← Blockspace: AI & Bitcoin
Blockspace: AI & Bitcoin
Why Bitfarms Rebranded to Keel, and What This Means for Its AI Plans w/ Ben Gagnon
00:45:30
Ben Gagnon joins the show to discuss Bitfarms' rebrand to Keel, its redomiciling to the U.S., and its AI plans for 2026.
Loading summary
Transcript45 lines
- [00:00]
Host
Welcome back to the Blockspace podcast for the bonus show brought to you by CleanSpark. Bitfarms is no more. The company didn't go away, but it did rebrand to Keel Infrastructure. And it re domiciled into the US As a part of the rebrand in an effort to consolidate its focus on AI and HPC and bring the company home to the country where it has
- [00:22]
Co-host
the majority of its operations.
- [00:24]
Host
To discuss the rebrand for today's show, we have CEO Ben Gagnon on to talk about the redomiciling, the benefits it brings for bit farms, index inclusion and capital markets access, as well as its AI pivot and when investors should expect a lease announcement. We also touch on what all of this means for Keel's bitcoin mining operations
- [00:45]
Co-host
as it moves into AI and HPC infrastructure.
- [00:48]
Host
You're listening to the Block Space pod
- [00:50]
Co-host
and we'll be right back.
- [00:56]
Host
Ben Gang Yon. Welcome back to the Blockspace pod.
- [00:59]
Co-host
Sir, it's been a minute. How you doing?
- [01:01]
Ben Gagnon
I'm doing great. Always great to be here and really happy to be able to talk about our Q4 earnings, our US redomicile or rebrand Today. We've got a lot of really exciting things to talk about.
- [01:13]
Co-host
Yeah, a lot of irons in the fire. And we'll lead with that rebrand because we're fresh off that news. We're recording this on April 2nd, I believe y' all, finalized yesterday or two days ago. The rebrand and Bitfarms is now keel Infrastructure. So what was the driving decision behind this, Ben? I mean, when you're looking at where bitfarms was a few years ago, still mining bitcoin, now looking ahead towards an AI expansion, what was the driving decision behind this rebrand?
- [01:41]
Ben Gagnon
So, you know, we've always been known as, I think, the international bitcoin miner. And as we transition off into this new business, we really just want to be focused on doing what we do best, which is developing power, land and infrastructure. And when we look through kind of the options that we had, you know, HPC and AI is obviously the highest value option you have on the energy infrastructure side. That's where all the value creation is. And we don't want to be a bitcoin company anymore. Right. All of our bitcoin mining exposure, we're going to continue to wind down all of our bitcoin. We're looking to reinvest into HPC and AI. We actually made this decision a while ago, but because we have built up such a sizable portfolio across the Americas, it takes time to Go through things in a disciplined way, evaluate the assets, sell off assets, starting to wind down exposure. So we've been at this process for about 12 months now to redomicile, to transition, to wind down the bitcoin mining exposure, sell off things like Iguazu and really, really just focus the company down to one thing. You know, we don't even want to operate the infrastructure, we just want to develop and own it. So it's a very, very clear focus for the business. And you know, with that we want to get away from the name Bit Farms because it has such a strong bitcoin connotation. And that connotation really doesn't reflect the business that we've rebuilt and who we are now and what we're doing rolling forward.
- [03:19]
Co-host
So before I get into what this materially changes about the company beyond the name change, I do have to ask this. I was looking through Yalls press release about the rebrand and it mentioned retiring Bitform share on the Toronto Stock Exchange and on the NASDAQ for Keel shares. Was that a difficult process in terms of having to get that change? Another way to ask this, what was the thing that gave you the biggest headache during this name change in terms of the procedure to go about cutting through all the red tape to make sure that this is above board for markets and for regulators.
- [03:53]
Ben Gagnon
So, so doing a rebrand relatively straightforward, it's not a difficult regulatory process. But you know, there's a lot that goes into the creativity and the thought behind the new brand and how you want to position yourself and how you want to present yourself. Really the challenges are around the regulatory for the redomiciliation and that is incredibly difficult. We believe we're the fifth company in 15 years to redomicile from Canada to the United States. That's what our lawyer said. That's what AI confirmed with me. It is a very, very rare thing for Canadian companies to read domicile into the US because it's hard. Right? But there's reasons why you do that and there's, there's really good reason on our, in our case because it gives us huge access to US investors that we really didn't have as a Canadian company. And it really gets us that broad index inclusion. So things like Russell, which is, you know, a huge part of most of our, our peers cap table we weren't even eligible for and because of the re domicile we're now eligible for all these different kinds of indexes in the US which really brings in a lot of high quality passive institutional investor money into the business. And the last thing is really on the accounting standards. So we were previously under IFRS as a Canadian company, while all the other peers are under gaap. Makes it really difficult for the investors to compare apples to apples, the two different companies, because the, the fundamental math is pretty different. You know, one of the things, for example, under IFRS is we don't follow fasb. So this kind of quarter to quarter markup, markdown of bitcoin price and what that means for the company's net operating income and adjusted EBITDA and all these other things, that wasn't being reflected under ifrs. And so our numbers always look so different. They were so difficult to compare with our peers.
- [05:53]
Co-host
So when you said it's difficult to redomicile, why is that? Is it that Canadian regulators don't want to give approval for that? I saw that y' all had to get approval from British Columbia. Or is it on the other side in the US where they don't want you to redomicide? What makes it difficult?
- [06:09]
Ben Gagnon
All of the above. It's US regulators, it's Canadian regulators, it's going through taxation implications. There's all sorts of challenges associated with moving from one country to the other. So it's a very, very rare thing that people do. You know, there's no one particular thing that you can point to. It's 15 months of complicated thing after complicated thing. So super huge, heavy lift pulled off by the team, you know, our legal team, our finance team, our marketing team. It requires everyone in the business to, to pull off a re dom and a rebrand, especially doing it all at the same time. Incredibly complex, even small things like the ticker change. You know, there's, there's challenges there because the US it could have been effective April 1, but in Canada it takes two business days. Right? So even there you have to, you know, adjust. Right. So now the step ticker is going to be effective on Monday, April 6, because that's the first day that it can be effective on both exchanges at the same time. But it's not possible to do it at the same time on the US schedule because the Canadian rules just follow a different pace. So there's, that's just one small example of hundreds of things that you need to work through with the lawyers, with the accountants, with the auditors in order to go and get signed off and get something like this through. But we pulled it off. I can't say it was, you know, without a hitch, but we did it on time as planned and incredibly Happy with the results. You know, we're now going to be eligible for a lot of different indexes. And as we look forward at this year, we've got a number of catalysts that we're looking forward to execute on. And so having this structural foundation in place really just enables us to move forward a lot faster, a lot easier and a lot more efficiently.
- [07:59]
CleanSpark Announcer
We are CleanSpark, America's Bitcoin miner. A publicly traded company with the largest operating hash rate powered entirely by self operated infrastructure across four states. This is our proof of work. We are setting the standard for what's next. Learn more about the intersection of energy and bitcoin@cleanspark.com so you already mentioned, I
- [08:31]
Co-host
think, some of the material benefits of this rebrand outside of just the name change. Just then you talked about index inclusion. You also talked about capital markets access. Are there any other benefits to this rebrand that will functionally change how Bit Farms operates?
- [08:48]
Ben Gagnon
No, not really. I mean the rebrand is not really changing who the business is. I think it is more reflecting the changes that we've already made in the business and who we've become. Right. So just changing the name doesn't really change the underlying, it's a reflection of the work that we've put in for the last 12 to 15 months to, you know, rebalance our portfolio to the US scaling backlight, am dramatically increasing the megawatts in the United States, you know, cleaning up our cap table, raising hundreds of millions of dollars, you know, bringing in the right people from the right different teams so that we've got the right people across power permits, engineering, you know, all these different areas. It's really a reflection of what we've done. I don't think fundamentally the name change really changes much. It's just a recognition of all the different hard work and accomplishments so far.
- [09:44]
Co-host
What about the redomiciling? Any other benefits outside of capital markets and index access or I mean really
- [09:50]
Ben Gagnon
it's, it's a lot of simplication there, especially with the regulators. There's some things that you can't even do from a financing perspective in Canada that you can do in the United States. And so, you know, there's great things with investor access, there's great things with our ability to finance. There's, there's a lot of simplification there for us as a business and as a management team. And you know, we think it was, it was well worth it. It was certainly, certainly not something that I'd want to do a second time. But I'm really glad that we pulled it off this time.
- [10:21]
Co-host
Yeah. And it makes a lot of sense to me given, you know, Panther Creek, Sharon and the Moses sites. I mean, all of these are in the US doesn't make any sense for you all to be operating out of Canada if your infrastructure is actually here. We'll get to those in a second. And some of the plans y' all have for them, for repurposing them for AI workloads. But before we get to that point, I want to double tap on the bitcoin sale really quickly. Makes a lot of sense. We've seen this from a number of miners. Cypher's doing it. Mara just sold like 15k bitcoin bit deer, already divested of all their holdings. Can you speak to this decision, though, of why you're divesting and exactly why it's happening now? And if anything specifically has changed in the market, either for bitcoin's price or in the market generally for how investors and traders are approaching bitcoin miners that drove this decision.
- [11:11]
Ben Gagnon
So, yeah, we've actually been looking at this since Q3 and Q4 of last year. And to clarify, we haven't actually sold any bitcoin so far this year. So we're Sitting on about 2,400 bitcoin right now, and that's continuing to grow a little bit every day as the mining operations continue to produce bitcoin. Our goal is to sell off the bitcoin and reinvest those dollars this year. But because we've got such robust financial liquidity, we're sitting on over $500 million, most of which is cash. We've got far more than what we need to move forward over the next couple of months through closing out the final permits, moving through the final stages of lease execution, and getting to the point where financeable lease and continue the construction with that extra financing at hand. So we've got a lot of robust liquidity, which means that we can take a really disciplined approach to selling bitcoin. We're not going to be selling 2,400 bitcoin tomorrow just because we have 2,400 bitcoin and we want to move on. We're taking a disciplined approach because we have that robust liquidity. And so when there are opportunities to sell into strength, we're going to be looking to do that. How? And exactly the cadence, we still haven't decided. We're looking at all sorts of different options. And, you know, bitcoin is something where the market can change on you on a Dime we still remain optimistic on, on bitcoin price but it's not where we want to position the business. So really it's now just about, you know, do you want to sell your bitcoin today at 70k when you know, 5 months ago was at 125 just to sell it or do you want to say look, let's set a discipline approach, let's put it out there that we do want to sell the bitcoin into strength. Let's look for those opportunities to do so but not feel pressured to do so. And you know, we may do something like sell the bitcoin off kind of on a monthly basis. Over time we might sell the bitcoin all in one, you know, or two goes. If the price has really, really nice action, we might do something like sell off some of the bitcoin and buy some of the. Some long dated calls to get us, you know, retain some of the upside if bitcoin continues to rally further. So no exact decision has been made on how we're going to sell our bitcoin and when. We don't have a specific date in mind, we don't have a specific price target in mind that we need to hit. Just an overall objective that we would like to continue to wind down the bitcoin mining. So you know, in the second half of this year bitcoin mining is really going to start trickling down as we're executing in sites construction to the point where we probably are going to end the year with maybe around four or five exahash and that will all be located within Quebec and at our scrubgrass location and everything else in the United States should be, should be offline. So that'll continue to trickle down and we'll continue to work down the bitcoin until we've, we've invested all of it in HPC and AI.
- [14:08]
Co-host
I want to earmark what you just said about the wind down of bitcoin mining and return to that particular probably to close your. You're getting to all my questions before I can ask them. Ben, you're a little clairvoyant over there and I'm glad that you mentioned these sites though because that's what I want to tackle next. In the press release for the 2025 earnings, y' all mentioned an active go to market process at Panther Creek, Sharon and Moses Lake. What does this mean exactly? Can you unpack that? What, what does the go to market process look like? At what stage are y' all in the go to market process? All Those things.
- [14:41]
Ben Gagnon
Yeah, that's probably the one thing everyone wants to know the most about is, you know, what is the plan to get leases signed at all of these different sites? And what we've said over the last, you know, 12 months or so is, look, we haven't been in a rush because we've still been executing against the things that we need to get done on site before we're in a position to sign a lease and before customers can feel comfortable to sign a lease. And really that's about having high confidence and high certainty on your timelines. And in order to have high confidence and high certainty on your timelines, you have to have good visibility and confidence on your permits, because clearing all your permits determines what is your T equals zero day. Right beyond there, you can have the plan that you know from when you start, you put that first shovel on the ground, what's it going to look like? 3, 6, 9, 12, 18 months down the road. But you need to have that confidence on your T equals zero day. And, you know, for us, you know, we're developing sites in Pennsylvania, Washington, Quebec. These are high value markets, but they have high barriers to entry. I think one of the things that investors probably don't really understand because they're not developing and they're not in this world. Texas has historically always been the easiest state in the country to get power, land and permits. You can get those three things done in maybe three or four months where in every other state in the US is likely going to take you 12 to 18 months. And so, you know, comparing progress in Texas versus comparing progress in, honestly, any other state is not really a great comparison because they're very, very different regimes. You're still going to have the three same categories of permits. It doesn't matter if you're building a banana stand or you're building an HPC data center. You have the same three categories. You've got zoning, you have land development, and you have environmental. Zoning tends to be the most controversial. That's the part where the community really comes out in droves. And it gets very personal within the community as to whether or not zoning is going to be approved. We have cleared zoning at Panther Creek and sharing in February. So the most controversial part of the permits have been done successfully. What remains is executing against the rest of the land development and the environmental. And those generally tend to be a little bit more kind of data oriented with environmental surveys and engineering plans and wetland assessments and all these sorts of things. So it's less personal and it's More data driven. And you know, well, anything can happen in permits. We wouldn't be providing that kind of communication unless we were highly confident in our ability to secure permits rolling forward and our plan to do so. And frankly, if you aren't confident in that and you don't have a clear visibility, none of these customers are going to give you the time of day. If you go into a conversation with, I say, yeah, I'm working on the permits, but you don't have a very clear roadmap of exactly what needs to be done, what is going to be done, when and how the phone line just goes dead. And so for us, there were kind of two things. One, we're in high barrier to entry markets. We also started probably a little bit later than some of our peers because we didn't even close in on Panther Creek and scrubgrass until March 14th of last year. So we were working on closing the assets before we could even start. And then we're dealing with a complex permitting environment in Pennsylvania, unlike Texas. But we have learned as bitcoin miners over the last eight years how to engage with communities, how to go in there and understand their needs, how to go in there and communicate about our objectives, our intent, our strategies, and be transparent. And so we've got robust support from the very grassroots level like the principal, the superintendents, the police chiefs, the fire captains, the mayor, the state senators, the state representatives, the senator senators, the governor, the fast track office and the doe. So, you know, we do a really, really good job of going into the communities, listening to their concerns, responding to their concerns with, with solutions. And that really enables us to go in and prove that, you know, we're not in here to just get a permit and fly away. We're investing in the community so that we can be successful over the next few decades. This isn't, you know, just to be, you know, charitable. There's also some, some great benefits for us as a business because I want to attract high quality talent to the area. And in order to attract high quality talent to the area, you know, they're going to want to have good schools, they're going to want to have good infrastructure. And, and so for us to be successful, the community also needs to be successful. This goes hand in hand. These are win win scenarios. We know those really, really well. And when I was just at the Pennsylvania Data center and Energy Summit last week, we were hearing this from senators and county commissioners and state reps and all these different people that they've heard about the job that we've done in Pennsylvania on the community engagement. And they were blown away, just showering praise about how we've gone in there, transparently communicated, responded to the needs of the community and are really being responsible developers.
- [20:18]
Co-host
I think it's a piece that's overlooked when we see these headlines a lot. What has to go on in the background to convince or at least get the community on your side when you're building one of these things? Because there's so much misinformation. There's been a huge backlash against AI for a number of reasons, depending on what camp you're in. You know, there are the Luddites who just think that it's bad. There are the environmentalists who think that it's going to destroy the earth quicker than even bitcoin mining, which I never thought I'd see that. But we've seen a bu. A lot of pushback from, for instance, Logan county in Illinois. Hut 8's proposing a data center there. I believe the local board is not on board with it. I've also seen, you know, some. How should we say, some noise in Granbury, Texas, for one of Mara's bitcoin mines out there. Kind of seems like it's one of those things. If it's not actively managed, you can have problems with it. One other question on this permitting and regulation front, Ben, I'm curious. Have there been any uniques, maybe not the right word, but have there been any special considerations or difficulties with the Panther Creek and scrubgrass sites specifically because of the waste coal power plants? I'm wondering if any sort of environmental or ESG regulations have made that especially onerous to get approval for or zoning or if that's not even a factor.
- [21:32]
Ben Gagnon
I'm just curious, honestly, that's. That's a completely separate bucket. It's not even a factor when it comes to the data centers themselves. So like at Panther Creek, the data centers are all actually on the land that we own that's just directly adjacent to the power plant. So it's effectively the same property. Technically, it's two parcels of land. One of them has a power plant, one of them has a data center. We own both side by side. But you're getting the permits for that piece of land that's not attached to the data center. So it's actually a completely separate deal. But at this point, we're high confidence. And to clarify for everyone listening on the call, for us, that was the last piece. That's the last milestone that we as a management team and our potential customers need before we'd be comfortable signing a lease and before they'd be comfortable signing
- [22:27]
Co-host
like approval to operate those power plants specifically is, is that's the last piece
- [22:32]
Ben Gagnon
or what the, the remainder of the environmental permits and the land development permits for the data. So you know we've got good visibility on those in the coming months. We expect it to be done kind of in the mid late summer time frame. Could be a little bit earlier, could be a little bit longer, but that's currently what's on schedule with the meetings and the cadence that we have in place. Place. That doesn't mean that if you clear permits the very next day there's going to be a signed lease. But it does mean that that's a really important milestone for investors to pay attention to because that's the last milestone that we need to clear before we're comfortable signing a lease. And same with the customer.
- [23:08]
Host
Should we expect a deal soon?
- [23:10]
Co-host
Next quarter, sometime later this year? Do you have a rough timeline you can give our listeners?
- [23:15]
Ben Gagnon
So as I just mentioned there with the last milestone being permits, you know you could expect a deal sometime after we get the permits. Our goal as a business this year is to clear the permits on those three sites and to execute leases on those three sites. And we've got high confidence that we're going to be able to execute on those goals. Beyond that, we also have an objective around securing our additional expansion capacity at Panther Creek and at Scrubgrass. And so when you're looking at kind of what is know bitfarm's objectives and priorities this year they're highly focused around what will be or what we believe are going to be major catalysts for the stock in 2026. Clearing permits, signing leases and securing power. That's where the company is entirely focused on execution.
- [24:08]
Co-host
Got it. So we'll return to the thing we put a pin in earlier, which is the bitcoin mining operation. You kind of already spoke to this, I think you said what, four to five exahashes by year end or once you have the AI and HPC leases locked and you can start that transition. You're targeting roughly that. So should we expect though in the midterm future at a complete wind down of bitcoin mining across all bitform sites? Is there any desire to do hybrid, some mining on the Pennsylvania sites and then mostly AI and hpc or is it just like what you said? I think you mentioned the Quebec site will have it and then did you say Scrubgrass will also have it for some time afterwards?
- [24:54]
Ben Gagnon
So you Know, our goal is to be pure play developer ownered infrastructure. We're not going to be an operator when it comes to the assets and conversion. You know, effectively. We did an analysis about a year ago, actually longer than that, about 18 months ago, as to which sites would be suitable for conversion, which sites weren't. And you know, we did this with two different HPC and AI advisors. Basically what came back was our US sites are great for conversion, our Latin American sites are not, and our Sherbrooke site is really, really good in, in Canada and some of our smaller sites in Quebec are good. Some of them are going to be high hanging fruit. So what we've done is we focused on disposing of the assets that were not suitable for conversion, then focusing on the lowest hanging fruit on the tree and working our way up. And so the, excuse me, the reason why Quebec and Scrubgrass will still have bitcoin mining going on at the end of this year and rolling into next year is really just has to do with the timeline around us getting everything in place to convert over those assets. Because until they're converting over, there is no real compelling reason to get rid of the bitcoin mining outside of maybe a spike in bitcoin price and a spice in bitcoin miner pricing that we might want to take advantage of. Beyond that, the, the optimal scenario is just to keep those sites running as long as possible until they need to be developed.
- [26:26]
Co-host
Going back to the Quebec site, is there anything on the regulatory front that makes y' all want to wait until you put time into developing that site? Will it be, will that be slower moving than even in Pennsylvania or is that not a problem?
- [26:40]
Ben Gagnon
You know, too early to tell as to what exactly is going to happen in Quebec. Hopefully we have further progress and we can speak to it on the Q1 call in about six weeks. But you know, we've got a good path forward and a plan for the first 96 megawatts in our city of Sherbrooke. So we've got 170 megawatts in Quebec spread across eight sites. Sites 96 megawatts of that is in one city, the city of Sherbrooke, which is the next biggest city outside of Montreal. It's about 90 minutes east. There's a big university town there. We've got that 96 megawatt spread across three sites, a 48 megawatt 30 and an 18. What we're doing there is we're working to consolidate all of those 96 megawatts onto a new Development and a new location. And so in that scenario we're going to be able to, you know, continue to operate the bitcoin lines right up until the point that that site is being commissioned. Right. Even all the way through construction, we should still be able to operate those, those bitcoin mining sites. And right at commissioning we'd effectively be turning off the power at the bitcoin mines, turning it on at the new HPC data center campus. That's our big focus right now is the 96 megawatts there because it's over half the portfolio in Quebec. And we can do it at one single location with a city that's incredibly supportive, with the hydro municipal network who provides us power, which is incredibly supportive, with a university that trains people specifically for data centers and it in the area for good quality labor. That's going to be our focus. And then we'll work on the smaller sites which are higher hanging fruit.
- [28:14]
Co-host
One last question on this front Ben, before we move on to financing and then we'll get you out of here. How do you manage a wind down like this? Because I've thought that there's kind of a chicken and an egg problem for not just bitforms but any bitcoin miner that is expanding into AI and hpc, which is you're repurposing these sites that have active hash rate on them and you want to make sure that you can manage this construction in such a way where you're not taking all that hash rate offline at once, I would assume, and hamper your cash flows as a result. What's the game plan to make sure that these windows are managed in such a way to where you, you don't have a huge revenue shortfall during one of these quarters.
- [28:52]
Ben Gagnon
So you know, one of the interesting things and one of the things I love about bitcoin mining is the economics are so easy to model. And the bitcoin miner value is really determined by his future expected cash flow. And so you can take a look at a bitcoin miner or a full building of thousands of bitcoin miners and you can say here's what I expect this, this miner or this facility of miners to generate, you know, today, next week, next month, the next six months, the next 12 months, you can associate a value with that, right? It's also really predictable on the cost side. So we can predict revenue cost margin from there. You now can make a choice. You can say would I be better off with the cash flows from the bitcoin mining or would I be Better off taking a present value sum by selling off the miners today. And that's the dynamic that you look for or that you're working through as you're doing a wind down, you are looking at the opportunity cost. Either way, is it cheaper for me to or more, more value maximizing for me to keep these cash flows going or is it more value maximizing for me to just take the lump sum pay payment today? That's going to be different for different operators, different sites, different equipment, you know, different views on market forecast. But there is a really simple, you know, logical framework that you follow. And at the end of the day when you're looking at what we're executing against and you look at the value that we are getting out of the cash flows from the bitcoin mining, you know, those are really not what's funding the capex for the business. Rolling forward. Forward, right. The, the bitcoin mining revenues. Regardless of If I owned 1% of the hash rate or 10% of the network hash rate, they're not going to be a meaningful contributor to my HPC and AI CapEx. And so it really just becomes a function of, you know, I think more bandwidth constraint and what's the easiest value maximizing way that preserves, you know, optionality and enables the, the management team to focus as much as possible on HBC and AI where we're going to create the most value. And so for us it's really a constant evaluation. We're going to continue to make judgments as market conditions change. I still remain optimistic on bitcoin price this year. I think everyone was kind of going into this year expecting we might be starting to lower rates in kind of this time frame around March and April. I don't think anyone was really anticipating the Iran conflict to be sparking off. And so there's been a lot of geopolitical scenarios. Generally speaking. I always like when stuff like this happens. It should be good for Bitcoin because it should be considered as a safe haven asset. But you know, bitcoin has a funny way of functioning differently than you might expect it to when, when the rubber hits the road. And I think this is one of those situations where you would think that global uncertainty and rising costs would also make Bitcoin a safe haven asset. We haven't seen that so far, but I think we're still pretty optimistic that we're going to be able to get more value out of our bitcoin and out of our mining later on in the year.
- [32:16]
Co-host
I'm glad you mentioned Capex because that transitions well into the final topic I have today. And I want to start by leading in with some news from some of the private credit funds. We've covered this on the show in the past. Blue Owl, Blackrock, Apollo, they've all capped or changed the redemption guidelines or they've capped redemptions or changed their redemption guidelines for these private credit pools. Curious if you have any takes on that and whether or not you think, you know, AI CapEx deals and funding is slowing down, if y' all have seen anything like that on your end.
- [32:51]
Ben Gagnon
So there's a tremendous amount of capital out there for, for these projects. I, I think there's a lot of negative headlines because I think negative headlines generate more clicks and more views. And so there's certainly an element around kind of the fear mongering and the uncertainty and when conflicts brush off and when everyone is talking about hundreds of billions of dollars in capex, that gets people concerned. But I think the underlying economics are pretty compelling. And the reality that you see on the ground is that demand is just rampant. It's just, it's just, you know, in it insatisfiable how much demand there is out there. And you know, one of the things that we've been doing as a business, right, is we're working on integrating AI into different departments and different teams, work streams and workflows. And I mean, it's a powerful, powerful tool that I think we're just scratching the surface of what it can accomplish. Like I have an agentic chief of staff, right? So I have an AI system that is just a chief of staff. And it helps me keep track of things, synthesize information. It gives me, you know, morning briefs, it gives me evening briefs, it keeps me up to speed on all the different developments. It keeps track of, you know, what was said in meetings and what needs to be kind of followed up on. Like that's just one example of many ways that we use it. We have AI integrated into our minor management software. So, you know, operations techs can query the management system in a natural chat function which has dramatically improved response times on repairs. There's a lot of different ways that we're integrating AI into the business. And I think like when I was at Nvidia gtc, you're seeing all these different applications for it. Like there was, you know, window cleaning robots, there's robot robots, there's dog robots, there's humanoid robots, there's robotic arms, there's a robot's first sorting, recycling, there's a DJ robot, there was A brain surgery robot, there's protein folding simulations, there's all sorts of different things and different applications for it. Like the amount of applications is just, it's just endless. And what I think is happening is, is actually a paradigm shift in the world. I think we have an opportunity here with AI where it really improves the interface between a person and a computer. And it makes it so that you don't need to understand how computers work or how they think or how they process. Because now they understand how you think and how you process. You don't have to learn how to program. They understand what you mean when you communicate in the chatbot. Right. And they can create the code and they can create the programs. I think we're moving to a world where the ability to do work is no longer going to be measured in calories or joules, but in tokens. And the reality is that that's really a game changer. We haven't had this since we created electricity. That was the last time that you had some sort of fundamental paradigm shift like this taking place. And so the demand for HPC and AI is going to continue to go up. The applications are going to continue to go up and it's going to get exponential because the AI is now improving the AI. And so the better the AI improves the AI, the better it's going to be at improving the AI. And there's an escape velocity that happens in that dynamic. Right. It will become an exponential efficiency curve. It has to. And then there's this other aspect too, around just like as that efficiency goes up, Jevons paradox kicks in and people just start using it more. Like, I'm not sure if you follow Gary Tan at Y Combinator, but he outsourced or Open sourced on GitHub, basically 30 decade or 30 years of startup experience into 20 agentic AIs. And so he's got, you know, an agent for corporate finance, he's got an agent for reporting, he's got an agent for marketing. He's got all the different components that you might need to run a business. Now you've got 20 AIs that are all speaking to each other. And so you've even removed the bottleneck of the human. These AIs can be talking to each other at AI speed. And they're going to do so 24, 7. It's going to be significantly more use than one person. And so you start thinking, okay, this is the demand from one person. What if one person has 30 agents? What if those 30 agents are talking to each other 24, 7. You know, now that one person's demand becomes like the equivalent of like a hundred people's demand because those you've got 20, 30 agents that are talking two to four times the speed. Right. The demand is never going to go away. It's just going to increase. And the applications there are pretty powerful. So what that means for us as a business is, you know, you are going to be solving a very, very important bottleneck problem. The one thing that's, that's fundamental is, is the power. You're not going to be able to get away with it even. You know, we've heard, we've heard a lot of concerns around what's happening with, with Iran and the Strait of Hormuz and how that might be impacting global supply chains. And there are some fragile supply chains out there. But the priority around kind of Nvidia and Apple at TSMC means basically everyone else is going to be shut off before they would ever be shut off. I would be much more concerned if I was Bitmain about my TSMC chip allocation than if I was Nvidia. Right. Bitmain will be turned off far faster and far earlier than Nvidia loses one wafer.
- [38:53]
Co-host
Yeah, we've actually heard some rumblings about some troubles in that regard, but that's for another time. We won't turn this into a rumor mill podcast specifically with Bitman and tsmc. Just a sidebar. I really like this concept of AI as a translation layer for humans and computers. I think that's a really good way to look at it and I've never really heard someone put it in those terms before, but. Ben, last question and then we'll have final thoughts with regards to financing and sourcing fundraising for the CapEx spends on these data centers, which they're by far and away much more expensive than Bitcoin mines. What is on the table for bit farms in terms of financing for these AI builds? We all do project level financing, equity, convertible notes. Is everything on the table? Basically just whatever makes the most sense. What's Yalls approach to the CAPEX spend here year?
- [39:45]
Ben Gagnon
We have completely restructured the business along with the redom. Every single asset that we have has now been wrapped up into a consistent and uniform SPV structure across the business in a very traditional infrastructure manner. So, you know, we are well positioned to raise capital either at the project level or at the top co. So you know we're going to be opportunistic. We think that there's going to be significantly more opportunities at the project level than we've ever seen as a bitcoin mining business. We think it's going to be way more availability. We think it's going to be way lower cost. And it's probably the, the ideal structure is to raise the vast, vast majority at the project level. That doesn't preclude doing some capital raising at the top co but we have restructured the entire business to enable the project financing because we do believe that's going to be a big part of, of of our strategy. But all options are on the table. You know, we have about $520 million of liquidity as I pointed out too earlier on the call. And that's far more than what we've budgeted to get us through permitting and the lease negotiations. And so we've got a lot of flexibility. There is no, you know, pressure for us to be issuing equity right now or raising capital right now. We've got everything that we need to execute over the, the, you know, the next three to six months. And you know, we've got really, really high confidence that we've got the right team, the right capital, the right structure in place for us to be able to, you know, clear out the permits and move through lease execution and then on to financing and construction.
- [41:26]
Co-host
Then any, any final thoughts before we hop off either on yalls earnings for 2025 forward, looking for the AI pivot? Anything that comes to mind, Anything you want to say before we turn this one off?
- [41:39]
Ben Gagnon
You know, I think the, the big things to focus on here is that, you know, we drew a line on the sand yesterday between the company we are, we were and the company we are now. The bitcoin will still be a part of the business this year, but you know, it's not material. It's going to continue to get smaller and smaller and smaller. Not because we're being forced to do so, but because we're taking a disciplined path towards a pure play. HPC developer and operator I think the important thing for most of your audience is probably what we put in our deck on, on slide 7 and what we put out there was, you know, a really, we try to make as clear as possible. Look, we're probably valued fairly if you think of us as a bitcoin mining company. But we're no longer a bitcoin mining company. Bit Farms was a bitcoin mining company. Keel infrastructure is an HPC and AI development company. And if you look at Slide 7, you see just from signing leases, companies have a big re rating and a big Catalyst. I'm sure every investor in this space is very well aware of what it means to sign leases. And that's not delivering facilities or creating the actual revenue and cash flows, it's just signing the leases. We fully aware of what that means for investors and value creation, it being a catalyst and it's our, one of our top priorities and our top goals this year was high confidence on those three sites. Beyond that, the other two catalysts are securing the expansion capacity, which is our other big priority this year, in which we've had a few different hires specifically engaged just to advance on the power side of things because we know that's a huge, huge value driver for, for the investors. And beyond that, we have delivery in 27. So you know, if you're one of these investors who are saying, I'm really interested in the hbi, HPC and AI story and the, the industry, I want to get exposure, I want to get access. And you're looking at Nvidia and how it's outperformed Bitcoin over the last 10 years and you're saying, do I get into Nvidia or where else can I get this exposure? Exposure? You know, the, the peers in our industry have had a lot of success last year for good reason. And I think they're going to have continued success as they execute against their deployments. But when you're looking at what has, I think, you know, the most asymmetric upside and the most opportunity for investors, it's going to be in companies like Keel and the other companies who haven't actually signed leases yet, who have the most upside. But there's tremendous opportunity in Keel and there's tremendous opportunity in our other peers. There's tremendous opportunity still in, you know, the peers who have already announced leases as the execute and as they deliver those megawatts. This is a really, really exciting space that I think the dynamics are such that we have a really unique opportunity to capitalize on this bottleneck unlike anything in recent memory. And so everyone is probably going to do well so long as they execute. It's not, you know, lost on us that we're a show me story and a lot of other companies in our space are showing me stories. You know, even the companies who've signed leases are still show me stories because most of them have not delivered a megawatt yet. Most of them are not delivering the revenues yet. But I think for the investors who can see, see that, see that opportunity and they find the right teams and the right projects to invest in, I think there's just tremendous upside in the space.
- [45:10]
Co-host
Ben Ganon, thank you so much for joining, man. We'll have to get you on block stream. Excuse me? We'll have to get you on block space live once y' all have some more news about those leases. But, hey, appreciate your time, man. Godspeed with the rest of the year.
- [45:24]
Ben Gagnon
Thanks, Khan. Always a pleasure.