WEBVTT

1
00:00:00.060 --> 00:00:03.840
The biggest news this week is Block laying off forty percent of its workforce.

2
00:00:03.940 --> 00:00:14.200
We'll get the, uh, tweet up here from Jack Dorsey, and he basically goes into that this is a preemptive move where they're not hurting, they're, they're comfortable, they're making money.

3
00:00:14.220 --> 00:00:25.920
But he's just saying, "Look, now it takes maybe one person to do the job of two or three people formerly because of these AI tools." Block's stock did exactly what you would expect in this scenario. It popped, ripped.

4
00:00:26.280 --> 00:00:32.509
It, it ripped like twenty-plus percent after market hours. [chuckles] Obviously, from a shareholder perspective, like, a good decision.

5
00:00:32.740 --> 00:00:44.540
Well, depending on where you fall on the, the white pill versus black pill for AI, this is either important for mass layoffs to come or something that is just part of the, you know, natural technological progression.

6
00:00:47.500 --> 00:01:00.640
Holy smokes, AI is coming for your job. It already took forty percent of Block's workforce. Yes, this is the Blockspace live show. We do this every Friday at noon Eastern.

7
00:01:00.650 --> 00:01:10.770
Make sure if you're on YouTube, hit that bell so you get a notification. If you're on Twitter, just turn on notifications for the Blockspace account. This show is brought to you by CleanSpark, so thank you very much.

8
00:01:10.960 --> 00:01:20.700
Tuning in, we have a banger show. There's actually too much news, so we had to pare some stuff down last minute. We got Jack Dorsey, Block laying off four thousand, that's forty percent of their workforce.

9
00:01:20.780 --> 00:01:37.240
We've got Jane Street conspiracies. We've got our Magic Eden story, which went viral last night, and Sailor's the most shorted stock. [upbeat music] Hey, Charlie here. Guess what?

10
00:01:37.480 --> 00:01:44.960
We just announced our next Bitcoin technical conference, OpNext. That's right, y'all. OpNext is back for twenty twenty-six.

11
00:01:45.000 --> 00:01:55.270
We're running it back after a successful event at Strategies HQ in Tysons, Virginia, last year, and this year we are bringing it to the Big Apple at the iconic Times Center in Midtown Manhattan.

12
00:01:55.580 --> 00:02:07.280
We're hosting the big names and projects that you recognize, like Robin Linus of BitVM, Nick Jonas of Blockstream, Antoine Ponce of Chaincode Labs, and Calay of BitChat will also be present.

13
00:02:07.360 --> 00:02:19.420
And this isn't just for the devs. We have institutions talking with the developers. That's what OpNext is all about. We have Robert Michnik, head of digital assets for BlackRock, in the building.

14
00:02:19.460 --> 00:02:29.150
We've got folks from mining pools, investor funds, Bitcoin startups, and other groups. With a ticket, of course, you'll get access to all the high signal programming and networking you could want.

15
00:02:29.180 --> 00:02:40.400
You'll also get coffee, catered lunch, and access to the after-party at Pub Key. If you wanna go VIP, you'll also get access to the speaker dinner following the event and an investor brunch on Friday.

16
00:02:40.640 --> 00:02:52.040
Tickets are capped at three hundred, and early bird tickets are already sold out. If you wanna save yourself a spot, go to opnext.dev. That is O-P-N-E-X-T dot D-E-V.

17
00:02:52.720 --> 00:03:01.120
Use code PODCAST to save twenty percent off a GA ticket to the event. Ticket prices go up every few weeks, so don't wait, y'all. Lock in that ticket today.

18
00:03:01.360 --> 00:03:13.580
We'll see you April sixteenth at the Time Center in New York City. I'm gonna bring up Colin. Colin, red hat for a red week in the market.

19
00:03:13.660 --> 00:03:19.380
I'm entering my, I'm entering my Jack Dorsey era with my drug dealer hoodie and my beanie.

20
00:03:19.780 --> 00:03:31.910
[chuckles] I, I had to, I had to christen this episode with something 'cause like you said, Charlie, there's a lot in the news this week. We're back in scoop city with a- Yeah... fat scoop from you on Magic Eden. Oh.

21
00:03:31.910 --> 00:03:41.380
I'm really excited about that one. All right. Before we get to the news, let's do our difficulty hash price and hash rate update from Hashrate Index.

22
00:03:41.500 --> 00:03:53.700
This is Luxor's premier Bitcoin mining dashboard, and it is looking freaking ugly, Charlie. [laughs] It's just- Oh, it's so bad. It's just languishing down there. Yeah.

23
00:03:53.800 --> 00:04:02.320
So hash price is sitting at a very depressing twenty-eight dollars and seventy-one cents per peta hash per day.

24
00:04:02.340 --> 00:04:11.240
During Bitcoin's drawdown at the beginning of February, and actually just recently in the last week, it set an all-time low.

25
00:04:11.300 --> 00:04:20.550
Yeah, looking at it right here, on February twenty-fourth, hash price hit an all-time low, b- below twenty-eight dollars per peta hash per day as, uh, Bitcoin's difficulty increased.

26
00:04:21.120 --> 00:04:30.550
That is setting a new all-time low from the all-time low that it hit back in February when Bitcoin tanked to about sixty K. Right now, difficulty is...

27
00:04:31.120 --> 00:04:43.600
or miners are reeling from a fourteen point seven three percent difficulty adjustment. This followed a string of downward difficulty adjustments that were spurred forth by, you know, some, a number of factors, right?

28
00:04:43.680 --> 00:04:54.980
There's some terrible weather in the US. Uh, hash price has obviously just been suffering, so fewer miners are actually able to stay profitable at the current levels given Bitcoin's price decline.

29
00:04:55.060 --> 00:05:03.760
And, you know, there's also some headwinds for hash rate with regards to... or sorry, tailwind. Yeah, headwinds for hash rate with regards to the AI pivots that we're seeing.

30
00:05:03.790 --> 00:05:09.890
A lot of these big public miners are starting to actually liquidate some of their fleets so that they can make way for AI workloads.

31
00:05:10.440 --> 00:05:25.300
Hash rate back above one zeta hash, one thousand exahashes after the winter storm fern impact. I mean, look, look at this canyon right here. This is an absolutely massive drop-off.

32
00:05:25.760 --> 00:05:31.340
We can see hash rates at about one thousand five, uh, one thousand fifty exahashes on January twenty-third.

33
00:05:31.780 --> 00:05:44.260
Winter storm hits, uh, and it looks like some miners were kind of perspectively, uh, curtailing, you know, on the twenty-fourth and twenty-fifth, and then it just craters as the storm makes impact.

34
00:05:44.300 --> 00:06:01.740
It, it really, uh, hit, uh, most of the states that would've had hash rate, you know, in the southeast and Midwest, I believe on like the twenty-sixth, twenty-seventh of January, and it reaches a nadir on January thirty-first at eight hundred fifty-three exahashes, which is the lowest we've seen it since peak curtailment in the summer of last year.

35
00:06:01.780 --> 00:06:18.736
Now, it's recovered somewhatBut we- we're still nowhere near-- I shouldn't say nowhere near, but we're, you know, almost, uh, roughly, uh, 10% off, almost 10% off from the all-time high that was set in October of 1,157 exahashes, uh, when Bitcoin was roughly at its all-time high, too.

36
00:06:18.936 --> 00:06:32.036
So number of factors coalescing against Bitcoin's hash rate in terms of, you know, inclement weather this winter, the AI re- uh, the AI expansions, and also Bitcoin price just absolutely suffering.

37
00:06:32.516 --> 00:06:43.296
Quite frankly, I- I- I have a hard time believing that we will see it grow very much at all this year. This is something that I talked about with Rafa Zguri, the CEO of Electron.

38
00:06:43.336 --> 00:06:47.676
They manage, uh, a large portion of Tether's Bitcoin mining fleet.

39
00:06:48.236 --> 00:07:09.276
We spoke about this with him this week for a pod that we'll be r- releasing next week, and a- all that being said, it's pretty amazing to see how much the dynamics of Bitcoin's hash rate have shifted over the last year as these AI pivots take off, as hash price continues to suffer under, uh, Bitcoin's bear market and, and what little hash rate is being added currently.

40
00:07:09.376 --> 00:07:22.176
Yeah. And I mean, maybe it was very obvious. It's there in the name, the difficulty adjustment. It only gets more difficult. I should have known years ago that this is a business that just only keeps ramping up.

41
00:07:22.256 --> 00:07:31.436
I actually kinda wanna share something, 'cause I think I have a, uh, a page that doesn't get enough love on the Hashrate Index website here, which is the global hash rate heat map.

42
00:07:31.476 --> 00:07:42.416
Like, you don't really-- There's no one else that has this. Like, the Cambridge stu- the Cambridge research has to, like, try- tries to do something like this, but this is, like, the best global view of hash rate.

43
00:07:43.006 --> 00:07:49.936
And of course, only Luxor and Hashrate can do this really, but, uh, you know, it shows you, like, where the hash rate in the world is located.

44
00:07:49.976 --> 00:08:00.456
Again, this is an estimation based upon Hashrate Index's, like, where they see rigs being sold and their intel, but it's the best we can do so far. And we see the United States still in the clear lead,

45
00:08:01.656 --> 00:08:13.635
over a third of the hash rate that we believe, uh, exists in the world. The really interesting story is Russia, with 175 exahash, 16.4% of the network according to this. This is Q1 of 2026.

46
00:08:14.135 --> 00:08:20.516
If we go back to '25, Russia had 15% of the hash rate with 125 exahash.

47
00:08:20.636 --> 00:08:36.356
Now, it's like Russia's, you know, famously opaque, but this, you know, it's interesting to see s- like, the actual, like, story we, like, we're watching happen, which is AI is eating into Bitcoin profits.

48
00:08:36.796 --> 00:08:49.936
Domestically in the United States, the hash rate producers, the miners, are pivoting to AI. Where's the hash rate going? Um, this is a big open question. There are suspicions it's going to Russia.

49
00:08:50.036 --> 00:09:01.716
So I'd be looking at, like, this map and other stuff Luxor produces on their Hashrate Index to kinda monitor this trend, 'cause I think this has big geopolitical con- you know, implications and also plays into the AI narrative.

50
00:09:02.436 --> 00:09:07.356
Yeah, I would agree with that, that we're gonna see an incredible hash rate restructuring over the next few years.

51
00:09:07.436 --> 00:09:24.496
And as Tom Messero of Cathedral pointed out on a pod we did with him recently, I actually did it in Nashville for the Nashville Energy Mining Summit right before S- uh, Fern hit, that he feels like the current AI shift has echoes of the China mining ban in terms of how impactful it will be for Bitcoin's hash rate.

52
00:09:25.056 --> 00:09:34.036
But we'll leave that there, and we'll go ahead and hop on over to news. The biggest news this week, undoubtedly, as you led with, is Block laying off 40% of its workforce.

53
00:09:34.076 --> 00:09:50.876
So Block, before these layoffs, had 10,000 employees. We'll get the, uh, tweet up here from Jack Dorsey where-- In, in which, i- in which he does lowercase for, uh, for firing people. Not, not a single capital letter.

54
00:09:50.936 --> 00:09:56.316
So I don't know if that's, if it's just broken- I have to say, I, look, I like Jack. I gotta say this though.

55
00:09:56.936 --> 00:10:10.676
This reeks of, like, an Iowa Writer's Workshop MFA candidate trying to be differentiated in the most banal way possible with their literature. I mean, I, I just-- For, for text communications, do it.

56
00:10:10.716 --> 00:10:21.895
For official comms, come on, man. You're leading a multi-billion dollar company. Button it up a little bit. You're about to- Hashtag, hashtag... fire nearly half your workforce. Yeah. Four, four pound signs.

57
00:10:22.036 --> 00:10:32.776
That's not a markdown symbol. That's not actually anything in markdown. I think it might be a, a line break, but whatever. He says- So- "We're making, we're making blocks smaller today," and then he explains.

58
00:10:32.836 --> 00:10:39.476
And he basically goes into that this is a preemptive move where they're not hurting. They're, they're comfortable. They're making money.

59
00:10:39.956 --> 00:10:50.416
But he's just saying, "Look, now it takes maybe one person to do the job of two or three people formerly because of these AI tools." And so what we're getting here is the Citreni s- thesis.

60
00:10:50.476 --> 00:10:56.836
I don't know if any of y'all read that really-- that article that apparently tanked the stock market earlier this week. But Citreni Research- Yeah...

61
00:10:56.856 --> 00:11:15.536
put out a kind of perspective fictional account of the stock market and the labor market two years out, and he talks about specifically this, about how these companies are going to start laying off people, as most, a lot of people have prognosticated, because AI is going to be able to fill shoes of workers who are superfluous at this point.

62
00:11:15.636 --> 00:11:20.796
And for Block's case specifically, they're going from 10,000 people to just under 6,000.

63
00:11:21.256 --> 00:11:30.156
Uh, it comes with a very lush or very comfortable severance package, 20 weeks plus one week of tenure for, for each week- Yeah, that's a lot. I mean- Yeah.

64
00:11:30.476 --> 00:11:40.736
Also equity vesting through the end of May for the employees laid off, six months of healthcare. They keep all their corporate devices, and $5,000 of basically a have fun bonus.

65
00:11:41.016 --> 00:11:51.636
[chuckles] So, and, and, you know, incredibly, as you might expect, I mean, I don't think that this was that, uh, strange to anyone. I'm gonna share my screen now, Charlie. Yeah.

66
00:11:52.396 --> 00:12:04.726
Block's stock did exactly what you would expect in this scenario. It, it popped- Ripped. It, it ripped, like, 20-plus percent after market [chuckles] hours, uh, when, when Jack Dorsey announced this, and it makes sense.

67
00:12:04.836 --> 00:12:16.174
You just cut your expenses by, I think I saw someone run the numbers where it was like-Somewhere in the hundreds of millions, if assuming an average salary f- of, of, you know, roughly six figures maybe for these, for these employees.

68
00:12:16.804 --> 00:12:28.083
Obviously, from a shareholder perspective, like, a good decision. Of course, X did not spare any, any vitriol, right? Or invective against Jack Dorsey. I saw a lot of takes about this.

69
00:12:28.144 --> 00:12:38.804
One I wanna pull up here just to debunk it really quickly. Of course, Zero Hedge comes out and says... It's, it's a really- It's a zinger, really. It's kinda funny. "Spend $68 million on a party in Q3.

70
00:12:39.004 --> 00:12:49.944
200 days later, fire 40% of your workforce. Blame AI." Okay, so that's $68 million for a party. That's not true. Wait, is that true? I, I didn't dig too into... No. I saw this as like some S- like a G&A expense.

71
00:12:50.224 --> 00:13:03.794
Yeah, exactly. It was. Yeah. And so this comes from their 10-Q for Q3, where it says, "General and administrative expenses increased by 68.1 million, or 14%, uh, for the three months ended September 30, 2025."

72
00:13:04.444 --> 00:13:16.444
And then the ellipsis that then follows this in the Zero Hedge tweet goes into, "These expenses increased re- These, these expense increases resulted from an in-person company event held in the third quarter of 2025."

73
00:13:16.944 --> 00:13:24.144
Oh, so that must have been 68.1 million. Uh, no, as well as restructuring costs, including severance and other related expenses.

74
00:13:24.224 --> 00:13:34.944
So no doubt, whatever that company event thing was, was probably an on-site that cost them a lot of money. No doubt millions of dollars probably, right? I mean, that's not uncommon- 10,000 people. It's 10,000 people.

75
00:13:35.004 --> 00:13:42.504
Tech companies. For one of the bigger tech companies, yeah. Yeah, but also, like, do your research, guys. $68 million in one day is crazy. Like, that clearly didn't happen.

76
00:13:42.564 --> 00:13:52.744
That, that's just a clickbaity tweet from Zero Hedge. It's funny, you know, unless you got fired from Block, then maybe it's not funny. But it, it, it didn't actually happen. And just a few more notes on this.

77
00:13:52.804 --> 00:13:58.554
You know, we, we've been talking to some of our sources within Block, and it does seem like this has touched basically every single vertical.

78
00:13:59.184 --> 00:14:09.384
Um, we're not sure about certain Bitcoin product lines, like how this affects BitKey. Um, I, I, I think that you said, Charlie, that it seems like Spiral is mostly unaffected. Yeah.

79
00:14:09.704 --> 00:14:21.284
W- this is actually alpha here, so this is new news that I don't think anyone else reported on. We're saying first here live on this show, mark it. Under the Block umbrella, there's multiple entities.

80
00:14:21.424 --> 00:14:25.584
Some are, like, Jack Dorsey passion projects, some are in, in between.

81
00:14:25.644 --> 00:14:41.624
Um, I would say Spiral, which is, like, the Bitcoin developer Seal Team Six that Jack Dorsey funds, you know, a dozen or so, you know, cracked devs, they are untouched. So nobody lost from Spiral, uh, as per our source.

82
00:14:42.264 --> 00:14:54.584
There's other umbrellas. There's BitKey, which is the wallet, uh, the, the hardware wallet that, uh, is under Block. I don't think I've heard anything from them. I don't know if you have, Colin, either.

83
00:14:54.644 --> 00:14:57.024
We're s- if you were- No. If you got info on BitKey, DM me.

84
00:14:57.704 --> 00:15:07.024
Um, then we have Proto, the hardware, like, the mining, uh, side of Block, the actual physical ASIC that they've been producing, the modular thing, swap in, swap out.

85
00:15:07.184 --> 00:15:17.064
I, I think they didn't have as many layoffs proportionally as the other, like, Cash App and Square, uh, parts of Block, but they did have some layoffs.

86
00:15:17.244 --> 00:15:28.284
But, uh, Proto's still going with maybe a small head reduction, headcount reduction. Unknown about BitKey. Spiral untouched. It's mainly, like-- 'cause Jack tweeted about it more this morning.

87
00:15:28.804 --> 00:15:39.334
He said that they grew, quote, like, grew Cash App and Square as, like, different structured businesses, when in reality they should be kind of the same corporate structure.

88
00:15:39.524 --> 00:15:47.524
So that makes sense why the majority of the headcount was lost in those two areas. Yeah, and just a note on that, it does kinda seem like a confusing corporate structure.

89
00:15:47.644 --> 00:15:52.124
You have, you know, Block, which is the parent company, and you have Square, point-of-sale terminal.

90
00:15:53.024 --> 00:16:03.224
You have all of this Bitcoin stuff that's kind of in its own silo, and then Spiral is a part of that Bitcoin stuff, but also kind of separate. Yeah. You know, it, it's, it's kind of dizzying. But- Yeah...

91
00:16:03.264 --> 00:16:13.684
anyway, you know, depending on where you fall on the AI doomer or gloomer angle or bloomer, I'm gonna, I'm gonna- I know that. [chuckles] I'm gonna, I'm gonna coin that, bloomer.

92
00:16:13.764 --> 00:16:28.264
[chuckles] Depending on where you fall on the, the white pill versus black pill for AI, this is either a portent for mass layoffs to come or something that is just part of the, you know, natural technological progression and, you know, adapt or die.

93
00:16:28.384 --> 00:16:42.534
But anyway, that- Okay... that fir- first big, huge AI layoff I think that we've seen, at least explicitly naming it in, in, in tech. But- Yep... now to the big scoop this week from- Yeah, this was a banger...

94
00:16:42.564 --> 00:16:49.764
from CB Spears. It really was a banger, and damn, how the mighty have fallen. Charlie, what's going on with Magic Eden?

95
00:16:50.704 --> 00:16:58.924
So if you're not really in, like, the Bitcoin asset, uh, trading world, you probably don't know about this. This is very big in, like, broader crypto, though.

96
00:16:58.964 --> 00:17:03.184
So we've, we're kind of a, we come at this from a Bitcoin lens here at, uh, at Blockspace, but, like,

97
00:17:04.384 --> 00:17:17.994
Magic Eden is a mainstay and a big name across the, the world of trading NFTs, digital assets, and, uh, one of the leading marketplaces in the NFT and, like, crypto trading space.

98
00:17:18.684 --> 00:17:35.824
Um, and they are shuttering their, or allegedly, according to our sources, imminently going to announce they, uh, are going to shutter their Bitcoin, uh, Ordinals trading and, uh, EVM-related marketplaces.

99
00:17:35.844 --> 00:17:46.444
For those of you who aren't in the know, EVM stands for Ethereum Virtual Machine, any kind of, like, system, uh, that uses the same kind of VM language or basically plugs into Ethereum.

100
00:17:46.484 --> 00:17:53.244
So this would be Arbitrum, Base, uh, I think Berachain, but basically all those, like, other chains of which there's a ton of volume on.

101
00:17:53.294 --> 00:18:06.844
So, uh, this is, uh, according to our sources, gonna happen over the next month, and, uh, this is big news everywhere because this is, like, a, like... You know, a year and a half ago, Magic Eden was doing--

102
00:18:07.804 --> 00:18:25.232
They took the number one spot in terms of, like, overall trading, uh, across, like, the, the platforms of their type.Uh, beating out OpenSea and Blur and did something like 700 million-- $734 million ah, dollars in trading volume in March 2024 alone.

103
00:18:25.692 --> 00:18:36.692
So this is significant, and, uh, Magic Eden had a very notable like Bitcoin pivot in early 2023 at the rise of Ordinals, and they absolutely dominated that market.

104
00:18:36.792 --> 00:18:43.822
So, you know, this does relate to what happens in Bitcoin. If you think, oh, it's the Bitcoin trading platform, that's fine, you can think that.

105
00:18:43.852 --> 00:18:51.812
But like, if you look at like the transaction landscape of Bitcoin, a lot of this is related to stuff which is then bought and sold with Magic Eden.

106
00:18:51.872 --> 00:19:02.072
So it's very notable that they allegedly, according to our sources, are sunsetting this. And, and what are they doing?

107
00:19:02.092 --> 00:19:11.172
They're probably-- It sounds like they're going prediction markets because that's what everybody's doing. You either... It's, it's kind of like, you know, if you're in Bitcoin long enough, you start a podcast.

108
00:19:11.592 --> 00:19:19.312
If you're in crypto long enough, eventually you go into prediction markets or stable coins. That's kind of what- Let, let's, let's be real about our definitions here.

109
00:19:19.332 --> 00:19:32.762
If, if you're in the casino world, if you're in the gambling industry, you move from flipping JPEGs to prediction markets. No, but in all seriousness, just a few other things to highlight from this.

110
00:19:33.732 --> 00:19:49.512
Looks like this is going to be a multi-tiered shutdown in the sense that according to Charlie's sources, they're going to discontinue their cross-chain wallet, uh, shortly after they announce this, which should be within the next week.

111
00:19:49.572 --> 00:19:57.912
And then after that, uh, sometime in mid-March, they are going to completely remo-- or in April, they are going to, uh, completely remove that.

112
00:19:57.972 --> 00:20:04.192
It's going to be export only in mid-March, and then they are going to just get rid of it entirely sometime in April.

113
00:20:04.312 --> 00:20:12.052
Um, the marketplace, uh, might be shutting down a-as soon as the first week of March or roughly around there. And- Correct.

114
00:20:12.112 --> 00:20:22.521
It-to what you were saying, Charlie, looks like apparently they're going to still keep their Solana-related, you know, class- Yeah. So- Which is where they- Magic Eden- How they started, right? Like they were on Solana.

115
00:20:22.521 --> 00:20:36.972
Yeah, Magic Eden started on Solana, I think it was, if I remember correctly, it was like mid '21, and they quickly became the number one trading platform for NFTs on Solana, which obviously was the darling child of the last couple years in like non-Bitcoin blockchains.

116
00:20:37.152 --> 00:20:45.152
And yeah, I, I missed-- I didn't talk about the wal- the wallet's kind of interesting because they only launched it like two years ago, maybe like a year and a half ago, I forget.

117
00:20:45.752 --> 00:20:52.732
And, uh, so just for like the average person, like this doesn't mean you can't access your funds necessarily.

118
00:20:52.792 --> 00:21:02.852
If you have your private keys, theoretically, you can go, you, like click-- you know, put those private keys into a different wallet and, and sign for those assets.

119
00:21:03.292 --> 00:21:15.432
But like the actual like software, that is the Magic Eden wallet, we expect they will announce that it will be deprecated, so they won't service that anymore. This often creates a lot of like challenge, like,

120
00:21:16.432 --> 00:21:27.852
you know, using and claiming your assets, 'cause it-with these EVM chains, with these smarter blockchains, it's never that straightforward. Like it's, there's always like way more goofy like things that are going on.

121
00:21:27.932 --> 00:21:42.832
So, but, uh, that's kinda crazy 'cause they really like pushed their wallet. Um, it was like a flagship product under two years ago that they launched, and so sunsetting it is pretty crazy to me. Oh, also,

122
00:21:44.012 --> 00:21:57.672
they-- this-- I... You know, they raised, what is it, hundred and fifty-seven million dollars in venture funding over the past three years. Wait, hang on. This is correct. Yeah, this is correct.

123
00:21:58.172 --> 00:22:03.792
So, uh, this is, this is correct. This is a source-- Yeah, sources from TechCrunch and other places.

124
00:22:03.852 --> 00:22:16.832
So they did a couple rounds, one of the, the culmination of which was a hundred and thirty million Series B in June '22 that valued it at one point six billion, including Paradigm, Sequoia, Greylock, Electra Capital- Okay, and what's-- Okay, hang on.

125
00:22:16.892 --> 00:22:27.692
What's really insane about that is that is two years before they had the seven hundred million dollar trading volume during the Ordinal splurge. I mean- Yeah.

126
00:22:28.672 --> 00:22:36.632
So imagine if they did that fundraising now, it probably would've been even more. And you know what? I'm-- I don't-- You, you know me, I didn't think any of this was sustainable anyway.

127
00:22:36.652 --> 00:22:43.872
But raising that much money, if you have seven hundred and like almost seven hundred and fifty million in trading volume, I could see that, right?

128
00:22:44.052 --> 00:22:53.092
I mean, like, I could see why an investor would be like, "Holy crap, you guys are moving volume. We're gonna give you money." But before you even get to that point, right?

129
00:22:53.212 --> 00:23:08.172
When, when-- I, I get the Solana NFTs were kind of riding the like, you know, the, the 2021 Ethereum NFT wave as well, and that whole ecosystem had its own huge, you know, uh, splurge moment and, and, and, and meteoric rise.

130
00:23:08.292 --> 00:23:18.462
But that, that's insane to me. O-one more-- Unless you have anything else to add, I just wanna- No, go, go on. Yeah...point out one more thing before we get Mike on to talk about GPUs.

131
00:23:18.832 --> 00:23:31.272
After we did the-- or we posted this article, the- [laughs]...number of markets on Magic Eden, the bet. Oh. I mean, all of the major Ordinals collections, except for a few

132
00:23:32.212 --> 00:23:37.992
smaller ones, and I, I think just people flipping things, are red over the last twenty-four hours.

133
00:23:38.572 --> 00:23:47.831
And just, you know, I wanted to point that out because in this case, some more market-moving news from Blockspace, but in this case, for a market that no one really cares about.

134
00:23:47.912 --> 00:23:56.211
[laughs] Yeah, and, you know, I'm in, I'm in a lot of the Degen chats, and, uh, they're like, "Why didn't you tell us?" I'm like, "Journalistic integrity, guys."

135
00:23:56.532 --> 00:24:08.512
I'll also say, you know, for those of you who are in like, in the trenches with me, um, this had been a rumor for more than a week in like the, the know-- the who knows who of Discords.

136
00:24:08.552 --> 00:24:12.352
Like everyone-- I thought it was kind of a meme. I wondered if it was actually verified.

137
00:24:12.392 --> 00:24:25.632
So this is like kind of a memed rumor in, uh, that, uh, we now believe is pretty credible.And so it could be, you know, maybe they're announcing it right now. We'll see. So that is that story.

138
00:24:25.692 --> 00:24:41.032
That's a-- uh, that was just for anyone who's listening, like, uh, that was, uh, we are seeing record numbers on o-our website impressions and, uh, some of our, uh, social media. So that story really took off.

139
00:24:41.332 --> 00:24:42.912
Um, you never know how a story's gonna do.

140
00:24:42.992 --> 00:24:52.532
Sometimes you wonder if you're just screaming into a void, you put it out, um, but this one- Well, and it's funny, if you had told me this one would perform that well, I would have said, "Yeah, it's probably gonna do okay."

141
00:24:53.512 --> 00:25:03.912
It's bi-- I mean, it's big news specifically for that s-corner of the market, but I think it goes- Yeah... to show you just how plugged in and devoted the, the Ordinals heads are.

142
00:25:03.972 --> 00:25:14.352
You know, people who got involved in this market are still fiendishly devoted if, if, if they've stuck around this long, so. Yeah. Yeah, but- People are like, "The, the market's totally dead."

143
00:25:14.392 --> 00:25:25.352
Yeah, the market's dead, but the people are still there. Everyone's just zombies now. Yeah. A thousand yards, thousand JPEG stares. The market is waiting for a Lazarus moment, a miracle to bring it back to life.

144
00:25:25.472 --> 00:25:40.612
[gentle music] 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.

145
00:25:42.412 --> 00:25:58.392
This is our proof of work. We are setting the standard for what's next. Learn more about the intersection of energy and Bitcoin at cleanspark.com. [laughs] GPU news. Welcome, Mike, to the show. Morning, Colin.

146
00:25:58.452 --> 00:26:02.952
Morning, Charlie. How are y'all doing? Fantastic. Yeah, you're with Luxur.

147
00:26:02.992 --> 00:26:10.312
We love to have Luxur folks on regularly, and we're gonna switch from the news of the day, and we're gonna talk about, I would say, more substantial stuff.

148
00:26:10.532 --> 00:26:17.092
So AI, b-but we gotta anchor this in, like, the physical hardware side. I'll toss it to Colin. Yeah.

149
00:26:17.172 --> 00:26:25.112
If you're interested in the esoterica that are, are GPU market dynamics, hold onto your seats, y'all, 'cause that's what we're gonna be covering here.

150
00:26:25.272 --> 00:26:30.572
Mike, before we get into this, can you just give people a little bit of background on yourself and what you do at Luxur specifically?

151
00:26:31.472 --> 00:26:40.352
S-so, uh, here, I'm the, uh, principal CPU and GPU, uh, sales associate handling all things AI, HPC, and GPU-related.

152
00:26:40.752 --> 00:26:53.062
Uh, so that includes site development, hardware selection and acquisition, uh, monetization of that hardware, deployment, financing, all over the place, uh, for anything AI, GPU, cloud-related.

153
00:26:53.812 --> 00:27:03.972
Uh, before that, I was with Dell for about seven years as a solutions architect and sales executive. I've also been with, uh, Cogent. That's one of the world's largest ISPs.

154
00:27:04.352 --> 00:27:11.952
So to set the stage for this, a lot of the folks listening to this who are interested in mining will be somewhat familiar.

155
00:27:11.992 --> 00:27:17.712
You know, there are varying levels of familiarity, I think, but pretty familiar in a lot of respects with the ASIC market.

156
00:27:18.282 --> 00:27:28.552
And when I think about miners going and, and sourcing ASICs, I know I'm gonna be dumbing it down a little bit, but the i-- the considerations are pretty simple, right? It's like, what's your power cost?

157
00:27:29.092 --> 00:27:34.142
If it's higher, uh, usually it means you wanna go for a lower joule per terahash, more efficient machine.

158
00:27:34.712 --> 00:27:41.572
If it's lower, you might be more comfortable moving up the curve to a higher joule per terahash machine and saving on CapEx, right?

159
00:27:41.932 --> 00:27:51.952
And so the basic idea is just like you're kinda balancing your operating cost versus your capital expenditures for hash rate when you're expanding. You're looking at hash price to see what you can actually sustain.

160
00:27:51.972 --> 00:27:56.892
And then obviously you're thinking about depreciation and which model. You know, w-some models are better than others, yada yada.

161
00:27:57.952 --> 00:28:03.272
But the general idea is that you really only have to worry about one form of compute, SHA-256. You're just producing numbers.

162
00:28:03.892 --> 00:28:12.542
What layers of complexity does the GPU market add to these equations for bitcoin miners who are expanding into this product line that they didn't have to think about with ASICs?

163
00:28:13.772 --> 00:28:20.752
So the first thing to understand is that it's a different business model. ASICs, you plug them in, you boot them up, and it starts generating hash rate.

164
00:28:21.412 --> 00:28:35.092
Uh, with GPUs and anything cloud-related, you're really more of a, uh, like a landowner. You're essentially becoming a cloud operator, uh, cloud service provider, and then you have to find a way to bring that to market.

165
00:28:35.272 --> 00:28:38.052
So, uh, that's where monetization gets to be important.

166
00:28:38.312 --> 00:28:47.472
You know, the big thing that everybody sees are these pubcos that are signing these, you know, hundred million to billion dollar deals, and everybody wants to win the lottery.

167
00:28:48.232 --> 00:28:57.552
Uh, but the reality is, is that there's, uh, a large spectrum there, and, uh, there's a lot more complexity to the hardware. There's a lot more complexity to the sites.

168
00:28:57.842 --> 00:29:10.012
And all of those factor in to how you'd actually procure the GPUs and how much money you're actually gonna make off of it. So, uh, with, with ASICs, you, uh-- if you wanna upgrade, you unplug the rig.

169
00:29:10.112 --> 00:29:24.252
You just sell it often direct or through third-party brokers, um, 'cause, you know, ASICs have fewer variables. What's the secondary market look like for GPUs? I mean, is there much of a robust secondary market?

170
00:29:24.972 --> 00:29:37.212
That's, that's a huge question. Um, the tail end of GPUs is kind of, of unknown. You're still seeing A100 GPUs being at ninety to a hundred percent consumption rates from a lot of the NeoClouds.

171
00:29:37.992 --> 00:29:47.592
So, uh, the actual utilization rates on those and the depreciation isn't quite known yet. Uh, H100s and H200s are still very much in demand.

172
00:29:47.912 --> 00:29:56.752
H200 is in a interesting space, um, where it's kind of been end of sales life, it kind of hasn't.

173
00:29:56.872 --> 00:30:05.612
Uh, but you're still seeing these things trading, and you're seeing the values go up of some of these end-of-sales-life systems that have been on the market for five years.

174
00:30:05.752 --> 00:30:11.452
Uh, B200 and B300, due to some of the supply chain crunches, are in a lot of demand.

175
00:30:11.492 --> 00:30:22.260
You have to get allocation, and the prices keep going up.Uh, there's a huge RAM shortage right now that's affecting both used markets as well as new markets.

176
00:30:22.640 --> 00:30:30.840
So the tail on these, and because of the capital being invested in them is, isn't quite known, but they are still profitable, even the older generations.

177
00:30:31.160 --> 00:30:37.260
So is the consideration there basically whether you get ol-older GPUs, you just need to find the right form of compute for them?

178
00:30:37.840 --> 00:30:46.940
Put, put another way, a GPU might be older, and it might not be able to keep up with new demand or demand from new AI services or new HPC services.

179
00:30:46.960 --> 00:30:57.540
But if you can find a kind of older form of compute or a, uh, less intensive form of compute for that machine, you can basically keep it active as long as you have a customer for that. Is that the idea? Essentially.

180
00:30:57.620 --> 00:31:06.940
You know, it's, it's important to think about compute in AI as being a way that's gonna be relevant to that market. So there's training, right? LLM training that everybody's familiar with.

181
00:31:07.500 --> 00:31:16.750
Then there's inferencing, which is running those models that you've trained, and then there's agentic, which is basically using inferencing models to run other inferencing models.

182
00:31:17.460 --> 00:31:31.070
So, uh, not all of those workloads are the same. Um, you don't actually have to know a ton about the models that are being ran on these, but you do have to know sort of where they would be located.

183
00:31:31.120 --> 00:31:35.100
A lot of that can factor into your fiber connectivity between sites.

184
00:31:35.340 --> 00:31:49.020
You know, a real-world example would be if you're running an inferencing model that's being used on self-driving cars, there's a latency to send data back and forth between the car and the server, so that has to be very physically close, within a hundred miles.

185
00:31:49.040 --> 00:32:01.240
If you're doing training, you know, if you're, you're entering ChatGPT and, and querying it, thirty seconds to a minute isn't gonna change anything for you realistically in getting a result.

186
00:32:01.840 --> 00:32:05.360
But if you're driving a self-driving car, you, you need that down to the millisecond.

187
00:32:05.520 --> 00:32:16.160
So the site location can sort of determine the workload that you'd run on it and determine the hardware that you'd want to procure in order to run that.

188
00:32:16.460 --> 00:32:26.720
With, with Bitcoin mining, like, the miner just has to turn on and run efficiently. You have someone who's going to purchase your compute at a globally, at a, you know, price that's equal to all your competitors.

189
00:32:27.060 --> 00:32:40.820
But with GPU or just AI compute, it seems to me like your counterparty and, like, who you're selling to is equally as important to the deal. Am I correct? And maybe qualify or tease that out a bit. Hundred percent. Yeah.

190
00:32:40.920 --> 00:32:46.340
So there's a couple different ways that you can, you can sell that, and scale gets to be an issue here.

191
00:32:46.380 --> 00:33:03.520
So typically, you get higher demand and longer contracts for consuming your compute if you've got a higher number of, uh, GPUs. So let's say you get a, a contract for a hundred H100 GPUs for six months to a year.

192
00:33:03.560 --> 00:33:13.030
Well, you've sold essentially all of that compute, hundred percent dedicated to that one end user. You've got a contract for it, and you can sort of negotiate the terms there.

193
00:33:13.080 --> 00:33:19.180
Maybe you're getting eight dollars per GPU per hour. So that can be very profitable.

194
00:33:19.240 --> 00:33:26.960
Now, if you're selling it, you know, on spot on an open marketplace, you might only get, uh, a buck fifty to two bucks for that same GPU.

195
00:33:27.810 --> 00:33:37.950
Uh, so finding that end user or working with a monetization partner that can help you do that or pair you with a contract is extremely important.

196
00:33:38.190 --> 00:34:00.410
This has always been what's blown my mind about miners who are expanding into this is you already mentioned some of the networking complexities with making sure that sites are communicating properly to make sure that a cluster is operating or to make sure that the, the computations are doing what they're supposed to be doing, and but then y- y- we didn't even touch on then, but you just did, this idea that you actually have to go out and do business development when you're running one of these clusters.

197
00:34:00.930 --> 00:34:10.800
So on that note, what is-- what are your opinions on miners who are deciding to go the power shell approach, i.e., they're saying, "We're gonna get the power. We're gonna build the shell for a data center.

198
00:34:10.920 --> 00:34:22.410
You bring the compute, and you run it," versus a miner like Iron or Hive that's saying, "Actually, we're gonna run the clusters as well." Do you have a preference? Yeah. Yeah. That's a really good question. Y-

199
00:34:23.320 --> 00:34:36.180
One of the, the big OPEX concerns that, uh, people don't think about coming from the mining side is about network uptime. Um, and having a network admin can really affect your OPEX.

200
00:34:36.460 --> 00:34:50.500
Uh, setting up as just a colo can be really good, but you're not gonna pull down big contracts unless you're Tier Three certified as a location, and you can prove that you're, you're capable of keeping those places up and running.

201
00:34:50.760 --> 00:34:59.619
So, you know, an Iron, Bitdeer, some of these other large companies, Crusoe, when they made the pivot more towards AI, um, they're able to prove that.

202
00:35:00.200 --> 00:35:14.600
Now, with, uh, just spinning up the, the site, you know, yeah, you can probably find someone, uh, who will put their systems into your location, uh, and I've met people that have done that successfully.

203
00:35:14.980 --> 00:35:26.440
But where a lot of these companies are making the majority of your margin is running the actual hardware, selling that compute, and then, uh, getting at least a revenue share off of that rate.

204
00:35:26.600 --> 00:35:36.020
So it's higher risk, higher reward in some ways. And sorry, Charlie, before you jump in, just one last question with that. Is that one way to read Iron's pilots?

205
00:35:36.160 --> 00:35:45.380
'Cause they did, like, much smaller deployments before they started moving to these bigger, you know, b- before they started saying, "Okay, we're gonna outfit hundreds of megawatts for this compute," right?

206
00:35:46.220 --> 00:35:54.220
I- is that one way to look at it as like they're, they're basically running these, uh, these pilots to prove that they can have the Tier Three uptime? A hundred percent.

207
00:35:54.440 --> 00:36:02.720
And, and that's something that, uh, I work with customers every day around is helping them grow to the point where they can actually prove that.

208
00:36:02.860 --> 00:36:30.376
Um, it's very easy to go out and buy containers and to pay someone to dig some fiber trenches and, you know, it's, uh, it can be tough to find enough power for a gigawatt, but you might be able to find a few megawatts there, which-Is great, but proving that you can actually run these systems is what's gonna help you build your, your business's reputation to where you can start talking to some of the hyperscalers or getting some of these larger neo cloud contracts.

209
00:36:30.656 --> 00:36:43.716
You know, it pains me like deeply to hear you say it's fairly easy to buy a container and do that 'cause like as a former Bitcoin miner, oh man, that was some of the [laughs] most difficult stuff I've ever done.

210
00:36:43.976 --> 00:36:52.356
Um, but I think it was a skill issue on my part. Um, okay, so I was listening to a podcast, I think with the Microsoft CEO Satya Nadella.

211
00:36:52.396 --> 00:37:12.316
Anyway, he had a really interesting comment that stuck with me about like why Microsoft is pulling back investment on some of their like giant data centers, and it's not because they're not bullish on AI compute, but rather like it's really con-- it's, it's like it's very hard to bet on the specific type of AI compute and how you build out these data centers.

212
00:37:12.576 --> 00:37:24.696
Like to me, when I look at like Bitcoin miners, it's like scale efficiency and you just-- the-- and the compute itself doesn't structurally change. This is not the case in AI. Uh, I'm curious your thoughts on this.

213
00:37:25.176 --> 00:37:40.886
Like, yeah. Yeah. So a lot of, you know, the-- a lot of the GPUs maybe a couple generations ago were very specialized. Right? Maybe you could get a, a GPU that was really good at inferencing- Uh-huh... like an A100.

214
00:37:41.396 --> 00:37:49.736
Uh, sorry, A100 would be good for training, or L40 might be good for inferencing. They're really good at only one compute or the other.

215
00:37:49.746 --> 00:38:04.996
And since then that's been unifying, and with like the Blackwell generations of B200, B300, they're both really good at doing both of those workloads. Um, they are extremely expensive, uh, to get a single host of B300.

216
00:38:05.416 --> 00:38:16.516
Uh, marketplace is insane right now, but maybe five hundred thousand. Probably by the time I get off of this it'll be five fifty. So it's, it's a lot of capital for those.

217
00:38:16.916 --> 00:38:28.236
Uh, not every site that you pick, and this is where miners are gonna be more comfortable with it, not every site is gonna have a really competitive energy rate, uh, 'cause it takes a lot of power to run these.

218
00:38:28.516 --> 00:38:41.996
Uh, you also have to factor in cooling. You know, with Microsoft's idea of, hey, it's hard to bet on this, true, and I also spent close to a decade looking at, uh, compute sizing from, uh, a lot of different companies.

219
00:38:42.116 --> 00:38:48.376
In most cloud workloads, people are not taxing their compute particularly hard.

220
00:38:48.676 --> 00:39:00.756
So until we see the market not having a place for those older generation GPUs that are a little less powerful, I think there's still gonna be money to be made. Uh, workloads are changing all the time.

221
00:39:01.156 --> 00:39:10.256
You know, I've just received an update on, uh, how we're using AI internally, and it's completely different from where we were a month ago. Uh, so

222
00:39:11.256 --> 00:39:20.906
I think all of these are gonna be fully taxed from an investment standpoint. Yeah, I think-- Uh, I'm-- I, uh, I think that does it for my questions, Charlie. I, I have one last question- Okay... which is- Go for it.

223
00:39:21.356 --> 00:39:34.976
I'm running a thirty-ninety founders edition. It's been great. Uh, should I, should I wait to upgrade another couple NVIDIA cycles, uh, or do I just shell out for the five-ninety or just the latest AMD? I don't know.

224
00:39:35.126 --> 00:39:47.416
Y- uh, any tips for us gamers out here who are sweating hard? Uh, and al- also my RAM. I can't afford RAM anymore. Holy smokes, dude, my DDR5 has like quadrupled in the past couple months.

225
00:39:47.796 --> 00:39:58.306
Uh, Mike, uh, tips for the, the gamers out there. Yeah. Fifty-nineties, you're, you're-- That, that isn't-- Every time I think that price inflation's gonna stop, it goes up.

226
00:39:58.756 --> 00:40:09.116
So, you know, if you've got a fifty-ninety, uh, you could probably sell that. Six months ago it was, you know, twenty-two fifty a unit for bulk price, now it's, it's closer to four thousand.

227
00:40:09.436 --> 00:40:21.536
Yeah, I'm not buying-- So I'm, I'm not gonna shell out 4K for a fifty-ninety. I-- My games run just fine on my thirty-ninety, but oof, it's rough out there. And RAM, RAM is even worse.

228
00:40:21.796 --> 00:40:25.876
Uh, you know, the word on the street was that OpenAI and, and, uh,

229
00:40:26.856 --> 00:40:37.895
xAI bought all of, on the same day, uh, all of the RAM from the three manufacturers that are out there for enterprise-grade RAM for the next eighteen months. Uh, so that's cascaded.

230
00:40:38.036 --> 00:40:47.926
A lot of your consumer-grade, uh, manufacturers are shifting over to enterprise, and most OEMs, if you go to them, those prices are good for forty-eight hours.

231
00:40:48.056 --> 00:40:55.256
Maybe they'll give you a new higher price when your product's ready to ship. Holy smokes. Okay. Well, Mike, thank you so much for coming on the show.

232
00:40:55.556 --> 00:41:06.676
Uh, it feels like we can run this back, uh, every time the GPU market spikes, uh, fifty percent, which is probably every other week for the next few months. So thanks for coming on the show, everybody.

233
00:41:07.006 --> 00:41:20.406
Mike, catch you later. Thank you, Mike. Thanks. All right. We got a couple more news items. We do have a couple more news items. Before we get into our Cry Corner, which is kind of a dual Cry Corner actually- Yeah...

234
00:41:20.406 --> 00:41:26.186
'cause, uh, I want-- The Jane Street stuff has a lot of tinfoil involved in it. So- Yeah. Maybe we'll like- We're gonna-...

235
00:41:26.196 --> 00:41:35.856
read through the Jane- We're gonna throw that in the microwave and let some, uh, sparks fly in a minute. But before we start that, I'm gonna pop up, well, if I can do it. There we go. There we go.

236
00:41:36.246 --> 00:41:42.266
The-- So, uh, we're gonna do a, a little bit of earnings coverage. I'm not gonna go over numbers. I'm not gonna bore y'all with that.

237
00:41:42.316 --> 00:41:54.306
If you wanna check out our coverage of all of the major public miners that released earnings this week, specifically Mara, Cipher, American Bitcoin, Hut 8, I might be forgetting one in there, TeraWulf, if you wanna look at those- TeraWulf...

238
00:41:54.376 --> 00:42:05.336
head over to blockspace.media, keyword those in the search bar, you can look at the financials for those companies. We covered all of them in our daily coverage this week. But I wanna f-focus on a few qualitative things.

239
00:42:05.936 --> 00:42:13.536
Uh, first we'll, we'll start with this Mara earnings call. Uh, they reported nine hundred and seven million full year revenue. Lot of money.

240
00:42:14.056 --> 00:42:18.616
Uh, w-we'll, you know, go, go dig into it to see how much of that led to any profitability.

241
00:42:18.656 --> 00:42:30.916
But the biggest part of this news item was concurrent with theirEarnings, Mara announced that they are leaning hard into AI and HPC for the first time ever.

242
00:42:30.996 --> 00:42:39.576
So they've established a joint venture with Starwood Capital Group to develop AI and HPC sites across existing infrastructure.

243
00:42:39.756 --> 00:42:46.336
The, the partnership targets roughly one gigawatt of near AI capacity with a pathway for more than two point five gigawatts over time.

244
00:42:46.736 --> 00:43:00.576
It looks like they're going to be focusing on their sites in Texas, mainly Granbury and Garden City. And for some background, Starwood group is not a neo cloud or a hyperscaler or anything like that.

245
00:43:00.616 --> 00:43:08.036
They're specifically an investment and development firm that focuses on digital infrastructure, so they invest money into basically building out data centers, right?

246
00:43:08.576 --> 00:43:22.036
So this isn't a deal in the sense that Mara now has a partner lined up to do AI services and do compute services. They do have a partner lined up to develop their existing sites for that in the future.

247
00:43:22.096 --> 00:43:29.596
So just wanted to make that distinction because w-when, when you read these headlines, sometimes the market treats it all the same. It's like, "Oh, they're jumping into AI."

248
00:43:30.076 --> 00:43:42.116
Uh, this is the most definitive step they've taken, I think you could say, at least in the US market to do this. Of course, they also recently closed on at-- their, uh, bid for Exion.

249
00:43:42.196 --> 00:43:44.866
Uh, that closed, I believe, last Friday.

250
00:43:45.076 --> 00:43:57.136
Last year, they announced that they were going to take a sixty-four percent controlling stake in Exion, which is the data center subsidiary for the Électricité de France, which is the national energy company of France.

251
00:43:57.176 --> 00:44:06.256
There was a lot of, of pushback, or not a lot of pushback, but the deal kinda got drug along. Originally, the French treasury approved it, then officials said, "Hang on. Wait a second.

252
00:44:06.526 --> 00:44:12.356
This American company is coming in, gonna take a controlling stake in one of our data center companies. We don't wanna make sure there aren't any sovereignty concerns."

253
00:44:13.006 --> 00:44:26.096
The French brought in billionaire Xavier Niel to sit in on the board and also take a stake in the, um, in Mara France, which is, you know, a subsidiary for the [chuckles] takeover of Exion.

254
00:44:26.116 --> 00:44:30.195
There's this whole web of ownership. All that being said, it-- that was completed recently.

255
00:44:30.296 --> 00:44:46.036
So within the span of two weeks, you have two big news items from Mara for their entrance into AI and HPC by taking that controlling stake of Exion and now this announcement that they're having a joint venture with this company, Starwood, to develop AI and HPC data centers here in the US.

256
00:44:46.516 --> 00:44:56.125
And as friend of the show, Reggie Smith, analyst at JP Morgan, said in a note that was released last night on this, "Better late than never to the party." He said it on an earnings call too.

257
00:44:56.556 --> 00:44:59.656
He congratulated Fred and the team and said, "Welcome to the party."

258
00:44:59.676 --> 00:45:07.316
Because Mara famously said last year, they didn't think that they were going to make a j- a leap into this market, or at least not in the way most other firms were doing.

259
00:45:07.356 --> 00:45:14.916
They were kind of cagey on the language as to what that would look like. Now it seems like with all the other miners, they're running full bore into it.

260
00:45:15.036 --> 00:45:27.536
And speaking of running full bore, last point on earnings calls, Charlie, and then I'll shut up. Another one that I wanted to highlight was Cipher Digital, formerly Cipher Mining, now rebranded to Cipher Digital.

261
00:45:27.936 --> 00:45:47.196
Two things: They sold one of their sites to Canaan this week, um, as they continue to divest away from Bitcoin mining, but they are also committing to selling their entire Bitcoin treasury through twenty twenty-six as they look at AI, um, and HPC business lines.

262
00:45:47.376 --> 00:46:00.236
And that-- I mean, that puts-- That's every-- That's like-- That's two big ones dropped this week for their treasury, the Cipher and Bitdeer. Yes. Bi-Bitdeer already sold all of theirs. Yeah, Bitdeer already sold? Yeah.

263
00:46:00.516 --> 00:46:05.796
Yeah. Bi-Bit-Bitdeer divested completely of their Bitcoin holdings, and Cipher is saying they're gonna do the same thing.

264
00:46:06.276 --> 00:46:16.296
And just to highlight this point too, as we point out in this article, Canaan purchased Cipher's joint venture stakes in its Bit- in some of its Bitcoin mining operations.

265
00:46:16.386 --> 00:46:24.656
And so, you know, once again, just kind of pointing to the fact that these Bitcoin miners are exiting Bitcoin mining entirely.

266
00:46:24.736 --> 00:46:37.696
And if you-- You know it's getting serious when they're deciding to sell their Bitcoin treasuries, because one of the early, you know, bull cases for these Bitcoin miners in twenty twenty and twenty twenty-one was that they were kinda these proxies to Bitcoin exposure.

267
00:46:37.716 --> 00:46:47.476
You know, Strategy was, was not a household name at that point. They were just booting up their strategy as a Bitcoin wrapper for Wall Street. The Bitcoin ETFs, though, came and totally shattered this.

268
00:46:47.596 --> 00:46:56.076
If you look at correlations between Bitcoin mining stock prices and Bitcoin's price, it degrades by the time-- from the time that the ETFs come out until now.

269
00:46:56.156 --> 00:47:03.296
I haven't run the analysis in a while, but you can see the correlation breakdown. It used to be super tight. Whenever Bitcoin would rip, they would rip.

270
00:47:03.316 --> 00:47:14.216
Whenever Bitcoin would, um, which the bed, they would also crap their pants. But now, uh, that, that correlation, it's still there, it's just not as strongly defined as it used to be.

271
00:47:14.276 --> 00:47:17.236
So for these Bitcoin miners, it, it really does make sense.

272
00:47:17.296 --> 00:47:30.276
If you have this Bitcoin treasury and you're not leveraging it by lending it out and making some yield on it, and even if you are doing that, you're leaving money on the table probably for, uh, for these expansions and, and, and, uh, pivots into AI.

273
00:47:30.716 --> 00:47:39.496
Like, that's capital that you could be deploying otherwise to grow your business into this new sector. So it makes sense to me. Um, I'm kinda surprised it took this long, honestly.

274
00:47:39.556 --> 00:47:45.946
I'm, I'm sure a lot of these miners are probably kicking themselves for not doing this last year. I mean, we kind of have seen this before, though.

275
00:47:45.976 --> 00:47:58.416
In twenty twenty and twenty twenty-one, some of the Bitcoin miners actually... Or the last cycle, they actually purchased Bitcoin, you know. Um, and when they ended up selling it at a bear market. Yeah, I like...

276
00:47:58.456 --> 00:48:10.996
Okay, we've talked, we've kicked this dead horse so much about, like, buy treasury strategy for miners. Like, you don't see other commodities producers really do this. They hedge, they don't stockpile. So yeah.

277
00:48:11.096 --> 00:48:20.576
Uh, I will say I was kinda disappointed in other media. Gotta pat ourselves on the back here because other media will be like, "Bitdeer, Bitcoin treasury company, sells." I'm like, "That's... No, that's not the headline.

278
00:48:20.636 --> 00:48:30.196
They're not a Bitcoin treasury company. These are..." You, uh-- Anyway. Yeah. And I think that's something- bitcointreasuries.net, which classifies them as a treasuries company- Exactly...which is very misleading.

279
00:48:30.256 --> 00:48:38.880
But shout out to bitcointreasuries.net. Just like, come on.Anyway, that, those are just a few, I think the more salient points that came out of earnings calls this week and earnings releases.

280
00:48:38.940 --> 00:48:43.820
If you wanna look into the actual numbers, go check out our coverage on the website. You can get into all that boring stuff.

281
00:48:45.140 --> 00:48:53.900
So if you've been paying attention to Bitcoin Twitter this week, you might have seen people insinuating that Jane Street is why Bitcoin has been selling off since October.

282
00:48:54.660 --> 00:49:06.380
And the reason for this is they're basically connecting the dots between the current market action and this lawsuit that the Terraform estate has levied against Jane Street.

283
00:49:06.440 --> 00:49:17.140
So this week, one Todd Snyder, who, uh, is the administrator of the Terraform estate according to the Delaware bankruptcy proceedings for Terraform.

284
00:49:17.680 --> 00:49:26.850
For those who don't know, Terraform is the parent company for the Terra blockchain and the Luna and UST ecosystem. Uh, I recommend going- Ghosts from crypto past come to haunt us again. This, this...

285
00:49:26.980 --> 00:49:31.900
You sound insane trying to explain this, 'cause like, there, there are so many different- Yeah, I always sound insane... actors.

286
00:49:32.000 --> 00:49:42.010
Basically, it's Mic- It's like if MicroStrategy's Stretch product were a crypto degen thing. [laughs] That's actually not a terrible- I'm right. That's not a terrible analogy. You're welcome.

287
00:49:42.080 --> 00:49:57.380
So in 2022, a algorithmic stablecoin, UST, blew up spectacularly, and depending on who you ask and what angle you wanna take, this was the, the f- the spark, lit the fuse that eventually blew up FTX and destroyed the market in 2022.

288
00:49:57.920 --> 00:50:11.660
But the way this ecosystem worked is that it's algorithmic in the sense that market incentives and this dynamic between UST and a, uh, companion token called Luna would basically keep the peg at roughly $1.

289
00:50:12.440 --> 00:50:25.060
So, you know, if UST traded above $1, traders could, uh, burn $1's worth of Luna to mint one UST and then maintain the peg, and it works the other way on the way down.

290
00:50:25.120 --> 00:50:37.420
If UST fell below $1, they could burn one UST and receive $1 of Luna as a result. What the lawsuit alleges is that this former Terraform intern,

291
00:50:38.460 --> 00:50:57.160
named Bryce Pratt, who ended up joining Jane Street in September 2021, he was a intern, a software engineer in the summer of 2021 for Terraform before he joined Jane Street, basically used insider information to trade against, uh, the Terra ecosystem.

292
00:50:57.840 --> 00:51:02.460
And I, I recommend you go and read the full article, 'cause I'm giving a truncated version here.

293
00:51:03.060 --> 00:51:19.980
The basic idea is right before the depegging event in May that basically destroyed UST and wiped out the ecosystem, the Terra team moved or announced it was going to move all of its UST, that's the stablecoin, out of a Curve liquidity pool.

294
00:51:20.460 --> 00:51:25.940
Now, this liquidity pool is not endogenous to the Terraform ecosystem.

295
00:51:25.970 --> 00:51:35.700
These liquidity pools were DeFi pools where traders would deposit currencies, and then you could then- It's how they generated yield to pay out, uh, the Terra- Exactly. Yeah.

296
00:51:35.890 --> 00:51:49.220
A- and it was a place where traders could trade UST for other stablecoins. It was a way for them to generate yield, like Charlie said, to pay yield on the, uh, Luna token or, or UST staking, which was, like, 20%.

297
00:51:49.800 --> 00:51:58.440
W- w- It's just like a crazy interest r- first of all, this is why I'm still talking about other people. That's why, yeah. That's why they loved it, 'cause of the, the absolutely juicy interest rates.

298
00:51:58.600 --> 00:52:05.580
So that's something that's really important to, to point out, is you could stake UST at this time, and you would make 20% interest, uh, from staking it.

299
00:52:06.240 --> 00:52:17.880
Crazy return, but in Apr- in mid, uh, in, uh, mid-April, Terraform announced that it was reducing that interest rate, and it was also migrating the UST it had staked in this Curve pool to a new pool.

300
00:52:18.200 --> 00:52:22.600
They didn't say how much they were moving, and they did not say when they would be moving it.

301
00:52:23.000 --> 00:52:44.560
On May 7th, they ended up withdrawing 150 million from the Curve pool to move it to a new Curve pool, and within, like, 10 minutes of them withdrawing that, Jane Street then sold 85 million UST from the pool that Terraform migrated their 150 million out of, and it tanked the UST price to about, like, 80 cents on the dollar.

302
00:52:45.070 --> 00:52:55.460
And then you basically had this kind of tortured multi-day depegging event where it almost recovered UST, and then it ended up crashing to the ground.

303
00:52:55.530 --> 00:53:07.200
But in this lawsuit, uh, Snyder, on behalf of the Terraform estate, alleges that Jane Street used insider trading to make that sell because, you know, Jane Street...

304
00:53:07.240 --> 00:53:20.500
or Terraform didn't announce that they were withdraw- when they were withdrawing and how much they were withdrawing from the Curve pool, but then Jane Street hopped on right after they withdrew their money, and then, uh, it ended up tanking the price.

305
00:53:20.920 --> 00:53:29.200
Now, two things here. They say that they profited on this, and it's heavily redacted in the lawsuit, so it's never really made clear exactly how they profited.

306
00:53:29.600 --> 00:53:36.400
Like, did they buy back lower, and then when the UST price, uh, bounced back, did they sell right before the thing completely crashed?

307
00:53:36.880 --> 00:53:53.420
The other thing I would say, I'm not saying that Jane Street's innocent, I'm not saying they're guilty, but you don't necessarily need insider information to have bots that are tracking fund flows from a pool that is complete, that is to- all the data is on chain.

308
00:53:54.000 --> 00:54:04.160
So, like, Jane Street, I'm just playing devil's advocate, Jane Street could have had a monitoring system set up to where they saw the funds leaving the Curve pool, and then they decided to execute a sell.

309
00:54:04.220 --> 00:54:13.480
Now, that being said, the weird question, the... and this is a weird question, why would you do this if you... 'Cause these are quant traders. They're not stupid.

310
00:54:13.900 --> 00:54:25.070
You know that if you sell $85 million of this highly, uh, how should we say, feeble asset in the sense that this algorithmic stablecoin- Real... there were a bunch of problems with these- Yeah...

311
00:54:25.120 --> 00:54:26.410
in terms of the market incentives.

312
00:54:26.860 --> 00:54:43.456
Any sort of sell pressure like this, even on a pool like, uh, the Curve pool, is going to exert pressure on the UST price.So why would you market sell $85 million into an extremely thin liquidity pool unless you're trying to mess with the price of that algorithmic stablecoin?

313
00:54:43.576 --> 00:54:49.756
Um, that's playing devil's advocate on the other side. There's no really good reason to do that unless you think two things.

314
00:54:49.896 --> 00:55:00.236
One, I can profit off of destabilizing this, or two, oh crap, maybe the insider information is they knew the ecosystem was close to collapsing anyway, and they wanted to get out.

315
00:55:00.676 --> 00:55:20.496
Um, total speculation, we don't actually know that, and I wanna leave that there before I get into the thread that people pulled for this for what this means for, uh, Bitcoin currently, but I wanna give you a chance to- I mean, I will say, you know, Ja- didn't India ban Jane Street from trading in the country of India because of shenanigans or something?

316
00:55:20.536 --> 00:55:34.516
Like, they extracted a lot of money from, like, [chuckles] Indian retail traders. Uh, and then also, uh, you know who, uh, just, just puttin' it out there, you know who's a ex-Jane Street? Sam Bankman-Fried.

317
00:55:34.616 --> 00:55:46.066
And so this is where the conspiracies come from. Yeah. Also, Jamie fact-checked for us, and by Jamie, I mean Claude. Yeah. India's market regulator did ban Jane Street. [laughs] Yeah.

318
00:55:46.076 --> 00:55:52.676
There's a million jokes there- And so-... and we will let you make those. Yeah, and so what Charlie said is salient about

319
00:55:53.696 --> 00:56:07.636
Sam Bankman-Fried specifically because this whole lawsuit has led to a kind of flurry of conspiracy theories that culminate on the idea that somehow Jane Street is behind Bitcoin's current anemic price action.

320
00:56:08.496 --> 00:56:12.516
To the best of my understanding, here's where this comes from.

321
00:56:12.676 --> 00:56:32.736
After the lawsuit was filed, Bitcoin surged at the exact same hour or roughly around 10:00 AM ET, which, you know, these theorists say is roughly the timeframe when you see Bitcoin market dumping every single day, or at least trading downward as the market opens in, uh, New York.

322
00:56:32.796 --> 00:56:46.836
The idea here is that Jane Street is a, um, is a qualified participant in, in a number of Bitcoin ETFs, which mean they can make and redee- they can make and redeem, uh, ETFs, Bitcoin ETFs at the source.

323
00:56:46.916 --> 00:57:00.316
So Jane Street can go to BlackRock and they can say, "Here's 10 Bit- here's 10,000 Bitcoin, we want 10,000 Bitcoin sh- uh, worth of, of, of ETF shares," and they can also redeem ETF, uh, shares for that Bitcoin.

324
00:57:00.856 --> 00:57:07.936
The idea here is that they're basically creating false claims to then pull out Bitcoin that they never actually owned or don't...

325
00:57:08.176 --> 00:57:19.056
uh, and never, uh, uh, you know, traded, uh, traded for, and they're then taking that Bitcoin and then market dumping it and selling the actual spot Bitcoin on market. That's all conjecture.

326
00:57:19.116 --> 00:57:34.716
There's nothing to actually link these two things, and there have been a number of rebuttals against this in, in the sense that this really is one of those kinda wacky conspiracy theories, you know, kind of getting in the realm of Alex Jones here, where Jane Street is some deep state ploy to destroy Bitcoin.

327
00:57:35.476 --> 00:57:37.316
I, I just think it's worth addressing here.

328
00:57:37.456 --> 00:57:48.036
If you want a pretty cogent takedown of this with someone who knows how these market dynamics work pretty well, I would, I would recommend go checking out Jeff Park on ProCap, formerly Bitwise.

329
00:57:48.156 --> 00:57:51.116
He has a pretty good analysis of, of this whole thing.

330
00:57:51.156 --> 00:58:03.836
But just to recap on that, this lawsuit basically led to this, this [chuckles] pretty crazy conspiracy theory that Jane Street is creating paper claims on Bitcoin to drive the price lower. And how do we know this?

331
00:58:04.296 --> 00:58:14.436
Well, Bitcoin pumped after the lawsuit was released, and also they've done shady stuff in the past allegedly with Terra, so maybe they're doing it now. Oh, and also Sam Bankman-Fried.

332
00:58:14.836 --> 00:58:26.216
I mean, it is a little bit of Charlie from Always Sunny in Philadelphia next to, you know, the board and it- It was like, it is like, I would say you're not wrong. It absolutely does affect Bitcoin's price.

333
00:58:26.816 --> 00:58:35.916
My response is you're trying to connect dots where just, like, the, like, the simple answer's way, way clearer.

334
00:58:36.396 --> 00:58:44.686
People selling Bitcoin, there's better trades out there, it's not like the darling child, it's not some grand conspiracy. BlackRock isn't suppressing the price.

335
00:58:45.116 --> 00:58:54.475
People are probably trading iBit options and, like, doing some kind of weird hedging shenanigans which affects your intraday or maybe week-to-week price, but, like, that's not why Bitcoin's at 65K.

336
00:58:54.656 --> 00:59:07.316
No, and we don't see any of the levers in, like, the precious metals market, like with margin requirements at Comex, that people usually point to to say, "See, you know, paper claims on precious metals are keeping the price suppressed."

337
00:59:07.356 --> 00:59:13.156
We're, we're not seeing that within the Bitcoin market. Doesn't mean that things can't be done to play with the price.

338
00:59:14.076 --> 00:59:24.356
It, it's just, it's usually more boring than people make it out to be if that is happening, and it has something more to do with maybe incentives in the market rather than, like, one actor pushing the price down.

339
00:59:24.736 --> 00:59:33.536
But fun c- we love conspiracies on the pod. Fun, but you know- We have, we have Cry Corner. We, we, you know, we should probably add a Conspiracy Corner. That'd be way more fun. Yeah.

340
00:59:33.606 --> 00:59:44.936
Okay, so the end of this, the end of this, we're gonna wrap up with strategy. Um, the most shorted stock of the week, absolute banger Cry Corner here.

341
00:59:45.416 --> 00:59:56.836
Sorry to all the strategy owners, but, uh, Wall Street hates your stock right now. I thought this was a rumor, but it's true. [laughs] I did too. When you brought it up in the chat, I thought there's...

342
00:59:56.936 --> 00:59:59.676
Is that really, is that true? Yeah. How does anyone actually know this?

343
01:00:00.136 --> 01:00:15.476
This is data from FactSet compiled by Goldman Sachs Investment Research, and if, uh, y'all have to squint to look at it a little bit, but if you look here, this is a ch- this is a table of 50 stocks, over 25 billion with largest short interest as a percentage of market cap.

344
01:00:15.856 --> 01:00:23.916
Currently, the short interest as a present- percentage of strategy's ma- uh, market cap is 14%, and it, it's the highest on this list.

345
01:00:24.576 --> 01:00:37.059
So if you're looking at this, you can kinda read this one of two ways: short squeeze incomingAll the haters and disbelief as Strategy goes to the moon, or, "Holy, it's starting to get real," Strategy is getting absolutely pummeled.

346
01:00:37.300 --> 01:00:45.000
I don't have anything else to add to this. I don't really have anything else to add there too except- But the- Um, I guess the one thing is that Strategy did have its, like, Strategy, uh, what is it?

347
01:00:45.380 --> 01:00:53.180
Um, Strategy World- Strategy Conference. Big corporate corporation... Strategy World, Strategy Universe. Uh, well, it's like Disney, you know? Strategy, like- Strategy World is amazing.

348
01:00:53.240 --> 01:01:04.380
It's like Epcot except, like, the [laughs] except the all the rides are just Saylor, like, ranting at you about why you should sell your house and mortgage it or whatever to buy Bitcoin. Yeah.

349
01:01:04.500 --> 01:01:13.780
They, their, their Imagineers are working overtime to come up with new financial products. [laughs] Adding new, new points to their, uh, star, which is totally not a pentagram, bro.

350
01:01:14.260 --> 01:01:23.820
Okay, on that note, [laughs] on that note, we're gonna wrap this up before the Strategy goons come and haul us off. Uh, Michael Saylor, come on the pod, please. We'll be way nicer to you.

351
01:01:23.880 --> 01:01:35.280
We won't, we won't be so harsh as Danny from What Bitcoin Did. We'll, we'll, we'll, uh, we'll, we'll suck up to you, and we'll, we'll only give you lauding praise, so come onto the pod. Hope to see you soon.

352
01:01:35.760 --> 01:01:38.920
Like and subscribe. Make sure to hit up all the Blockspace stuff. Bye.

353
01:01:39.080 --> 01:01:54.900
[upbeat music] Hey, this is Charlie and Colin from Blockspace Media, and you're listening to the Blockspace Podcast, a show about emerging tech in Bitcoin, AI, energy, and markets.

354
01:01:55.180 --> 01:02:07.880
We publish two interviews weekly with CEOs, investors, analysts, and anyone else of consequence within these spaces. Plus, we have a weekly news roundup for all the important stories you might have missed from that week.

355
01:02:08.180 --> 01:02:18.780
The show is perfect for retail and institutional investors, analysts, and really anyone who wants to keep their finger on the pulse of the stories that are moving Bitcoin, energy, and data markets.

356
01:02:19.120 --> 01:02:27.560
So if you've stumbled across us, make sure to search Blockspace wherever you get your podcasts, and on YouTube, hit the subscribe button, give us a rating.

357
01:02:27.600 --> 01:02:32.400
We produce bonus podcasts and other content on our main feed, so you don't wanna miss that.

358
01:02:32.410 --> 01:02:42.199
And if you have any feedback or comments to give us or shows that you would like to see and topics you would like us to cover, hit us up at hello@blockspace.media. [upbeat music]