One small step for drones, and no steps needed for mankind.

Autonomous drone transport for humans. Why not?

We’re eventually getting autonomous cars so you don’t have to do the driving, and drones are all the rage—although Amazon CEO Jeff Bezos’ vision of drones delivering packages to your doorstep hasn’t come to fruition yet. So combining the two makes perfect sense.

Relaxing in a drone on the way to work, or home from a bar, is a fanciful, but not all that unlikely, futuristic scenario posed by Marik Brockman, corporate strategy executive at Walnut Creek, Calif.-based CSAA Insurance Group, an AAA insurer.

Brockman took some time to speak about the impact of technology on the insurance industry and how the industry will be shaped by technology in five, 10 or even 20 years. The conversation was edited for brevity and clarity.

There seem to be new technologies and ways to do things popping up all over the place. How well is the industry handling this innovation? Can you give the industry a letter grade and then tell us why you gave it that grade?

Brockman: We always refer to it as the quickening pace of change, and now the application of these technologies is coming at insurance, and really it’s attacking multiple parts of the value chain. I think that’s because technology is largely an enabler until it’s a disruptor.

Now we’re seeing technologies brought together attacking parts of the value chain, and you can call it literally the quickening pace of disruption now. We’re paying close attention, and as far as a letter grade for the industry, I like to think this is going to sound unduly harsh, but in some respect I’ll give it a big fat C, for a couple of different reasons. It gets an A in angles of investment in core technologies that help companies run.

I don’t think the outside world always appreciates just how complex insurance is and how many systems are at play. These systems have been used and added on to for four decades, five decades in some cases. It just depends on which systems you’re talking about. Meanwhile, we’re managing millions of policies which are actually contracts, and they all have actually dozens of rating factors blended into the pricing, and all these things that make it a pretty complex industry.

So I give us a lot of credit for constantly trying to modernize. But then there’s dealing with the customer experience and bringing new technologies faster in the way that customers interact with us, and that’s when the grade starts to fall a little bit. That’s where as you talk to millennials, and I heard one millennial describe it–who by the way is launching a startup aimed right at insurance–saying, “I felt like it was the darkest part of the Internet. I didn’t understand what was happening, what language was spoken, why the experience was so antiquated, and I just wanted nothing to do with it.”

Then you talk about how quickly we can adapt to some small startups that are really coming after us now. I’ll just put us right around a C right now, maybe a C+ on a good day.

Just how is insurance handling or going to handle autonomous vehicles?

Brockman: Insurance, just like financial services, are partners in helping develop not just economies, really countries’ economies in an industry, and then really people’s opportunity. We’ll always be there to help partner in the development of, in this case, the evolution of this industry. What’s tricky is the shift in risk from personal liability to product liability. It’s really which carriers are involved or how any one particular carrier tries to adapt the kinds of risks that it would like to cover.

So I’d say we’re all trying to surround this, and understand it, and partner with everyone from car makers to dryers to other technology makers, and just understand what’s the best place for us to play, and everyone’s finding their way through this great sea change.

What other innovations do you feel will impact or maybe significantly change the insurance industry?

Brockman: If we talk first about the enablers, we think about for the insurance industry. I spend a lot of time in property and casualty now, working for a P/C carrier. I spent some time in life as well in prior years in consulting. I think some fun ones to talk about—I’d say wearable technology.

That’s really useful on the health and life side of just things that help you monitor your activity and perhaps your heart rate, and things like that. It’s keeping people active, and it kind of turned fitness into gamification. Also much more information to use for both improving your own behavior as well as companies using it for perhaps rating you more appropriately or consulting you more appropriately.

On the property/casualty side, we’re looking more and more closely at drone technology. On the underwriting side, you can survey properties much more efficiently, much more accurately, and you can survey more dimensions than a human could on their own—temperatures and structural elements, as opposed to just what the naked eye can see. On the claims side, you can assess damage in certain territories where disasters have occurred, which can immediately result in not just more accurate information but faster claims payout, which is what we want for all of our insureds.

How will insurance look in five years or 10 years? Say, for example, in 20 years will a small business owner be buying workers comp on Amazon with his iPad Air 25 while riding to work in the passenger seat of his autonomous vehicle, and behind him is a paying rideshare passenger?

Brockman: That’s good. I like that scenario there, I like that a lot. Quick answer is going to be “Yes.” Actually, yeah, it just depends on if iPad’s still just not integrated into their headwear now. Just to get to your question, I think for five years it won’t be ridiculously different, but there will be some changes that probably came sooner than we think. So how many levels of cars are allowed on the road, or are they confined to certain areas?

Just to clarify for folks, level 1 is some level of autonomy. Level 2 is two systems at once, so it can steer, and brake, and accelerate for you. Level 3 is you could go for a long period of time, pretty much put in an address and the car will take you there, but the car has to be able to give us back control. We expect that will happen in places where weather becomes inclement, or the conditions aren’t what the maps said it was, etc., so the driver really needs to be there, so the car would still look a lot like what it does today.

Level 4, it may not look like a car. It could take on a lot of form factors. It doesn’t need anybody to drive it. It’s understood that it can handle the environment in any condition for the use that it’s been deemed. So that’s all exciting; we’ll be somewhere on the spectrum farther than we are right now. By the way, 0 is just the human driving, no real automation at all, and that’s a lot of what most people have is some version of 0 or 1. So we’re already at 0, 1 and 2. Within five years we’ll have some versions of 3, absolutely. It’s just will 4 be allowed, and if so, how much and where?

Then also use of drones will become much more commonplace. Machine learning and artificial intelligence will be (where) most of the top carriers will have worked out some application.

We’re already doing predictive analytics and some aspects of artificial intelligence, but machine learning is really interesting—where you can describe a new routine to a machine, and then in terms of the inputs that it needs to learn, you can teach it by telling it which inputs are what you mean, and which are not, and then work the routine until it’s doing exactly what you thought, and then it doesn’t need you anymore. You just taught the machine. You didn’t have to code it; you used the code that was already there.

You string those together into forms of artificial intelligence, and then there’s a whole bunch of jobs we think that don’t have to happen. In five years it will be just brushing the surface for certain applications, but we’ll certainly be getting closer. Ten years? You’ll start to see level 4 all over the place—cars, autonomous. You’ll start to see drones doing a lot of work that people used to do, and then artificial intelligence will be embedded in a lot of different things, such that you won’t always need to deal with folks for routine interactions with companies and services.

Now, 20 years? It might be a fool’s game to try to get into that one too deeply. I liked your scenario pretty well. I’ll throw one out there that’s kind of fun to think about. If we have autonomous cars already starting, and drones are coming, we like to think about autonomous drone travel for humans. That’s something that we could think about—why not take it to the next dimension and get us out there?

There’s some various experiments already on the horizon, so the degree to which that could be commercial or even personal in 20 years is quite potentially likely.

(This article first appeared on InsuranceJournal.com. Insurance Journal and Carrier Management are both part of Wells Media Group.)