2026 04 28 tftc 740 hard asset lending 99 percent sadi khan

[MUSIC]

You've had a dynamic where money's become freer than free. When you talk about a Fed just gone nuts. All the central banks going nuts. So it's all acting like safe haven. I believe that in a world where central bankers are tripping over themselves to devalue their currency, Bitcoin wins. In the world of fiat currencies, Bitcoin is the victor. I mean, that's part of the bold case for Bitcoin. If you're not paying attention, you probably should be. Sadi, welcome to the show. Thank you for joining me. Like I said, before we hit record, it was like we're recording the podcast already. So I figured let's just get into this. A 100 percent. Well, thank you so much for having me, Marty. I've heard so many great things about you. We have so many mutual friends. So I'm just excited to have this conversation with you, and get to know each other, and talk about new products, and talk about Avon. Yeah. Well, I think more excited because I've been telling your team that a product like you're launching today is something that many people, myself included in Bitcoin, have been waiting for, and it validates a theme that I think is going to be predominant for the next few cycles, which is this intersection of consumer finance and Bitcoin as collateral. That's what you guys are launching at Avon today. But before we get into the Bitcoin specific product, I think getting a background on Avon, how it came to be, and really leaning into and starting with your mission, which is I think very ambitious, but also very virtuous, which is to reduce the cost of capital for consumers, which is very important in these days especially. A hundred percent. As you know, well, Avon was founded on this very simplistic mission, which is our mission of reducing the cost of capital for consumers. We are almost overly serious about it in some ways, and very unserious about anything else. We are extremely focused on this mission, and we believe that the way we can reduce the cost of capital is first by thinking about it from first principles. What is the cost of capital to consumers and to anybody? It's the cost of the risk-free rate, plus the cost of the risk, plus the cost of the transaction. Our strategy at Avon is to reduce the cost of capital by reducing the cost of the transaction, and driving that down to zero by inventing, and developing, and deploying technology. In 2022, we launched the Avon Home Equity Card, which helped us drive the single largest change in the cost of capital on credit cards in US history. We've now originated over $4 billion in lines on our home equity products, and saved consumers over $300 million in interest payments actually. The way we were able to do it is by attacking the largest source of unsecured capital, just credit card debt in the United States, which is over a trillion dollars. Consumers today in the US are paying over 20 percent APL on that, which is over $200 billion a year in interest payments. Our goal is with asset-backed lending, we can reduce the cost of the transaction of borrowing against the assets you already own, and by doing that, reducing the cost of the risk, and passing on that savings in the form of a lower APR to the consumer. Going back again to that point of the cost of the capital is equal to the cost of the risk-free rate, which as you know and I know, we don't control. Today, Jay Powell controls that, and hopefully in the next year or so, our friend Kevin Warsh will have more influence on it than not. But you and I, and by and large, we are not in a position to influence that. The cost of the risk is an information theory problem, where the more we understand about a consumer and the more we understand about the world, we will be able to price that risk appropriately. But the pricing of the risk isn't a strategic advantage for anyone in the long-term. It's an advantage of the core problem that is an information theory problem, which is how much information can we have about the underlying asset, an underlying person, and the world at large. Then finally, the cost of the transaction is where we can invent a lot of technology. So we started off by approaching the problem of home equity, which is the largest asset class owned by Americans today, and extremely inefficient. We developed and built a lot of technology to make that go from four to six weeks down to as fast as 15 minutes, and reducing the cost of that origination by an order of magnitude. Which by derivative then decreased the cost of the risk itself. We can bundle those savings on in the form of a 50-70 percent lower interest rate for consumers, and put that in the form of a credit card, which is both convenient and familiar for most consumers. In our philosophy of unsecured debt being expensive and irresponsible in many ways, because we want to give people credit for what they already own, so we can drive down their cost of capital. We think that also applies to what I, and I think hopefully you also consider the most important future asset class for the world at large, which is Bitcoin and digital assets in general. But Bitcoin specifically, I think I've heard from Garrett that you are similar to me an old-time Bitcoin holder, perhaps longer than myself. I've been a hodler since 2014, not a huge amount. Mind you, probably not as big as yourself, but I've been a believer in the structural value of this. Largely because I grew up in Bangladesh, Zambia, where inflationary currencies are the norm. I think in a lot of circles, there's a lot of comparison between Bitcoin and the value of Bitcoin as an asset class compared to USD, and I think that's a red herring actually. I think the value of Bitcoin is against the Zambian kwacha, and another inflationary currency, and we can go through many of them. I think for many of these people in many of these countries, having a cryptographically constrained asset to help them fight against arbitrary monetary policies is extremely valuable and important in the world, and that's why I've been a believer and a holder of Bitcoin for a very long time, and I could not be more excited about launching and building the product that we're building today, which I'm excited to talk to you more about. Before we get to the product to really humanize the problem here, I think it would be, I'm just curious, having lived through high inflationary periods, lived under currency regimes with the debased currency, what does that do to your psyche? How does that affect your day-to-day living? More importantly, what does it prevent you from doing? Great question. I think there's a few elements to it. One, it enforces and it emphasizes the imports of building systems that do not require you to trust in other human-driven systems in and of itself. I think that's one of the most powerful things where the reality of it is in these politically volatile regimes, the incentive structure for a governing body to make decisions that are inflationary to currency is quite high, frankly. You as an individual in these countries who aren't often governed in the most responsible matter, is often subject to the whims of arbitrary, sometimes in monetary policy, sometimes desperate monetary policy, frankly. It can lead you to a place where not only can you not trust your government, you can't trust even the dollar bills you've saved under your mattress for years. People lose their life savings, people lose generational life savings sometimes, not just what you've been able to save for yourself. I think there's a real opportunity in the world for us to provide an asset class that consumers globally can hold that can give them an asset that they can trust in its value over a long enough time horizon. Historically, as you know, it's largely been gold. If you look back in world history and even today, for example, one of the largest purchases of gold in the world is the Indian subcontinent, right? And in many of these countries, in Africa, India, and subcontinent, gold is the asset class in which you store value. And the reason for that is obviously we've stored a long duration of time, but it's extremely inconvenient. And oftentimes, people will buy and sell and trade in jewelry, whether it be buying it with your savings and keeping that gold in a form that you can wear, which is a convenient way of storage and safety, but also in how they can exchange marriage vows and gifts, right? In Indian subculture, a huge amount of the kind of the workings of a marriage is upon the exchange of gold between husband and future wife or wife and future husband. And the reason for this is gold is oftentimes the store of value and the store of net worth, effectively. And to me, the power of Bitcoin is to be able to do that in a digital form. That removes the kind of the physical security needs. Obviously, there's other challenges we have to solve on physical security and crypto that we can talk about in some more detail. But it solves the kind of need to store, manage these physical assets and be able to borrow and transact with them in some form that is sufficient, that I think is super valuable by having an asset class like Bitcoin. And I think to me, that is like the core of what I value a lot in Bitcoin as a digital asset. And I think that value is tremendous in much of the world. And I think it's easy for us to underestimate the value of that while we are in the United States or a Western democracy, which has a strong and valuable regulatory regime that we can depend on. But in many parts of the world, people can't. And in those parts of the world, I think having asset classes into Bitcoin are extremely valuable. And I think we'll increasingly see that in the world. I agree. And I think one of the beauties of Bitcoin is the fact that not only is it permissionless, distributed, but it's accessible on different scales. You can save a dollar, you can save a million dollars. And I was telling your team, I think one of the big themes that I think is going to really take hold. And it started a little under a decade ago, but it's beginning to gain steam as this intersection of Bitcoin and traditional credit markets, Bitcoin being used as collateral, as a hard asset to get credit on it. And I think in the notes that I have after talking to your team, like really bringing the tools that have been sort of siphoned, not siphoned off, but only for the uber wealthy, the 1% where they borrow against their assets to realize liquidity without having to take on a taxable event or sell assets that they want to hold on to. And bringing that to everyday consumers and making that a widespread practice is an incredible wealth multiplier for people who are holding these assets, don't want to sell them, don't want to take the tax hit. But I've done a good job of being right and doing the judicious thing of saving in a hard asset over time, understanding that it's scarce and solving a problem for them, which is, hey, I have this wealth. I don't want to part ways with this asset specifically, but I would like to tap into it. And that is what you're launching at Avon today. And I think this is a good time to jump into the specifics of the product, because I think you're going to make a big wave because one of the biggest, two of the biggest problems that people have highlighted with Bitcoin collateralized lending to date is the rates, the cost of capital and the relatively short duration that a lot of these products offer. And it looks like you're bringing both lower rates and longer duration, which is going to be a well welcomed advancement in this particular part of the market. That's wonderful to hear that you're feeling as positive as we are about it. A quick one liner on the product that we're looking to launch is we're looking to launch what we hope to be the best in class Bitcoin backed lending product, which is a credit card secured by Bitcoin with the ability to draw 10 year fixed rate plans with an APR as low as 7.99% interest rate. No re-hypo, fixed rates over a long time horizon with LTVs that go to 30%, 50% and 70% of the Bitcoin that you hold with a five year interest only period. We think this should be the best in class Bitcoin backed lending product in the market today, covering the span of cases, large purchases where you'll be able to draw down without a taxable event against your Bitcoin holding for large amounts and a fixed rate with a very low interest rate, all the way up to smaller transactions where you can use the convenience and the cash back and rewards of a credit card. And we think that combination should holistically fill a lot of your needs of having to borrow or utilise and get liquidity on the Bitcoin that you hold. We think that should be a very valuable product to hopefully many consumers. Our own research indicates that about today, people who hold Bitcoin in the US, which is significant already and growing, obviously, about 20 to 30% of people have already tried or have a desire to borrow against their Bitcoin today. And we think we want to build the best in class product for these consumers. So they're able to get some liquidity on the responsible decisions that they've made, which is to buy and hold Bitcoin for a long period of time. And we're excited to bring that to the market. And we think and we hope you will agree and others will agree that this should be one of the best in class products in the market. Let's touch on the cost of capital again. This is your mission is to lower the cost of capital. If you have a 30% LTV, you're going to be able to get a 799% rate out of the gate in the space. That is, I think, the lowest that I've seen for these types of products outside the world of DeFi, which has its own sort of risk, which I think the DeFi world is finding out over the last few days specifically. But what is giving you the confidence to go out at the market at that rate specifically? What's allowing you to do that? That's a great question. Just tactically, there is, you know, one or two things, and I'll touch on that, and then I can touch on the long term view that I have about being able to borrow against Bitcoin and digital assets in general. There is two really important things about Avin. So first is we've built a reputation of building kind of world class or best in class asset backed learning products in the United States over the last, you know, over half a decade. We built the world's best in class home equity backed credit card program. We've had billions of dollars of origination. We build a world class capital markets function that sits behind it and helps us power and grind down that cost of capital for consumers who are borrowing this home equity. We've taken a lot of that expertise in capital markets and legal and compliance, and we're now bringing it to bear on this market, on the asset class, which is Bitcoin. And this has allowed us to navigate the market's ecosystem and the regulatory and compliance ecosystem to bring down our cost of capital in a way that is safe, secure and sound and pass that as a low cost of capital to consumers. We're going to be borrowing against our Bitcoin backed credit card. So I would give a lot of credit to our capital markets team, to our credit team and to our legal and compliance team for putting down this foundation over the last few years actually to enable us to come to market with a product at that price. That's 7.99% interest rate. We're not losing money on that product. And we're making enough to pay our bills and add value to consumers at the same time. And we think that's valuable to the world. And in the long term, I believe this is just us getting started, Marty. If you think back about how we think about risk and how we think about the cost of capital, we actually think in the long term our Bitcoin backed credit card should be the lowest interest rate product and lending product that we offer. Because in many ways, we're taking the minimal amount of credit risk relative to our other products. The ability for us to underwrite this, the ability for us to take position and liquidate that in case the market moves against it allow us to take much less risk on this product than some of our other products. Which in the long term, I believe, will lead to lower and lower cost of capital for consumers who are looking to borrow against the assets that they own. Going back to our theory and our philosophy is we give people credit for what they already own. And if you think about the asset classes we operate in, home equity, now Bitcoin, and other forms of assets we'll work on in the future, of all the assets that we, you and I, can conceive of, the digital asset class is probably the lowest risk asset class to borrow against for the consumer and also to lend against as a lender. And if you think back to that equation of the cost of the risk is the risk-free rate plus the cost of the risk plus the cost of the transaction. The transactional cost should be the lowest and the risk should be the lowest. In which case, if the risk-free rate is equal across all asset classes, the cost of capital should be the lowest for a Bitcoin-backed lending product compared to any other lending product. And we should be driving towards that. Now, I think there's some gap between our-- I would actually even argue our home equity product actually has a lower EPR right now as a headline rate than our Bitcoin-backed products. But I think it's a matter of time before we grind that down. And I'm excited to work towards a world where our Bitcoin-backed product has the lowest headline EPR of all the products that we offer to consumers. All right, freaks. You know me. You know I don't take sponsor money from products I wouldn't use myself. So listen up. The Avon Bitcoin Visa card is one of the most interesting things I've seen in the Bitcoin lending space in a long time. Here's the deal. You can get a line of credit up to a million dollars backed by your Bitcoin without selling a single sat. No gains, no annual fees, no minimum draws, and your Bitcoin is custody by BitGo, which is one of the most trusted names in digital asset security. Avon never lends it out. There's no rehypothecation. You stay in control. And guess what? You can lock in a fixed rate for up to 10 years. That's 10 times longer than most lenders out there or go interest only for up to five years. Rates start at 7.99% APR for a product that lets you keep your stack and still access liquidity. It's hard to beat. I mean, the duration and the rates is the best I've seen in the market to date. You also get 2% unlimited cash back every time you use the card. Spend fiat, keep your Bitcoin, the whole game. If you've been stacking for years and you need liquidity without triggering the taxable event, this is worth a serious look. Go to avon.com/Bitcoin, that's A-V-E-N.com/Bitcoin. Check it out. Sup, freaks? This rep at TFTC was brought to you by our good friends at BitKey. BitKey makes Bitcoin easy to use and hard to lose. It is a hardware wallet that natively embeds into a two or three multisig. You have one key on the hardware wallet, one key on your mobile device, and Block stores a key in the cloud for you. This is an incredible hardware device for your friends and family or maybe yourself who have Bitcoin on exchanges and have for a long time, but haven't taken the step to self-custody because they're worried about the complications of setting up a private public key pair, securing that seed phrase, setting up a pin, setting up a passphrase. Again, BitKey makes it easy to use, hard to lose. It's the easiest zero to one step, your first step to self-custody. If you have friends and family on the exchanges who haven't moved it off, tell them to pick up a BitKey. Go to BitKey.world, use the key TFTC20 at checkout for 20% off your order. That's BitKey.world code TFTC20. - You're speaking my language, Sadi. I've been banging the desk for years because you look at a Bitcoin collateralized loan and if you look at the risk profile compared to something like a HELOC, where you're looking at a piece of collateral, it's a piece of real estate, you have to get it appraised, God forbid, somebody goes into bankruptcy or foreclosure and you've got to figure out how to offload that asset, what price are you going to be able to get for it and you compare that to Bitcoin, which trades 24/7/365. You're using BitGo for your custody, so if you have an agreement between yourself, BitGo, the end customer, for some reason, God forbid, they're unable to make their payments and you need to liquidate that collateral, it can happen within minutes. - And I'll add to that, right? - You can do it on the weekend if you need to right away. - Exactly, and by the way, I just want to add that we take our regulatory compliance extremely seriously as a company. We built our reputation on that over the last five to six years by building and serving some of the most regulatory complex products with a spotless track record on home equity. Our home equity backed credit card is arguably one of the most complex, most regulated products on earth in the financial services. We fall on our mortgage regulations and we fall on our card act and all the credit card regulations and we navigate the superposition of both of those set of regulations and we've done that with great help from our partner bank and from our legal and compliance team in, I think, a very smooth fashion to date. And we're bringing all that expertise to bear. So to your point, Marty, around like the risk pieces, we are going to take this risk piece extremely seriously. We do KYC, we do ability paycheck, we're actually going to do a soft pull on credit as part of our application process, which is a higher standard than many of the kind of lending products that have existed in the market today. And the reason we do that is so that we can bring the kind of advantages of that safer and sounder practice, that more compliant and highly regulated practice to bear so that we can drive down that cost of capital for consumers. This allows our capital markets partners to feel comfortable and safe with what we're offering so they can offer us a lower cost of capital that we can pass on to the consumers. And that is like a very real decision that we kind of intentionally made when we were designing this product. Because again, to our mission, we really want to drive down that cost of capital. And we want to drive it down to be lower than anything else in the market today. And hopefully for for a long time against any other asset class that exists to your point. Why do you think it's taken this long for Bitcoin specifically? Do you think that's actually another way to phrase that is why for Avon is the timing right now? That's a great question. Two pieces. One is we have actually thought about this product since we started the company. So it's not like a new kind of idea to us that came to us at some point in time, we actually thought about this product for a long time, we knew there was a timing element to us being able to do it. There are two elements to it. One is we wanted to establish ourselves first by building, frankly, one of the most complicated products that one could build, which is the home equity backed credit card, which would enforce kind of the full stack built across every inch of the stack from processing to origination to advertising, to servicing, all the way to the capital markets. And we need to build that so that we can build more asset classes that are, you know, hopefully a little bit easier to build over time, but at least gives us kind of a strong foundation to upon which to build. The second piece is there did need to be kind of an increased degree of clarity in the market about, you know, where lending against Bitcoin and other cryptocurrencies exist. You know, I know there was, you know, great debates over the last few years, as you know, Marty better than most around, you know, is this a commodity? Is this a security? How is it going to be treated? And as a company, I'll tell you, at Avon, we generally do not take regulatory risk as a company. And so until we felt that there was clarity in the market around how this would be kind of treated by kind of the appropriate regulatory regime, we were not comfortable with launching and building a product. As that clarity has now emerged over the last year or so, we felt more comfortable about building and launching this product. And the third piece is, you know, we wanted to be able to kind of leverage the infrastructure on capital markets and compliance that we built out over the last few years, and obviously, the technology infrastructure that has come together over the last few years, to be able to build, you know, the best in class product. In a lot of what we build sometimes, in fact, many times, many of the products we build, we're not the first to market, Marty. Our home equity credit card, we're not the first people to build a home equity credit card. 20-25 years ago, Capital One and others actually built a credit card backed by home equity. We like to be, we don't like to necessarily be first, it's great to be first, we just want to be the best, we launch a product. And for the Bitcoin backed credit card, a number of pieces all needed to kind of come together, we need to have the right team, Ceesan and Ani on the team that I think you've met with now Gerrit, or you know, crypto natives, as I like to call them. They both, you know, cut their teeth and building wallet startups and going deep on on L1, L2 chains, and they're the right people to build a product like this. Number two, we wanted to make sure regulatory kind of clarity existed in the market. And number three, we wanted to come in with an advantageous product to your point, we want to have the lowest rates in the market. That's who we take great pride in. We wanted to be able to come in with a product that has the best rates, has the best terms, and has the safety and soundness behind it in order for us to know that we can sustainably and continuously deliver this value to consumers, not just today, but over the next few decades, and build out that infrastructure and provide that value to consumers for a very long time to come. So there were a few pieces that all needed to come together. I wish we could have done it faster, Marty. But here we are today, and I hope, I hope, you know, the world loves it. And I hope everyone uses it and will love it as much as I do. So freaks, when you take Bitcoin seriously, you start with custody, you want to control your keys, avoid single points of failure and make sure your savings cannot disappear because you or someone else screwed up. That is what Unchained has been focused on since 2016. Unchained is the leader in collaborative multi-sig custody of Bitcoin financial services that keep you in control. They secure over $12 billion in Bitcoin for more than 12,000 clients. That means about one out of every 200 Bitcoin sits inside an Unchained vault. Their motto is simple. You hold two keys, they hold one key. It always takes two keys to move Bitcoin, meaning their single key can't access your Bitcoin on its own. Just resilient, shared custody that gives you institutional grade security while keeping you sovereign. Unchained also lets you trade straight from your vault, access Bitcoin-backed commercial loans, open a Bitcoin IRA where you hold your own keys, and set up personal, business, trust, or retirement vaults. They even offer inheritance solutions built for long-term hodlers. Or opt for the highest level private client service with Unchained Signature and get a dedicated account manager, discounted trading fees, exclusive access to events and features, and much, much more. If you want a partner that helps you secure and grow your Bitcoin without giving up control, go to unchained.com and use the code TFTC10 at checkout to get 10% off your new Bitcoin multisig vault. That's TFTC10 at unchained.com. it was $6,157. CrowdHealth negotiated down to $2,309 and we only paid $500. The rest was crowdfunded by the CrowdHealth network. If you're sick of health insurance premiums and having to pay deductibles and getting ripped off at the hospital, join CrowdHealth. It's an alternative way to pay for your health care. It's not health insurance. It's CrowdFunded health care. As you can tell, they negotiate prices for you. You pay in cash. It's much cheaper. Overall, we're much happier. They have incredible perks. Go to joincrowdhealth.com/TFTC to sign up. Five years on CrowdHealth, not looking back. Joincrowdhealth.com/TFTC. Use the promo code TFTC. Once you set up your account, you're going to get $99 a month for your subscription for the first three months. All right, back after some slight technical difficulties, but I think the sidebar conversation is a good addition to the question I was going to ask, which was surrounding how you pick your team. I was telling you that I watched an interview that you did earlier today in which you said you guys hire for intelligence specifically, which many people will say they don't do, but you think is a non-starter when hiring. While we were going through our technical difficulties, we were talking about AI and how we're implementing it at our businesses. I think you made the comment it's going to replace middle management, which I think is becoming very clear to people. I think Sequoia and the block Jack and Olaf wrote that paper together. It's becoming clear that it's going to make the smart humans, the intelligent humans, even smarter and more productive. Yeah, I think there's a few elements to this, right? One is the massive accelerant. It provides intelligent people with a superpower to be more intelligent. I think we talked a little bit about this on the sidebar, but it's worth talking about how we look at hiring talent. A couple of thoughts. One is that we run a very lean team. Today, for example, we're only about 88-89 FTEs in the company. I think we've shared this broadly, but the last milestone we shared publicly is that we generate more revenue per FTE than most technology companies do as a company ourselves today already. We expect that to accelerate, not decelerate as we grow. A lot of that boils down to how much water we expect each of our individuals to carry. That requires them to be highly intelligent, very hard working, and mission aligned in what we are trying to achieve in the world. These three things are the kind of pillars that we look for. I know the intelligence piece can be sometimes controversial to some folks, but it is uncompromising in how we run our hiring process and our hiring bar. The reason for that is you know a lot of these tools, AI tools, we can call them, and I consider LLMs to be the last chapter in a long series of chapters in AI tools that we've provided people for probably over half a century at this point. I think this just feels kind of incredible because of the way process is natural language and that can feel very closer to how we communicate between humans. Therefore, it feels more magical than perhaps an XGB Boost model with a large ensemble of trees in the background, which is finding a node in the web that retrieves your query results. But nonetheless, these tools require, I think, a certain level of intelligence and diligence to use properly. I think we're seeing the trajectory of the intelligent being super powered with these tools. I think the second thing is actually turning out to also be very interesting, Marty, which is those with willpower now are actually unleashed at a degree that was not possible before in the past because they were blocked by their access to information, their ability to gain knowledge and ramp really quickly in a complex domain. The joke that I've said to some people is the only mode left is human willpower. But wasn't that always the case for a long time anyways? I think we may just accelerate that to a certain extent with modern tooling. Another thing that I think is interesting, and I'm not very confident in this theory, but it's an interesting theory, is I wonder if the tools will enable us at scale with sufficient distribution to actually increase the kind of intellect of the average human across the world, across all spectra of IQ by some significant amount. And even if it's like 10, 20 points, that's not something to sneeze at. If you were to take the entirety of the human race and say, we're going to lift up the baseline intelligent quotient SAT score, whatever metric you want to use to assess intelligence and give it like a 10 point, 20 point boost across the entire world, would that not be a much better world? And I think that that would be kind of incredible if it were to happen, and I'm optimistic that maybe it can happen. I'm not necessarily high confidence in this theory, but I think it would be a really powerful, powerful place in the world. I think it's contrarian because I think a lot of people are expecting these tools to make us dumber because we're not going to learn how to do things from first principles. But I think I agree, and it's in a similar vein, but a bit different is this idea. I was telling you I've been leaning into AI. We've been leaning into AI heavily here at TFTC, and I'm lucky to have really good friends, really smart friends who have been playing around with these tools for a while and have watched the progress over many decades now at this point. And one of the best pieces of advice, one of the, I think, the best mental frameworks, particularly when it comes to prompting, is you have this Napoleonic view of Claude code in the terminal. And what I mean by that is, my friend described is, when you're trying to get good output out of these models, you basically have to treat a Claude code terminal like a scribe in a war tent, like Napoleon is at war. He's done battle for the day. He goes into his tent and he's got 10 scribes, one who is sending a message back to refill armaments, one who's writing down his thoughts about war tactics, and another one about philosophy and politic. And it's really, if you use these tools the right way, because to get the output, you need to know the context and what to feed the LM in the first place. And if you're going to get good at these things, you need to have that first principles understanding of what you're actually doing, what you're trying to achieve. Yeah, I've seen this myself as well. So I have like, you know, three quarter sessions constantly up. And I'm probably a fairly heavy user of both actually Gemini and Claude. I find Gemini to be more optimal for queries that require recent web data that I think crawler is just superior today, at least. Whereas I find Claude to be quite valuable when it comes to just like navigating code. I still do a diff review every couple of weeks in the company. Debatably, you know, valuable to the folks on the team. But I find it valid to myself to understand, you know, what like little pieces of code in the company do. So it gives me a sense of where the architecture is, is weak, or it's strong, where there's kind of challenges for an engineer to like write code at velocity with high quality. And one of the interesting things I found is that almost every query I've tried to do with Claude has been okay, not every query, but majority of queries over the last week or two, for example, I'll just pick that I've done with Claude has been like, subtly incorrect in some like, weird but important way. And the most valuable thing I found in my practice of using it is decomposing the problem. Not so that Claude necessarily gets it right, but that I can like manually review the code and be able to identify the mistake, because it has written not too much code, or at least like it's a sufficiently small enough amount of code that I can manually review it and go like, Oh, that's actually what the problem is. And it's very valuable to me, to be honest, because I am obviously not as familiar with the entirety of our code base anymore. And so being able to kind of snipe into an area and be able to, you know, you know, produce a query of some sort is very challenging for me today at this point. So I need, you know, tools like this to be able to dive into an area and understand, okay, if I were to ask this question, how would the query be generated, but I have enough context to be able to look at a small decomposed piece of that query and go like, Oh, actually, that that's probably the wrong column, actually, what this join doesn't quite is not going to quite hit what I'm looking for. And correct it. But yeah, that decomposition piece and I think you I mean, I'm sure you know this, but Napoleon was notoriously famous in his instruction set in that how direct and how short they were his instruction sets. He was known for being very quite voluminous in the amount of instructions he was sending out the low latency of his instruction to delivery and the brevity of his instruction themselves is extremely famous for that actually. And one could argue that is actually the most effective way to use these tools today is to have short instruction sets that are extremely direct and precise. And and I think that your metaphor is I think quite apt for for the current AI systems. I will agree with that, actually. Yeah, I can't take credit for for creating it. But as soon it was as a good one. As soon as it was related to me, I was like, Oh, this makes a lot of sense. And it changed the whole way I interact with it. Yeah. And I think it's it's been like, quite, I think the big shift for me was actually Q4 of 2025. I think that was when I first saw MLMs do math correctly. Prior to that, I think it was just like, wrong and math all the time. And, you know, I just don't trust people or systems that don't do math correctly. I still think it, you know, obviously has, you know, mistakes that it makes. But I do think it, you know, was like a step function improvement in the quality of its like logic and math trees, which I think was like very meaningful and its ability to be used in environment for us at even, I think the the challenge we navigate, and we'll continue to navigate, frankly, is that LMS are largely high recall low precision systems. You know, like, chat GPT will always give you an answer to a question, it's just gonna be wrong a lot. That doesn't work for a highly regulated product in a highly regulated company like ourselves. So our investments are a lot about evaluation systems, and how do you constrain the system to operate in high precision and low recall, and we sacrifice recall to kind of get it to be, you know, answering fewer questions, designing these questions correctly. And then the second thing that we focus a lot on is using it instead of as, you know, direct to consumer directly facing to external parties, as internal tooling for our own, you know, employees and then and human agents. And I think we were speaking about this earlier is that we I call it we invested an enormous amount of energy and engineering infrastructure on building what I call like a really fancy I don't know engine. And we spent an enormous amount of time building out engineering infrastructure that effectively allows our agentic system to say, actually, I don't know, instead of having a bullshit answer, it says I don't know. And this, you know, is effectively what you know, the fancier way of saying it is that we sacrifice recall for high precision on the recall P&R curve. And we operate at a very high precision point in the threshold of the system. And we sacrifice a lot of recall, and then we deploy it to our human agents to help them get feedback. And you know, our evaluation of our agents shows that we're usually right in our answers about, you know, 85% of the time as a human, and our agentic systems are in operate at about a 95% position, obviously, a lot less recall or humans, but for the recall that we get, which is, you know, we sacrifice a lot of recall for that, but we're able to get to about 95% position and give our human agents feedback to help them be more accurate and precise and improve. And in our deployment of our machine agents that are giving our human agents feedback, when we deployed it that month, we saw the single largest improvement in our human agents quality. And I think this, you know, is kind of the building blocks that we even are building towards, you know, how much of a consumer financial banking platform can we can we run with machines and make it more efficient, so that we can take that efficiency savings and pass to the consumer in the form of a lower cost of capital. The theory that we have around this is something I call, you know, us seeking the Carnot efficiency of an industry. Are you familiar with Sadi Carnot, the physicist? So we obviously share the same name, Sadi, and Sadi, but he has a longer, you know, first name, but Sadi is I think somewhere in his middle name, but I call him Sadi Carnot, other people call him anything. And he produced this theory called the Carnot efficiency of an engine. And it basically, you know, is a fairly simplistic formula that calculates what is the maximal thermodynamic efficiency an engine can run at. And that is kind of the upper limit you can ever build on any engine, like a thermodynamic engine, that has the maximal efficiency you can run that engine at. So you kind of constantly as a mechanical engineer or an engine builder trying to, you know, even remotely approach that degree of efficiency. And in some ways, that even we're trying to find the Carnot efficiency of banking, which is, you know, what is the upper bound of like how efficiently you can build consumer financial products? And how much of it can truly be machined? And if we can machine it, do we get more observability? Do we become more compliant? Are we actually obviously lower costs to consumers? But are we also higher quality consumers by doing that? And I think all these things need to be true. And, you know, we take great pride in, you know, we're one of the highest rated on trust model for any HELOC lender in the United States. We take great pride in our NPS scores, we take great pride in, you know, how much of a clean kind of historical track record we've maintained on our compliance fronts. We're a very regulated company. And so we go through, you know, something like two dozen state examinations in the last 12 months. And we kind of navigate that with with, I think, a stellar record. Many thanks to our compliance and legal team. And many thanks to also our engineering team who invest a lot of, you know, technical firepower on building a highly compliant piece of infrastructure for us to be able to operate these products at scale. And I think we're getting at here to something that I've really come to appreciate as I've been doing research about even how you built the company, it's taking something like that's existed in traditional finance as relatively boring. But to your point, like has all these inefficiencies that exists because it was built, the primitives were built in an era by companies that were built in an era that wasn't really back forward and have all this institutional regulatory sort of calcification built up over time. And you have these behemoths in traditional finance that are oil tankers that can't, or warships that can't be as agile. So starting from scratch, and bringing a tech lens to this part of the market, starting with HELOCs, or an asset lending, more broadly, seems like it affords you guys an advantage in terms of what you can actually do from an efficiency standpoint, then that efficiency passed on to the consumer with lower rates. 100%. If you think about like, the core of what we've done, right, is kind of not rocket science. Please, you know, HELOCs have existed for over half a century. Credit cards have existed for three quarters of a century. Neither of these are new products in the world. What we've really done is grinded down the efficiency of HELOC origination, and put it in the form factor of an accessible, you know, product that consumers are familiar with. And we've taken the risk savings of a HELOC and passed it to the consumer as a lower interest rate. I would actually argue for the first few years of AVEN, our risk models for our HELOC were really no better than any other HELOC lender in America. We used FICO, debt to income ratio, and loan to value ratios as the core of our hypothesis of pricing and line sizing. And sure, over the years now, we've built much more sophisticated models and statistical derivations of how to price consumers and how to line size them. And that has reduced, you know, credit risk tremendously. But our initial kind of like versions of our product were actually fairly basic and work really well, actually. And, you know, the real magic, the real secret to AVEN was the origination efficiency of a HELOC, not actually, you know, significantly or a magnitude better credit risk underwriting. It was actually the fact that we were able to originate a HELOC really efficiently and service that HELOC really efficiently, that structurally reduced the risk, and thereby structurally enabled us to reduce the cost of capital for consumers. And that's, again, a similar kind of vein in our Bitcoin backed product that we're bringing, which is like this asset class is structurally lower risk for the consumer. So how do we enable a really great capital markets function to operate on top of this asset class with all those kind of legal compliance frameworks that enable us to kind of continuously grind down that cost of capital, and then bring that value of a lower interest rate to the consumer in the form of lower APR, and put it into the form factor that the US consumers are familiar with and understand, which is, you know, credit card and, you know, to similar to our home equity product, you know, the credit card just happens to be one of the access tools for this line of credit that we give a consumer. Just similar to our home equity product, we have the ability, you can literally click a button and get a 7.99% loan over 10 years with a five year IO in a click of a button, and that you can get to your deposit account of your choice. I would recommend an Avon deposit account that is now available to consumers as well. We offer great yields and great cashback on deposit account with a beautiful metal cart, I might add. And so we think the credit card is just one of the access tools, I think it's a very powerful access tool, it allows you to use it on a day to day convenient fashion. But by no means, the only way to access this line of credit, and you should use it, you know, for large purchases, you know, oftentimes being able to draw down a line with a fixed rate of a long time is super valuable. And I think I hope to see consumers use it for both of those cases. So you have the HELOC Beachhead product, now you have the Bitcoin lending product as well. What is your long term view of where Avon's going? What is the end state of this business? That's a great question. I mean, in the very long term, we want to serve all the customers in the United States with the best financial products that we can build for them. And I think that should span the majority of their financial needs. And I think there's four or five major needs that a consumer has. One is financial products that are based on their home as the core asset class. That's mortgages, it's HELOCs are the two kind of major products in that space. I don't know if you know this already, but in just the last few months, we've started rolling out our tests for for our mortgage products. And we have one of the fastest cash out refinance products in the world today. And that's in testing, we're very excited about it, it's doing really well. And so there's a lot there's a lot more to come on that on that front. I think other asset classes, as you can imagine, we're starting with Bitcoin, we're very excited about it, because we've been believers in the long term on this asset. And we'll bring it to more assets, both digital and non-digital. And for example, I think, you know, as you can imagine, if we can do it, if you can borrow against your Bitcoin, why can't you borrow against your stock in in a way that is accessible to to us consumers? You know, the 1% are able to borrow against their stock, Elon Musk was able to borrow against his stock to buy Twitter. Why can't why can't you and I? And how do we bring that that product to consumers in a way that helps them reduce their cost of capital, while helps them and encourages them to actually invest and build in assets that accrue value over over the long time and be able to borrow against it at low much lower costs than they would otherwise? Yeah. And so there's a lot more to do. And I love to hear it because I'm gonna choose my words carefully because I don't want to. I don't want to besmirch anybody specifically, but I think in the world that we live in today, where if gambling apps are top of mind for a lot of people, they're every I feel like any sports game you put on or any billboard you see it's trying to push you towards gambling. In the broader world of crypto, there's a lot of what I would need to be degenerate speculation that is pushed on on people. And I think what people really just need is better financial products that that actually makes sense. And I think hard asset lending to me at least, could be wrong. Don't think I am. Yeah, makes a lot of sense. You own the asset, it's not unsecured. People are adults, if they want to take risk by taking out a loan, they can. But there is sort of an ethical framework in my mind that people should operate within seems like, correct me if I'm wrong, that's how you're approaching it too. Let's do sensible lending with sensible structures. That's exactly right. And look, I think we happen to be fortunate to live in a country that has sensible regulations around this. For example, in in asset back lending, unlike unsecured lending, by the way, where if you're unsecured credit card unsecured personal loan with 20-25% interest, you know, the irony of this is that unsecured credit cards, unsecured personal loans, none of them do any income verification. None of them are checking, do you actually have the ability to pay back this loan? That's something that's really important to us. Every single customer given the day we started, goes through income verification, which we must make sure that you have the ability to pay this this loan on our home product, that's actually a regulatory requirement. We will also do ability to pay a sting for our for our Bitcoin backed product as well to make sure you aren't in this position where you're just going to have to liquidate your your Bitcoin, we want to make sure you have the ability to pay back this loan that you're taking, so that you can be responsible in how you borrow with with the even Bitcoin product. And we think these kinds of elements make asset backed borrowing much more responsible, much lower cost for consumers. And I think they operate in I think, healthier regulatory regimes for us consumers. Agreed. Well, thank you for your time and for putting up with the technical difficulties. I am excited for you guys excited for Bitcoin, excited for Bitcoiners. Like I said, I think many of us in the industry and Bitcoiners more broadly have been looking at this sort of Bitcoin collateralized lending part of the market and saying that this is number one, we think going to be big and number two, it would make sense if the cost of capital was lower because of the risk involved with Bitcoin backed lending compared to other asset lending structures that exist out there. I think even coming to market with this product at this cost of capital is going to raise the eyebrows of many in and outside the industry and hopefully a rising tide lifts all boats and you get more sensible products across the board, more sensible rates across the board because of this. So thank you for putting in the work and for building this out. I'm excited to see this at the market. Thank you, Marty. It's been a pleasure. Thanks for having me on the show. And many thanks, frankly, to our team, who would have made this happen and the hard work of CSUN, Garrett, Ani, Gerald, and so many people on the team that helped build this product and bring it to market. And look, Marty, all I can promise you is that we're just getting started. We think there's a lot more to do. We think that 7.99 is just where we're starting. My goal is to keep grinding that thing down with hard work and and perseverance. And hopefully our Bitcoin backed products will have the lowest interest rate, not just amongst our portfolio, but in the world. You heard it here first. Brilliant. Marty, a pleasure. Thanks for having me. Thank you, Sadi. Peace and love freaks. Thank you for listening to this episode of TFTC. If you made it this far, I imagine you got some value out of the episode. If so, please share it far and wide with your friends and family. We're looking to get the word out there. Also, wherever you're listening, whether that's YouTube, Apple, Spotify, make sure you like and subscribe to the show. And if you can leave a rating on the podcasting platforms, that goes a long way. Last but not least, if you want to get these episodes a day early and ad free, make sure you download the Fountain podcasting app and go to fountain.fm to find that $5 a month gets you every episode a day early ad free helps the show gives you incredible value. So please consider subscribing via Fountain as well. Thank you for your time. And until next time.