Ivan on Tech analyzes Bitcoin's position, Ethereum's new roadmap, and AI market dynamics
Ivan on Tech solo Bitcoin and crypto market analysis stream covering macro conditions, Ethereum's development roadmap, Solana's growth, and a colorful Silk Road story.
Summary
Ivan on Tech presents a wide-ranging solo analysis stream covering Bitcoin's current price action, macro conditions pointing toward a Fed pivot, Ethereum's newly announced development roadmap, and broader AI market dynamics. He argues that Bitcoin is currently within a buy zone but carries meaningful downside risk to the $40,000–$50,000 range before a Q4 cycle bottom. On Ethereum, he gives qualified praise to Vitalik Buterin's "Lean Ethereum" roadmap — strongly endorsing the privacy and quantum resistance priorities while questioning whether the three-to-four year timeline and researcher-driven process reflects actual developer and user needs. He also raises pointed concerns about quantum computing as the single largest long-term risk to Bitcoin, one he argues Michael Saylor conspicuously omits from his own risk analysis. A Near Protocol ex-core developer's claim that Vitalik's roadmap replicates features Near already shipped (quantum safety, sharded state, WebAssembly) is cited by Ivan as evidence of Ethereum's "lost decade." On AI, Ivan highlights Palantir CEO Alex Karp's critique that OpenAI and Anthropic's model is not scalable, contrasts the struggling end-client AI companies (Meta, Nvidia) with chip manufacturers (Micron, AMD) that are still in bull trend as the last-man-standing beneficiaries of AI capex, and argues a significant correction in AI company valuations is coming — with resulting capital likely rotating into crypto.
Key Takeaways
FULL TRANSCRIPT
Bitcoin Price Action and the 200-Week Moving Average
Ivan on Tech: Bitcoin is doing something very interesting right now. We have a very nice pump back towards the 200-week moving average. We are now actually above the 200-week moving average, and we still risk this being just a retest — because we broke down from here, we went down, and now we are retesting. There's a big risk for Bitcoin that this is going to be a rejection down towards the 50s, down towards the 40s — very, very nice levels down here.
But at the end of the day, we are so deep into the bear. We are so deep into the bear hole, so to speak. We're swimming in it. Could this be the bottom for Bitcoin? It's not impossible. That's why we have this buy zone — this green, beautiful buy zone right here. If you DCA here, it's fantastic. We're within the buy zone, but there's still elevated risk that we go down further into the 50s and into the 40s. And that's happy days. That's more discount. That's exactly what we love.
Macro Conditions: Jobs, Oil, and the Fed Pivot
Ivan on Tech: Something is happening on the horizon. We do see the job market getting very, very weak — the first weak job market in a very long time — such that the Fed actually went out and said that inflationary problems, inflation risks, have diminished, and we may actually see deflation. If this continues, we may actually see deflation, also with AI and the growth of AI and how AI affects the workforce. It's very interesting to see how the Fed is changing in real time.
What happened was that the new job numbers came out worse than expected. The job market is not doing that well. But we live in an economy where bad news is good news, because when the job market is failing it means the Fed has to come in, they have to print, they have to ease, they have to support the market — and that's bullish. Bad news is bullish. You need to understand that. That's the first thing many people don't understand. How can bad news be bullish? It's because the Fed is going to step in.
Kevin Warsh came in and said we need to be careful, we cannot just cut, we need to be serious — that was a few months ago. Now that he has established that credibility, that he's not just going to go nuts cutting, that he's actually responsible, he now has this trust capital, so to speak. The markets trust him. He has this credibility capital that he can now spend cutting. Very, very simple.
Inflationary risks are coming down. Payrolls are the weakest in four months — very, very important. Also, you have to realize that 720,000 people withdrew from the workforce. They didn't even try to find a job anymore. They said there's no job, and they sat at home. That's the situation we're in right now.
All in all, you have the deflationary pressure mounting quite nicely. The Iran conflict is ending, the ceasefire is holding for now at least. Oil prices are coming down. More oil supply will be released onto the market in August from OPEC — very important. Oil price is below $72, holding up quite nicely after this dump in oil.
You remember when the conflict was happening and there were all of these ceasefires — there was a ceasefire per day — but the oil was not coming down. We said don't trust these headlines that the conflict is ending. The conflict is not ending. If oil doesn't come down, it means the markets are not buying the ceasefire. Now oil has come down. So the markets are buying the ceasefire. They are saying that likely this ceasefire is going to hold.
Energy shocks are the most mechanical form of inflation there is. When the shock reverses, as we see now, the inflation reverses with it on a lag. So just keep it in mind — we now have deflation pressure, oil price low, and core PCE pass-through follows with a two-to-three month lag. When you look at inflation, we'll likely see within the coming months a big fat inflation drop. Because this is the lag we speak about. When things change with oil, we will have a few months where no change is seen in inflation, and then — bam — months three, four, five, inflation is going to drop a lot.
And guess what? This corresponds with our Q4 thesis for Bitcoin bottom for the next cycle. It all plays out so nicely. If you don't over-complicate it, it plays out so nicely. We're now in a position where Bitcoin is down a lot from all-time high — 50-plus percent. We can buy here cheaply. The new picture, the new dawn, is being painted in front of our eyes. Literally Bob Ross is painting the new dawn for Bitcoin that is going to happen within the coming months. The Fed is on board. The Fed will have to cut bigly.
The Anatomy of a Fed Pivot
Ivan on Tech: Here is the anatomy of a pivot. First we had hawkishness. When Kevin Warsh came in, he was hawkish. He needed to establish himself as a serious person who is not just going to cut because Trump wants to cut. That's done. Next, you have very bad payrolls — that's done. Next, you have rhetorical softening. You have to be soft, silky smooth in your rhetoric. And this is what's happening right now. The Fed is saying, listen, inflation risks have moderated, all good, no need to cut — and the markets are adjusting in real time. The likelihood of a rate hike has decreased after this dovish statement.
If you look at the prediction markets, the odds of a rate hike are going down. Next is reframing and cut. He's going to say policy is sufficiently restrictive, and then in a few months, big fat scissors are going to come out and they're going to cut very, very nicely.
Does this mean we may not see the 40s or 50s? I don't think what's happening with the Fed affects our downside in the short term to the 40s and 50s, because this is more of a multi-month road map towards Q4. Always, when people look at macro, they become so impatient. People were saying the Fed is going to pump us all the way back in October when we said to be risk-off, sell Bitcoin, sell altcoins, be very, very careful in Q4 in October. Many people said, "Oh no, but the Fed — the Fed hasn't even started to cut." And listen, they're correct. It just took another year, and your coins are down 99%. Bitcoin is down more than 50%.
All of the macro guys are going to say they were right because they said in October that the Fed is going to cut, and then the next October they cut. So they're always going to claim credit. That's why following macro guys is very dangerous. Following pure macro guys is the worst thing you can do. You need to follow the price trend for trading. But with macro stuff, you need patience. Patience.
So all of this thesis with the Fed pivot — it is fully correct. But it's not going to be tomorrow. It's not going to be next week. Likely it is towards Q4. This is just a reinforcement that we're here on a path within the buy zone. In Q4, expect us to reset the cycle and then have a very, very nice 2027 for Bitcoin. 2028 very nice. 2029 potentially nice — but be careful in 2029. That's one message to future self: in 2029, be careful. No matter what the macro guys are saying, be careful.
The funny thing is, in 2022 we said we're going to sell at the end of 2025, towards the summer, towards the autumn. We said that's the danger zone already in 2022. All of the OGs in the chat remember it. We said the danger zone at the end of Q3, Q4 2025. And it was simple — just follow it. The trends turned at that time as confirmation. So if you don't over-complicate your life, you're going to have a very happy life in all regards, not only in markets. In 2022, we said end of 2025 we need to be careful. Now we're saying in 2029, no matter what your favorite macro influencer is saying, you need to be careful.
Cathie Wood and Raoul Pal Turning Bullish
Ivan on Tech: Someone who is always instantly bullish is Cathie Wood. She now has a new chart. She pulled it out of somewhere — I don't know where — but she's always having bullish charts. Even right before we go down 90%, she always has a bullish chart. But in this case, she's correct because we're going to go up starting in Q4. Let's check what she says.
Cathie Wood: Here's Bitcoin to gold. Gold is going down. Bitcoin had been going down. It seems to be in a bottoming process. We believe that it has bottomed on this basis and will resume the very volatile but broad uptrend that you see on this chart.
Ivan on Tech: Bottom out. So Cathie, would you say that we have bottomed out? At least against gold. But with Cathie Wood, you have to be careful. We don't really listen to her for what to buy or sell. She's not too good at that. And her incentive is also very different from yours. If Bitcoin goes down 50%, she doesn't care. If shitcoins go down 50%, she has the fund. Her fund, by the way, has very bad performance if you compare it to the S&P. They say they are innovation, but they haven't really been a part of AI. So yeah, be super careful with Cathie Wood. But in this case, now that we're also turning a bit bullish and we are in the buy zone, why not?
Even Raoul Pal — I mean, even Raoul Pal. We're going to be this close with Raoul Pal because now I'm bullish in the buy zone. He has macro charts that are yet again saying we're going to go up, which he always has had. He now has another one — correlated global liquidity. What happened to it in Q4? There was so much liquidity. We were drowning in global liquidity in Q4. But we still went down 90% for alts, Bitcoin 50-plus percent, with max global liquidity. Anyway, he has another thesis that we're going to go up basically. Find me a historical correlation — here are two overlapping lines. Nice.
Now Raoul and I are so close. We're the best. She's bullish, I'm bullish. That's the problem with macro guys. You have to be so, so careful.
Michael Saylor's Bitcoin Risk Article — and the Missing Quantum Risk
Ivan on Tech: Michael Saylor wrote a big fat article about Bitcoin. You can read it. It's kind of half AI, kind of half not AI, but it's a big thing. Let me go down to the risks, because the bullish case everyone understands — he's bullish, I'm also bullish, everyone is bullish now. So let's discuss the risks.
The biggest risk is not that Bitcoin disappears. The biggest risk is quantum. But he's not saying that. So it's very bad. He needs to be saying quantum is the biggest risk. The biggest risk he lists is that bad ideas compromise it, that custodians obscure it, leverage distorts it, or that political actors attempt to control it. And quantum is an even bigger risk.
The first risk he lists is protocol corruption — Bitcoin's monetary integrity depends on hard consensus, changes to the base layer should be rare. Okay. But is he pro the data increase or not? Is he pro BIP 110 or not?
Second risk is paper Bitcoin. Well, he is the paper Bitcoin. MicroStrategy is paperized Bitcoin. What do you mean paper Bitcoin is a risk? It's like this interview we watched last Friday where he's speaking to Channel 4 and he's saying, "Yeah, Bitcoin is the new money for a new generation. New generation. They don't trust the banks. They don't trust any company." And the interviewer, a smart lady, says, "But you're a company." What do you mean don't trust companies? You're a company. The same thing is here. The risk is paper, man. You are paper Bitcoin. You do all kinds of debt and instruments in paper Bitcoin. If intermediaries create more claims to Bitcoin than actual Bitcoin, the market will suffer.
The third risk is custodians — and you are the biggest custodian, man. You have the biggest bag of Bitcoin and you put it in a custodian. So you help with that also. The fourth risk is regulatory capture — they can regulate exchanges. Yeah, they do that. The fifth risk is fee market uncertainty. Yeah, that's true. Who the hell knows how miners are going to get paid long term with fees not really going up and the block reward going down.
But what about quantum? No mention of quantum. Quantum is the biggest risk. It's the biggest risk to inflows. We discussed it last week a lot, because if quantum is not solved, what happens? You cannot put money in Bitcoin long term. Simple. You can buy gold. Gold is a pet rock. If you have gold, your kids in 50 years have the same gold — gold is gold. Bitcoin in 50 years without quantum? You simply cannot say whether it's going to be okay or not. That's the ugly truth. They need to fix quantum.
Listen, I'm the biggest Bitcoin bull. I've been here since 2013, full-time crypto since 2017. I'm telling you — thinking long term, without quantum resistance, I cannot put my money long-term in Bitcoin. I'm a computer scientist. I know what I'm speaking about. Most people who are so sure that quantum is not a problem are not computer scientists. They just listen to a Bitcoin podcast where some dude told them not to worry, and they don't worry. And then they try to educate me — just like all of the macro guys tried to educate me in Q4 last year that it's not the top. I told them it's the top. Non-technical guys tell me, "Oh no, Bitcoin quantum, it's not a problem." What do you mean? Where are your credentials? Are you a computer science master's degree? What are you speaking about?
So I think it's a big risk. Being a technologist myself, I would not put my money in Bitcoin long, long term. For the next bull market, yes. For the coming maybe decade, yes. But long, long term, if they don't get their act together, they need to move as soon as possible.
Vitalik's "Lean Ethereum" Roadmap
Ivan on Tech: Let's move on from Bitcoin to Vitalik, because Vitalik has been making some moves this weekend. While people were watching football, Vitalik was not watching football. He was using his brain — his wonderfully alien child brain — to figure out a new road map for ETH. Very interesting. I have a lot to say about this as a technologist with many accolades. Computer science degree verified.
So what happened? Two weeks ago, Ethereum researchers met in Berlin. The first thing I hate is this word "researcher." I hate when he says "Ethereum researcher." What do you mean? Elon Musk recently eliminated the job title "researcher" at his AI startup xAI. He orders that all technical staff be called engineers, stating that traditional academic distinctions are pretentious. They think, "I'm a researcher, I'm smart." No, you're not smart. You have no responsibility. You don't build anything. You just write some stuff and create unnecessary hierarchies. Musk's stance on technical roles stems from his approach at xAI and SpaceX. He has expressed that calling someone a researcher rather than an engineer implies a two-tier system. Instead, he believes that true innovation comes from practical, hands-on engineering that delivers tangible results. Exactly. So that's the first feedback to Vitalik — remove this "researcher." There is no high priest. They are not super smart anyway.
They met in Berlin to continue charting the protocol's long-term trajectory. You have a bunch of researchers coming out charting the protocol's long-term trajectory. How many DEX teams are involved? How many decentralized exchange teams are involved? How many lending protocol teams are involved? How many users of the chain are involved? How many wallet teams are involved? You understand the problem? A bunch of researchers are deciding the long-term trajectory. Just like a few years ago, they decided to do L2s. Why did they decide to do L2s? Because they're researchers. Because they're smart. Because they are researchers.
I actually commented: hopefully these requirements are really needed by dApp devs and users before committing three to four years. Because basically he then goes into what they're going to build, which is — I mean, it's not bad. I'm just not sure it's what's needed. But then he says this is going to take four years.
You remember a few years ago there was the merge, the purge, the splurge — all of these purges. There was like a five-step road map with the merge as number one. They no longer do that. Instead they have a new road map. But I think they're making the same mistake here, because how many prediction market teams were here? They say that some client teams were involved. Well, the client teams are not dApp developers — they're implementing the ETH protocol. ETH researchers do the road map and then these guys implement it. So having worked in tech for a very long time — soon ten years in tech, both building tech and building startups — I can tell you that when you build a startup or a business, the number one problem is building something that no one cares about. Because they haven't involved any users, and it's not clear exactly what the target is.
What the New Ethereum Roadmap Actually Contains
Ivan on Tech: So what are they going to do? This is my favorite part. Privacy is no longer an afterthought — it's a first-class goal. This is my favorite. This is my favorite. Thank you very much for doing a smart thing with privacy. That's very nice. And quantum safety has shifted up a lot. That's good. I fully agree. Hats off. For real, the best part of it — privacy and quantum. Good. This is the best. Everything else, I don't know if they checked with anyone. They're going to do all kinds of other stuff here.
But the conclusion is that less is going to be on L2, more is going to be on L1. That's it. But privacy is very good.
For those who don't speak Vitalik — because yeah, he speaks, but actually here he explains in a very nice simple way. Normally he writes these blog posts no one can read. Everyone just pretends they're smart enough to read them. But this was pretty clear technically. Basically: faster, quantum resistant, and private. This is good. This is good. Faster — I guess good, but how are you going to compete with Solana? I don't know. Is it going to be at the level of Solana? Let's see. But private, guys. Private is good. I like private. I love quantum. Those are the two things. And hopefully they checked with the devs that the way they implement it is actually what's needed.
The next Ethereum upgrade will rival the merge and take three to four years. That's the only problem, man. Three, four years. Why does it take so long? How many vacations do you take? Elon Musk — look at the SpaceX progress in three, four years. He made all kinds of rockets in three, four years. xAI wasn't even around three, four years ago. We have such long road maps. Now, to be fair, you need backwards compatibility, a lot of money is at stake. So I guess it's good. But imagine Solana in three, four years. Imagine all of the other stuff in this whole industry in three, four years. What's going to happen?
Vitalik describes what he's calling "Lean Ethereum" as the blockchain's third major revolution, saying almost every major piece of the protocol will be replaced. I'm excited. I think it's good. I think ETH needs to do something. They have a big problem. The branding is that it's like — Solana is fast moving. We're not fast moving. Why? Because we're kind of a store of value. We're money. But guys, Bitcoin is money. Be real. ETH is not money. If you want to compete with store of value, you compete with Bitcoin. Just realize that. That's also in terms of product thinking, in terms of competitive analysis — ETH ecosystem, they're very bad at it.
Okay, you say that you are so responsible, so slow, for a good reason, because you're the grown-up. Solana is the kids. You're the grown-up. Okay. So what's your KPI? Your KPI then is store of value. Okay, store of value — you lose it fully to Bitcoin. What else are you? Are you for institutions? Because no institution will touch Solana? Literally all institutions do only Solana. This new stablecoin by institutions is natively on Solana and on Stripe's chain. Tokenized stocks — they're on Solana. Robinhood's chain has their own stock. So it's a big problem.
That's why I think they need to do something. But they need to also be clear what KPI they're going after. Because privacy could be one — a fully private smart contract chain. That could be one. If it's a store of value, you're competing against Bitcoin. You're cooked. It's like trying to vibe-code Amazon. Are you going to vibe-code Amazon's marketplace? No, you're not. Then what are you? Are you a smart contract platform? Solana is way better in all directions — a million times better. Everyone is there. Everyone is doing coin making, money making, tokenized stocks. And then Solana itself has issues competing with Hyperliquid. So how are you going to compete?
I think there needs to be some product thinking there. And I'm excited that they're doing something — that's good. I love privacy and quantum. That's it. Everything else is a big question mark. But I would love ETH to be fully private. That would be amazing. Maybe that's the niche. Really find some niche. Maybe private. Let me know what you think, especially if you are a licensed product person — meaning you've worked as a product manager, you have experience building technical product — because that's the biggest question: how do you position it?
Technical Feedback on the Ethereum Roadmap
Ivan on Tech: A start-up guy has some feedback. His take on the new road map — TLDR: many good things, a few problems. Number one, the good: recursive STARKs. As soon as he sees STARK, he gets excited. Huge progress since the early days. Okay, good. STARKs — I don't know. Do you build a DEX with it? How many DEXes do you have? How many prediction markets? Where is the perpetual DEX? Where are memes? Where is everything people do on chain nowadays? These tech guys are so focused on their thing. STARK in and of itself — I don't know how it's going to lead to more DEXes, more volume, more liquidity. But the tech guys love it. SNARKs and STARKs. You have both of them. Nice.
Privacy — excellent. I love that. I agree.
Formal verification — excellent. This one is interesting. Cardano, as you know, is at the forefront of formal verification. It does not help their price. For all of you who are non-technical, let me explain what it is. Formal verification is when you use math to verify that the code has no bugs. For example, let's say you build a train routing system. You need to ensure that under no circumstance are two trains routed against each other. You can analyze the code line by line and do a mathematical proof that ensures this never happens — you mathematically prove that your code has no bugs.
The issue is that it's super expensive. It is also very hard to do, and easy to make a mistake, because your code needs to be formatted and structured in a way that you can actually do formal verification. Most programmers cannot do it. If you ask the average dev to do something that can be formally verified, they cannot do it. The way normal programming works is that you just write tests. You have a bunch of tests and they test that whatever spaghetti you wrote passes the test. You test the functionality. That's how normal programming works.
But in some use cases — life and death use cases, like train routing, or a space mission — it's good if you don't have any bugs in the code. When you go to the moon, you spend billions. Good that there is a formal mathematical proof that there are no bugs.
How are they going to formally verify everything? It's going to be super expensive, super slow. I need to double check exactly what they're going to formally verify. You cannot formally verify everything. Come on. How are you going to formally verify everything? Now, although I am a highly educated computer scientist, this is slightly beyond my practical experience. I've done some tests on it but never done something proper with it. If you are the mega geek in the chat who can enlighten us how the hell they're going to formally verify the whole of ETH — because my gut feeling from my experience in tech is that for a large system like ETH, open up the ETH GitHub. For the average dev, you're not going to understand anything. How are you going to formally verify it? I don't know. Maybe if you rebuild the whole thing from scratch, maybe you can do it.
But here's the problem. You formally verify things. How are you going to change it in the future? You understand? You formally verify, but your product has no product-market fit and you've formally verified it. Okay, it's good, but you need to move, you need to compete. Formal verification is for when you don't have to change anything in the future. How are you going to compete? If anything, it will make further development very slow. Anyway, it's interesting. Tech stuff is interesting. I don't know if it's needed from a product perspective. Is it actually needed for product development to compete with Solana, to compete with other things? I don't know. Let's see. Even Bitcoin is not formally verified.
Near Protocol Claims Prior Art on Ethereum's New Features
Ivan on Tech: A Near Protocol guy jumped out of nowhere and said that Vitalik copied the Near road map. Let's see this guy. He's an ex-Near core developer and he's saying that Vitalik published the Lean Ethereum straw man — almost every major piece of the protocol will be replaced over the next three or four years — and he's pointing out decisions Near shipped years ago. Quantum safety, now marked urgent — Near shipped post-quantum ML-DSA keys on testnet last week. He added one to his account. Good news.
So just take Near and call it ETH. Why do you have to rebuild? Call it ETH. But yeah, you need to port all the apps, etc. That's a problem.
New state types where nodes store only a fraction of everything — with Vitalik's open question: why do they store more than 1% and why are they willing to serve it? Near answered that at Genesis in 2020 with sharded state.
So all in all, I think ETH is entering a more modern era, but it's going to be there in four years, and many chains are already there. Solana is already there. Near is already there. So yeah.
There's also the EVM to compiler-level feature while the protocol only uses RISC-V — this is something they discussed years ago. I remember in 2018, 2019, we were so excited when we read about WebAssembly. Do you guys remember WebAssembly? What happened to it? Nothing happened to it because they did L2, and now they are coming back. Now they're going to do some other compiler. Maybe WebAssembly, which Vitalik calls his ideal but still far away. Wasm plus Aurora live on Near — exactly. But to be fair, Vitalik spoke about Wasm back in 2018, 2019, but then they — I don't know. It's a lost decade. Kind of seen in the price chart also. Lost five years at least. Maybe even a lost decade. Let's see the highs here. Yeah, almost a lost decade. If we go to one point — yeah, you see it. So that's that.
Dankrad Feist — he's quite experienced — he's a researcher at Ethereum Foundation. He also says three to four years is very slow and that we should code it in one year. I agree. I think this is realistically possible now with LLMs. I don't know if it's good with LLMs here. Maybe it is, because especially they do so much research at the research stage. LLMs are probably good for research, ideas, mathematical proofs. I don't know if AI is good enough now to do mathematical proof. Before, it couldn't even tell you whether you should walk or drive to the car wash 50 meters away — it said walk. You ask a simple question: you have two or three birds on a tree, two flew away, three flew back, five disappeared, how many birds are left? It said like a hundred or something. So I don't know if they can use it for real math. But to brainstorm, to get stuff in, maybe it's possible to do in one year. Possibly.
But you need someone to double-check each and every single line, especially if you're going to do formal verification. If they're going to formally verify each and every single line, I guess you can do it with AI, because everything is still mathematically verified afterwards. Maybe it's possible. Maybe that's how AI is used. You give AI the restriction that everything it does needs to be formally verified. Why not? That'll be pretty interesting actually.
Coinbase AI Hallucination Incident
Ivan on Tech: Coinbase is sending out fake alerts, fake news alerts. This is what happens when a crypto company uses AI to generate sports prediction markets. Coinbase is hallucinating results for a World Cup game that hasn't even been played yet and sending factually incorrect notifications to its millions of users as breaking news. Dangerous and irresponsible. Exactly. People see this and then they make trades. It's very, very bad.
If it is formally verified, maybe it's okay. Maybe it's okay.
Robinhood Launches Agentic Crypto Accounts
Ivan on Tech: Robinhood — what's happening? They're going to do agentic crypto stuff. Let's see in their wallet. Let's see what they say.
Robinhood Presenter: We're announcing the launch of Agentic Accounts for crypto trading.
Ivan on Tech: If the agent is going to be fed fake news from Coinbase AI — that, let's say, Brazil won over Norway, which they didn't — this is bad, man. Imagine all the hallucinations that are going to happen here. But I'm actually bullish overall. More good than bad, but you just have to be careful a bit in the coming years because we're still early.
Robinhood Presenter: Why is this important? As I'm speaking, on the screens you're seeing just an example of the pricing data, news, order book for Bitcoin that could happen over the course of just a few minutes. Institutional giants have the ability to use algorithms and rooms full of analysts to scan all of these data points 24/7. But the reality is no single trader has the bandwidth to process this volume of data and have the same advantage. However, where humans drown, that's where agents thrive.
Ivan on Tech: Anyway. And what the hell is this aesthetic? I mean, this beautiful Vlad is here. There's another Vlad. What is his name from Romania? The Impaler. Is that the aesthetic? You go to Transylvania to check on Robinhood. Anyway — agentic. Agentic. Impaler. Exactly. But he's from Bulgaria, this Vlad. Kind of the same vibe though. I love it. Very nice. Instead of doing another copy of Apple — people pretend to be Tim Cook, Steve Jobs, they do all of this Apple-like presentation — let's do freaking Transylvania. Fantastic. Big shout out.
Solana's Real-World Asset and Stablecoin Growth
Ivan on Tech: Solana was on ESPN. No one cares. But here's what people care about: Solana hit $3.6 billion all-time high when it comes to real-world assets. That's good. Stablecoin crossed $200 million in a few weeks. That's good. I like this. Focus on the end result. Focus on the products. Focus on what's happening. I like this. I don't like discussing STARKs, SNARKs, and some other tech fugazi. Focus on this. This is prediction market. Bam. This is this, this is that. Focus on the end client. Not "oh, we're going to have SNARKs and STARKs and we're going to store state" — that's a 2017-level discussion. How state is stored? It doesn't matter. Everyone has passed that. We know how to store state fast. Look at the end result.
Tokenized asset spot volume grew from $2.69 billion in Q1 to $5.7 billion in Q2. This is the kind of stuff. How this SNARK and STARK is going to bring it to ETH, I don't know, but it better.
Someone is saying asymmetric returns come from taking risk before ideas become consensus — Solana is down 75% from highs with meaningful developments across the ecosystem, will be obvious in hindsight. I disagree. Because it's down 70%, it can still go down another 70%. I'm not buying it before it goes to $30 or it goes to bull trend at $93. This is a very dangerous way of thinking. While from a product perspective I love Solana and what they're doing, from a price perspective it's very bad still. I'm very bullish when it goes to bull trend or to $30, but to say that it's cheap just because it went 75% down and has tech improvements is wrong.
The case study from the last bear market: it went down 75% from the peak, and then from there to the bottom another 90%. So yeah, I'm not buying it. Being in altcoins, 75% from the highs is not cheap. It's kind of expensive still. It can easily go down another 80%. Easy. I'm not saying it will. And I agree — as a product, Solana is way better than ETH. ETH is still trying to figure out whether they're going to do STARK or SNARK and it's going to take five years. So as soon as Solana goes bull, fantastic, happy days. Now, not happy days still for the price. Still bear, still kind of retesting this support as resistance and likely going to go lower. But if it goes to bull trend, we're bullish at $93. We're going to go to $300 to $1,000. We're going to go very high when it goes to bull trend. Now it's still bear trend. That's it.
AI Market Dynamics: Overhype and the Revenue Problem
Ivan on Tech: AI — very interesting. Alex Karp is doubling down that AI, in its existing form, or more correctly, the companies behind AI — OpenAI and Anthropic — the way they promote and the way they do business is not scalable and it's not going to work out long term, because they steal your IP and they sell you expensive tokens. This is what Alex Karp is saying. Anthropic and OpenAI give no value and take your IP.
I find it fascinating, because you have all of these models which are great, and now there's a bit of a fight over who is going to own the client. Is it going to be the AI labs that own the client, or is it going to be all of the implementers? For Alex Karp, it's good if nation states tell him to install local AI instead of plugging in Anthropic or OpenAI. It is interesting how this is playing out. This innovation with AI — which is great, it's not as great as it's marketed, but it's super great anyway — it's marketed so insanely that there's no way reality can match it.
Nvidia's CEO is saying you need another hundred-plus billion to build some AI factory. Chamath is saying the price of poker is going up. It's a bit of a poker game because no one can say, "Hey guys, we overinvested a bit here. We don't know how we're going to get revenue for all of this." The end client has to pay at the end of the day.
Now, to be clear, I'm bullish AI stocks now. Why? Because they're in bull trend. At least some of them. Many of them are now in bear trend. Nvidia went bear trend. Broadcom went bear trend. Tesla bear trend. Meta bear trend. But the ones that are bullish — like Micron, etc. — we're still bullish because they're bull trend. They get paid from all of these investments that Meta, Google, all of the big companies are making into AI. So the chip companies pump because they have revenue in the pocket. Now everyone's happy. That's why they're the last man standing. Micron is still bullish. AMD still bullish. Everything else turned bearish — the end client companies turned bearish. And Nvidia, which was too overhyped.
At the end of the day, these companies that are investing all these hundreds of billions need to get paid from the end client. The end client has to pay for it. And where is the end client? Alex Karp is saying, "Hey, it's for guys like us — give us the open source model, we're not going to use the AI lab, we're going to run open source." Another thing is that you don't need to use the frontier models for most use cases. The frontier models are ten times more expensive than the average open-source ones, but they don't give ten times more value. They give maybe a bit more value, but not ten times. So anyway, fascinating stuff. At the end of the day, this will mean rotation into crypto. We're bullish right now. Trends are bullish, but it's going to rotate so hard.
Mark Zuckerberg admitting Meta's $145 billion AI bet is not paying off. Where did the money go? It went straight into Micron and Intel and AMD. Meta is now in bear trend. Zuckerberg told employees that Meta's AI bets haven't come to fruition yet and the AI development has not accelerated as expected. Meta has committed up to $145 billion mostly for AI strategy this year while cutting 10% of the workforce. Zuckerberg said AI spending puts strain on the company. And Meta's stock price would be higher, man.
So guys, if you got laid off because of AI, you may be getting hired again soon. Because I think people are understanding the AI models. They're good, but they're a bit unreliable also. They can be the smartest person in the room and then do something completely wrong. You cannot really give it responsibility. The bigger the context, the more it's forgetting things, mixing things, making mistakes. The hope during the last two years was that this is just temporary, that AI labs are going to solve it. But each and every new model that AI labs release — which costs ten times more to run, needs all these billions in data centers — they don't solve it. They are just a bit better but nowhere close to solving the core issue. So I love AI. AI is going to be great. It's going to be fantastic. But it's overhyped, over-marketed, and we will see a correction in AI companies bigly to account for that.
The Silk Road Bitcoin Exploit Story
Ivan on Tech: Let me tell you a bit of a story about a guy in Georgia. In 2012, a college kid in Georgia found a bug on Silk Road, the biggest drug market on the dark web. He clicked to withdraw five times and got more Bitcoin than he deposited. He walked away with 50,000 Bitcoin. Nobody caught him for ten years. By 2021, that stash was worth $3.4 billion. Imagine — you want to buy some illegal substance and then you walk away with 50,000 Bitcoin because you clicked a button a few more times.
What happened? Here is what separates him from every other thief. He did not touch it. He just let it sit while Bitcoin ran. When the Bitcoin Cash fork happened in 2017, he sold the forked coins and bought 3,500 more Bitcoin with the proceeds. The stolen pile literally paid him dividends.
How he sold it without KYC — it sounds a bit strange how he sold it on-chain. The money was never his problem. Being alone was. This is the tragic story, because he started buying friends private jets to football games, $10,000 shopping sprees for people he barely knew, a lake house with a stripper pole. He kept $700,000 in a briefcase because he wanted a case full of money like in the movies. When anyone asked him where it came from, he said he mined Bitcoin early. No one blinked. The lie worked because it could have been true.
Then in 2019, one of those friends robbed his house, took the briefcase and a USB drive. And this student, sitting on $3.4 billion in stolen crypto, panicked and called 911. Told the operator he was having a panic attack, filed a police report about the theft of his own stolen money. Come on, man. That report landed with the IRS, who had been hunting the missing Silk Road coins for years. A few months later, he moved $800 worth of Bitcoin through a KYC exchange. That one transaction mixed coins from the 2012 hack with a wallet tied to his real name.
Undercover agents visited him pretending to help with his burglary case. He was so happy someone finally took him seriously that he opened up his laptop and showed them $60 million in Bitcoin. The raid happened. They found the keys to his 50,000 Bitcoin stash inside a popcorn tin under a pile of blankets. Second largest financial seizure in US government history. His sentence: one year and one day. Not bad. Not bad. SBF got 25 years. The only victim was Silk Road itself. He lost to a broken window and a panic attack. Good ending. Good ending. I mean, listen — likely maybe he has a bit more stored somewhere. Why put everything in a popcorn tin? Maybe put 100 Bitcoin somewhere. Let's see.
New Twitter Seed Phrase Scam Warning
Ivan on Tech: You may see on Twitter a new kind of scam. Someone posts a seed phrase, kind of like by mistake. You think it's by mistake. You try to take the money, which they want you to do, and they expect you to send some money for gas. The moment you do that, they sweep everything including your gas. There is no free lunch. There's no free lunch.