Full Feed

Otakar Hubschmann on AI In Reinsurance

episode link: https://www.buzzsprout.com/126848/episodes/6184525

youtube channel link: https://youtu.be/w99YUTQAaLQ

Why did I do this show?

I have to admit, I’m a bit skeptical of AI in Reinsurance! I find Reinsurance to be very much about medium sized datasets and lots of human judgment, can we do anything better than that?

What did I learn?

One of the everpresent lessons in software is that there are always problems to solve. Every question is one of prioritization of resources and even relatively modest improvements in processes with big $$$ can be very valuable!

What was my favorite part?

Talking real about AI and reinsurance with someone who has legit experience!

Three Puzzles In Insurance

In my career I noticed three strange things about insurance that made it stand out among other industries:

  1. Why does nobody ever want to buy insurance yet it is a massive industry? 
  2. Why do companies and individuals crave outside guidance (regulators, peers, rating agencies, etc) when making decisions about risk?
  3. Why was the discourse in the business so moralistic? 

I used to think insurance was strange, now actually I’m coming to think of insurance as the most human of all business pursuits and it’s all other commercial activity that’s weird.

As a result, I think the deep insight we’ll gain into insurance by answering each of these questions will also reveal much about human nature and, I think, a view into how our culture will evolve over the next decade, century and even coming millennia (assuming we don’t kill ourselves!).

So here are my answers as they stand today!

1. Why insurance exists despite nobody wanting it

I sometimes am surprised I don’t get more pushback on the premise of this statement. In some sense insurance is buying nothing (protection for some imaginary thing that hasn’t happened yet) for something (money). No right thinking ape would ever do such a thing and we are right to hate insurance and be very skeptical of its purchase. Even when we think it’s a good idea, do we really know? Daniel Kahneman won the Nobel Prize for his work showing how clouded our personal judgement is about risk.

But insurance exists because in another sense we actually know very well how poorly we understand risk and our culture has evolved systems for compelling us to protect ourselves. Insurance as we think of it today is just one feature embedded within a whole universe of risk management techniques. They mostly all, however, boil down to one simple concept: our cultural compatriots, our community, protect us from risk. It does this by your neighbor offering a couch if your heat conks out in the winter, by your spouse forcing you to buy life insurance or by the nation state mandating liability insurance for your use of the most dangerous tool humans have ever created, the automobile.

In short we buy insurance because it’s compulsory. The amazing thing about that is that we tolerate being compelled; we are forced to be our better selves by our community.

2. Why do companies and individuals crave outside guidance (regulators, peers, rating agencies, etc) when making decisions about risk? 

I noticed this when negotiating with insurers about how much capital they should hold (in a mix of money and reinsurance). Nobody knows how to answer the question, really. How do you tell how much risk you should take? How do you even think about risk? Who takes it? Wouldn’t it be better if someone just told you the answer?

There are exceptions to this rule in the form of some folks who have unusual conviction in their own judgement. Those types often wind up as entrepreneurs and make these kinds of decisions all the time. The rest of us, however, freeze in the face of this question, and seek answers elsewhere.

The answer to why is that we have literally evolved to think this way. By think I actually mean *not* think, as we usually define ‘think’, and instead mimic authority figures when confronted with uncertainty. Here I’d point the studious reader to a book by Joe Henrich or perhaps my podcast with him.

The key idea is to realize that the world is literally inconceivably complex and a single human mind cannot figure everything out for itself. We use culture as a playbook for thriving in the world, letting us use all kinds of technology we don’t understand by just copying others. This is quite literally the secret to Humanity’s dominance and risk decisions are but one teeny, tiny application of it. Even the most contrarian and independent minded entrepreneur is a mindless, mimicking automaton in the context of the amount of true understanding he/she has for how to navigate the world.

3. Why are insurance people so moralizing?

I do feel the need to defend this statement for a sec. One thing I noticed fairly early in my career (and it always made me bristle) was how viciously insurance underwriting teams criticize their competitors.

Competitor turning a profit? Must be cheating. Competitor lost money? Must have tried to cheat and failed.

Both success and failure are the result of moral inferiority. Insurers live in a continuous test of character. Are you sacrificing the future for a quick buck today? By our nature we are critical of others we don’t know but in insurance this human quality is dialed up to a white hot intensity.

The investigation into morality is probably one of the more surprising undercurrents of my podcast given its insurance focus, even to me! But this behavior really puzzled me and it was a while before I finally started locking onto an answer. And the beginning of the answer came from the work of another anthropologist, the little known Mary Douglas, in particular in her book with Aaron Wildavsky, *Risk and Culture*. I devoured that little book, it was like finding a barrel of water on a desert island.

The key here is that questions of uncertainty are resolved through social decision making, as I mentioned above. We have a term for social decision making processes, it’s called politics. And politics, in its purest form, is rooted in morality. As Jonathan Haidt teaches us in the excellent *The Righteous Mind* the fiercest political debates reduce to competing moral frameworks, to differences of emphasis of values we all mostly subscribe to. The piece that I could fill in after reading Mary Douglas was that morality is the foundation of our technology for making decisions about risk. Moral systems are what guide us in figuring out which risks to focus on and how do deal with them. In ancient societies this means rain dancing and food taboos and in modern societies it can mean what kind of insurance to buy!

I have much more to say on these topics, in particular about how we can use these insights to make a better society. Stay tuned!

Alex Lazarow on Frontier Startups

Episode Link:https://www.buzzsprout.com/126848/4424837

Youtube channel link: https://youtu.be/Axy-4ilhRKw

Why did I do this show?

Alex is an interesting guy! Not only from Canada but also trying to set out a logical structure for an alternate universe for how innovation happens in software. This is no easy task!

What did I learn?

Studying other cultures is studying alternate versions of one’s own universe. It’s always exciting to learn what works elsewhere so that we can bring those lessons home.

What was my favorite part?

That you can be from a small town in Canada and still come to understand startup ecosystems in Africa?!

Robert Hoekman on The Tao of User Experience

Episode Page: https://www.buzzsprout.com/126848/8328832-robert-hoekman-on-the-tao-of-user-experience

Youtube Link: https://youtu.be/-lyxd9E4lJM

Why did I do this show? When I joined the software world I was trying to figure out what made excellent software companies so much more excellent than… well everyone else because everyone makes software in their organization. The best answer I’ve come to is that amazing software companies have an exquisite understanding for what to build and why. That product skill, I think, is most heavily concentrated in the domain of product designers. Robert Hoekman is a designer extraordinaire and has so very much to teach us about building amazing things!

What was my favorite part? As I mentioned in my favorite clip from this podcast, he is relentless and unforgiving in the book. Behind it of course is an excellent guy!

What did I learn? There are deep and simple truths about great design that we can all say and even all believe. But unless we go through pain from not really, really knowing these truths we’ll never understand them.

Samir Shah on Innovating Capital

episode link: https://www.buzzsprout.com/126848/8171216-samir-shah-on-innovating-capital

youtube link: https://youtu.be/diERmzI7YSc

Why did I do this show? Inside the reinsurance world the idea of finding more ‘efficient’ capital carries an almost religious fervor. In my opinion there is a lot of silly thinking on the topic and Samir struck me as someone who combines idealism with a pragmatic and experienced flair. That is rare! I got the chance to take the gloves off and talk about all the things that really bother me on the topic!

What did I learn? learned about the underlying principle of Samir’s strategy, which I have to admit is one I agree with. It’s smaller and much more powerful than I had appreciated going into the conversation.

What was my favorite part? Oh, take on all manner of sloppy ideas about how capital really works in insurance and Samir and I were much more aligned than I had appreciated before. It was a fantastic conversation!

David Soloff on OTT Risk

Episode page: https://www.buzzsprout.com/126848/8216830-david-soloff-on-o-t-t-risk

youtube link: https://youtu.be/QwBFpkxTDs4

Why did I do this show? I love talking to startups run by smart people who think deeply about their new businesses. David was an easy one to talk to!

What was my favorite part? Listening to David deliver a pretty thorough analysis for why triggers for some complex insurance products are incredibly complicated. He gets it!

What did I learn? It’s passion that carries you through difficult times. I mean, this is hardly novel insight, but it’s one of those things that you always need to relearn. David is a passionate as well as very smart, that’s what it takes!

Jessica Leong on Who Actuaries Are

Episode Link: https://www.buzzsprout.com/126848/8725339

Youtube Link: https://youtu.be/9uI7NVo2t0Y

Why did I do this show: I think the actuarial profession is at a very interesting transition point and also caught between new and old school data science. Tradition is their source of institutional strength but they are at a branding deficit compared to mainstream “AI”. Jessica embodies a way forward and she’s the president of the CAS!

What did I learn: I learned that there is a plan at the CAS! Jessica is launching a deliberate effort to move the profession forward

What was my favorite part: I got the chance to touch on the debate that I never got to have about the potential merging of actuarial professions. Someday to revisit I am sure!

GA Bartick on How to Sell

Episode link: https://www.buzzsprout.com/126848/9658359

youtube link: https://youtu.be/TK0ZNnK-od0

Why did I do this show? Sales is underrated! I love thinking deeply about sales and I have learned quite a lot studying GA’s work.

What did I learn? My favorite lesson was in thinking about how sales has changed over time (or not!).

What was my favorite part? I loved GA’s story about how he named his book. What is the silver bullet for sales?

Doug Hubbard on How To Measure Anything

Episode Link: https://www.buzzsprout.com/episodes/8425601

Youtube Link: https://youtu.be/3S2KtvqyBKU

Why did I do this show: Doug’s work is an astounding depth of highly practicable advice on how to make decisions. Quantifying the value of information (you can do that!), overcoming cognitive bias (you can do that!), using actuarial techniques to apply to any decision making process (you can do that!), and on and on. Read Doug Hubbard!

What did I learn: Ok, here’s the real thing about Doug’s work: nobody has heard of Doug Hubbard. The implications of that are mind boggling to me. We have in our hands the way to really improve the world if we all respected and admired Doug the way I do. Yet we don’t! Figuring out why is becoming an obsession of mine.

What was my favorite part: Doug dropped a great phrase on me, analysis placebos. Love it

Bryan Caplan on the Myth of the Rational Voter

Episode Link: https://www.buzzsprout.com/126848/8862288

Youtube Link: https://youtu.be/x6hT6d2qP30

Why did I do this show: I have been studying politics of all things to better understand how we make decisions under uncertainty. Anthropologists that study risk make the link between decision making under uncertainty and the political decision making process. Bryan is such a fantastic person to study because he is an incredibly thorough researcher. You get an integration of almost all thinking in a field with you read one of his books and I wanted to understand politics.

What did I learn: I learned about Bryan’s concept of irrationality and how that captures the real political decision making process that underpins all decisions about uncertainty. There is much to continue learning here!

What was my favorite part: I think my favorite part was linking the organizational political process to Bryan’s work which isn’t something he’d thought about.