never heard that insurance angle in addition to that, and that its sort of was in some ways a catalyst and the pandemic was also a great perspective. I know you’re a great thought leader and you’ve very closely been following the development of autonomous vehicles at least since 2009. Maybe even earlier because I heard you say that.

You know, I’d love to hear for you on the thoughts on where we’ll be in a few more years from now, maybe 10 to 15 years before self-driving cars are widespread. What are you seeing, where are we going with the technology, and do we think we’ll really see autonomous vehicles become ubiquitous?

Guy: It’s a great collection of questions. Let me start by answering another question which is how in the world does a career insurance guy who doesn’t have a technology background end up so dedicated to autonomous vehicles? In 2008, I’m sure several people have heard this before. But in 2008, I was sitting down at the dining room table and my son comes running down and he’s 6’5, with a wingspan that’s 6 inches longer than its height. So, you can imagine he comes running down he reaches around me, drops the laptop in front of me, and says, “Look at this.” And it was the very first New York Times videos, they’ve been stalking the Google’s car team.

So, these are the first broadcast videos of Google self-driving cars and I’m instantly kind of got dizzy. I was getting ready to found this innovation team, had been looking at transportation changes. And he looks at me and says, “What do you think?” And in a moment of rare parental clarity, I didn’t answer him. My answer back to him was a question, “What do you think?” and he points at the computer screen, and he says, “Dad, I get to have a car someday.” He’s autistic. He had immediately gotten what would have taken me a year, if ever, to realize that autonomy in vehicles doesn’t just clear up a slew of societal challenges that we face.

It also, again, democratizes access to independent transportation for people who are currently dependent on social welfare systems because they don’t have access to that kind of transportation.

Now, rideshare help, we’ll talk about that. So, I entered into this never thinking about really trying to change the world, I was just trying to get my son a car. He had been robbed by three taxi drivers. So, over the course of years that followed, it’s been close to 9 years, probably 90% of my capacity is dedicated to this weird intersection of risk managers, regulators, and the vehicle tech. So that we could come to terms and embrace and propel this technology.

I think during these COVID days this year, people have doubled down on accelerating the pace of the development. Here’s what the technologists and engineers underestimated that the insurance industry has long known and is actually built around. We, humans, have an infinite capacity for stupidity. This trillion-dollar auto insurance industry exists because we find the most creative ways to wreck stuff.

And by the way, just a passing message to Tesla drivers who submit videos to YouTube of near misses, let me make a plea here. Please edit the audio out, we don’t need to know what you were doing while the car was doing the driving. I’m always amazed by these near-miss videos.

The technology is going to be ubiquitous when it comes in the form of two sources. One, when you can take your car to any AutoZone and have an aftermarket kit installed. Retrofit of existing vehicles is the great wildcard that’s also...I can verify an active moon race among some tier-one tier-two providers in startups. The pundits generally ignore when they make these protracted forecasts about fleet turnover. That capability is well under development.

If you look at a picture of one of Google’s original vehicles, or any of the DARPA cars from way back in the day, you see mountains of technology. Where the computer processing equipment itself to be able to handle autonomy is down to two decks of cards embedded in a trunk. So, the technology is there, the wrestling issue that engineers have underestimated is the object recognition and the complexity of our interaction with the machine.

That was verified down here in the state of Florida by public policymakers who have taken a position, if you want to bring a car to Florida, that is fully level four, by the engineering definition level four control, without human intervention, you need a million dollars liability insurance, come on down. No other regulation, no other registration our roads are open. If you want to bring a car to Florida where it’s partially autonomous where there’s a constant transfer, of exchange or control between the driver and the car, then we got to talk.

So, there’s a state that really set precedence by saying we trust the technology more than human drivers. As more and more states and countries adopt, and in countries that have already adopted that perspective, you’re seeing a greater deployment. There are no issues out there that cannot be resolved and the reason I make the comment about the near-miss videos is a good example.

The insurance industry is not prepared to measure the non-event. If they could... And telematics is certainly bringing them this capability. They’re not quite sure what to do with it yet. But the near miss, the accident that’s narrowly avoided with a human driver is considered by the insurance industry as a precursor to an inevitable accident. It’s an indicator that this person is going to wreck this car at some point. In the autonomous world, a near-miss is an indicator of accident’s avoided. It’s a huge swing.

And for every time there’s a near-miss that’s recorded and you make a software programming update that’s transmissible to the network of vehicles with that technology. Then you’re eliminating...the presumption is...and so far, the facts have played out. The presumption is that that’s an accident in the future that won’t happen.

Now, let me bring up one other important point about autonomy and vehicles and back to the question of transformation. One of the great barriers to entry into insurance, particularly auto-insurance world, that has been fertile ground for many in insurtech, is accident frequency. Servicing claims from human error is a full-time industry, that’s full-time job and requires, networks of relationships and service providers, etc. What if we don’t eliminate...because 100% elimination of excellence is a non-starter. But what if we reduce accidents by 75%, 80%? Therein falls one of the great barriers to entry.

The result is an auto manufacturer, who embeds insurance in the purchase of the vehicle and becomes the insurance provider, can now own and control the entire lifecycle of that vehicle from design to recycling without a third-party intervening.

Many people change auto brands after a total loss even if they walk away from it. Most people don’t say, “Oh, thank God I was in a Subaru because it rolled downhill, and I was able to drive it back up.” They’ll for whatever reason switch brands. Auto manufacturers know this, but they don’t know who people go to.

With the advent of these technologies and the data capture capabilities, the whole new industrial sectors or financial sectors open up for auto manufacturers. And believe me, they are taking advantage of it during this COVID period. In terms of developing and deploying not just the technology, but the other ancillary services like embedded auto insurance, or real-time auto insurance. Where you’re not paying your full premium when the car is sitting and parked. You know, the variable premium in real-time because they don’t have this historic financial structure that they have to rely upon.

Initially, we’re seeing some insurance companies embrace this and partner with them. Reinsurers are certainly far more open to sharing that risk directly with auto manufacturers. And so, the business model itself is already broken ground, broken precedent. Here’s a real-world example of moving from rating and recovery to prevention and prediction. And having that be more efficient model for new entrants, who now can service that lower claims frequency with their own in-house body shops with the advent of chatbots and artificial intelligence, that can help make all of this much more understandable to consumers. I think you’re going to see...over time you’ll see an erosion to the historic brands within the insurance industry.

So, it’s kind of a long answer to a question that has a lot of moving parts. But it is playing out this is not theoretical. I know of a couple of LIDAR manufacturers, who have taken up the charge of helping their auto manufacturer partner who’s adopting the technology, figure out this insurance model. Because they can recoup both the owner and the auto manufacturer, can recoup the cost of that technology, through reduction in accident frequency and severity. Many of these technologies may not be able to prevent the first loss.

A deer jumping off a highway overpass and, injuring or killing somebody in the front seat, no technology in the world is going to stop that. What autonomy can do is prevent that car from then going across the median, causing a secondary head-on, and not just a second, but maybe a third and a fourth and that’s especially true in trucking and freight. So, it is both frequency and the severity of the loss and a far more efficient use of data. That’s a very powerful collection of tools that would transform any industry. And the insurance industry is not prepared to make the change fast enough to keep up with people who want to really make progress.

Ron: Do you think the public is ready for self-driving, you know, fully loaded Tesla semi-truck barreling down the highway with nobody behind the wheel?

Guy: That’s a great question. Let me give you a bit of a heads up on that answer. Near where I live in the Florida Keys is Big Pine Key, and speed limit is the law of the land for a variety of reasons. It’s ferociously protected. So in the wake of a couple of hurricanes, the police were short personnel, and they parked a cruiser on the side of the road which for a lot of the tourism industry that come...tourists visitors, has the impact of slowing them down, just seeing the car parked in the grass off. But for a lot of other people, you finally get used to the fact that there’s nobody in it.

What the sheriffs down here did is they took a CPR training dummy and put a wig and a county police hat on the dummy in the front seat of the car. And at night, in particular, the effect is dramatic. Seeing somebody sitting in the car, or be it a crash dummy has had, you know fairly significant impact on safety. I would encourage the trucking industry to think about that and think about that example. You’re far better off of the mannequin, or some other representation of the driver in an 18-wheeler barreling down.

There have been several occasions not well-publicized in Nevada, where an autonomous semi-truck, you know...because they’ve been testing them driving across Nevada for years now. Will pass a slow-moving vehicle and there’s nobody in the seat. And even if it’s only momentary, the person just happens to be out of the seat when they go by, it invariably results in a 911 phone call. And kids are pulling down like they want the driver to pull the horn on the truck and there’s nobody in the seat. That alone could cause an accident. So, I think the public is absolutely prepared.

I’ll share with you my other favorite case study on this. From 2012 to 2017 once a year I’d get a bunch of people together in a focus group. Now, the only requirement was you couldn’t be under 70 years old. The focus group was always entirely people 70 and older. And it would start out with a single question. I’m going to give you a car, it’ll take you wherever you want to go, it will take you home. If you don’t remember where home is, just say home, it’ll take you. It is 85% safer than a taxi or a human driver. Here’s the catch, though there’s not going to be anybody in the driver’s seat.

Well, you can imagine in 2012 - 2013, I started to see a change in 2014. The answer was, “I hope there’s a lunch with this focus group because you’re wasting my time.” “Hell, no, I’d never get in it. Absolutely not.” The phones really had a dramatic impact. One of the fastest-growing demographic groups on Facebook, for example, is over 70.

Once people in their mind in their definition became technologists or adopters, then all of a sudden, all these other technologies starts to make more sense. So, if I can talk to the car, I don’t have to use a keyboard, yeah, I’m okay with that. And by 2017, the answer was, “Oh, absolutely, when, thank God because everybody on the road around me is going to be safer than they were today when I drove myself over here. You bet I’ll take that car.” And that’s not a minor opinion I mean, it was forceful.

So when I compare where I started with the last time I did it, I would have to say not only is the public ready, but I believe they understand all of the benefits enough. Reduce congestion, more fuel efficiency, greater adoption of electric vehicles, air quality, run down the list. They understand it’s an opportunity to be both a feel-good, about contributing to the health and welfare of their community. And they understand we have this aging demographic group that does not want to lose independent mobility, and highly motivated to accept these technologies. You talk to the rideshare services about the growth in this demographic group.

So, where it gets adopted and how it gets adopted may surprise people. But you know, without question, I think people are waiting. I don’t think they’re sitting around waiting for the other shoe to drop and for it all to go away there’s far more hope that it comes sooner than later.

Ron: I love that focus group that’s a unique data set. I don’t think anybody else has that type of data set. When you think about commercial fleets though maybe multiple trucks in a convoy potentially, or something like that. How do you think commercial insurers are going to react? Like if you’re saying everything is going to be insured, let’s just say by the auto manufacturer, like, you know, are insurance companies going to be happy with that because that’s maybe taking money out of their pocket. Are insurers ready?

Guy: I think insurers are...they run the gamut from still believing in science fiction to totally embracing it. And the commercial fleet is a really interesting question. I’ve always hypothesized if you will, that that’s the tip of the spear, that that’s where we’ll see the adoption first. Now, I was challenged once at a conference in MIT by the President of the Teamsters, who wanted to make it very clear that they had no intent on letting technology replace their drivers. And my answer to this person who was fairly persistent in front of a large audience was, “Let me be very, very clear about this nobody is interested in replacing your driver. One, the truck can’t protect itself from criminal intent, hijacking, etc. Two, the truck can’t load itself, it can’t unload itself, and it can’t refuel itself.”

I’ve spent time with families of truck drivers where the truck driver is alive, they were not the victim of a fatality. But they caused a fatality. And my interest was in the family impact on helping that professional very good driver recover from the loss of another human life. I asked the president of the Teamsters at that time if he had the opportunity to spend time with families in discussions like that. And he said no, obviously, their focus was on the drivers, they didn’t necessarily get together with the family. I said, “I don’t want to replace your driver nobody wants to replace your driver. We want your driver to go home, in a safer office, with fewer health challenges, fewer temptations to augment their capabilities, and still have a good professional career.”

So, I think convoying has some challenges in that, if you have anybody in any of the vehicles, and they’re not the lead vehicle, there’s inertia problems. We have to work through the distance between vehicles for the people that are in them. But I do see freight and short-haul delivery, long haul and short-haul as being very, very fertile ground.

I believe the commercial insurance sector, which relies on what’s called the surplus lines practice to launch new products and services, more so than personal line companies do, I believe, by and large, they’re ready to embrace it.

They understand it’s not becoming...the technology does not make insurance irrelevant. It just makes it far more efficient, helps the margins, and is still going to be necessary. Ultimately, the determination of fault when something goes wrong, we laugh about this, right but it is essentially going to reside with the software developer. We’ve never had the greater capability of determining fault than you have in a vehicle that is capturing between 16 and 20 gigabytes a second and processing it in real-time. So, there won’t be a question of who’s at fault? There may be some internal arguments, but they’ll all be behind the scenes.

Ron: I’m curious, though, do you think there’s any fear around like you mentioned...I’m with you on the theft piece, although I would say you could probably just get thicker glass, right? Especially if there’s no humans and the windows don’t need to roll down. And then on the refueling piece, there are, and I believe Tesla even manufactured like an arm that can come out of the “gas pump” assuming it’s a gas vehicle, or diesel, or electric, but I think it can be customized. And basically, sort of is like a snake and slithers its way into the gas tank.

Guy: So, I tend to look at these things through the lens of a behavioral economist which is academically my background. And you add to that the design concept of the job to be done. So, who are you doing the job for? What’s the problem? What’s the solution? Those are the kinds of constraints that generate breakthrough results. We’ll talk more about constraints in a minute. But the more complexity you introduce, the more you protract the time before deployment. Do I think we’ll get there someday? Absolutely. Is it a critical dependency to adopting an autonomous truck? No.

I think it’s going to be...this is true with artificial intelligence and risk underwriting and pricing science of X-ray, there is still a place where human judgment is necessary. It may not be in the driving, but it may be in the care and maintenance of the vehicle. It may be in the care and maintenance of the load. But we’re a long way away from eliminating humans from the driving equation but that does not mean we’re a long way away from the benefits and deployment of autonomous technology.

Ron: That’s a great way to look at it. We’re going to take a quick 20-second break to tell you where you can find more information and insights about insurance innovation. We’ll be right back.

[If you liked this episode of AI Wisdom, subscribe to our blog, Writing the Future: AI in Commercial Insurance at for feature articles, interviews, opinions, and more.]

Ron: We’re back with our featured guest, Guy Fraker. Let’s jump right into the next question. I’m curious to hear your thoughts on ridesharing for a bit. How are insurers tackling the rise of ridesharing?

Guy: There’s no question that we’re at the point of embracing it’s an economic juggernaut. Now, I had the good fortune of being in on the earliest developmental stages of some of the rideshare applications. I remember when Lyft was actually a free vanpool servicing a single college campus. And the reservation platform wasn’t fully developed. And by the way, the a plug to the ecosystem in Toronto, the original reservation system for rideshare and car share, largely came out of Toronto. It was really the breakthrough ecosystem for the technology platforms that have become what we experience today it was all developed in Toronto, Canada.

The insurance industries absolutely embraced rideshare and have partnered with reinsures. So, you’re seeing mainstream big-name brands standing behind these rideshare platforms now. Now, where I was called in to if you will, offer an opinion on autonomy of vehicles being developed by certain rideshare platforms. And there was little doubt that a need for fast economic return was outweighing safe development on a couple of companies that don’t need to be named. This was in the state of Pennsylvania. They were really wrestling with whether or not to continue to allow one of the cities as a test market for the development of autonomy by a rideshare platform.

Where we get those two together, there’s no question the insurance industry is not ready to embrace that. But for those companies focused on the expertise of moving from point A to point B in a rideshare platform, it’s well entrenched in the mainstream auto insurance industry. There was a time and the big stumbling block...this is where ridesharing is...history will look back on ridesharing as a pivotal point in the transformation of insurance. Throughout the history of the industry, there’s been this almost old Berlin Wall-like divider that was immovable – might as well have been the Himalayan mountain range – between personal lines and commercial lines. And that enormous immovable barrier got blown up by the mixed-use of ride-share, where it was part commercial part personal.

As it has evolved the industry realized that they had to figure out how to go back and forth between personal use and commercial use. And that enabled a lot of technology breakthroughs, a lot of efficiency breakthroughs for the companies that adopted the platforms. Same with car-share. My first meeting with Getaround was before they launched and helped relay rides now Toro through its first round of venture capital gathering, which included insurance industry money even on round one.

The stumbling block was the insurance industry members who were investing in carshare and ride-share platforms couldn’t provide a product to the company that they were investing in. There’s what you call a moment of clarity. You can do six months of due diligence or what have you and make an investment but a year and a half later, you can’t develop an insurance policy? That was a wake-up call. So that helped motivate the industry to embrace that change in vehicle use.

I think rideshare platform is one of the most transformational technology platforms since, you know, outside of social media. Simply because now what our manufacturers realized right away is people who drive for rideshare...and certainly, COVID has only helped the rideshare industry as people have evacuated trains and buses, realize people burn through cars faster. So, now I know there’s the spread of a new business model, where subprime auto loans are granted to people who don’t own a car so that they can acquire their first car, despite whatever background put them in the subprime market. The earnings go directly to pay back the loan and anything over that each month goes to the pocket of the driver.

And it’s actually enabling. Now the downside is it can enable congestion, but it’s actually enabling vehicle ownership. Again, the core mission of insurance and innovation to democratize access by spreading risk. And it’s enabling people to be employed, who for whatever reasons, were not able to generate their own income prior. That’s been...seeing that business model spread and embraced by insurance partners, has been I think very gratifying to see take place. That gets back to the history of insurance enabling people by taking risk.

Ron: That’s a great example. I think for sure, like a lot of people are benefiting from vehicle ownership especially with public transit being reduced, or people just simply not feeling safe to go on it. Vehicle ownership is definitely something that people are probably striving for and happy to get. On the business side, I’m curious to hear your thoughts around business interruption. It’s something that’s a rapidly changing story legislation has been asked to weigh in. We’d love to hear any thoughts you have on the topic.

Guy: So, business interruption in the context of COVID-19, economically is break the bank laws. Ultimately, it’s going to be decided in the Supreme Court. But let’s demystify some of this. I had the great honor of representing many in the industry, in 2003 to 2006 with a small group that was facilitated by the Department of Homeland Security and the U.S. Treasury, on pandemic planning. When it looked like avian influenza H5N1 was going to make the jump which would have been catastrophic.

So, we ran national simulations. We did a simulation on World of Warcraft on social distancing. We ran a national insurance pandemic simulation by asking companies to voluntarily send a third of their workforce home and tell them not to work. And then send another third of their workforce home and have them work from home. We needed to find out if the last mile connection to Wi-Fi was viable with that load. Business interruption in the small business owner industry, so small business owners have been dramatically impacted by the COVID-19 protocols.

If ultimately the Supreme Court rules that a government-mandated lockdown or government-managed store occupancy is somehow equivalent to civil unrest, then essentially what is going to be happening is we’re generation shifting the cost. The reason the insurance industry is labeled critical and essential infrastructure equal to utilities is because roughly 105% or 100% of every premium dollar paid flows through the insurance industry back into the economy. It may be in the form of jobs, salaries. In the crash economy, we talk about autonomous vehicles, there’s roughly 105 sectors of the economy that are paid because we wreck stuff.

The insurance industry is a highly tuned organizational structure for cash flow. And that has been impacted by COVID. There isn’t a margin to suddenly cover a national scale equivalent to a cat 4 hurricane. Which is what commercial insurance would be facing if business interruption is mandated to be covered because the premium dollars aren’t there, they’ve never been collected. So that leads to a government solution. The dollar volume would be so enormous that it’s essentially moving the cost of recouping that loss at least one if not two generations forward. This is along with workers’ comp.

So, most workers’ comp carriers nationally, have never been willing to provide coverage for people who work from home, not in the United States. Independent contractors, freelancers, and work from home employees have been considered outside the scope of workers’ comp, where everybody is organized in a location. Same thing here as governors and legislators have seemingly randomly decided COVID-19 for stay-at-home employees is an occupational disease all sudden the workers’ comp industry is in the crosshairs. Again, on exposure and loss that’s already occurred that they’ve never collected premium for.

So either way, unfortunately, the reality is unless federal governments nationalized governments are prepared to help pay that bill at a national scale then either way, either in the form of recouped premiums that were never collected in the first place or paying the government back because they come in with a solution. You know, I’m afraid this one central question on one sector of the insurance industry literally becomes a future tax that everybody is going to have to share.

Ron: Talking about future generations, you recently participated in a panel discussion, titled, “The Future of Insurance: Innovation in a Changing Industry” at Gamma Iota Sigma’s virtual 49th Annual International Conference. What do you see as the role for innovation and the adoption of technologies as a way to attract talent into the industry and to help me bridge the talent gap?

Guy: Well, I think it’s essential. Now, it’s a different kind of employee, it’s a different kind of recruiting. Let me give you a concrete example. I did a fairly in-depth research piece three years ago on being able to quantify the difference between an entrepreneur and an entrepreneurial executive. Because most innovation somebody has to start with the question, can an incumbent innovate faster than a new company or a startup can scale? And 99% of the time the answer is resoundingly no. Which is why you see these tremendous startup ecosystems like you have in Toronto. Because the basic belief is that an incumbent cannot match the pace. Well, that’s a choice not a law of nature.

One of the big misunderstandings about innovation is how many people does it take and how much money does it take? So, insurance companies need to be recruiting top talent that is technology-oriented and is solution-oriented. That often involves employing people from backgrounds that they never envisioned recruiting from chemists, biologists, physicists. The life sciences, and engineering sciences, understand a fundamental issue that people who are very focused on a business background don’t necessarily understand. And that is how you connect the dots horizontally can lead to a solution nobody thought was possible.

So if I want to help a company found an innovation team, it’s not looking for the best internal horizontal political, or relationship manager, generally, you need three personality types and you only need three doesn’t have to be cast 1,000. This is central to the startups that I’ve tracked.

You need somebody to be an evangelist, who is very good at relationship building and that’s their role. You need...So that is a recruiting field to put on the maps of HR departments. You need somebody who understands how to bring disparate disconnected parts together to solve a problem. That’s usually in the life sciences, biology, chemistry physicists, to some degree economists.

And then you need somebody that’s going to keep the other two in check. You need a third tie-breaking opinion, who understands the technology or the solution or the problem set. And then get out of their way. Most entrepreneurial executives and incumbent companies go to their sources for consumer data, to develop consumer profiles or develop constituency base for some new solution.

Entrepreneurs, go to customers, they’re face to face, or Zoom to Zoom in the current environment. But they’re not relying on third-party necessarily as much third-party processed and if you will digitized data, to find out if they’re on a problem that matters versus on an interesting problem that’s got no economic viability. So, recruiting those types of people are very important and often that may require a different incentive system than most incumbent HR systems are set up to handle.

When I advise a company on either building from the ground up, or more often than not breaking up a logjam on an innovation practice. I tell people all the time, you have to have a couple of departments onboard inside of the incumbent company for innovation to be successful.

You need HR and you need corporate law. Because the people that you’re going to be recruiting are going to challenge your existing incentive systems and your talent review processes. You should listen to them not force them to conform. That requires a flexible procurement process and a flexible employment process.

Typically, there’ know, I’ve had CEOs literally say to me, “If you’re telling me that my corporate law department or my HR department has to be the most innovative department in this company, I’d say you’re talking about one of the seven signs of the apocalypse.” Once they get that out of their systems and we kind of have a good laugh about it, they very quickly realized that people are dying to be innovative and dying to exercise talents that they’ve not been able to use yet. Just set up the right framework for them to be successful at that.

There’s no question in auto insurance there’s a need for engineering talent to be institutionalized in the claims process. There’s no question in residential dwelling insurance, that you need architectural talents, you need engineering skills, and you need people with material science. I know several startups that have created almost a Kevlar drywall that could take a 2 by 4 at 10 miles an hour and not be damaged by flood. Well, that whole claims process and everything it takes in the industry to support construction with those types of materials means you’ve got people who understand the chemical properties of a lot of the materials going into buildings now that say were not applicable 5, 10 years ago.

Ron: I would love to hear your thoughts around constraints. Because I think constraints is something, we’ve touched on several times it also ties into this topic. What are some of the constraints you’re seeing and how can companies either innovate, or invest, or whatever is required to get around those constraints?

Guy: Great question. Hopefully for the answer will bust a couple of the great myths that hold innovation back. I tell companies all the time, “You can’t tell people to just think out of the box and stop there. What you have to do is build a box that they can think in, you just want to be the one to design the box.” What I mean by that is one of the quickest ways to kill innovation is to throw a blank check at it. Without constraints, you don’t have focus. So, quality set of constraints that have unleashed innovation in several companies.

One very publicly known and recognized, by now it’s fairly established, is the Munich Re Mobility Solution Center, that is kind of wide swath with autonomous capabilities, ADA systems in public transit. That started out as an experiment with three buses in Washington.

So, a set of constraints is you sit down with an executive group and they agree on what’s considerable, what’s core to their current business model and product line, and what’s out of bounds. Now, some of the out of bounds things may seem ridiculous. I’ll never forget, I asked one executive team “Do you want to support autonomous vehicles?” The answer was yes. And I said, “What about mining vehicles? Autonomous, these enormous mining vehicles being autonomous?” The answer was, “That would be a bit of a stretch, but yes, we’d be interested in that market.” And I said, “Okay, what about autonomous mining vehicles off Earth?” Dead silence.

One of the gentlemen kind of laughed, and I said, “So that would be out of bounds” with a straight face. They all laughed and said, “Yeah that would be out of bounds for now.” Companies have a very difficult time deciding what to do, because they tend to focus more on what they might be missing, which is cloudy, and boiling the ocean. So set a budget, limit the team size, give them permission to act faster than the culture can say no to. By setting up these boundaries around the innovation on what the company is willing to consider, and where the line is drawn, nobody needs to waste their time.

Once you have rules of the road established, a small team that can move quickly, then the job becomes helping the other departments accept the ideas that are developed by innovators. The handoff of a concept that’s proven viable and now it needs deployment and scale is where...that is the essential weak link in incumbent company business models. That handoff for adoption, deployment, and scale to a department that’s already overstretched, already over budget, and wasn’t their idea in the first place is a challenge.

So, you have to have the executive support, you have to have a set of rules of the road to guide the development of new concepts. You have to have a budget process. Then you have the talent management side, those are the four sides of the box.

Now, you can focus in and open it up to the entire workforce, or to your customer base. One of my favorite innovation examples was people were constantly complaining to Starbucks about spilling their coffee going over speed bumps. It would be very easy for Starbucks or any other corporation to look at the consumer and say, “This is kind of your problem, not mine.” And, you know, what are you going to do, put a shock absorber in a cup holder? No. So they put out globally, an idea campaign asking for answers to stop spillage of cups and cup holders. A guy pulls up at a drive-up window with a popsicle stick he’s carved into a stopper and says, “Here, this is for your idea campaign.” The rest is history.

They had defined what the problem was, they had defined that it was a problem that mattered to enough of their customers to take it up. Then they trusted their customer base to help come up with a solution. A guy at a drive-up window gives them a stopper. That’s when you have the constraints.

Another constraint is put a time limit on how long you’re going to let a concept be researched and developed and converted into an MVP timebox it. For example, no ideas should linger for longer than six weeks. Clock starting if it can’t be done in that timeframe or nine weeks, if it can’t be developed into a prototype or an MVP in that timeframe, probably not something that’s going to be scalable even at a minimalist beta.

So, you have to build in these mechanisms and constraints to help narrow what could be thousands of ideas down to actionable ideas that matter and are commercially viable on an autopilot basis. What constraints do is they put your innovation system not spinning out of control, but keeping it focused, and on this autopilot almost management system, so that you’re constantly intentional about balancing incremental ideas for moonshots to you know, everything in between. But it’s serving a purpose.

Here’s the number one thing I tell companies that call about help with innovation. The conversation goes something like this. “We want to become a more innovative culture we need help.” And my answer is, “Now you’re talking about a law of nature.” The great sage Peter Drucker was right, culture will always win. So how do you be innovative and create an innovative culture without making the culture the primary objective because that is a non-starter? The answer is you get a little success.

Start small, get some success, it could be...success could be a concept that was fully developed, and gets killed because, in the end, it turns out to not matter. Companies have to find a way to stop going to insurtechs or going to technology companies and ending up with a great 2015 solution in 2021 by the time it’s integrated. Well, get around that by taking these constraints and allowing a team or a unit to move fast, place expectations on them regarding how the resources are spent, what’s the return on the investment. Then with even the smallest amount of success that it enables growth.

The culture eventually transforms unintentionally, it’s a byproduct. Success breeds change. Change is extremely hard to accomplish in incumbent cultures. So, reinvent it from the inside out by having success in the marketplace, then the culture changes. Now you’ve used constraints to achieve both economic success, minimize the resource impact on the organization, and at the same time as a byproduct, become a destination for new talent, become a destination for new skill sets. Also become a destination for possibly the customers in a market that you’ve created.

It is not unthinkable by any stretch to have a goal of creating a new market and being the first end on that market before everybody else jumps on board. You can generally start with three to five people and there has to be a gender mix. We’ve done deep research on the growth of startups over a period of time. We analyzed 78 companies a year later...we started with 82 a year later, 78 were still operational, a second year later was 73. But there was consistently the same group of half a dozen that eventually grew to 10 startups that were seeing a 20, 30, 50, 100x growth year over year.

The differential was gender diversity among the executive teams in the startups. I would not invest in an early-stage agency that did not have a good gender mix at the top. That’s equally true with innovation teams. I can’t necessarily explain the science behind it. I can just tell you, it’s an economic reality.

Ron: I think that advice was so powerful, and part of me is worried because I think you in some ways outlined what I would think of as the startup playbook. Startups typically only have three to five people and they only have x amount of money or time before they run out of resources. They’re very constrained and they are not self-imposed constraints, which I think is the difference. These large companies need to self-impose these constraints, because they can invest millions of dollars for two years and if they have a large enough balance sheet, write it off as a small mistake, for a startup, that’s not possible. So, I think you hit the nail on the head. But I’m also worried that the people that listen take it to heart.

Guy: I had the great honor of being involved early on with a ratings agency that built innovation rating into their rating of insurance companies. It’s in the deployment process now. But when we first started the big question was doesn’t innovation favor the giant? And my response was innovation is size agnostic, it’s actually the tool for giant killers. Innovation has a far greater likelihood of being successful in a small insurer than a multi-jurisdictional giant because of the constraints.

So, the ratings agency to their credit and I kind of held my breath while they did this. But you know, spent six to seven months, going around talking to insurers of all sizes. And came back and said, “Okay, we’re convinced, you’re right, anybody can do it.” That launched the development of this innovation rating. It was a huge validation for me, because I’ve worked with companies of all different sizes, and in continents where insurance is established and where it’s not established and have seen it play out. But that’s still just my clients I mean, it’s relatively small base.

So, I have great optimism, I have tremendous optimism for early-stage companies like Chisel AI, and the solutions that you bring to market, being integrated far faster than the earlier...I hesitate to say generation. But the earlier entrepreneurs who attempted to break into the insurance industry and faced these tremendous challenges. It’s not that those challenges aren’t there now. But I think the industry is far more willing to recognize the value of what you bring and the focus with which you bring it.

I do have one small piece of advice for insurance. This is both out of economic reality and dignity. Many insurtechs have faced the question of well, can you please provide us with a free proof of concept? And they do it because they need a proving ground. I would encourage insurers to stop asking that question. If you are committed to change, then put some skin in the game, pay for the proof of concept.

Ron: I agree, and I’ve personally heard that asked many times. I think you hit the nail on the head. Having skin in the game on both sides is what makes for a successful partnership and I don’t think there’s any other way. So, I’d love to ask you, as we wrap up if there’s one piece of wisdom for the next 60 seconds that you’d like to share with our listeners, what would that wisdom be?

Guy: Have faith in the inevitability of change. If you truly believe and your experience tells you for early-stage companies and for incumbents that there is a problem that matters to enough people, then don’t be afraid to learn. Before you make a decision to act, there’s tremendous responsibility with once you know you can’t unknow and what I remind people of all the time is, the worst that can happen is you’ve learned something.

Be absolutely ferocious in protecting your sense of curiosity. Be fearless in curiosity, then you’re making informed decisions. But always be open to acting because that is how the people you’re competing against are approaching it.

Ron: That’s so wise. Guy, thank you so much. Where can people find out more about you and IE Advisory?

Guy: So, there’s, and then there’s also I’m on Twitter, Facebook, LinkedIn, and certainly open to connections. Equally, it’s, or I answer questions all the time. Sometimes they lead to something else and most of the time they don’t, but it’s okay I want to see this industry make the future that it can make.

Ron: I love that. Thank you so much for your time Guy, and stay safe, everyone.

Guy: Thank you, Ron.

Ron: That’s a wrap for this episode of “AI Wisdom” hosted by Chisel AI and me, Ron Glozman. Thanks for listening.

If you like our podcast and want to hear more, check us out at or tune in and subscribe wherever you get your podcasts: SoundCloud, Spotify, iTunes, Google Podcast, or Stitcher

Join us next time for more expert insights and straight talk on how AI and insurtech innovations are transforming the insurance value chain. See you on the next episode!

Browse different topics

Recent Posts