In this episode of Hot Topics on the Edge of Show, a leading Web3, tech, and innovation podcast, Christian Davies of Relm joins hosts Richard Carthon and Josh Kriger to explore how the hard-won lessons of Bitcoin mining are shaping the future of AI infrastructure.
From gamified mining platforms like GoMining’s Miners Wars to the energy arms race driving AI data centers, the conversation dives deep into what resilience looks like in today’s decentralized, high-stakes digital economy. The hosts and Christian unpack how power strategy, risk modelling, and infrastructure placement, once niche considerations in the world of crypto, are now critical for emerging tech builders.
Christian also shares how Relm is redefining insurance for Web3 and AI operators, offering custom solutions that go far beyond legacy coverage. Whether it’s business interruption in BTC, slashing insurance for validators, or future-facing products for AI data risks, Relm is building the safety net that innovation actually needs.
Below, we’ve captured the key questions Christian answered during the episode
What kind of frameworks do you think need to exist on platforms like GoMining to scale responsibly?
“You know, I think there is a limit as to what you can potentially gamify in life. It’s an incredible kind of theory, gamification. When I saw about GoMining’s Miner Wars, I was scratching my head, right? But then when you look at 165,000 players, maybe I shouldn’t have been scratching my head.
I guess we just have to really look to see what the longevity is, right? It’s how do you keep people interested. You remember when Axie Infinity first came out, that was global and there were millions and millions of players. And we know how that’s kind of slowly fizzled out. There’s still huge rewards to be made in that. But when you’re looking at the resilience of some of these companies.
Look, let’s do a virtual underwriting session on GoMining’s Miner Wars, right? What is the infrastructure? Is it their own bare metal infrastructure that they have set up that they then have built a gamification tech stack on top of? Is it another third party’s infrastructure that might be running underneath that and they’re gamifying third party infrastructure? Is it already existing mining pools on some of the very large pool — you know — some of the very large mining pool creators out there that then they’re putting that gamification level on top of that? It’s like, how far away are you actually getting from the underlying mining itself? So, you know, a little bit of an insight there as to the process that we’d be going through.
And when you’re looking at the operational resilience, it’s all well and good having 165,000, but you’ve got to keep those people there. If anything goes wrong because you don’t have control of the underlying infrastructure, or the underlying process, the underlying technology, then you very quickly see that 165,000 people very quickly move to something else that’s a little bit more fun, a little bit more interesting. And yes, it’s mining. Yes, it’s a new way of earning revenue. Yes, it’s fun. But if it doesn’t continue to be fun, then where’s the revenue going to come from?”
Christian, for you as someone that’s focused on emerging infrastructure risk, how do you view the convergence of AI data centers and Bitcoin mining? And how do you think that power-based risk modeling is going to evolve this hybrid future?
“Well, for starters, I’m glad you said Kevin O’Leary and not Jim Cramer, because if you said Jim Cramer, this — we wouldn’t be talking about it because it would probably just go to zero if he said it was great.
Let’s just talk about the insurance industry to begin with, right?
It was very, very difficult for crypto miners to get insurance because everyone was like, ‘Oh, ESG, ESG, ESG — nightmare — absolutely can’t touch it, can’t touch it.’ Yet AI’s come and no one’s paying any attention to the ESG side. Bitcoin miners had to work incredibly hard to find as many different renewable resources as possible to kind of get rid of this ESG stigma that they had.
Not to mention, Bitcoin mining didn’t have a huge amount of government support or investment or state support, federal support, global support.
Yet AI infrastructure — you know — the British government, the U.S. government, every other government seems to be YOLOing millions and billions into AI infrastructure and AI data centers. So when you look at the risks we associate with it…
One of the cool things that we did at Realm is we built the first ever Bitcoin-denominated business interruption insurance product for Bitcoin miners. Because what’s the point in paying out a Bitcoin miner in USD when all of the profit they make is in BTC?
You get that massive kind of FX arbitrage problem. And what they need to do is they also need to stop putting themselves all in exactly the same places.
So Richard, you and Josh, you mentioned, you know, like Texas, North Dakota, and Alberta.
I don’t know what this obsession is with putting themselves in hurricane areas and flood areas and earthquake areas, etc., and they’re all relatively close to each other. You get the age-old problem that everyone has had in the property market with hurricane season.
If you look at California now, just from an insurance perspective, it’s close to impossible to get property insurance in California because of the wildfires. And they’re going to need to spread this infrastructure across multiple different jurisdictions across multiple different states because of the operational risks you’re going to have with natural disaster. So that’s something that we obviously look at.
You know, we’re able to build these custom products like I mentioned on the BI — you know — the Bitcoin-denominated business interruption. And they are enormous overlapping categories.
We see a lot of Bitcoin miners now transitioning to AI miners because of a lot of the grants associated with it. But people forget — Bitcoin uptime is like 80%, maybe even 70% — and they’re still making money.
As soon as you’re outsourcing any of your power and infrastructure to AI, if you’re looking at generative AI, that’s like 99.9999% uptime that is required. So, when we talk about operational risk on that, we’re starting to look at SLAs with third parties. It was great whilst you were just kind of mining and printing your own money from a Bitcoin perspective, because you could go on, you could go off, the grid could turn off because of power savings, because of the solar flares or something else that might be going on in the state at the time. Or everyone’s got their central heating turned on because they’ve got this weird freeze going on in the middle of Texas.
Now you can’t do that with AI, because you’re beholden to all of your third-party service providers for providing this AI compute power. And those uptime SLAs are ruthless as well.”
So, how is Realm helping Web3 operators and miners manage risk, and what’s your approach — how is it different from legacy insurers who are still trying to figure out how to put these square pegs in the round holes?
“Let me just talk about our history very briefly, right? We’re six years old. And that doesn’t sound very long for an insurance company. But when you look at the industries that we’ve been insuring — we’ve been insuring crypto for six years. As we know, crypto is a bit like dog years — that technically means we’ve been doing it for about 70 years right now.
We’ve seen in excess of 12,000 companies in over 38 different jurisdictions. And we’ve been insuring Web3 companies for six years across all different lines.
It’s not a case of, ‘Oh well, we only do this part of crypto or only do that part of crypto.’ No, we cover the whole of the Web3 gamut: layer ones, layer twos, DeFi, mining operations, mining rigs, all the full ecosystem and everything that comes with it — the directors and officers of it, the theft of digital assets, the infrastructure, the theft of physical miners, the destruction of mining shipping containers.
When we look at what Realm is — the industries that Realm is insuring — it’s not just AI, it’s not just crypto. It’s alternative medicines, it’s the sharing economy, it’s everything that is going to further the world in what we do — from either a finance perspective or a mobility perspective.
And quite frankly, that’s our interest. That’s our passion. Because the traditional insurance market isn’t really worried about it. They don’t want to learn about this new stuff when they’re already very happy making money on the old stuff that they cover.
The other problem is, you know, insurance has a cyclical nature where you get soft markets and hard markets, right? It’s the difficulty rate for hash rate and then lots of people pile in and lots of people pile out when they start to get burnt.
In the six years, we’ve had people come in and then very quickly leave after they thought this is really easy — have not taken into any consideration things like tokenomics — and then got immediately burnt by massive claims that have come in, like shareholder class action lawsuits against them.
And now we’re getting to a point where we’ve got more markets coming in, and we’ve already seen that within six months some of them are immediately getting burnt. We’ve handled in excess of 300 claims just on the crypto side now. And so we know what works and how it works and when it works on a global basis. So that’s kind of really at Realm’s core, right? It’s all tailored underwriting. We have a massive jurisdictional awareness. We work very heavily with regulators.”
How do you do that in an impactful yet cost-efficient way? Like, how does that work? Because all these new products and services involve analysis and bring a lot of complexity to the table. So, how do you balance all that?
“The cool thing about Realm is obviously we’ve got our global broking network. But our brokers come to us and they ask us how we can help their client.
And we have amazing interactions with our clients — that, again, traditionally classical insurance doesn’t really get a lot of that — because the broker is the one broking the risk to the insurer.
So we have these relationships where we will build completely custom bespoke products because the client is like, ‘Realm, here’s all of our information. This is what we’d like to do. Let’s figure out a way to do it.’
It could be one month, it could be six months, it could be a year of a load of information and data sharing — and we then build a completely bespoke product.
That actually happened on the BTCBI product for crypto miners. It happened three and a half years ago, I think it is now — on the first ever regulated slashing insurance product that we built.
Blockdaemon came to us and were like, ‘Do you think we could insure this?’ We were like, ‘Let’s share some information here and let’s figure this out.’ And we built an insurance product — that is slashing insurance — that exists today.
And most of the staking-as-a-service companies now purchase that. So that’s why we like to be hands-on. We like the information sharing. And ultimately — this is going to sound really strange — but it’s fun. Insurance is not classically a fun thing. And this is really, really, really fun.
It’s constantly innovative.
I referenced the six years of existence. One of the fascinating things about Realm is the first two years of its existence — actually, I believe three years — it was entirely its own balance sheet.
So we really had to stick behind all of our underwriting and our knowledge gathering and our processes and procedures. Now, ultimately what resulted in our prowess of that was — we understand what you guys are trying to do. And if we don’t, we’re here to actually learn from you so we can better support you. So that’s the cool thing.”
