WEBVTT

1
00:00:00.100 --> 00:00:17.120
[upbeat music] Hey, what's up, y'all? Welcome back to Blockspace Live, brought to you by CleanSpark. Today, for our top story, big day for agentic payment enjoyers because AI is coming to Robinhood. That's right.

2
00:00:17.180 --> 00:00:27.240
You thought you were gonna use AI to automate your workflow. It's actually going to make you trade better or make you poorer, depending on the decisions that it makes. Not financial advice.

3
00:00:27.280 --> 00:00:36.080
And then for our interview today, we have Asher Gnutt of Hut8 on to talk about their new Beacon Point AI data center that they announced earlier this month.

4
00:00:37.140 --> 00:00:49.800
On to other news, Iren is buying Blackwells for its Nvidia rollout in Childress, $1.6 billion purchase that just hit today. We will cover that as our second news item.

5
00:00:50.200 --> 00:01:02.220
And then for our third news item, we have Core42 seeking out half a billion in financing to build data centers in the US and, surprise, Europe. Perhaps the demise of Europe has been much overstated on social media.

6
00:01:02.710 --> 00:01:14.240
And to wrap it up, a little bit of a juicy story here, Mera's security expenses in 2025 and why that's actually not out of the ordinary these days for crypto executives.

7
00:01:14.620 --> 00:01:25.360
As much as the Financial Times would like to berate them and make you feel like this is a weird thing to do when wrench attacks are happening all over the world, it's actually kind of normal and becoming the standard for some of the top companies.

8
00:01:26.760 --> 00:01:36.270
That's right. Blockspace goes live Monday, Wednesday, Friday at noon, soon to be weekdays at 1:00 p.m. Eastern, featuring quick hits on the latest in Bitcoin, mining, AI, and emerging tech.

9
00:01:36.400 --> 00:01:40.100
Make sure to like and subscribe, follow on your favorite platform.

10
00:01:40.120 --> 00:01:49.640
Make sure you hit the notification bell to get the push notifications to your mobile device while you are scootering to your slop bowl somewhere on one of the coasts.

11
00:01:51.230 --> 00:01:59.080
If you like the stream, it turns into a podcast right after, anywhere podcasts are found, RSS feeds. If you're listening on CoinDesk,

12
00:02:00.060 --> 00:02:13.680
stop what you're doing, get off the scooter, navigate to the Blockspace feed on, uh, Spotify or Apple, and subscribe to the Blockspace feed instead of staying on CoinDesk. And

13
00:02:15.160 --> 00:02:29.000
we still write a newsletter, drops on Fridays, newsletter.blockspacemedia.com. This show is brought to you by CleanSpark, NASDAQ listed, ticker CLSK. More on them later on in the show. Let's kick it off.

14
00:02:29.560 --> 00:02:45.630
Colin, I'm very excited to figure out a way to get poor faster by letting my AI agent run amok on Robinhood's MCP server. So here's the story. Robinhood just

15
00:02:46.700 --> 00:02:53.040
today announced that they are rolling out-- they're open to agents.

16
00:02:53.080 --> 00:03:00.940
So specifically, they're rolling out the agentic trading and agentic credit card, quote, "Enabling agents to trade and make credit card purchases on your behalf."

17
00:03:01.740 --> 00:03:12.000
Basically, the TLDR is Robinhood set up an MCP server, a Model Context Protocol server, to let your AI agent plug into their platform and make trades on your behalf.

18
00:03:12.120 --> 00:03:19.540
It's got some guardrails around it, like you can turn it off, you can give it, like, a budget and an account, and it's live only for stocks right now.

19
00:03:19.800 --> 00:03:27.180
Can't do crypto yet, but there's a little bit more under the surface here that we can get into.

20
00:03:27.190 --> 00:03:35.480
Are you excited to finally figure out how to set up OpenClaw and how to-- and let OpenClaw go long and short Nvidia upon earnings? What are your thoughts?

21
00:03:36.360 --> 00:03:48.200
You said guardrails, and that's the first thing that I thought about with this because you had all these horror stories when, um, OpenClaw came out of people using it to automate workflows, and it would, like, delete their entire email.

22
00:03:48.360 --> 00:03:56.580
[chuckles] Yeah. It would, like, delete full databases from their computer. I'm sure that that is very much the exception and not the rule.

23
00:03:56.620 --> 00:04:08.020
In fact, we know that's true because when one of those stories pops up, it goes viral because it's crazy. And then, of course, all of the AI doomers are like, "See? It's actually really stupid, bro." You know? Yeah.

24
00:04:08.440 --> 00:04:15.180
Um, and so my, my first thought is, like, what, what is the liability here for, for Robinhood? [snorts] Right?

25
00:04:15.280 --> 00:04:26.860
Like, it's-- do, do you have to, like, do you have to get a defense fund going in case the AI decides to, like, ape your entire 401[k] or your portfolio into some shitter that ends up tanking? You know?

26
00:04:26.880 --> 00:04:33.020
It's awes- you know, so you say guardrails, and I will say that this particular product appears to have a lot of guardrails out the gate.

27
00:04:33.620 --> 00:04:44.420
Um, first of all, it's a separate trading account, so it's not like your primary trading account. It's a separate trading account, which means they're gonna give you fewer knobs and a lot more, uh, shutoff switches.

28
00:04:44.560 --> 00:04:58.580
So specifically, it'll send you real-time notifications and have a big instant shutoff option. Um, I would love to see, uh, you know, really tight leashes that you could put around this, these AI agents.

29
00:04:59.160 --> 00:05:12.300
Um, and, uh, what's-- you know, the crazy thing is, like, this-- If we go back in time, and we go back to the original, like, launch of Robinhood, um,

30
00:05:13.560 --> 00:05:25.110
Robinhood's, like, whole point was, like, put, like, letting you, retail, get m-more, like, hands-on, direct interaction with the stock market. You could buy fractions of stocks.

31
00:05:25.620 --> 00:05:32.920
You would have a lot more, like, granular touchpoints. They-- I think they added futures, um, pretty early on.

32
00:05:32.940 --> 00:05:44.469
And so these were, like, things which are not really broadly accessible, and, um, to the average person and- And probably shouldn't have been. Yeah. [both laughing] Because, you know- I mean, like- Yeah...

33
00:05:44.469 --> 00:05:52.800
straight up stock picking, one thing, but, like, honestly, bro, like, your average retail investor doesn't need to be messing with futures. They don't need to be messing with options, you know?

34
00:05:53.250 --> 00:06:04.744
Um- Yeah, but, but we've seen these average retail, uh, investors become very, very, uh, advanced, um, in maybe an asymmetric way, and at what cost?We can debate that some other time.

35
00:06:04.764 --> 00:06:13.504
But this is, you know, a- in contrast to the more sophisticated, the citadels of the world, your Hudson y- r- Yard, Hudson River, whatever they're called.

36
00:06:13.544 --> 00:06:23.304
Like, these, these really sophisticated institutions who basically, you know, are at least one to two to three years ahead of you in their technical prowess.

37
00:06:23.924 --> 00:06:35.203
So, um, the irony is, like, now Robinhood's letting you offload a lot of your responsibility onto an AI agent, if you will. And, uh,

38
00:06:36.144 --> 00:06:48.064
it- it's interesting because this means that your little AI agent can compete more directly with all of the other retail people's AI agents. But what does this do to reflexivity in the s- in the markets?

39
00:06:48.144 --> 00:07:00.944
What does this do to, like, the, the speed of trades and, um, you know, are the AI agents going to create their own little, uh, GameStop cult? I do wonder these things. [laughs] Uh, can you imagine?

40
00:07:01.064 --> 00:07:07.574
Do you re- remember what was the, like, AI social media platform that Facebook bought for some absurd number? Yeah, Moldbook. Okay.

41
00:07:07.624 --> 00:07:17.864
So imagine Moldbook but with- Just sitting there scheming, "Hey, we could, like, this, you know, y- you could have a GameStop moment. The stock seems to be undervalued." Yeah. "The volume is low.

42
00:07:18.284 --> 00:07:24.123
Let's, let's instrument a short squeeze." O- one thing, though, just to go back to, like, what Robinhood did for retail trading.

43
00:07:25.524 --> 00:07:40.284
It seems to me, and I don't really use Robinhood, but it seems to me that what Robinhood really did and where it shined is it took, like, the UX and UI from, like, a crypto exchange, and it made brokerage better.

44
00:07:40.404 --> 00:07:49.344
So, like, all of these things that you can do on Robinhood, you could go to your Fidelity dashboard and do, and you could go to, uh, any number of these financial c- um, services companies.

45
00:07:49.864 --> 00:07:53.704
But it- it's very boomer-fied, you know? Like, it's, it's, it's...

46
00:07:53.744 --> 00:08:05.764
It makes my eyes bleed whenever I open up my Fidelity account because it's not- The, the, the buy button on Charles Schwab is so, so small, I can't even find it. The buy button needs to be half the UI.

47
00:08:06.304 --> 00:08:07.854
May- maybe there's a lesson in that. Yeah.

48
00:08:07.884 --> 00:08:19.544
And, and going back to, like, what I think this, what, you know, ramifications for this for retail, I think a lot of people are gonna fool themselves into thinking that they can spin this up and the AI is going to make genius stock picks.

49
00:08:19.564 --> 00:08:30.684
But I would imagine that a lot of the benefit here is, like what you said with automatic trading, setting parameters, explaining what portfolio weighting you want. Do you wanna go high growth?

50
00:08:31.014 --> 00:08:37.764
Do you wanna play it safer with consumer staples? And you're gonna have to factor all of that in, and then it's going to make trades for you.

51
00:08:38.104 --> 00:08:52.564
Like, if you don't already have a thesis and you don't already know what you wanna be allocated to, if you don't al- if you haven't already done the homework to see what you're comfortable with in terms of portfolio allocation to all these different sectors, like, is this gonna help you?

52
00:08:53.084 --> 00:08:57.944
You know, you could just say, you know, YOLO into whatever or just, like, give it loose parameters.

53
00:08:58.304 --> 00:09:07.534
But I think a lot of retail folks are probably gonna trick themselves into thinking this is some magic trading bot, when really it's probably just going to make it more efficient for the people who actually know what they're doing.

54
00:09:08.084 --> 00:09:15.804
Now, the one interesting thing here is copy trading. This is going to make copy trading very easy for people. Yes.

55
00:09:16.193 --> 00:09:30.384
And so you'd have to imagine, like, a almost, you know, uh, secondary effect of this is the, uh, stock trading gurus are probably going to have a very easy time selling their trades because now people can copy them super easily.

56
00:09:30.444 --> 00:09:37.064
Oh, yeah. Whereas before you had to set up all these bots and do all this other crazy stuff. The AI can just do that for you, so. It's, yeah, it's very clear.

57
00:09:37.124 --> 00:09:44.824
Everybody needs to set their OpenClaw to connect to Robinhood MC- MCP server and have it copy trade Nancy Pelosi and inverse Jim Cramer.

58
00:09:45.244 --> 00:09:56.244
Like, imagine a 60/40 portfolio where it's just copy trading, you know, long Pelosi, short Cramer. Uh, you know, there's your... Now I actually get the whole AI agent thing. Now... [laughs] Yeah, right?

59
00:09:56.444 --> 00:10:07.204
I mean, it, it would be interesting to see, uh, if the Pelosi tracker comes out with some sort of, like, affiliate thing with Robinhood. Yeah. A Pelosi tracker MCP server. [laughs] Yeah. Okay.

60
00:10:07.244 --> 00:10:21.784
That's, uh, that's it for Robinhood. Um, we have our first, uh, guest. We have Asher Gnut, CEO of HUT8 here in the wings. We're gonna transition to talk about data centers. Let's bring him on up.

61
00:10:22.704 --> 00:10:32.104
Asher, welcome back to Block Space. How's it going? I li- uh, we should keep the topic on copy trading. I liked that. That was interesting. [laughs] Yeah. What, what, what, what's, what's your...

62
00:10:32.244 --> 00:10:38.934
Well, we'll, we'll go ahead and do that. Um- [laughs] We'll, we'll get to HUT8 here in a second to the real interesting stuff. But, like, what do you think about this, uh, Robinhood news?

63
00:10:39.024 --> 00:10:49.404
And, and what do you think about the implications for retail kind of given a bazooka where maybe, uh, you know, [laughs] may, may, may, maybe a BB gun would be more, uh, appropriate for most of them?

64
00:10:50.404 --> 00:11:02.654
I, I, I think it's fascinating. And I, I think to your point, these influential investors, it actually makes their trades even more powerful, right? Because a lot of times you see a trade, people fall into the thesis.

65
00:11:02.664 --> 00:11:07.744
And so now when they actually move, they'll have a whole train of folks that follow them.

66
00:11:07.764 --> 00:11:19.784
And if you think about, like, the kind of most OG person that was, like when War- Warren Buffett made a conviction play, people usually kind of followed him after, and you had the Warren Buffett effect where the stock would go up because he was involved and had conviction.

67
00:11:19.804 --> 00:11:28.964
And I think that's good because people can just, like, bet on investors that they believe in, and you don't have to pay two and 20 for a fund, honestly.

68
00:11:29.584 --> 00:11:40.864
[laughs] I think it's bad because, like, who are the investors that you should believe, right? Who have, who are momentum traders versus who actually has deep, like, fundamentals and is doing the diligence and so forth.

69
00:11:40.924 --> 00:11:51.394
So I feel like people should be excited by it, but also have a certain level of risk threshold and not just assume, okay, this person has this track record,

70
00:11:52.314 --> 00:12:00.624
so I'll guarantee make money, and then put in 90% of their wealth and potentially it, it loses. 'Cause I think right now the markets are very healthy, to say the least.

71
00:12:00.704 --> 00:12:07.704
And, like, you see a lot of kinda stocks appreciating and doing well. But we're in a very volatile environment, as everyone knows.

72
00:12:07.724 --> 00:12:23.780
If we think about, like, the stock market volatility over this last year compared to the decade beforehand, I mean, the amount that you see these kind of, like, multi-trillion dollar companies trading in terms of single percentage changes kind of week over week or day over day-Is, is really unprecedented.

73
00:12:23.800 --> 00:12:31.880
It used to be a lot more stable back in the day. Um, and so I think it's, it's exciting 'cause it's the natural trend of where things are going and people are doing it in an analog fashion anyways.

74
00:12:32.340 --> 00:12:40.710
Um, I think it just requires m- a bit more kind of thought and sophistication on like, what am I actually getting myself into and like, what is the risk I'm actually taking?

75
00:12:41.040 --> 00:12:55.180
And, and going back to your point about which investors are actually worth following, I would imagine that following this release, a lot of Twitter traders are going to end up having some sort of paid subscription service like they already have, but where you can basically just copy paste their trades.

76
00:12:55.200 --> 00:13:00.400
And it might lay some of them bare, right? And that is, I think, where it gets tricky for a lot of people in retail.

77
00:13:00.420 --> 00:13:09.070
They follow some of these guys thinking that they're going to find a diamond in the rough, but it's like a dude with an anime profile picture. You have no idea what you're getting day in and day out.

78
00:13:09.100 --> 00:13:16.990
Yeah, but the, the new, the, uh, the paid API endpoint for your favorite influencer's trades is the new gated Telegram group. This is, there, there's your...

79
00:13:17.480 --> 00:13:26.379
And then the question is, are you trading now on momentum or are you trading on fundamentals, right? Because a lot of- Yeah... people have made a lot of money in recent years trading on momentum.

80
00:13:26.879 --> 00:13:37.400
And I mean, GameStop is like the perfect example of like a momentum trade that a lot of people piled in. It's gonna be easier to momentum trade behind like influential folks and say, "You know what?

81
00:13:37.600 --> 00:13:45.600
I trust you to go and like try to drive this." But that's a very different type of investing, and people just need to be eyes wide open to that. Yeah, that's a really good point.

82
00:13:45.660 --> 00:13:52.260
As my dad has said over the last few years, everyone's a genius in a bull market, right? And I just, that's the, my one caveat with this- Right...

83
00:13:52.270 --> 00:14:01.100
is I worry a lot of people are gonna think this is going to make them into stock picking gurus- Oh, wow... because the AI is just like, you know, a machine god. But- I'll give you an example.

84
00:14:01.120 --> 00:14:10.510
When people ask me about like what are my favorite stocks, I, I say, "I don't spend all my time thinking about investing in stocks, so therefore like there are people out there that are much smarter than me on this.

85
00:14:10.580 --> 00:14:16.220
And like I can maybe give you some insight on the data center space, what companies I think could do well, supply chain, things like that.

86
00:14:16.540 --> 00:14:25.620
But in general like there are people and there are multi-billion dollar infrastructure data centers that support different types of trading and different types of, of infrastructure."

87
00:14:25.650 --> 00:14:33.200
And so I, I think you're exactly right because when they're trending and it's in a bull market, everyone thinks that they're a guru.

88
00:14:33.580 --> 00:14:38.600
But when things turn, I mean, that's really where you see like who actually is a strong investor versus not.

89
00:14:38.640 --> 00:14:46.660
And again, a one year kind of downturn, like if the whole market's down, it is what it is, but it's how you perform over a longer period of time. Right.

90
00:14:46.680 --> 00:14:55.180
And I'm sure when people ask you what your favorite stock is, you gotta say Hut, and it's almost kind of a funny question 'cause it's like- It's up on the day, up on the year, so- It's like, dude, I'm-...

91
00:14:55.230 --> 00:15:00.780
it's good time talking to Asher. Yeah, I'm all in on Hut here. Like I don't e- I'm not thinking about other stocks other than my own.

92
00:15:00.840 --> 00:15:09.290
So w- we'll go ahead and start talking about some of the, uh, the announcement that y'all had earlier this month. It, it was really... it's nice to see. It seems like the deal flow is picking back up.

93
00:15:09.300 --> 00:15:15.060
There's kind of a lull in Q1, and now it's like we're right back to some of these deals. We're right back to the financing agreements.

94
00:15:15.120 --> 00:15:29.560
So can you give us a rundown of this Beacon Point deal and, um, you know, y'all's, uh, uh, kind of what, uh, led to it, um, what y'all's plans are for this site, all that stuff, just to give our listeners a, a rundown for those who might have forgotten or don't know?

95
00:15:30.700 --> 00:15:39.800
Our first deal we announced in December was a $7 billion deal in Louisiana, and that was with Anthropic and Google as, uh, support with FoodSack as well.

96
00:15:40.520 --> 00:15:49.980
The second deal we announced was in Texas, and that was a $9.8 billion deal, and so it was, uh, larger almost by 50% than the initial deal.

97
00:15:50.620 --> 00:15:58.680
It's a high grade investment counterparty we didn't share kind of the name of who, and I can kind of share why we are going through that on that strategy going forward.

98
00:15:58.720 --> 00:16:09.920
But it's phase one of a gigawatt campus, and it's really exciting because it's actually one of the most dense data center buildings in the world when it, when it comes online.

99
00:16:10.200 --> 00:16:20.389
We have 500 megawatts of capacity in a single building, and so we work closely with NVIDIA. They actually wanted us to share that they were a design partner on the campus because of how exciting and innovative it was.

100
00:16:20.420 --> 00:16:26.240
But we're really exciting that brings our total capacity of IT megawatts almost to 600 megawatts.

101
00:16:26.280 --> 00:16:36.400
It brings our total contractor revenue almost to $17 billion, and that's really over the la- course of the last six months from December till now. So I have to quick side note here.

102
00:16:36.420 --> 00:16:49.140
When I was going through prep for this, you know, we were talking about Core42 doing a half bill financing, and they're trying to build in Europe, and so that got me thinking, "Okay, well let's get projections for Europe versus US for data center build-out."

103
00:16:49.160 --> 00:17:00.740
And when you look at the projections for the US, what is still kind of awe-inspiring to me is you have all of these Bitcoin miners that are going to make up a significant chunk of that when they're making these, these transitions.

104
00:17:01.200 --> 00:17:09.630
So you're just talking about the density of this data center, one gigawatt. It, sometimes the numbers stop making sense to us in, in the sense that we just kind of become numb to it- Yeah...

105
00:17:09.640 --> 00:17:12.500
but that's just absolutely freaking massive. Yeah.

106
00:17:12.600 --> 00:17:35.240
Uh, a- and to go back, so to go back to the, the- I mean, just to put it into perspective, like those two deals we announced without like a future expansion on Beacon Point or anything, 'cause we have like a 3% escalator, so across the 15 years, it's about $1.1 billion a year in cash flow, and these are triple net leases where there's no real cost to operating them, so that's n- like all net operating income is like 99%.

107
00:17:35.800 --> 00:17:46.140
And so like it's really powerful, and to your point, like these are massive projects with massive numbers. Um, and, and we think this is like a generational opportunity to go build like an infrastructure business.

108
00:17:46.160 --> 00:17:55.140
What I'm really excited for is we're not building a data center company. Like we didn't pivot from crypto to AI. Like we still have 700 megawatts through American Bitcoin that we manage.

109
00:17:55.580 --> 00:18:08.300
And so like in my mind, Hut 8 is the business that rethinks and builds infrastructure at scale that supports next generation technologies, and AI is obviously the largest consumer and user, but over time there will be other consumptions.

110
00:18:08.340 --> 00:18:15.379
And if you look at like the history of humankind, our quality of life, our GDP has increased as our energy consumption has gone up.

111
00:18:15.700 --> 00:18:21.100
And so how do we build physical infrastructure at that intersection of power and technology? And like that's what I'm super excited for.

112
00:18:21.980 --> 00:18:30.302
So going back to the tenant for this one, I mean, the, the press release said they're of hyperscale tier.Um, it was an undisclosed... W- what's the reason for that?

113
00:18:30.392 --> 00:18:40.372
Just confidentiality, they requested it, um, uh, 'cause that, that's wh- I think the first one that we've seen where, um, tenant has not been disclosed, so. Yeah.

114
00:18:40.452 --> 00:18:47.412
So when we announced our deal in December, I think a couple of things came up when we announced it. One was people were like, "Holy shit, that was a great deal."

115
00:18:47.432 --> 00:18:52.262
We thought it was like a market deal, 'cause a lot of the deals that transact are actually in the private markets and not in the public markets, right?

116
00:18:52.272 --> 00:18:57.712
The public markets are more kind of the old crypto companies and, and so forth, but most of the deals happen in the private market.

117
00:18:57.772 --> 00:19:04.692
So we thought it was like, "Hey, this is actually a pretty fair deal with, like, institutional blue chip platforms that, that are data center developers."

118
00:19:04.832 --> 00:19:13.192
Um, but I felt like we added unnecessary, like, like, eyeballs to our tenant, because everyone then was like, "Oh, we like the Hut 8 deal, we want the Hut 8 deal," and so forth.

119
00:19:13.201 --> 00:19:22.432
And we said, "Our goal is to go get more and more sites done, and I don't wanna add, like, unnecessary kind of friction to our tenants if, if not necessary." Right? And so- I see.

120
00:19:22.602 --> 00:19:31.491
'Cause that does kind of paint, not a target on them, but it does create a maybe competitive disadvantage when they're trying to sign- Right... other deals, and they say, "Well, hey, this deal- Right...

121
00:19:31.532 --> 00:19:40.832
why can't we get these terms that he did?" Right? [laughs] Exactly. Exactly. That, that, that's right. And so, um, so I, I spoke to, uh, to kind of the tenants at River Bend. I'm like, "Hey, I have this idea.

122
00:19:40.872 --> 00:19:50.052
What do you think, like..." Because... And then I spoke to some of our key larger shareholders, and I said, "Look, you care about, like, who's the credit who commits to this contract, right?

123
00:19:50.072 --> 00:19:58.762
Because that determines your confidence level that they g- will pay the bill over the 15 years and it's still renewed. So if we tell you the credit, is that good enough where, like, there's only

124
00:19:59.872 --> 00:20:06.612
a select handful of people in the world that have that credit and that profile, and so you have a general sense of who it is?" And they're like, "Yeah, that makes a ton of sense.

125
00:20:06.702 --> 00:20:15.772
Like, tell us that, we can underwrite it, and then just go get more deals done." And so I think it's optimizing towards, our first deal was can we do this and can we build a data center platform?

126
00:20:16.352 --> 00:20:20.692
And then a lot of people said, "Well, was this like a one-time kind of striking luck, or can you do it again?"

127
00:20:20.732 --> 00:20:28.442
And a lot of people, like, were saying, "Oh, Hut can't do it again," like, "This was a one-time thing," et cetera. And then our second deal was the same but bigger, and I would argue better.

128
00:20:28.782 --> 00:20:35.102
And so, like, same structure, same economics, but a bigger site. And so I, I think we've shown that repeatability.

129
00:20:35.142 --> 00:20:41.372
And as we grow now, it's like we've already, I think, shown the market what we can do, and now all we have to focus on is scaling and growing.

130
00:20:41.732 --> 00:20:50.832
And so then I think we can be a little bit less kind of forward leaning on disclosing every single thing. We want to do that to kind of show credibility with our shareholders and what we can do.

131
00:20:50.872 --> 00:21:04.072
And, like, we did that not because, like, oh, I think we're gonna get a higher valuation, or we're going to go and, like, shareholders are gonna love for, for, for that, or we're gonna get an investment grade credit rating on construction phase, which allowed us to do, like...

132
00:21:04.292 --> 00:21:08.772
We did also, like, a kind of first time type of financing, which I can go into later if you want. But

133
00:21:09.772 --> 00:21:20.152
we did it because the question I ask is, I see the data center business as an ability to build our foundation of the company with, like, reliable and strong cash flows.

134
00:21:20.872 --> 00:21:25.852
And I am a deep believer that AI, we're only at the early innings and will only continue to grow.

135
00:21:26.372 --> 00:21:35.252
But running a large scale infrastructure business where we're deploying billions of dollars of capital, I have to prepare for what if, like, demand stops? What if demand pauses?

136
00:21:35.632 --> 00:21:43.472
And where are we, and what does that happen to our commitments, the obligations we make, and so forth? And, like, the reliability of those contracts, those cash flows are critically important, right?

137
00:21:43.532 --> 00:21:51.752
Because I want us to compound and grow as a business, and I care less about, like, where we're gonna be in, like, six months, a year. I care more about, like, where we're gonna be in 10 years, right?

138
00:21:51.762 --> 00:21:58.472
And so, like, how do we make sure, like, we do the right things for that? And I think as a result, we got a really strong deal.

139
00:21:58.852 --> 00:22:06.922
We got one that was financed really well, and, like, we were the only company that was able to do a 16 and a half year financing instead of a five-year financing 'cause we got an investment grade credit rating.

140
00:22:07.892 --> 00:22:15.782
So that, that financing you're talking about is specifically for Beacon Point or for Louisiana site? That was for River Bend. Um, and that was, like, the first one for River Bend. Okay.

141
00:22:15.792 --> 00:22:24.372
And now obviously we're looking at similar for, uh, Beacon Point 'cause it, it did really well. Yeah, so I, I wanted to touch on that. You know, you talk about planning for the future.

142
00:22:24.532 --> 00:22:30.232
The financing for, um, River Bend had this backstop from Google. You know, I...

143
00:22:30.742 --> 00:22:41.412
[laughs] It, it, it's gotta be so sweet as a company, uh, that used to mine Bitcoin pure play, and financing was impossible to get on terms that were at all favorable.

144
00:22:41.861 --> 00:22:59.072
And I'm curious, as you look towards the project level financing for Beacon Point, how, how does the quality of the contract or, you know, having a hyperscale tier tenant, what does that do in terms of obviously increasing your credit profile, but in terms of, uh, opening the doors to other access of financing that y'all didn't have in the past?

145
00:22:59.352 --> 00:23:13.292
Like, what options are available to you that did not exist five years ago? So see, credit of the tenant, which is critically important, but during construction phase, it's actually the stability of the execution program.

146
00:23:13.652 --> 00:23:21.912
'Cause there have been other Google, uh, s- financial supported, like, data center deals, but they weren't investment grade from S&P and Fitch day one during construction.

147
00:23:21.972 --> 00:23:31.192
Usually what happens is, like, you're sub-investment grade, you raise in, like, the high yield markets. Then when the data center comes online, then you become investment grade because the construction risk is gone.

148
00:23:31.752 --> 00:23:41.192
For us, we got an investment grade during construction phase, which is why it allowed us to go to the investment grade markets day one instead of during a refinancing in a couple of years when this data center stabilized.

149
00:23:41.232 --> 00:23:49.472
And that came straight from, like, the quality of, like, how we think about execution, the partners that we have with Jacobs and Vertiv, how we thought about the commercial contracts.

150
00:23:49.492 --> 00:23:57.372
So, like, that's what led to the day one. And then post-commercialization, we think we're gonna have an upgrade to even have a higher, uh, grade of investment credit rating.

151
00:23:57.392 --> 00:24:02.352
And so the tenant is important, but the execution is equally as important for this structure. And what that unlocked for us was

152
00:24:04.272 --> 00:24:12.432
we think there's gonna be a lot of financing, 'cause most of this financing is short-term, like two to five-year financing, which is essentially like construction financing, and you have to go and you have to refi the project.

153
00:24:12.822 --> 00:24:16.791
And so you're gonna have, like, 100 billion plus dollars of capital trying to get refinanced in that period of time.

154
00:24:17.092 --> 00:24:37.288
And so we said, "If we can just finance this for 16 and a half years, pay down the whole thing through our cash flows of the data center, not rely on any future financings, we completely take away refinancing risk and we stabilize the project."And we were only able to do that because we got an investment-grade rating day one during the construction phase because of how much we de-risked the project.

155
00:24:37.748 --> 00:24:53.888
And what was exciting, to your point, like, five years ago when we first kind of started the business and we were growing, the idea of raising $3.25 billion on a 16-and-a-half year financing with some of the most renowned institutional investors in the world, we had over $10 billion of demand with, like, the largest institutions in the world.

156
00:24:53.928 --> 00:25:01.168
I mean, it was like a dream come true, you know? It's like as a founder, as an entrepreneur, like, these are the things that you take a pause, that you pinch yourself, and you're like, "Is this, is this real?"

157
00:25:01.228 --> 00:25:14.168
And I think it goes to show, like, the team works really, really hard, and we wanna build a great business, and this is, like, a validation of the work that we spend, like, is getting recognized because we didn't, like, try to go get an investment-grade rating.

158
00:25:14.208 --> 00:25:18.668
We just tried to get, like, a deal that we were comfortable with, that we, like, really minimize execution.

159
00:25:18.698 --> 00:25:22.708
And then when we were going through the rating agencies, they're like, "Hey, we think you can be investment grade." And we're like, "Wait, what?"

160
00:25:23.048 --> 00:25:29.068
Like, no data center has really been investment grade during the construction phase, and it kind of unlocked this whole series of opportunities.

161
00:25:30.868 --> 00:25:48.068
So you guys, a lot of data center companies are really cooking, executing well right now. But against this backdrop, we have a rising pushback I see at the rural, grassroots, and national level against data centers.

162
00:25:48.208 --> 00:26:04.288
I'm in a very red state, and my, you know, rural Podunk buddies, like, are hating on data centers. Never anticipated this. Yeah. What do you think HUD8 and other data center operators can do to address this?

163
00:26:04.388 --> 00:26:19.518
And, um, what do you think is behind this, uh, this rising outcry? So much of what's driving this is fears that are not true, right? The biggest fears around data centers today are water consumption,

164
00:26:20.428 --> 00:26:34.668
rising energy pricing, aesthetics, noise, and then, like, general AI jobs. And if you go through those, water, I think all data centers today, and if not, like, they are, are going to soon, are a closed loop system.

165
00:26:35.108 --> 00:26:42.368
You're no longer using water to cool down the chips. And historically, I don't think people did that because it's more expensive up front to invest into, and the PUE is not as great.

166
00:26:42.388 --> 00:26:46.708
But, like, in a world where, like, water matters, you invest in that up front, right?

167
00:26:46.748 --> 00:26:54.988
And so it's interesting, in our Riverbend project, we're using a closed loop system, and in addition to that, we donated a $16 billion water infrastructure upgrade program.

168
00:26:55.248 --> 00:27:04.168
And, like, if you think about what the data center uses, we use water for sinks, toilets, and, like, showers on emergency use cases. Like, that's it. It's less than, like, a traditional office building, right?

169
00:27:04.208 --> 00:27:07.968
And so, like, the actual data center servers don't use water if you're using a closed loop system.

170
00:27:08.288 --> 00:27:17.068
Those are already, like, uh, like, those are, uh, fluids that you come and you ship in when you kind of commission the project up front. The second is, um, in terms of energy.

171
00:27:17.828 --> 00:27:27.868
If you think about building a data center today, data centers are asked to pay for all the upfront CapEx and upgrading all the infrastructure, transmission lines, electrical systems to allow to support that energy.

172
00:27:27.948 --> 00:27:31.527
So you're actually investing into a better grid for everyone else to use.

173
00:27:32.038 --> 00:27:42.748
And then the second is, like, the energy supply that you consume, you're having to commit to that, so if the utility goes and procures that power, you're committing to it on the other side, which is why the credit of the counterparty and the tenant matters so much.

174
00:27:42.768 --> 00:27:54.468
But in a lot of these energy utilities that I'm talking to, they're trying to change the narrative too, 'cause they actually think power prices are coming down from data centers because they're helping subsidize additional net new build and more utilization.

175
00:27:54.748 --> 00:28:03.258
So I think that, like, you have one-off probably small utilities that, like, brought in someone, didn't do this early on, and then, like, that creates it, but, like, that, that's not a real thing.

176
00:28:03.868 --> 00:28:10.088
The, the third is aesthetics. That's one thing we're focused on is let's spend an extra 10, 15 million bucks and make this thing look beautiful, right?

177
00:28:10.108 --> 00:28:14.088
Like, I've told our team, like, "I want these things to look like museums when you drive by, not like warehouses."

178
00:28:14.348 --> 00:28:21.208
Like, let's make them, like, beautiful things that people are proud of in terms of, like, infrastructure assets. Um, and then we make a big impact.

179
00:28:21.248 --> 00:28:29.108
Like, it really does make impact in the rural areas that we go and build. Like, the tax bill revenues are real. They make an impact on the schools. They make the impact on the communities.

180
00:28:29.148 --> 00:28:38.908
Like, they al- those are real things. Um, and so I think, like, why do we see all of this? One is when I s- tell people those things, they're like, "Oh, really?"

181
00:28:39.288 --> 00:28:46.708
Like, it, there's so much FUD out there, and then the, and I don't think the data center, the AI ecosystem has done a good job 'cause historically everyone kind of just builds quietly.

182
00:28:46.728 --> 00:28:52.908
They're not really proactive out there, and, like, the narrative has kind of taken its own form, right? And I think that's something that needs to be solved.

183
00:28:52.948 --> 00:28:58.578
And the other element is, like, I feel like this is a very easy thing to be politicized with.

184
00:28:59.308 --> 00:29:09.318
And, like, we had a data center we were working on, and we had people who came and protested from three states over in, like, a local, like, county. I was like, "What do you have to even do with this county?"

185
00:29:09.588 --> 00:29:12.368
And they're going, like, spewing all this FUD with everyone that, that's there.

186
00:29:12.398 --> 00:29:24.038
And so I think, like, I was talking to one of the large kind of AI companies, and we were, like, jamming on ideas and like, "You know, the day that, like, AI goes and solves cancer, people will see, like, the impact of what this is, like, possible of."

187
00:29:24.308 --> 00:29:31.088
Right now it's, like, going and trying to build company processes and systems [laughs] and this and that, but, like, there's so much, like, good that can come from it.

188
00:29:31.188 --> 00:29:39.268
Um, and so I, I think in general, like, I think about it as we need to b- allow more building of infrastructure for a couple reasons in the US.

189
00:29:39.808 --> 00:29:48.688
Number one, latency is becoming less of an issue, and so as a result, you can really place data centers anywhere in the US, but also anywhere, like, internationally.

190
00:29:48.697 --> 00:30:00.128
And so if it becomes really hard to build in the US, people are just gonna buy compute internationally, and then you're driving the hundreds of billions of dollars of investment overseas, which is, like, not something I think we want as a country.

191
00:30:00.188 --> 00:30:10.368
Like, I think these, like, jobs and, like, going and building this infrastructure, like, we have this revitalization of, like, investments of, like, infrastructure into the US, and I, I think we w- we wanna retain that.

192
00:30:10.808 --> 00:30:18.148
Um, the, the second element, this is more of, like, a macro thing, the US needs to continue to be the leading, like, front, at the leading frontier of AI.

193
00:30:18.748 --> 00:30:35.348
And as you all know, like, building the frontier technologies requires the compute, and we don't want a scenario where we're bottlenecked and we don't have enough compute, and China's building generation, and China's building infrastructure faster than us, and we say, "Shoot, like, we have the smartest people in the world, and we can't do it 'cause we just don't have the compute infrastructure."

194
00:30:35.908 --> 00:30:48.036
Um, I was with-Someone at AirCo, a m- very senior person there, and we were talking about this kind of, like dynamic. China's brought on as much energy infrastructure as a whole US grid just in the last four years alone.

195
00:30:48.066 --> 00:30:56.196
Like, to think about, like development and, and build. And so I think, like I, I, I'm a deep believer that like we need to come together as a country.

196
00:30:56.236 --> 00:31:06.556
We need to like clear up all the FUD that exists and like give people the truth, and like there are closed loop and open, uh, systems in terms of cooling. Like, let's just get everyone to do closed loops.

197
00:31:06.596 --> 00:31:16.876
Let's invest more money and let's, let's solve the water issue. Like, there's things that are already being done that we can educate folks on and, and like this is, this is not like self-serving. This is more of a...

198
00:31:18.426 --> 00:31:23.576
This investment I feel like is so good for the country, so good for the communities that it impacts.

199
00:31:23.616 --> 00:31:34.316
And for any kind of folks that like need to learn more, for like any bad actors, 'cause you have everyone and their money- mother trying to build data centers today as well, which creates a lot of like noise and stuff in the system as well.

200
00:31:34.336 --> 00:31:40.126
And so I think some of the things that utilities are doing, which is increasing like deposits for people to develop data centers, I think those are good.

201
00:31:40.436 --> 00:31:50.836
Because it, it, it pushes like the bigger developers who I think have a more mature foundation and like who care about like building a reputation and doing the right thing, go and build a lot of this infrastructure.

202
00:31:51.616 --> 00:31:58.706
Asher, I think you're the first person I've ever heard, uh, use the term data center aesthetics, so I'm excited- I, yeah, I like that. Yeah. I like that.

203
00:31:58.716 --> 00:32:07.336
For anyone who wants to go check out some of their mock-ups for, uh, I believe it's the Beacon Point, maybe it's the RiverBend campus, but um, either here- It was only the marketplace view.

204
00:32:07.396 --> 00:32:15.256
I'm trying to make it look better than that. [laughs] Yeah, I mean, well, you know, I will say this much, it, it looks a lot better than some of the other mock-ups I've seen or actual data centers.

205
00:32:15.296 --> 00:32:17.996
And last thing- I've been to some ugly data centers, so yeah.

206
00:32:18.156 --> 00:32:34.106
[laughs] Last thing before we get you off here, Asher, it's gotta be really frustrating to have come from Bitcoin mining where you got screamed at for boiling the oceans and thinking, "Well, now we're an infrastructure and energy company that's providing one of the most useful, if not the most useful technology in the world," and people still aren't happy.

207
00:32:34.636 --> 00:32:42.996
You know, at the end of the day, you're not gonna be able to please everyone, but, um- But you're used to it by now. We've been through this rodeo a few times. I think it's important to

208
00:32:44.136 --> 00:32:52.586
deeply understand what conc- the concerns are and like have your own like perspective of like are they real, are they not? And I think like data center aesthetics was a real thing.

209
00:32:53.136 --> 00:33:08.536
Water could have been a real thing, and I think as long as like we do the right thing and like we believe in what we're doing and we believe it's the right element, to your point, like there will be people who love us, there will be people who hate us, but if we actually solve the problems people care about, like that's how we, we solve this in the long term, right?

210
00:33:08.576 --> 00:33:19.196
Like, I believe actions are a lot more powerful than words, and so like why don't we just like walk the talk? 100%. Well, Asher, thank you so much for joining, man. Appreciate your time as always.

211
00:33:19.276 --> 00:33:29.496
Best of luck, uh, for the rest of the year for y'all. Thanks for having me, guys. See you later. Thank you, Asher. So s- just a quick riff on that before we go to the ad and the news. Yeah, data center aesthetics.

212
00:33:29.596 --> 00:33:38.276
Um, you know, I, I just keep coming back to this meme because Bitcoin miners really have been on the forefront of this for a long time. Yeah.

213
00:33:38.946 --> 00:33:48.826
[laughs] So it's like all these data center companies, which like to Asher's point earlier, this was not something that they ever felt like they had to do community outreach for 'cause number one, it just wasn't in the news cycle as much.

214
00:33:48.876 --> 00:33:56.996
It's not as exciting. You know, traditional data centers, cloud data centers, very boring. You know, there, there's no dazzling application that you can point to to say, "This is what this does."

215
00:33:57.036 --> 00:34:01.056
It's like, "Oh, I can actually, I can stream. I can, I can pull up my email."

216
00:34:01.065 --> 00:34:09.116
[sniffs] So but they never really had to do community outreach, which is ironic because per Asher's point, as I understand it, a lot of the traditional data centers are not closed loop.

217
00:34:09.496 --> 00:34:18.716
For the ones that are water-cooled, like they actually, most of them are, or a lot of them are actually using, you know, uh, water that is a, uh, open loop system.

218
00:34:19.196 --> 00:34:26.656
So those are actually theoretically worse for the water situation. But, you know, Bitcoin miners have been fighting this for a long time.

219
00:34:27.016 --> 00:34:37.096
And so I actually think that in some ways these Bitcoin miners who are now AI services companies are kind of more well-positioned maybe to at least take on the criticism. Maybe they have a thicker skin.

220
00:34:38.116 --> 00:34:39.056
[lip smack] But- Yeah.

221
00:34:39.096 --> 00:34:52.676
So we have here- I- Uh, we have a comment here from Yeehaw Bitcoin on YouTube, quote, "The Hut 8 Vegas site is absolutely the most aesthetically pleasing data center/Bitcoin mining facility I've ever seen." So, uh, I...

222
00:34:52.996 --> 00:35:05.296
Colin, I think, I think we should start... We should do a data, a, a Bitcoin mining data center, AI data center aesthetic tier list. S- Sexiest, uh, data center? That's a g- Yeah... that is a good tier list.

223
00:35:05.356 --> 00:35:12.596
And not to, not to glaze them too much, but is, does this mean that Hut 8's becoming the Apple of, of data centers right now? Yeah, the Apple of data centers. [laughs] It's all about the image.

224
00:35:12.626 --> 00:35:26.976
It's all about the brand, baby. Yeah. All right. We'll, we'll leave that there. All right. We're gonna keep going. We have, uh, Iron news, we have Core42. We have, uh, some fun Maru pulp at the end of this.

225
00:35:27.016 --> 00:35:38.456
But before that, we have a word from our sponsor, CleanSpark. [gentle music] We are CleanSpark, America's Bitcoin miner, a publicly traded company with the largest operating hash rate,

226
00:35:39.856 --> 00:35:51.956
powered entirely by self-operated infrastructure across four states. This is our proof of work, and we are setting the standard for what's next.

227
00:35:52.916 --> 00:36:09.756
Learn more about the intersection of energy and Bitcoin at cleanspark.com. All right. Iron is back in the news, and by back in the news, they never really left. They're in the news every week it seems.

228
00:36:09.836 --> 00:36:23.096
Iron just signed a $1.6 billion Dell Blackwell deal for Childress, which has raised its targeted ARR to 4.4 billion. That's billion with a B, ladies and gentlemen.

229
00:36:23.196 --> 00:36:37.836
We've left the millions behind with all of the S19J Pro Series ASICs. [lip smack] So few notes on this. This hardware is not for their Microsoft deal. They, they have a separate deal with Dell for that rollout.

230
00:36:37.916 --> 00:36:49.416
This will support their n- deal with Nvidia, five-year, $3.4 billion deal for 60 megawatts at Childress. Iron said is that it, it is advancing GPU financing in connection with the agreement.

231
00:36:49.726 --> 00:36:59.636
Would love to have a segment on exactly-The ins and outs of GPU financing, obviously a much different game than ASIC financing, so I'm not gonna draw any parallels to it.

232
00:37:00.096 --> 00:37:06.516
But typically, hardware financing has not worked out well for Bitcoin miners in the past, but different era, totally different terms, I would imagine.

233
00:37:07.196 --> 00:37:18.096
Like I said, the, uh, the ARR will-- uh, Iron's ARR will rise from three point seven billion to four point four billion as a result of this, according to their own estimates. Commissioning is targeted for early 2027.

234
00:37:18.516 --> 00:37:31.676
That's pretty close. That's right around the corner, man. And y- like I said, this is for the Nvidia rollout. They also have their nine point seven billion, five-year deal with Microsoft for 200 megawatts of IT load.

235
00:37:31.796 --> 00:37:43.396
So, uh, markets like this, Iron jumped on the news. Obviously, you like to see having the picks and shovels in place before you start digging for the gold. [sniffs] What are your thoughts here, Charlie?

236
00:37:43.416 --> 00:37:52.496
I mean, pretty standard. Yeah, pretty standard. I, I always try to come up with some color 'cause I'm not like the money or the finance guy. Th- you know, that's more you and other folks.

237
00:37:53.076 --> 00:38:00.456
Um, uh, I think a little detail about this is that, uh, this is a deal for air-cooled Blackwells.

238
00:38:00.496 --> 00:38:12.216
And, uh, as I research [chuckles] GPUs more and more these days, the, like, typical, like, configuration that these, that these Blackwells ship in is like a, like a water-cooled rack system.

239
00:38:13.016 --> 00:38:22.036
And I think it's notable because these air-cooled ones are a little bit easier to fit into, like, the existing, I think, Iron data center structure. So kinda notable there.

240
00:38:22.356 --> 00:38:30.336
Uh, that's really the only kinda color commentator, commentary. That's-- Well, that's kinda leads into my, my color for this one as well.

241
00:38:30.756 --> 00:38:39.656
Wondering, that obviously probably helps them on the CapEx for these builds, right? Yeah. Because as a reminder, Iron is going full-on vertically integrated in the O-Cloud.

242
00:38:39.696 --> 00:38:47.476
Unlike other Bitcoin miners that are building powered shells, where they're going to have a tenant come in and fill that space, they're just gonna manage the data center and power, Iron's doing it all.

243
00:38:47.696 --> 00:38:56.456
And that's part of the reason why Iron has been the largest gainer or one of the largest gainers in the AI race among Bitcoin miners, because there's more of a share of revenue.

244
00:38:56.616 --> 00:39:06.956
If you're gonna do that, obviously, you're not leasing out to a tenant. You're leasing out the compute to a company like Microsoft or Nvidia, right? But with that comes a massive CapEx burden.

245
00:39:07.016 --> 00:39:17.336
So this is from our megawatt report that we released [lip smack] last week, maybe two weeks ago at this point. The weeks blend together. And this is a Bitcoin mine down here.

246
00:39:18.556 --> 00:39:28.216
This is in, uh, thousands, uh, dollars, thousands. So, uh, this is per megawatt build. On the low end in Bitcoin mine is, like, 300K. This is probably gone.

247
00:39:28.276 --> 00:39:30.296
This is-- These are-- Some of these are a little outdated figures.

248
00:39:30.316 --> 00:39:39.746
They're based on- Probably higher now, but like a year ago- Probably higher now, but- Three hundred a megawatt was a good, like, efficient, like, uh, build-out price Good Chicken Shack build. Yeah. You know?

249
00:39:39.796 --> 00:39:48.336
And then mid would be a $600 me- a 600,000 a megawatt. High would be a million dollars per megawatt. Now we're getting into big boy numbers.

250
00:39:48.836 --> 00:40:00.376
If you're building a powered shell, you're looking at anywhere from eight to 16 million per megawatt. If you're just buying GPUs, you're looking at 25 to 32.5 a million per megawatt.

251
00:40:00.956 --> 00:40:12.916
And if you are doing it all, 35 to almost $55,000 per megawatt. So just a massive CapEx burden, right? And hence why, you know...

252
00:40:14.556 --> 00:40:20.656
[chuckles] Hence why the dollar figures are so big on these deals, but also why financing obviously is going to be instrumental here for Iron.

253
00:40:20.676 --> 00:40:23.836
They also have a decent cash position, although that's not gonna cover all of it.

254
00:40:24.456 --> 00:40:34.896
And going back to what I was saying earlier, I won't share this part of the report, but we do EV multiples per megawatt for some of these deals. And if you look at M- Iron's EV,

255
00:40:35.856 --> 00:40:53.096
uh, per megawatt capacity, projecting out into the future when this-- when all of these are built, because all these companies are being valued on future cash flows, Iron's EV to megawatt for future capacity as of May 12th was they had a, a multiple-- uh, they had a ratio of 80.

256
00:40:54.056 --> 00:41:04.506
And that is by far and away much larger than any of their peers who are just doing the powered shell approach. So obviously, the market is rewarding this kind of risk very highly.

257
00:41:04.936 --> 00:41:11.156
And I say risk because, like, there is more risk in ha- trying to do it all. But also with more risk, as I write in the report, comes more reward.

258
00:41:11.835 --> 00:41:22.116
So, you know, there, there's a reason the market's been loving Iron recently, and they, they kinda just keep trudging along. So I don't really have anything else to add. No, let's, let's keep rolling on.

259
00:41:22.176 --> 00:41:30.776
We're gonna go to Core42 news and then Mara. But before that, a word from our sponsor, Luxor.

260
00:41:31.296 --> 00:41:42.176
[static noise] This episode is brought to you by Luxor's Commander, Bitcoin miner management software for enterprise operation.

261
00:41:42.556 --> 00:41:48.776
Commander gives you real-time fleet monitoring, bulk remote commands across your fleet, and Intelligent Miner.

262
00:41:48.816 --> 00:42:00.936
That's an automated profitability engine that runs every five minutes against live hash rate and energy markets. ERCOT backtests show a 10% improved profitability with intelligent mining versus binary mining.

263
00:42:01.476 --> 00:42:11.376
Commander Pro is rough- $100 per megawatt or a 25 basis point pool fee adder, which is roughly half the price of competitors. But you can try it for free for 60 days.

264
00:42:11.416 --> 00:42:16.796
So if you're interested, go to luxor.tech/commander to learn more.

265
00:42:17.956 --> 00:42:34.456
All right, Core42, a company that we only really cover when we have news with them with regards to TeraWulf, but getting them on the show today because they have secured 550 million from HSBC for US and European AI infrastructure build-out.

266
00:42:34.466 --> 00:42:44.376
The European part is part of why I wanted to cover this, 'cause that's obviously not a market we tap into very much. The data center market in Europe is much smaller. They just don't produce as much energy.

267
00:42:45.116 --> 00:43:03.246
And ironically, the largest data center market in continental Europe is Germany, and I wonder how tenable that's going to be because Germany decommissioned, like, all of its nuclear, and now it's relying on coal and nat gas to fill in the gaps.In fact, its manufacturing base is really hurting from that.

268
00:43:03.256 --> 00:43:12.756
They've been in a recession on the manufacturing side of things for a, a year or two at least. And so you have to wonder what the downstream effects will be for their dominance in the data center realm. But

269
00:43:14.096 --> 00:43:21.396
that's a sidebar. To get back to Core42, some details from this news. Uh, 550 million, like we said, there will be...

270
00:43:21.456 --> 00:43:34.996
it'll come in two tranches, 240 million completed in February 2026, and 310 million completed in May 2026. Core42 said the facilities will support deployments for enterprise, government, and hyperscale customers. Now,

271
00:43:36.076 --> 00:43:47.516
this company is headquartered in the UAE, and it's been expanding on both sides of the Atlantic. Its European headquarters are in Dublin, but it has data center deployments underway in I-Italy and France,

272
00:43:48.576 --> 00:44:01.456
and it also obviously has this 70 megawatt critical I- or 70 megawatt gross IT, 60 megawatt critical IT with TerraWolf at its Lake Mariner facility that, that started booting up, I believe at the, uh, uh, last year.

273
00:44:02.536 --> 00:44:16.996
Okay. So Charlie, when I was looking at this, I had to run the numbers on Europe versus the US for, for data center capacity. Caveat here, there's obvious-- there's not, like, any one true source for this.

274
00:44:17.056 --> 00:44:24.836
There's no single source of truth. A lot of th-these estimates come from investment banks, from other think tanks, things like that.

275
00:44:25.396 --> 00:44:33.116
And most of the figures that I could get were only for like 2024, 2025, so take that with a grain of salt. But I think this does give us a rough idea of the ratio.

276
00:44:33.136 --> 00:44:40.356
It's like basically for every 10 data centers we have in the US or v- 10 megawatts of data centers we have in the US, Europe has like one megawatt.

277
00:44:41.076 --> 00:44:45.876
So the primary markets in Europe are Frankfurt, London, Amsterdam, Paris and Dublin.

278
00:44:46.556 --> 00:44:57.256
Uh, Newmark estimates these five data centers, uh, these five markets have 4.6 gigawatts of live capacity, which is roughly 40% of all of Europe's installed capacity at the beginning of 2025.

279
00:44:57.276 --> 00:45:09.876
For 2024 figures from Deloitte, the US had roughly 30 gigawatts, just give or take, of data center load, of which roughly four gigawatts was AI specific. Now that's 2024. That's ve- early days for the AI stuff, right?

280
00:45:09.936 --> 00:45:15.356
Like that number has no doubt increased today. Um, but that just kinda gives you an idea.

281
00:45:15.416 --> 00:45:23.696
In terms of growth, Bloom Energy prodect- projects US data center capacity to nearly double from 80 gigawatts in 2025 to 150 by 2028.

282
00:45:24.656 --> 00:45:34.256
Looking over across the Atlantic at Europe, McKinsey expects Europe to grow from 10 gigawatts in 2024 to 35 in 2030. So the European market's obviously going to expand.

283
00:45:34.316 --> 00:45:43.136
I think the opportunities are going to be fewer and farther between. I would imagine France is going to, uh, my, my, you know, I, I guess the one

284
00:45:45.096 --> 00:45:56.066
prophecy I would throw out here is that France, I would be shocked if they didn't end up becoming the largest data center provider in Europe because they are almost all nuclear now. All nuclear.

285
00:45:56.076 --> 00:45:57.456
They export a lot of that.

286
00:45:57.536 --> 00:46:12.376
So this is part of the problem actually with, you know, the, the EU's loosely confederated system is, you know, I mean, they could sell it without the EU, but you're only as strong in some ways as your neighbors, and France is probably gonna sell a lot of that energy, and already is.

287
00:46:12.406 --> 00:46:18.676
It's, it's, it's the largest exporter of energy in Europe. So that might throw a wrench into things.

288
00:46:18.716 --> 00:46:26.636
But there are some attractive jurisdictions like France for this, these data center build-outs in Europe, but still- Yeah... well behind the US.

289
00:46:27.236 --> 00:46:48.716
I mean, if you are this like AI machine god, AI and data centers are the single most important commodity as the world moves away from oil to compute, then like think about the implications for these like loose confederations of like the EU where, um, you know, Germany is the manufacturing and, and finance powerhouse of the, of the EU.

290
00:46:50.036 --> 00:47:02.956
E- E- England used to be its like capital and some foreign direct investment and, but if France is the, is the, is the, uh, power generation, like they're going to want to, uh, own that more directly.

291
00:47:03.096 --> 00:47:16.856
So this could put a strain on some of the, the pretty kind of like uneven distribution of some of these like assets, uh, across the EU as- Yeah. I would also be curious the un- the kind of

292
00:47:18.636 --> 00:47:29.716
intangibles of doing business over there. I mean, we can look at megawatts per- Or, or in, if you're in France, you say intangibles. Intangibles. With a bunch of extra E, U's and accents. Intangible. Yeah. Quel surprise.

293
00:47:29.786 --> 00:47:40.466
Yeah. [laughs] No one builds there anymore. But seriously, I mean, you do have to wonder, you, you look at these bubble charts of like billion-dollar companies in the US versus Europe. Yeah.

294
00:47:40.516 --> 00:47:46.536
And it's like comparing, you know, a freaking yoga ball to a gumball when you look at these things.

295
00:47:46.596 --> 00:47:55.676
I mean, it's like the, the, the size i-in the US i-is, is massive for these multi-billion, um, billion dollar companies, multinationals, things like that.

296
00:47:56.096 --> 00:48:07.036
Whereas in, in Europe, they just, uh, they're so choked by bureaucracy and red tape that it's hard to get anything done, and I do wonder if some of these projections are maybe not taking that into account.

297
00:48:07.116 --> 00:48:17.696
Or if Europe will have a sea change in opinion and start building things again. I, I'm not really holding my breath on that though, so I'm, I'm really curious to see how that race pans out.

298
00:48:17.736 --> 00:48:24.516
I mean, if you think anti-data center stuff here is bad, I would imagine that- Oh my gosh, can you imagine European anti-data center?

299
00:48:24.576 --> 00:48:33.516
Yeah, like someone does a Eurovision song about, you know, the American or like the, the Europeans fighting a gallant war against the AI overlords or something like that.

300
00:48:33.916 --> 00:48:43.476
But seriously, I do, like the, the difference I think mostly between Europe and the EU i-in terms of actionability or, um, actionability is not a freaking word.

301
00:48:43.756 --> 00:48:52.776
In terms of actually being able to execute on those fears is Europe actually does have a history of a much stronger bureaucracy.

302
00:48:53.196 --> 00:48:59.556
They c- they, they, they, they do hamstring things, and they seem to be able to wield more power over what can and can't be done.

303
00:49:00.176 --> 00:49:15.504
For all of its faults in the US with a federal system, if you don't like the federal law, you can go to a state, and then you can build there.And if the state has laws carved out for you, you can challenge laws that you feel like are draconian in the Supreme Court.

304
00:49:15.724 --> 00:49:25.284
I, I do wonder between the EU's regulations and all of these companies' respective regulations, how much wiggle room do they actually have to build these things?

305
00:49:25.944 --> 00:49:36.624
And going back to the data center aesthetics problem, um, not gonna lie, as a guy who loves to occasionally go to Europe and travel, it's, it's a more beautiful country than- It is...

306
00:49:36.744 --> 00:49:47.374
um, it- they have more beautiful, uh, countryside. Look, if a big metal Walmart-looking data center goes next to a Love's, nobody really cares.

307
00:49:47.404 --> 00:49:56.484
If you go next to a French villa that's been there for hundreds and hundreds of years, uh, I do think that, you know, the- there's a problem there.

308
00:49:56.584 --> 00:50:03.704
Yeah, or like a half a gigawatt data center right outside of Rothenburg- [laughs]... which is, like, basically a still medieval intact city in Germany. Yeah. You know?

309
00:50:03.724 --> 00:50:11.554
Yeah, it's a data center right outside Seine or- Yeah... you know, something like that. They're just gonna have to build them like cathedrals, right? Yeah. We'll just that- Think of it-...

310
00:50:11.554 --> 00:50:21.584
that's, that's the way we get around this. I actually think Asher, you know, like it, it seems kind of, it seems like kind of a fluff comment, but I think that there's some truth- But it does, yeah.

311
00:50:21.644 --> 00:50:31.144
Yeah, to- The aesthetic is the, the aesthetic is the, is the, the true real ethos. The question is how much wiggle room do you have to make a box in the middle of a desert look pretty?

312
00:50:31.224 --> 00:50:38.674
[laughs] Like I'm thinking of like the Utah data center, right? Cover it in weird- Like a gaudy... I don't know. You know? Well, uh- Yeah... we'll leave that there. Yeah.

313
00:50:38.684 --> 00:50:49.324
Keeping an eye on Core:42 and also just, uh, really curious to see about the European market and, um, also curious to see if I end up being proven right about France taking the market share from Germany.

314
00:50:50.084 --> 00:51:08.004
We'll have to cover that on this podcast. We got one more story on MARA and armored cars, but before that, a word from our sponsor, Ligos. [lightning crashing] Hedge funds are getting liquidated. Is your Bitcoin safe?

315
00:51:08.444 --> 00:51:19.824
It's not just Bitcoin's price drying up. Whales, trading desks, hedge funds are reeling from October 2025 and February 2026 liquidation events that drove Bitcoin's price down, down, down.

316
00:51:20.584 --> 00:51:33.074
Now, counterparty risk is rampant, so it's more important than ever to know who actually controls your Bitcoin, and with Ligos Finance, that is always you. Ligos Finance is Blockspace's preferred crypto lending partner.

317
00:51:33.814 --> 00:51:45.644
They have Bitcoin native smart contracts that ensure you're always in control of your keys when you're using their product. No rehypothecation, no bridging, just Bitcoin-backed lending with native smart contracts.

318
00:51:45.904 --> 00:52:00.984
You always hold your keys. If you wanna learn more and get competitive rates for as low as 10% APR, go to ligos.finance. Okay. Can I drive this one? Yeah, yeah. Okay, so- Drive that, drive that Mad Max car, baby.

319
00:52:01.224 --> 00:52:17.964
Yeah, so I'm gonna drive this figuratively. Um, the headline from the Financial Times is, "A Bitcoin miner spent $860,000 armoring vehicles for its bosses," a fact for which we have very little context.

320
00:52:18.584 --> 00:52:27.624
So- Okay, sorry, I, I'm, I didn't know you were driving. I gotta like- Okay, go, go, go. There's little context 'cause you're not paying attention. Yeah. People get kidnapped for their cryptocurrency.

321
00:52:28.144 --> 00:52:32.454
I- Okay, you're getting a- you're getting ahead of it, Colin. Sorry. Let me give you the story, okay? Sorry. I just hate ESPN.

322
00:52:32.504 --> 00:52:44.284
Here's the story from Financial Times, one of our favorite publications which serious coastal elites love, but crypto Bitcoin people know that they're just kinda shooting from the hip. So

323
00:52:45.384 --> 00:53:00.914
to counteract such dangers, as Colin described, of crypto folks being targeted, uh, one bodyguarding firm boss was happy to tell, uh, by the Financial Times, that crypto firms are enlisting bodyguards.

324
00:53:01.504 --> 00:53:14.704
For example, crypto exchange Coinbase paid its chief executive Brian Armstrong $7.6 million for personal security-related compensation, representing home security, family protection, secure transportation arrangements.

325
00:53:15.404 --> 00:53:24.444
The notable one that kind of gets in our neck of the woods is, from the latest DEF 14A compensation tables for Mara Holdings,

326
00:53:26.164 --> 00:53:55.444
Mara Holdings CEO Fred Thiel, pursuant to Mara's se- Mara's security program of with a budget of $4.3 million, included a one-time expense for vehicle armoring of $430,000, a one-time expense for home security installation of $58,000, an incremental cost to the company associated with Fred Thiel's personal use of company aircraft of $43,000, and a company contribution under the 401[k] $10,000.

327
00:53:55.644 --> 00:54:08.684
Additionally, the CFO, Mr. Khan, pursuant to Mara's security, Mara's security program of $3.946 million, included a one-time expense for vehicle armoring of 438K.

328
00:54:09.364 --> 00:54:20.584
So this is $839,000 across the CEO and CEFO of Marathon, Mara Digital Holding, Mara Digital Holdings- And-... um, for their executives- To be-... to get armored cars.

329
00:54:21.104 --> 00:54:34.244
And to be clear, the 3.4 billion is for Fred alone in 2025. Yeah, Fred, that- So that, yeah, that covers Fred, and then Khan has 3.95 billion, or million. God, I'm gonna be like, 3.95 million.

330
00:54:35.124 --> 00:54:45.184
So together you're looking at over $8 million for executive protection. Yeah. S- and I think that seems crazy to a lot of people, but- It sounds crazy.

331
00:54:45.224 --> 00:54:55.744
Like, and it, this seems, like if you're reading this as a normal person, you're like, oh my God, they're so caught up in themselves. They're, they're ex- they're blowing company resources. This is ridiculous.

332
00:54:56.764 --> 00:55:00.944
You and I, Colin, I read this and I'm like, yeah, that makes total sense.

333
00:55:01.004 --> 00:55:17.564
Right, 'cause we've, we've covered when we, we used to do Bitcoin season two, we covered, uh, Jameson Lopp had research on how many wrench attacks happened in 2025.And a wrench attack, for those who don't know, it's like, you know, the joke is, you know, yeah, you hold- not your keys, not your coins, but somebody can get your keys out of you with a $5 wrench.

334
00:55:17.764 --> 00:55:26.844
Basically kidnapping, you know- [laughs]... extortion, uh, torture, things like that. And just to underscore this point, Mara...

335
00:55:26.904 --> 00:55:40.944
Now, this is irrespective of the fact that they probably have multisig or it's custodied somewhere or whatever. Attackers don't necessarily know that. Mara holds 38,000 Bitcoin. 38,000.

336
00:55:40.984 --> 00:55:52.284
I think they might hold 35,000 Bitcoin. But anyway, it's- Is it, is it, is it 35 now? 2.5... I, I think so. Um, okay. But it's, it's, it's- It's a lot. It's over 30,000- It's $3 billion worth of Bitcoin. Yeah.

337
00:55:52.384 --> 00:56:05.064
They, they hold a, they, they hold, you know, two, roughly $2 billion worth of Bitcoin. So obviously they're gonna do this. And, and, and just to point out what Charlie was saying there too, this isn't uncommon.

338
00:56:05.184 --> 00:56:19.544
You know, we talked about Brian Armstrong getting 7.6 million for protective services. The Winklevoss twins did the same. In 2025, they spent $5 million, so about 2.5 million per twin.

339
00:56:20.384 --> 00:56:30.194
Now in a 2026 agreement, they're spending 400K a month, although there's a waiver there where, uh, this, if the services exceed one million, then the fees get waived for the first year.

340
00:56:30.294 --> 00:56:42.924
So I imagine this is a multi-year contract. But I'm sure that if we went through the public filings of Strategy, of Strive, of a lot of these other Bitcoin miners that still have Bitcoin,

341
00:56:43.924 --> 00:56:47.224
you'd probably see something similar. They may not break it out,

342
00:56:48.184 --> 00:56:57.084
uh, the, it may be couched in other language or lumped into something else like general corporate expenses, but I can assure you that it's there for a lot of these companies.

343
00:56:57.144 --> 00:57:09.484
And if it's not, honestly, me as a shareholder, I would like want it to be there because of the exposure that these people have. W- because you have normal individuals getting kidnapped.

344
00:57:09.524 --> 00:57:20.134
You have it now in the United States. You've had multiple s- scenarios on the West Coast of people- I mean, wasn't- Doordash being compromised. Look at the Ledger CEO. I was about to say. Former Ledger C- Yeah.

345
00:57:20.134 --> 00:57:33.584
Yeah, former Ledger CEO in France was kidnapped, finger cut off. So if you consider about like potential- Uh- Even if you just consider this as a potential disruption and bad like press for a company,

346
00:57:34.724 --> 00:57:43.524
what's a, a few million dollars when you have a market cap of several billion? And like, uh- Right. That- You know, it, it- That's a great, that's a great point.

347
00:57:43.584 --> 00:57:52.064
I mean, these numbers are big to normal people, but this is a drop in the bucket for these companies. Uh, I also have to say, I was laughing earlier- Yeah... not 'cause I think it's funny that he got his finger cut off.

348
00:57:52.104 --> 00:58:00.644
I think that's horrendous, but I was just thinking of The Big Lebowski, imagining the Ledger shareholders being like, "How do we know that's, how do we know that's h- his finger?" Yeah. [laughs] "You want a finger?

349
00:58:01.044 --> 00:58:13.884
I can get you a finger. Get you a finger later today." Yeah. Um, o- one last thing to note on this, just to remind everyone that the Financial Time is a gossip rag. It's not a serious financial publication.

350
00:58:13.964 --> 00:58:26.624
At bottom of this article, the journalist editorializes. She says here, quote, "We had further questions that the spokesperson declined to answer, chiefly what kind of vehicles were being armored.

351
00:58:27.204 --> 00:58:35.184
If they're just regular cars, wouldn't it have been better to just go whole hog? We say linking to, uh, commercially available tanks for the second time in a month."

352
00:58:35.224 --> 00:58:42.314
That's just, I get that they're having fun here, but like that just completely, uh, infantilizes this whole discussion.

353
00:58:43.044 --> 00:58:51.144
And, and it's like obviously they think this is absurd on the face, so if it's absurd, then why don't you get a tank? Um, because you can't drive a tank down the fucking interstate, guys.

354
00:58:52.024 --> 00:59:01.824
[laughs] Uh, if they are helicopters, who exactly has been threatening Mara? They're not helicopters. They're cars. They say vehicle. If you think this is a helicopter- But maybe they should get helicopters.

355
00:59:01.944 --> 00:59:08.873
I don't know. I don't know what kind of, I don't know what the, I don't know the threat profile, you know? Threat scenario. Yeah, apparently they think... Are, are they threatened by paramilitaries?

356
00:59:09.344 --> 00:59:16.854
And this is my favorite one- Probably... if it's boats, then why not just have boats? I- O- okay. Whoa.

357
00:59:16.904 --> 00:59:28.264
Like, yeah, let me just take my, you know, my, my freaking, uh, speedboat out to the data center in the middle of Texas. We'll see how that goes. Secondly, how can this be determined as a one-time expense?

358
00:59:28.723 --> 00:59:38.704
[laughs] I also, this one is, it sends me. It's like what if Theo and Khan fancy a newer model of redacted vehicle? What if, God forbid, redacted vehicle's newfangled defense capabilities were tested to destruction?

359
00:59:39.064 --> 00:59:52.064
Would replacing them then seem extravagant? If they do it again, they're gonna disclose it again, and then you'll write another hit piece about it. Yeah. Th- this kind of weird, you know, journalistic, you know...

360
00:59:52.284 --> 01:00:03.404
Like, like they, they act like they're dancing on a grave here, but really it just comes off as like the most unserious commentary about something that actually, as Mara points out, like the board approved this.

361
01:00:03.544 --> 01:00:04.784
They don't think it's crazy.

362
01:00:05.124 --> 01:00:16.944
And as you said, Charlie, as a shareholder of these companies, you might actually wanna see this because then you know that they're, the, the Bitcoin that a lot of these companies are sitting on is, is, is n- going to be hopefully not subject to any sort of extortion.

363
01:00:17.584 --> 01:00:17.924
So. Yeah.

364
01:00:18.544 --> 01:00:38.404
I mean, look, if you wanna know the, if you wanna get proper framing for this, the, the article, uh, shows a- as the, like header image a, one of the Mad Max, um, spiky armored cars, just to appropriately, uh, you know, bias you, the reader, towards their point of view.

365
01:00:39.424 --> 01:00:50.304
Also- Yeah... uh, Lewis Al- Ashworth is a man. It's, it's not a woman, so I apologize for misgendering the author. [laughs] But still, do a little better, Lewis. Come on.

366
01:00:51.564 --> 01:01:06.404
Yeah, so, uh, shout out to, uh, the Financial Times. You can find their, uh, you can find their, uh, writing, um, also available without the paywall, uh, on- On archive.is. [laughs] Archive.is.

367
01:01:06.984 --> 01:01:24.264
Okay, that, uh, that is our, uh, str- that's our stream for the day. You can catch us here on Friday again. We've got some great, uh, guests next Friday. Uh, we're talking both to Luxor, uh, and Udi Wertheimer- Ooh...

368
01:01:24.444 --> 01:01:34.423
if you can pronounce his last really bad. So- So get the tomatoes ready, everybody. We're gonna have Udi on Get the tomatoes. But yeah, um, all right. Well, thank you so much for, uh, watching Block Space Live.

369
01:01:34.524 --> 01:01:48.584
Make sure to like and subscribe. If you're on Coindesk, once again, immediately go to the Block Space, uh, channel on whatever platform you're listening to and sub to that one to remain, uh, to have continuity.

370
01:01:48.624 --> 01:02:01.814
Otherwise, see you all on Friday. [outro music]