← Blockspace: AI & Bitcoin
Blockspace: AI & Bitcoin
Why Anthropic Could Be Worth $6T by 2027, Inside CCP-Linked Data Center Pushback, Amazon’s $25B Bond
01:04:11
Loading summary
Transcript89 lines
- [00:00]
A
What's going on, y'?
- [00:00]
B
All? Welcome back to Blockspace live presented by CleanSpark. Charlie, we've got a meaty report to open with today. Semianlysis using their sleuthing skills to project that Anthropic may soon clear 1 billion in profit this quarter. And it's not on the backs of ordinary users like you, dear listener. It's through their B2B channels. That's our lead story for today. Following that we have a 30 minute interview with Bitcoin Policy Institute's head of research Sam Lyman on CCP linked individuals pushing anti data center FUD in the US and actually succeeding in derailing what they estimate is north of $23 billion worth of investment across various communities in the U.S. following Sam, we have Tom Massero of Sphere3D on to talk about the company's AI pivot and also their dark horse technology rebrand. We will close the show on news from CNBC that Amazon is looking to offer $25 billion in bonds as private credit does not stop in the AI capex boom.
- [01:12]
C
That's right, Block Space goes live weekdays, 1pm Eastern featuring quick hits on AI data centers, Bitcoin mining, emerging tech and markets. If you like what you hear, you'll love our newsletter@newsletter blockspacemedia.com and if you missed the live stream, you can catch the podcast shortly after we wrap up live. The podcast version goes out on all platforms where you get podcasts. This show is brought to you by CleanSpark, NASDAQ listed ticker CLSK or on CleanSpark later on in the show. Colin, we're kicking it off with Anthropic Now I'm not unbiased here. All my tokens go to Claude. These days I've got all my MCPs and Claude MDs all hooked up. But the big news is is that Semianalysis, probably the foremost like research outfit of this era of AI and tokens, Semianalysis says Anthropic is on track to be a $6 trillion company. So how do we get to that? Well, Semi Analysis put out a paywall piece. Again, can't read the actual unpaywalled side, but basically we look at their first, second and third quarter of 2026 financials, which are not fully published, but Semianalysis has predicted what they would be. And then as stories come out in the Wall Street Journal, Goldman Sachs and in their private IPO filing, they're able to back test this and say look, we Semianalysis were right about our prediction for anthropic's financials during 2026. And based upon these same predictions, that means Anthropic is headed towards a $6 trillion valuation by the end of 2027. This is big news. As their IPO approaches. Remember, SpaceX IPO'd at a little over $2 trillion if I'm not mistaken. So this would put Anthropic on track to be six times more valuable or higher market cap than SpaceX. So, Colin, I throw it to you before I get into kind of how the financials and how this projection arrived to where it is.
- [03:39]
B
Well, yeah, and it would make Anthropic the most valuable company in the world right now. That's Nvidia.
- [03:43]
C
Yeah. Assuming that Nvidia doesn't grow, which, who knows, but that would make Anthropic number one in the world. Dario becomes the CEO of the most valuable company in human history.
- [03:55]
B
Yeah, and then maybe we get another trillionaire minted on top of Elon Musk. Probably not, but it. You know, the other really important thing to note here is that Anthropic filed its, its prospectus for an ipo. It was not public, but they did that back in June, and OpenAI is actually pushing their IPO to 2027. So this is within context of the race for the first Frontier LLM provider to actually list on public markets, which I think is notable. And you got to wonder, Charlie too, with OpenAI pushing its outlook for its own IPO, how much of that has to do with some of the competitive positioning Anthropic has done here with regard to business to business revenue streams, which is a huge chunk of the revenue that Semianalysis projects in this report. So give us a breakdown. What exactly does it say? And give us some of the methodology too, because I think that's important. This is not based on hard numbers that have been reported by Anthropic. This is Semianalysis backing this out of Anthropics operations from some pretty incredible research, doing surveys, et cetera.
- [05:02]
C
Yeah. So at the heart of this is the Semianalysis tokenomics model, which basically they do a bunch of heuristics and testing and they can see how many tokens Claude is using, how many tokens the other LLM providers are using. And you can sign a value to those, and you can also figure out if those are coming from like regular subscription plans, they're coming from API access, and that actually really matters from where the tokens are being consumed. And so Semianalysis posts this chart here showing their Q1 prediction and their Q2 prediction, and then as per the Wall Street Journal we can actually get Q1 actuals and Q2 actuals which semi analysis puts against their project, their projections which were months and months and months ago and can point and say we were right about this. So based upon that, revenue for Clot for anthropic jumped 130% in Q2 to 10.9 billion. The first ever operating profit in Q2, around 559 million. Then the blended gross margin went up to mid 60%, most of which was the Anthropic's API business versus back in 2024. They're running at a negative 94% gross margin. And this is where their business model diverges from OpenAI because OpenAI, which is kind of like the Kleenex brand of AI and LLMs right now because it's all, you know, people go to GPT. OpenAI has a lot of like retail consumers 75 to 85% of, I'm sorry, OpenAI has 65% of the revenue from subscriptions whereas OpenAI pulls 75 to 85% from their usage based API. Now if you aren't really familiar with this, the API is you basically plug in directly into Anthropic and you figure out whatever kind of like front, front end you want to use. This is not like the app on your phone, this is plugging directly into Anthropic's intelligence. And a lot of people don't know this. When you buy a subscription from one of these companies, Anthropic OpenAI, you're basically getting a hugely subsidized use of their tokens and intelligence. When you pull from their API, you're basically paying full token price. And because Anthropic's revenue is largely from the API, they are getting more full price value for these tokens. So they're pretty much more profitable per unit of compute. And then over the past half of 26, 2026, OpenAI has taken the field as far as business use.
- [07:57]
B
Anthropic.
- [07:58]
C
I'm sorry, Anthropic. Oh my gosh. Yeah, thank you. Anthropic has taken the field for like business use. And that is pretty significant because businesses are the ones who want to pay for this token. Some other stats, Claude Code now accounts for over 7% of all code commits on GitHub. And 65% of Anthropic's annual recurring revenue comes from coding use cases. Semi analysis can get some rough numbers from Cursor, Cognition, Lovable and replit, which indicate 6 billion in annual recurring revenue. So if Anthropic keeps up this trend of $15 billion per month. Net new annual recurring revenue ARR could reach 300 billion by the end of 2027, which implies a $6 trillion enterprise value. Obviously the market will decide what the market cap is, but that is pretty incredible. Anthropics, the first, I think the fastest growing company. It's gone a 10x in gross revenue over the past three years. So this is.
- [09:10]
B
I wonder what they're basing that multiple on because that's still. That's a pretty rich multiple.
- [09:15]
C
It is, yeah, it is. As we know, 20 end of 2027 is a whole year and a half away. That may as well be eons. This business shifts on a dime. If you remember there was the clod moment like the opus4.5 moment in late November of 2025 and that pretty much kicked off the modern era of like these, the Anthropic OpenAI duopoly frontier model racing.
- [09:44]
B
But if you're doing it off of 300 billion revenue, that's a 20x multiple on sales. I mean admittedly SpaceX was 111.
- [09:53]
C
Yeah. Compared to everybody else right now.
- [09:57]
B
Maybe that's a little more conservative. But I don't want to spend too much time on nitpicking that.
- [10:03]
C
Yeah, there's only one other chart which I want to show here and that is this one. Friend of the show Small Cap Sniper on Twitter shows this chart from Goldman Sachs which is a chart showing token use by AI agents and its projections as per Goldman Sachs targeted to cross 120,000,000,000, which is an increase of 24 times by 2030. Anthropic processed 5,000,000 tokens in the last month of June and are on track and anticipate to process 13, quadrillion tokens by December. Basically a token is a unit of AI Compute. For those who I manage not to preface this for at the beginning that's all I've got for OpenAI. I think we should probably get to the fun Deep Dive report interview here.
- [11:05]
B
Yeah, I'm excited for this one. We will get Sam Lyman up here in just a minute. But first a word from our sponsor, CleanSpark. We are CleanSpark, America's Bitcoin miner. A publicly traded company with the largest operating hash rate powered entirely by self operated infrastructure across four states. This is our proof of work and we are setting the standard for what's next. Learn more about the intersection of energy and bitcoin@cleanspark.com
- [11:44]
D
if Bitcoin's actually the best money and it's the thing that people should accumulate. It's the best risk adjusted asset. I'd lose zero sleep about whether or
- [11:52]
B
not that's gonna happen.
- [11:52]
D
I just asked the question of when is literally matrix math that you're running
- [11:59]
B
on large pieces of data.
- [12:00]
A
The bitcoin miners can absorb that energy
- [12:03]
B
and in many ways this feels like a second bite at the apple to
- [12:06]
D
build a new Internet.
- [12:16]
B
Alrighty, we have Bitcoin Policy Institute head of research Sam Lyman waiting in the wings. Sartorial excellence with the suit and tie. Sam, welcome back to the show.
- [12:28]
A
Thanks guys. Great to be back. I had to rock the suit coiner look today because we had some meetings on Capitol Hill.
- [12:33]
C
You got, you gotta look, you gotta look presentable. I, I put on a button down shirt. That's the closest I'll ever get.
- [12:40]
A
I was gonna say I heard I was gonna be on Block Space today, so. I should have worn a tuxedo.
- [12:44]
B
Yeah, we're, we're more business casual here. You know, it's always Hawaiian shirt Fridays at Block Space. So, Sam, y' all put out this really meaty piece on part two on the foreign influence in the campaign against American AI. I'm going to pop it up here on the screen in a second. A few headline numbers and the dollar figure here, which we'll get into, I'm sure is conservative, but. 23.6 billion in delayed or blocked data centers and funding, 10 local moratoria, four scrapped or rejected data center projects, one permanent ban. And I would also add there maybe this is in that figure, a downsizing of Kevin o' Leary's box Elder county center and a partridge in a pear tree. So Sam, could you give us a quick TLDR of your research here before we dive into some of the details for sure.
- [13:33]
A
Great to be back with you guys. And I'm excited to talk about the sequel to our first report. So for your viewers who may not know about our first report, we were able to uncover a foreign influence operation from China where China State Media was broadcasting for years to American audiences about the dangers of data centers, how they would raise electricity prices and water prices. And we recognized in our research that working in tandem with China State Media was a non profit network which we called the Singham Network. The Singham Network is led by a man named Neville Singham based in Shanghai. He's a self identified Marxist and for years he's been pushing CCP propaganda to American audiences. And most recently he's been doing that with data centers. So in our SQL report part two here we were able to find 21 different case studies across the country where Singham's group has been boots on the ground in data center fights that have stalled or blocked $23.6 billion in AI build out here in the United States. So they're doing real material damage on the ground beyond just the propaganda that they've been pushing for years.
- [14:38]
B
So a lot of this report hinges on specifically Singham. And you mentioned, like you said, he shares an office in Shanghai with one of the Chinese state medias. In fact, his nonprofit organizations have actually echoed their own those CCP linked media's language in some of these data center pushes. But it also hinges on the psl, which is Party for Socialism and Liberation, quote, as you describe it here, a Marxist Leninist group with documented foreign ties who stated mission is to dismantle American capitalism and whose leadership is drawn directly from executives of Singham's nonprofits. Can you give us an outline of these connections? Because it seems to me that a lot of the report hinges on this idea or the documented evidence that a lot of the leaders of ESL are actually drawn from Singham's nonprofit network.
- [15:28]
A
Yeah, this was one of the biggest findings of our report. That finding is that the Party for Socialism and Liberation in the same network are one in the same. If you were to draw a Venn diagram between the two SIGM network here, PSL here, it'd be a complete circle, just direct overlap between the leadership and the PSL itself. So we were able to find five different nonprofit executive leaders that work for Cingham, that are paid by Cingham, who, coincidentally or not, are also the leadership of the Party for Socialism and Liberation. We realized that there's really no daylight between these two organizations. And we built out a graphic to make it clear as day. For anyone wondering who runs the psl, the answer is very clearly Neville Singham. But that's not just our telling. Neville Singham also has a lot of enemies who are on the far left of American politics. And they themselves have identified Singham as being the mind behind PSL and the person who calls the shots for PSL all the way over in Beijing.
- [16:27]
B
So go ahead, Charlie.
- [16:29]
C
Yeah, so when we talk about like astroturfing or propaganda or these efforts, like, what do they actually look like on the ground? Because I go to Facebook and I see what appear to be a pretty organic post, which, like everybody, there's a big meeting at the city council tonight. Data center. Got to oppose it. Is it just getting people to these. Is it circulating? What does it actually look like? What are they actually doing?
- [16:56]
A
That's a great question. And we want to be clear that in our research we're no way saying that the entire anti data center movement is Astroturf. Because very clearly there's a grassroots element of this. There's a lot of people who have organic, authentic and genuine concerns about the impact data centers might have on their communities. What we have noticed though, that running alongside that organic movement is an inorganic movement. And we want to be able to provide transparency for the impact that inorganic movement is having and exactly who is funding it and who's driving that messaging. And so this is what it looks like. I'll offer up a very tangible example that took place just yesterday. So we're based here in Washington D.C. here at the Bitcoin Policy Institute, 10 miles up the road in Prince George's County, Maryland, the city Council yesterday was able to enact a two year data center moratorium. Who was leading that fight? It was the party for socialism and liberation. And that's not according to BPI's own reporting. That's according to an NPR affiliate based here in Washington D.C. that did a whole cover story on the PSL a few months ago talking about how the PSL went door to door. It's called canvassing in activism. You go door to door and you raise awareness for whatever issue you're concerned about and you gather signatures. The PSL in Prince George's County, Maryland in this specific case study, was able to get over 20,000 signatures protesting this data center. And then they were able to rally hundreds of people to city council hearings where the city council members said they'd never seen this many people before. They were able to do that through viral posts on social media and again by going door to door through their neighborhood canvassing. And by so doing, they were able to create such an illusion of this mass opposition to data centers in Prince George's county that the council members really had no choice but to listen to what looked like a huge congregation of people protesting this construction. What was interesting though was last week there was a counter protest to the Singham Network led protest in Prince George's County. So I think a lot of people who are pro data center are starting to catch on to the importance of activity activism in these particular fights. But that's just one example of what it looks like in real time because of that two year data center pause. That's $5 billion in AI investment that will either be stalled or blocked indefinitely in Prince George's County.
- [19:13]
B
Sam, I want to return really quickly to two points. One, the pushback on the left against the left on this and then also the PSL leaders in Singham's orbit and their ties to adversaries of the us so we'll start with the first one you mentioned specifically in this piece. One Jeremy Ferguson. Fergie Chambers, who if I believe am I correct about this, is actually the heir of the Cox Communications fortune.
- [19:41]
D
That's right.
- [19:42]
B
So you know, I mean net worth estimated according to Wikipedia, $250 million. You know, communism for thee, but not for me, some might say. What is your feeling as to why he had a falling out with Singham? You probably can't speak to that very much, but I'm just curious as to what hunches you have for the motivation to push back against the psl. Reminds me, and you actually even mentioned Trotskyists in this article or kind of neo Trotsky views on what the PSL is doing. It reminds me of some of the power struggles between Stalin and Trotsky in the early days of the Soviet Union. What is your sense as to why there's fracturing in some of these communist communities over the PSL and this issue specifically?
- [20:27]
A
It's a great question. I went down the rabbit hole of modern day communism. A lot of people think that communism is dead and gone. What I learned over the past month or so is that it's very much alive and well and the internal spats between communists are just as intense today as they were over 100 years ago. You mentioned between Stalin in Lenin for example. I think Jeremy Chambers specific critique is that Neville Singham is not a true blue or I guess true red believer in communism. That he is more in the tank for China and the CCP than he is to spreading communism worldwide. That seems to be his critique. It could also be just sharp elbows within the communist movement. You have two multi millionaires who are the primary financiers of socialist movements across the world. So naturally they're not going to get along 100% of the time. It sounds like the impetus of their falling out was over the Israel issue. It sounds like Jeremy Chambers says that Neville Singham doesn't believe in advocating against Israel to the same extent that he does because all communists seem to hate Israel and the United States. And so it seems to be inside baseball within the communist movement, but looks like it boils down to personal differences, political differences and problems. Probably just some ego clashing. At the end of the day it
- [21:46]
B
seems like, you know, Trotsky isn't probably a good word for it here in regards to the view whether or not communism should be contained within a single state versus projected onto the entire world as a global workers movement. That was the clear delineation between Stalin who and Trotsky who preferred specifically just to focus on the ussr. And so going back to that, you mentioned the Trotskyist angle here specifically when looking at some of Singham's data deputies, if we can call them that, the PSL leadership and some of their ties to US adversaries, some of whom are actually still communist countries. You break down some of those ties. I believe it's, it's, it spans all the way from Singham's wife, who is the founder of Code Pink, as well as the PSL Party presidential candidate for 2024 and also the editor in chief for, for one of their magazines. Can you explain what some of these ties are and how concrete they are?
- [22:43]
A
For sure. And you mentioned a good point about the Trotsky is the Trotsky critique of Singham and his network right now is that the work they do is indistinguishable from a PR agency for the ccp. And so I think that tells you all you need to know about how other people on the left view the SIEM network. Now, the CM networks particular ties to other countries is what is truly damning. To begin with. The PSL is not an America first organization by any stretch of the imagination. This is a internationalist first organization. By that I mean their allegiance is to what they call the global working class, as opposed to the American working class or everybody else in America. Because of that, they've partnered heavily with Cuba over the years. In fact, making trips to Cuba has been a key part of their agenda. In addition to that, they've been able to build really close ties with the Cuban president. They've been able to build close ties with Venezuela, and they've partnered with an international communist organization called the International People's assembly in doing so. Who hosted the International People's assembly was the, was Nicolas Maduro, the now removed president of Venezuela. But these are the kind of dictators and communists that PSL and other organizations have been working with over the past several years to make it clear to the American people these are OFAC sanctioned countries that you have leaders of this particular political network working with over the past several years to undermine American interests right here at home.
- [24:14]
B
So, Sam, I would like to move into a few of the muddier aspects of this report with regards to how strong the CCP link is and also where the funding comes from for all of this, which is more or less a black box, we'll start with the CCP angle is the strongest connection here. As you have this New York Times report cited in the report that y' all did, that he works in the same office Singh does as the. The MACU Group, which is Chinese media firm. Is that the strongest angle for us to say CCP affiliation here? Is there anything more outside of that that we can grab on? Is there something. Are there facts that you didn't feel comfortable putting in the report because there wasn't enough of a link? How much can we attribute Singham's, you know, alleged affiliation with the ccp?
- [25:07]
A
In my view, you can contribute it to a heavy degree. Why is that? It's because beyond just sharing office space with the Macro Group, for those who don't know, the MACU Group is a company based in China, that its primary duty is to spread China's message abroad, to be a propagandist for the regime. So Singham's network, not only do they share office space, they've also worked very closely with that particular network. They've paid them about $9 million to help the Singham network do their own propaganda abroad. So there's that. But in addition, we know that Neville Singham has been documented at multiple CCP conferences that talk about telling China's message. Well, that's the CCP phrasing for how to share China's particular story with people outside of its borders. And so there's those connections, but again, there's the critiques that come from the left. This isn't just a critique coming from people in the center of American politics in the right. But Trotsky, people who support Trotsky themselves have said that they see everything that Singham and his network do as supporting the Chinese Communist Party. And it's very clear in all their messaging, in the. In the open collaboration that his network has done with China state media. You know, you have the head of Code Pink, for example, Jody Evans, and the head of Tri Continental, another single network nonprofit, Vijay Prashad. They've done exclusive interviews with China state media over the years. So for all intents and purposes, they are CCP surrogates who. Their primary role, whether it's deliberate or coincidental, is to spread Chinese propaganda here in the United States. But it's not just our own determination that there's a there there. It's also the determination of federal investigators. And this is key. Right now, Neville Singham is being investigated in the Southern District of New York for his ties to China. You also have multiple federal investigations from the DOJ and an interagency investigation into his ties to China. So what those specific findings are has yet to be released to the public. If there is a there there, we'll know hopefully in the next few months. But it seems to be the conclusion of federal investigators themselves that there's at least enough evidence there to bring federal to bring a federal investigation against Neville Sam himself.
- [27:21]
C
And maybe this is getting into how organizations are funded. Like why is there no documentation? Like I know typically if you have, you know, C3s or C6s or whatever, you've got to file forms and disclose financials. Why is there no documentation?
- [27:39]
A
I think that's the most important question of all because with Neville Sangham's 501C3s, there is some reporting that they have to do. But with the Party for Socialism and Liberation it falls just outside of the reporting regime for federal elections in a way that's really convenient for psl. So if you're a political party, you have to report who funds you when you're raising money for a federal election. For example, a presidential election committee and the PSL did that when they ran two different candidates for president vice president in 2024. So that's disclosed who funded those particular campaigns. But what's not disclosed seems to be the day to day activities of the PSL because under U.S. federal election law they don't have to disclose that information. So it seems as if they have been able to identify a particular vulnerability in federal election law where, where they're able to evade certain disclosures for who funds this activity. And that's the most important question of all is who is financing these protests that the PSL shows up at? Who is giving them money to make all these banners to pay people to canvas for them? Nobody can answer that question. Nobody can say with any degree of certitude that it's not China or that it's not Russia or some other foreign actor because they don't need to disclose.
- [28:56]
B
I'm glad you mentioned the canvassing part of this. As you know, we don't know where the money is coming from to actually fund these campaigns. And I would like to get into that a little bit more. Charlie asked a question about it early on, but I'm curious from Yalls research, what are the typical strategies that they've employed? Is it mostly just canvasing? Knocking on doors? Is it just harassing town council meetings? For zoning or for approvals for these things. What have you found have been the most successful tactics for the PSL and its protesting apparatus for getting these things stalled?
- [29:33]
A
I would say it's all of the above. Any activism that the PSL can do, they will do and they have done. So what we've seen that's been most effective are petitions. In particular, as I mentioned, in Prince George's county they were able to gather over 20,000 signatures that they were then able to deliver to the town council. But you also have a certain case study in Alabama that's an ongoing fight that the PSL is waging right now. They've already been able to raise over 3,000 signatures there. And how do they get those signatures? It's the canvassing. They make hundreds of flyers and they go door to door in a very old fashioned style of politicking and they knock on doors, they talk to people, they hand them the flyer. They then proceed to share a lot of misinformation about data centers. That's one thing we've seen in a lot of their social media posts. It's they spread a lot of FUD about water usage and electricity consumption data centers, but there's no one there to fact check them in real time when they're canvassing. And because that canvassing they're able to mobilize a lot of community residents to go to these particular town council meetings and create this illusion of massive protest against these data centers when a lot of this again is ginned up by an organization that has close links to China and is funded by a Marxist based in Shanghai.
- [30:46]
B
So I want to get into the specific areas that they've targeted really quickly. I'm going to rattle them off really quickly for our listeners. You have a number of case studies here. First one Charlotte, North Carolina second Prince George's County, Maryland third Tucson, Arizona four. Chandler, Arizona five. Franklin Township, Indiana six. Madison and Dane County, Wisconsin seven. D. Forest, Wisconsin eighth Athens, Clarke County, Georgia nine. DeKalb County, Georgia ten. Durham, North Carolina and eleven Cumberland county in Fayetteville, North Carolina. 12. Denver, Colorado 13. Monterey Park, California. And then there's also New Orleans, Louisiana, Box Elder County, Utah and Cleveland, Ohio and Milwaukee, Wisconsin and Wisconsin. Statewide. There are two campaigns going on there. Pritchard, Alabama. It's just, it's a laundry list. Why do you think they chose these areas? Did they identify a weakness or a chink in the armor, so to speak? And there was already enough, you know, disquiet about these data centers that they felt that they could move in. There are these places that already had a strong PSL presence, so it was just easier to mobilize. It's really notable to me that North Carolina and Georgia popped up a few times. I'm just curious why. Why do you think they chose these areas specifically?
- [32:08]
A
Yeah, that's a great question. Because as viewers will notice, a lot of those states are red states. They're states that you wouldn't typically associate with these kinds of causes. What we've seen that PSL did is they identified what was, for the most part, organic opposition. And they recognized that that was a great opportunity for them to latch onto a wedge issue and increase that wedge further to foment even greater anger and division among the American people. In fact, that actually is the textbook play of most foreign influence operations, is you can actually camouflage the foreign influence taking place if you're able to embed yourself within a particular movement that looks organic from the outside and maybe even is organic. But we recognize that in every jurisdiction they've been involved. There were a couple commonalities. There was a lot of what looks to be organic opposition, but in addition to that, there was a pronounced PSL presence. What's fascinating about PSL is they have at least 28 what they call liberation centers across the country. These are brick and mortar buildings where they're able to do a lot of their activism. They hold community events there. That's where they make their protest signs. And we did see a correlation between locales that had these liberation centers, which again, who funds them is anybody's guess, and where these data center protests were taking place. Place. So a combination of things. One, they were able to identify an organic movement. But two, there they are taking place in locales where there is already an established PSL presence.
- [33:38]
C
So I'm going to push back a little bit here. I mean, I live in one of the reddest states. I live in Oklahoma. And we've seen multiple data center moratoriums. Even in my own county of Tulsa, Oklahoma, we've seen Oklahoma City, we see a couple of suburbs of Tulsa. We see some other counties or municipalities, like regionally. And by some measures, I've seen that Oklahoma, second to like a part of Virginia, might actually be one of the most, like, moratorium pushed back areas of the country. Oklahoma is extremely red. I would say the reddest state in the union. And my subjective view on the ground is that this is entirely organic and bottom up. These are not typically people who are environmentalists. They're almost like more skeptical of the Government, big tech. And I know you try to be, I know you try to like differentiate. There's the foreign influence and then there's the natural occurring grassroots movement. But to me, in my own backyard, I see all this organic movement having happened regardless of whether the CCP or PSL is involved. Do you think this is a regional thing? Do you think that this was maybe gonna happen anyway and you're trying to shine light on foreign influence because you know, in my own experience, these are red blooded Americans who don't like data centers.
- [35:08]
A
Yeah, no, I think you're 100% spot on. Again, there's a lot of that organic pushback. What's interesting though is the talking points pushing that organic pushback for context. Before I was at bpi, before I was at Treasury, I worked for Riot Platforms, a bitcoin mining and AI HBC company. When I was at Riot, we were building a data center in Corsicana, Texas, a rural part of Texas. As red blooded as you can get, great people, strong conservative values. And for the first few months of our data center construction, there was really no local opposition whatsoever. And then out of nowhere, there was a huge amount of local opposition. And so we started to trace where that opposition was coming from. Coincidentally, it was coming from one particular activist who didn't even live in the town of Corsicana. She lived in a town nearby. And we started to follow the money with her activism. We recognized that she actually had partnered with Greenpeace at the time. For those who don't know, this was during the period that XRP Ripple had paid millions of dollars to Greenpeace to stage a FUD campaign against all things Bitcoin. And a lot of the talking points this activist was using and then spreading to the community on social media, specifically, specifically through Facebook, was that our data center was going to suck up all the water available to the city that would hike their electricity prices. All specific claims that we were able to refute one by one. But what we recognized was by following the money and by simply engaging with the community members themselves, we could clear up a lot of that fud, a lot of those misconceptions. So we held a series of meetings with community residents, including a town hall, where we were able to clear up these misconceptions. It more or less exposed the opposition moment for what it was, because again, these were authentic concerns that these community members were voicing, but they were spread through a astroturfed campaign from an environmentalist group that was funded by XRP Ripple. So it's a very complex web of influence that takes place really in any data center protest across the country. And again, a lot of that is organic, it's authentic, it's real. But a lot of it is fomented by people and parties that have other interests in mind, not the interests of either those local residents or the United States of America.
- [37:23]
B
Sam, I've got two more quick questions and we're running up on time. The first one is kind of a two parter. You mentioned a few case studies in this that actually resisted some of these PSL campaigns. A few jurisdictions that didn't end up buckling under the pressure. Can you speak to why you think those were effective? And as a follow up to that, are there any campaigns that you think might be happening under PSL's influence that are going unnoticed currently?
- [37:53]
A
Yeah, great questions, both of those. With the two case studies we found where PSL did not succeed. The common thread in both of those is that they took place before the national moral panic about data centers was taking place. So it seemed like they had less momentum at those times when they were pushing for that moratorium. It could also be in those specific jurisdictions that there was just a more pro business, pro capitalism community because that seems to provide some measure of resistance to a lot of this PSL activism. Now to your second question. Are there locations where PSL is mobilizing that we do document in our report? Yes, there are. One particular place is in Minnesota. We know that in the Twin Cities PSL is pushing for a permanent data center ban. We just weren't able to document that in our report because we're limited in our time and resources. And we believe that that's just one example of among many, Again, there are 28 liberation centers which are the PSL headquarters based across the United States. And the common thread throughout PSL's activity nationwide seems to be an anti data center push. So I think even now outside of our report and outside of Minnesota, there are likely other places where the PSL is mobilizing against data centers.
- [39:09]
B
All right, last question. We'll get you out of here. You mentioned that lawmakers are starting to take action. Believe you said New York is pining them or attempting to investigate his ties to China. What do you think will come of these? Can you give us a brief rundown of exactly what's being done within Washington to investigate these activities with the PSL and Singham and whether or not you think this will actually bear fruit?
- [39:35]
A
Definitely, I think it will bear fruit. And Washington has awakened to what's happening right now. And that's really encouraging. So a couple things have taken place in the last month alone. Number one, the House Energy and Commerce Committee wrote a letter to the White House and the FBI actually citing our report and our findings and requesting that they investigate the Singham network and their potential involvement with China in their anti data center activism. In addition to that, I mentioned earlier that the Southern District of New York, which is if anyone's seen the show Billions, that's where the Show Billions takes place. It's with the Southern District of New York and the U.S. attorney who works there. It's one of the highest profile jurisdictions for white collar crime in the country. So if you're getting investigated by the Southern District of New York, you are with, you are in the crosshairs of federal investigators. They're looking right now into Singham's ties to the ccp. In addition to that, you have Tom Cotton asking the DOJ for particular investigations. So federal policymakers are very aware now of what Singham's doing and specifically how he's targeted data centers. And so that's heartening. What I think will take place in the next few months is we're likely to see either important findings from these investigations, whether they come from subpoenas or just from discovery. In addition to that, I think you're likely to see the Congress itself, either on the House side or the Senate side, hold hearings into foreign influence within the data center campaign. I also think this is sort of new and developing. You're also likely to see particular Russian involvement in the anti data center fight. We spoke just this morning with a reporter from a very well known newspaper who said he was investigating the Russian influence aspect of these things. So there's more to come, but I think more and more our research will be vindicated as the months unfold.
- [41:24]
B
Well, to get you back on, if we have a testimony or anyone testifying on this, and who knows, maybe you will be sitting in a chair there. But Sam Lyman, thank you so much for joining, man. Appreciate you coming back on. Great work with the research and, you know, we'll reach back out if there's something to touch on on this angle.
- [41:43]
A
Sounds great. Thanks, guys. Great being on.
- [41:45]
C
Thank you, Sam. All right, we're going to keep on rolling with our guest interviews. We've got Tom Masier in the wings going to talk about Sphere 3D with him. But before we bring Tom up, a word from our sponsor, Luxor.
- [42:05]
B
This episode of Blockspace Live is brought to you by Luxor's commander, Bitcoin Miner Management Software for enterprise operations. Luxor's Commander gives you real time fleet monitoring bulk remote commands across your fleet. And Intelligent Miner, that's an automated profitability engine that runs every five minutes and tests your fleet settings against live hash rate and power markets. Ercot backtests show 10% improved profitability with intelligent mining versus binary mining. Commander Pro is a hundred dollars per megawatt or a 25 basis point pool fee adder and you can try it for free for 60 days. If you want to learn more, go to Luxor Tech forward slash Commander.
- [42:46]
C
All right, let's bring on up friend of the show, Tom Osiero. Welcome.
- [42:52]
D
Hey guys, how's it going?
- [42:54]
C
Great. Good to see you.
- [42:56]
B
Good to see you. Farmer Tom, you missed the communism segment,
- [42:59]
D
but I caught it. I caught the end there.
- [43:04]
B
But anyway, we're not here to talk about communism. We're here to talk about Sphere 3D. And Tom, I'd love for us to open up with this dark horse rebrand that y' all are proposing. Can you give our listeners a little bit of info and background into that and, and what is leading that push or why y' all have decided to push forward this reboot?
- [43:26]
D
Yeah, I mean, it's a, it's proposed right now. So our board approved this, I believe, last week sometime. And it goes to a shareholder vote, I think towards the end of this month or. Yeah, end of this month, and gets officially either voted down or, or not sometime in early August. I mean, the big thing is we're a merger of two companies, pretty much a merger of equals. Both companies have, you know, their own history, their own identity. And it became very clear once we became, you know, one company that we're going to need a new unified identity. And especially with kind of the timing of, you know, AI compute, kind of converging on sort of our segment in the market in terms of the size sites that we, that we own and operate. You know, this idea of like dark horse became sort of like, I don't know, it kind of got incepted and I think there's, you know, there's something to it as far as being like a symbolic aspect of doing things a little bit differently than maybe some of the other folks that are out there. And yeah, we're, we're really excited about the process. Working with a design and agency to kind of go through all these things.
- [44:37]
C
Yeah, working with a design agency. I mean, we saw what Keel did, the rebrand, and then they've got Ben Gagnon, like switching these. Captain Keel now, you know, Dark Horse. Do you have any logos in mind? I, I love a good logo and visual scheme.
- [44:53]
D
Yeah, we got a bunch of stuff. I actually teased some stuff yesterday. I will not be, I'll just, just to get out any controversy of you. I will not be rocking a love boat, Captain Hap. So we'll just put that to bed. But yeah, we're working on some schemes. I think the icon we'll see if it ends up playing out is, is based off of the, the chess piece, the knight. And you know, that's our way of saying we're gonna maybe like move on the board a little bit differently, but we'll see how it all ends up playing out.
- [45:25]
C
Love it. I'll tag in, Colin.
- [45:28]
B
Well, speaking of moving on the board, I mean, what does that look like with regards to an AI pivot? Tom, we've had you on a few times to talk about this. You know, y' all have got 53 megawatts operating 100 plus megawatt pipeline, which includes AI and HPC. After a week where anthropic deals are repricing some of the majors. Could you recap the potential for AI and HPC rollouts for a miner of Yalls scale? Because I think a lot of people look on it and think that it has to be a hundred plus megawatt data center.
- [46:01]
D
Yeah, I mean, we've talked about it the last couple times I've been on a call with you guys. It's happening very quickly. There are a lot of folks who are very interested in inference right now. So you've got this inference movement, you've got the agentic stuff that's happening. And so we're going through the same process that I think a lot of our larger peers have gone through, which is, you know, we have to set up new data rooms. And the data rooms are very different than, you know, a traditional bitcoin mining, you know, asset would look like, you know, they're, they're trying to make sure they get, you know, at the, every layer that they possibly can get. You know, you know, who owns the land, are there, you know, chemical issues with the land or any of that type of stuff is, is kind of at play. And then we're just talking through the four essentially legs of the stool of what makes an AI compute project work. First of all, it's infrastructure. Then you got your power and land, then you got your offtake and then you got financing. And so, you know, we're methodically trying to solve each one of those and find the right partners and to, you know, try to, try to take it down. It's just one of those things where it's, it's an execution thing. A lot of peers have been able to do it. It's not like, you know, we're not trying to go to the moon, you know, for the first time. It's, it can be done and it takes a lot of work. I can tell you I've never worked this hard in all of my years of bitcoin mining than I have in the last three and a half months in terms of, you know, just getting up to speed, trying to figure out how this all works together. You know, typically a bitcoin mining play, it's like you find a customer, you, you, you handle all the build out the structure all yourself. So the, it's usually very simple deals and not, not very complex. Whereas with these, because the capex is so expensive, you're talking 10 to $15 million a megawatt for infrastructure and then an additional 30 to $40 million a megawatt for GPUs if you're on the higher end. So you have to have like, you know, you've got to have the financial engineering side of things down. And very thankful to work with with Joel Block, our CEO who has a bunch of experience working with the team at that is now HUD8 when he was at US Bitcoin and feel really good about where things are going.
- [48:18]
C
So we've reported on a few, on a few pieces from semi analysis over the past week and notably Meta announced their cloud operations. If you saw that and that caused almost like a, it felt like a deep seek moment for like neocloud just compute in general. This had a lot of people calling like GPU glut, GPU compute constraint that out the door. But yet at the same time Alice says Meta is actually increasing above their initial projections earlier this year and at the same time this morning Semianalysis says Anthropic's on track to become a $6 trillion company. How do you think about this in terms of like you're building sites, you're setting up deals, like you're trying to think about enterprise value, the company and strategy over the next two, three, five, ten years.
- [49:14]
D
Yeah, it's really like you just got to ignore the noise on all this stuff. There's a lack of power ready capacity available for compute right now and that's very clear. The fact that Meta doesn't have the ability or that Meta is sort of positioning themselves in this way is, I mean this was sort of teased like a month or so ago. It wasn't even like the news Mark had mentioned it prior. And it's no different than Elon. Well, I would say the two distinguishing differences is with Elon, he already built, he already has his own infrastructure. And so when they realized that their model training wasn't being as fruitful or profitable as they thought it was going to be, it was very easy for them to go find customers. And I think with Meta, because they're not actually like owning a lot of their own rack space, they realize that maybe, you know, hey, our model training isn't getting where, where we want, so we might as well make sure we make, you know, we make the best use of these investments of the, the GPUs that they have and they're smart for doing it. And there's still plenty of compute that's, that's needed. So I don't think we fully understood how much compute people are going to need moving forward. And just, you know, I, I've talked about this like long tail theory of the Internet. You know, that's what made Google so profitable is like when you search, you can literally just go on and on and on and there's just niches upon niches. I think the same thing is going to happen with, with compute. Different verticals, different subcategories. You know, I've seen pictures of Chinese cheap chicken shack AI containers already being distributed to like small sites. And they're, they're all, they're all targeting like, you know, very small inference plays, but it's like happening, you know, at breakneck speed. And that's very different than maybe, you know, trying to tailor up something for, you know, a B300 deployment where you're talking, you know, 30, 40 million a megawatt deployed. So those folks have customers too. And I just, I don't see that stopping.
- [51:21]
B
Tom, quick question. Going forward with Dark Horse rebrand and whatnot, so obviously Sphere and Cathedra just combined. Is sphere 3D at this point a buyer of compute, or rather sites? Would you consider buying up more sites if the opportunity presented itself, or is there an exit strategy of being a target for an acquisition in the future?
- [51:50]
D
No, I would say right now we're evaluating every single site and profile that we have in terms of the power that we have, the fiber capability, all of that stuff is at like on a one by one basis and then determining what type of compute is the best and then figuring out, okay, which site do we want to focus on first with whatever deals that we have coming in because you know there'll be multiple deals that were were being shown and then we're trying to figure out and, and, and actively working on is this one should be slotted for this site, this one should be slotted for that site. And we do have sites that I would say are less attractive for you know, I would say more valuable compute but they absolutely would be for lower end compute. And I think that's one of the, the good things about having a lot of or just more than one site is you can have some optionality in your business model. I think I heard Mike Alfred talk on a spaces a couple weeks ago where he was like I, I like the fact that you know if you're, if you've got multiple sites you don't have to have the same exact business model. Rinse and repeat over and over and, and that way you could maybe blend in your mar. You were getting a little bit more access to compute. So yeah, we're looking at all those things.
- [53:04]
C
Yeah, I might double tap on that. Like we compare mining deals from the distant past three years ago and there's the prop mining, there's like the fixed margin mining. There's kind of a mix of that. What do you think some of the good optimal structures for powered land infrastructure provider to hosted client are like? What do you, do you think that like certain deal structures might be be better than others or preferable at this time? I'm curious your thoughts here.
- [53:43]
D
Yeah, I mean it really depends the ownership. So if you're like if you're leasing a site where you currently have an existing power contract that's going to dictate some of the terms that you would be more open to. If you own everything outright, obviously you have more leverage. I would say that I think I, I think that like Marathon's structure that they did with Starwood where they contributed their power and land at a premium into the JV for a really large, you know, potentially would be a really large offtake client is a, is a really simple one and as as clean as it could get and I think it aligns incentives for bigger projects. I also think that the financing side hasn't been completely ironed out with this segment of our market. Meaning yes, if you're getting some IG backed, you know, Fortune 500 company to be your off taker, it's easy. You can get your 15 or 20 year lease and but that's not most likely going to be the segment engaging with these sub 50 megawatt sites. A lot of labs, startups, et cetera. So there's going to need to be some creativity around it. I would think though that there's going to be some type of financial innovation or I don't want to say financial innovation more so there are going to be wrappers developed that will allow. Because at the end of the day all of these off takers, they want to be able to scale more. So they don't just want our 50 megawatt site or 20 megawatt site. They want to say where's a path to where we can continue to do more of these? And I don't think it's like the bottleneck isn't the, the power isn't there or there's opportunities not there. But like how do you quickly do that? And you've got to solve those four, you know, legs of the stool is making sure there's infra. Up until just recently in the last like month or not even a couple days, you saw Giga come out with their, their water or their, you know, more of like their upper end GPU compute infrastructure. Armada's got some stuff. Schneider Vertiv, there's a company called Orbital Industries that has them as well. They're all sort of now just deploying, you know, this upper end kind of deployable stuff and then there's lower end stuff that I was telling you about before. So at the end of the day though, it's still very expensive and there's going to be this mix of debt, credit or debt and credit as well as equity into some of these deals. And I think it's just going to, you know, determine at the actual site owner level how you want to participate. You just want to take cash and go or do you want to kind of ride along? You know, I think that there will be a lot of opportunities for that.
- [56:21]
B
Tom, it's great to have you on man. As always. Thank you for, thank you for joining. Hey, we'll be keeping an eye on the dark horse. Rebrand will no doubt have you on again sometime soon as developments continue to hit the wire for sphere 3D, potentially dark horse. So Tom Massero, thank you for joining man. Hope you have a good week.
- [56:42]
D
Appreciate you guys. Thanks so much.
- [56:43]
C
Thank you. All right, we're gonna keep rolling on. We have next Amazon 25 billion dollar bond sale. But before we go to that, a word from our sponsor. LY.
- [57:06]
B
Hedge funds are getting liquidated. Is your bitcoin safe? It's not just bitcoin's price drying up. Lending desk and Hedge funds are reeling with bitcoin just above 60k. If you're using a Bitcoin backed lender it's more important than ever to know who actually controls your Bitcoin. And with Lygos Finance that is always you. Do yourself a favor, don't become the next Celsius victim. Check out Lygos Finance if you are collateralizing your Bitcoin for a loan because they use Bitcoin native smart contracts to make sure that you always control the keys to your kingdom and no one else. With Lygos Finance you're always in control. There's no bridging, no rehypothecation, no wrapping, just self custodial Bitcoin backed lending and rates as low as 8.5 to 10% APR. Go to Ligos Finance to learn more. All right, got a Mag 7 bookender today Charlie. Yeah, specifically Amazon announcing a 25 or not announcing actually this is a Scoop. According to CNBC, Amazon plans to raise 25 billion in bond sale, but it's saying that it won't issue more debt in 2026. This marks the company's latest debt raise as it looks to buttress its massive investments in artificial intelligence. Quoting here from CNBC, Amazon plans to raise at least 25 billion through an eight part bond sale as it looks to continue its massive artificial Intellig, according to CNBC's David Faber. The company has also shared with its underwriters that it won't issue any more debt this year, according to people familiar with the matter who asked not to be named because the details are private. Amazon disclosed plans for the capital raise in a filing with the SEC on Tuesday, but it didn't disclose the dollar amount. This debt sale comes after Amazon raised roughly 54 billion in bonds earlier this year in the US and Europe, followed by a 10 billion bond raise in Canada in June. Amazon also raised 15 billion from a US bond offering in November. Take it while you can get it. That's my take here Charlie.
- [59:11]
C
Obviously the market's hot. Yeah, the market's hot. Grab it. If we're all hungry hungry hippos and bonds and ways to raise money are out there, go and grab it. I think Nvidia just did a $20 billion one of these. And then I mean for context. So like Amazon did 131 billion in capex spin last year and that they're on Track to spend 200 billion, meaning that the entire like mag7 sector is gonna blow past, I believe 700 billion in AI spend just this year in 20, 26. So the money's out there. Go grab it. Like didn't Google do one of these too, like another $20 billion few weeks ago? So you know it in the grand scheme of things, 20 billion might not be big compared to their overall 200 as well as 10%.
- [60:06]
B
But yeah, Google was an equity sale, $85 billion, which is pretty nuts, honestly. And I want to confirm this really quickly before I say it or I'll confirm it after. But according to Bloomberg reporting. According to Bloomberg reporting, investors sold existing hyperscaler bonds on Tuesday to make room for the new Amazon paper, which kind of which dragged down debt outstanding, attack debt outstanding and kind of signaled some fatigue with all of this AI financing. Yeah, which is pretty incredible when you think about, for instance, we covered SpaceX's debt raise recently at also $25 billion bond and it took a haircut on the first day. And that was largely because there were more speculators in that trade than not people expecting to flip it almost like the IPO. But it was down, I believe north of 300 million. Those bonds took in aggregate, took a 300 million haircut. And that's pretty incredible. Specifically the longer dated bonds, the ones that were 30 years out.
- [61:24]
C
Yeah. It is kind of weird that it seems like now bond sales are perhaps the subject of speculators and quick flippers. Maybe I'm naive and didn't notice that those were a thing before, but that seems pretty surprising to me. Tend to be like, you know, bonds, while a liquid market does not seem to be something that you'd want to like buy and flip the next day. So that Weimar vibes just, just gonna put it out there, you know, little,
- [61:59]
B
little, little Weimar Republic, little, little Weimar shuffle going on. That being said, credit cycle still in full swing. We've been covering private credit on the show for the last year or so and there are questions floating around now about how much systemic risk these private credit instruments could actually pose to the rest of the economy, especially if they collateralize other loans. But right now, parties not stopping. There are still chairs to grab if you want it. But 25 billion from Amazon, huge debt raise. And we will be keeping an eye on just what happens with this and if they upsize it and how much, how much interest there was in it because that was currently not reported. But we will follow up on this as new information comes through the wire.
- [62:50]
C
Yeah, I do get this vibe from all of the raises here. Grandy saying, I didn't hear no bell. So this is kind of what I see with a lot of the, you know, these bong tales, which maybe, maybe the fight's gonna keep going on. But, you know, music hasn't stopped yet.
- [63:15]
B
You know, a lot of debt floating around.
- [63:17]
C
Yeah. Why would you think about sitting down? Anyway, thank you so much for listening to Block Space Live. We go live Monday, Tuesday, Wednesday, Thursday, Friday, every weekday at 1pm Eastern. If you like what you hear, you'll love what you read at newsletter Blogspace Media in your inbox every single day. If you missed the live stream, it turns into a podcast shortly after. Wherever podcasts are found, Blockspace Live is presented by CleanSpark. NASDAQ listed ticker CLSK. Thank you, Clean Spark. I'm Charlie.
- [63:50]
B
I'm Colin.
- [63:51]
C
And we'll see you tomorrow.