It's worse than you think w/ Tulip King, Eli Ben-Sasson and Saeed Badreg
Summary
- Ansem is cutting risk while Tulip King calls this his “retirement cycle,” making asset selection and time horizon the episode’s central disagreement. Ansem’s October 22, 2026 view is that weakening memory and semiconductor stocks could drag crypto into another pullback; he initially expects the summer lows to hold, while later entertaining Bitcoin tests of the mid-$70,000s or mid-$60,000s. Tulip is looking across a decade at businesses that could absorb much larger on-chain markets: “Stop buying shitty coins.”
- Tulip’s portfolio framework separates coins that are money from coins that make money, leaving ETH and SOL under pressure to justify their valuations. His argument is that stablecoins increasingly supply collateral and trading denominations, while scaling reduces the fees native tokens capture. Ansem disputes the assumption that existing holders will suddenly abandon monetary and ecosystem valuations, but agrees that profitable applications might become money more easily than base layers become better businesses.
- Hyperliquid and Derive anchor Tulip’s argument that investors can own emerging financial infrastructure without predicting every new winner. His HYPE “100 bagger” case explicitly assumes Hyperliquid wins as assets and trading migrate on chain; his options thesis rests on convex payouts, volatility trading, and Derive’s claimed 90% share of options premiums. He would reconsider competitors when they establish durable share, rather than constantly trade between prospective challengers.
- Saeed Badreg frames tokenization as a distribution and market-structure opportunity measured against a roughly $160 trillion asset pool. Capturing even 1% would exceed $1 trillion, but moving tokens between chains is insufficient: Sunrise needs market makers, working mint-and-redeem arbitrage, and distribution through wallets, exchanges, and DeFi. “We have 8,000 stocks” is unimpressive if “you have zero markets”; the objective is liquid, programmable assets that unlock uses unavailable through conventional brokerage apps.
- Eli Ben-Sasson says Starknet’s community has supported exploring an L1 because Ethereum scaling demand and Ethereum-dependent security no longer settle the strategic question. He describes a potential post-quantum “ark,” with programmable privacy and account abstraction, while acknowledging that he cannot predict a quantum breakthrough. Asked when he would worry most, he says “over the next two years”; that is his concern window, not a claim that a practical attack will arrive then.
- Zcash and Pearl represent different attempts to capture demand for alternatives to Bitcoin’s existing design. Tulip categorically says Zcash reaches a trillion-dollar valuation but offers no timing, while Pearl’s appeal is its proposed substitution of matrix multiplication for SHA-256 to connect mining with AI infrastructure. He qualifies the miner-migration evidence and calls quantum a “pretty bad narrative,” creating a clear disagreement with Ben-Sasson’s emphasis without supplying a detailed rebuttal.
- Trading media’s commercial opportunity rests on audience distribution and trading activity, while its credibility depends on distinguishing evidence from mythology. Tulip describes sponsorship economics that favor customers who retain capital and keep trading, plus a clipping operation whose reach can far exceed live viewership. Banks sees Kimchi’s mainstream breakout as valuable, but also acknowledges manufacturing an exaggerated story for a viral clip; the hosts’ discussion does not establish the widely repeated $40 million trading-profit claim.
- The practical research edge described here is persistent investigation beyond headline dashboards, followed by enough patience to let a thesis develop. Tulip uses protocol testing, source-code activity, holder analysis, and AI-assisted data retrieval to ask why revenue should persist; GMX at roughly 1.2 times revenue is a research prompt, not a completed investment case. His largest stated equity exposure is Eli Lilly, and his reluctance to buy AI-stock dips reflects both concentration concerns and a preference for opportunities where he believes he has an edge.
Deep dive
1. Ansem’s bearish turn begins with equities failing to reward good news
Ansem explicitly reverses the view he held two weeks earlier. He expected another advance in AI-related equities, memory stocks, and semiconductors, with crypto extending its Q3 leadership. “I was wrong,” he says about expecting fresh highs. His concern is that stocks which have dominated since late 2022 and 2023 now contain substantial unrealized gains, creating room for sustained profit-taking rather than another shallow interruption.
The backdrop adds several pressures simultaneously: continuing war, sharply rising oil, still-high interest rates, and AI discovering software vulnerabilities faster than humans previously could. Ansem cites a significant PHP vulnerability as an example of processes being disrupted. He does not reduce the selloff to that technology story; the central call remains that equities might be approaching the “tail end” of their macro cycle.
The uncomfortable detail is that business results can remain strong while stock prices stop advancing. Ansem points to Micron’s substantial Q3 earnings beat, after which the stock failed to extend higher and effectively marked a local high. “The revenues are matching up, but they’re still not making new all-time highs.” He also sees memory stocks forming a second lower high after their parabolic advance.
His crypto interpretation remains a question, not a settled diagnosis. Bitcoin weakened in Q1 and Q2 while equities rose, then bottomed around the same late-June week that memory stocks peaked. Q3 strength might therefore reflect an earlier oversold condition rather than a fresh bull market. Tulip’s later challenge—why Bitcoin has remained “pretty damn strong”—gets this answer: relative strength can coexist with vulnerability to another equity-led decline.
2. Taking profit is a different decision from initiating a short
Ansem says he is both reducing exposure and shorting, but immediately distinguishes his own trading from what he recommends to viewers. BTC, ETH, and SOL are already near important support, so a new short would be “shorting into support.” Timing matters: getting stopped out before the eventual decline produces the particularly frustrating outcome of losing on the short and then missing the move.
For people up 2–3X on positions such as Pump, Zcash, or VVV, he thinks taking some profit makes sense. His initial expectation is a higher low, with the roughly $70,000 Bitcoin area offered as a possibility, and he says he does not think the summer lows break. Later, he entertains mid-$70,000s or mid-$60,000s tests while saying, “I’m not sure.”
The cleaner relative-value expression he offers is long Bitcoin against short memory stocks. That separates his belief that Bitcoin can outperform memory equities from the harder question of whether Bitcoin rises outright. He also recalls SOL around $60 near the earlier summer opportunity: someone who bought near those lows can still have substantial gains to protect without needing to declare the entire crypto recovery finished.
Banks pushes back with a familiar contrarian rule: widespread CT pessimism usually makes him more bullish. Ansem’s own experience complicates that instinct. He became bearish in September 2025, absorbed criticism, bought Bitcoin after it subsequently reached highs, and then had to sell into a decline. The exchange preserves the distinction between being directionally right, being early, and changing position under social pressure.
3. Paper Trade shifts its rewards from early farmers toward continuing traders
Ansem describes Paper Trade, launched by his friend Jez, as a synthetic perpetuals platform offering up to 1,000X leverage using Hyperliquid midpoint pricing for BTC and ETH. Positions are against the house rather than a matched opposing trader. Losses accrue to the house; winners are paid from its funds. When the house does not have ample funds, Ansem describes it as “kind of like a pool that collects over time.”
He describes a controversial change to the token distribution. At the beginning, he says, a large majority of the supply would be distributed based on how much the house was making. He recalls the maximum supply as 12.6 billion PAPER, and says it would be reached after a certain amount of time—“something like $500 million in the LP.” He is uncertain about the precise calculations. The supply has since increased; now there is no maximum supply and more PAPER is minted, though he says he does not know the exact calculation.
Ansem thinks the change benefits traders who actively use the exchange. “Before, it benefited farmers more than traders,” he says. He farmed early himself, but says the platform has changed how PAPER is distributed and people have been mad about it. His defense is that the revised approach better rewards active exchange use, while acknowledging that farmers benefited more under the previous setup.
Transferability has also been delayed: the token will be non-transferable for longer. Ansem says PAPER can currently be staked on the exchange to receive a portion of its revenue. He likes the platform and the changes, while recognizing that the altered distribution and delayed transferability are sources of user frustration.
4. Pump’s missing airdrop is a legitimate grievance; falling coins are another argument
Ansem separates two complaints that CT has bundled together. Anger that Pump is paying Kimchi, KOLs, and other promotional rewards before delivering its anticipated airdrop is, in his words, “justified.” Banks agrees that the airdrop was central to earlier expectations discussed on the show. Their defense of Pump’s business does not erase that unresolved distribution promise or the resentment around who has received money first.
The complaint that Pump should somehow make launched coins appreciate gets a different response. “Pump cannot make coins go higher. They’re just the launchpad.” Ansem says higher starting market capitalizations might produce different outcomes, but that remains an unproven counterfactual. Comparing today’s market against an imagined design that supposedly fixes everything does not establish that the alternative would actually work better.
His explanation for Pump’s resilience is its existing low-cap ecosystem. He says the platform can still make around $1 million daily during a bear market because traders and developers repeatedly use pre-bond and very small-cap markets, even without $500 million runners. Removing that activity to favor a different kind of participant could damage the users and behavior responsible for the platform’s continued revenue.
Banks and Ansem reject treating all successful exchange revenue as uniquely objectionable “extraction.” They point to Coinbase, Uniswap, Raydium, and Axiom as businesses that also charge for services. Their argument is competitive: users keep returning, and rival launchpads have failed to displace Pump. That supports their view of product demand, while leaving the airdrop criticism intact rather than answering it through fee comparisons.
5. Kimchi’s attention is observable; the trading fortune remains disputed lore
Tulip King repeatedly distinguishes Kimchi’s actual content from stories circulating around it. He says the 22-year-old has not personally confirmed the widely repeated claim that he made $40 million trading Trump. In one vlog encounter, a fan asks about that number; Kimchi laughs, says, “Good to meet you, bro,” and moves on. Banks had earlier summarized the persona as “hit big and logged off,” rather than publicly documenting a trading process.
Tulip also describes an on-chain thread presented as proving that Kimchi made “only” $6 million on Trump. He says the wallet was subsequently confirmed as Kenzo’s, Kimchi’s friend who travels with him, and therefore was not Kimchi’s wallet. Tulip says there was no comment from Kimchi and no confirmation one way or another about the $40 million claim.
Their strongest positive evidence is cultural reach: the vlog, luxury-brand attention around Bottega, and interactions with major streamers who appear genuinely interested in Kimchi. Banks argues that CT saturation is different from a mainstream ceiling. Someone can already occupy nearly all the conversation inside crypto while still rapidly expanding outside it; criticism within the smaller audience may become progressively less important to the overall business.
Banks sees a self-reinforcing commercial opportunity: attention itself could eventually allow Kimchi to make $40 million through partnerships and future projects, regardless of the original rumor. That is a prediction, not confirmation of the earlier trade. He thinks Kimchi’s advantage is having no obligation to debate CT: travel, friends, and compelling vlogs can continue working without turning the brand into a market-commentary service.
6. Tokenization’s market size matters only if assets become usable
Saeed Badreg explains Sunrise’s origin through an inadequacy in interoperability alone: “Just bringing an asset over from one blockchain to another doesn’t really solve the problem.” Internet capital markets require healthy, liquid markets and useful financial activity after the bridge transaction. He puts equities at approximately $160 trillion against crypto around $1.5–$2 trillion; even 1% migration would represent more than $1 trillion of value.
His stablecoin analogy is about demand invisible to people already well served. An American with easy dollar access can miss the importance of that access elsewhere. Badreg invokes stablecoins’ growth and Tether’s Treasury purchases to argue that making an asset globally accessible reveals demand that incumbents underestimated. The mechanism is distribution: “You completely have no idea how much demand is out there for a particular asset.”
Badreg says tokenized equities were below $1 billion “last year” and have expanded roughly 30X over about a year and a half. He describes the growth as a hockey stick and thinks even his own team underestimates the eventual scale. The hosts later loosely restate the $160 trillion figure as broader tradable assets; Badreg initially introduces it as the equity-market opportunity rather than supplying a reconciled market taxonomy.
7. Global access and cheaper capital formation are separate tokenization opportunities
Badreg’s response to Americans asking why they need tokenized US stocks is blunt: “It’s not for you then.” Their opportunity might instead be exposure to DeepSeek, Unitree, or SK Hynix. He says Sunrise has been meeting people and regulators internationally, including during a Shanghai visit, and hopes Chinese equities can reach these rails. “Hopefully soon” remains an aspiration, not a launch announcement or guaranteed access timetable.
Banks supplies the investment puzzle that made DeepSeek interesting to him: a valuation around $10 billion versus the roughly $300–$500 billion figures being discussed for major American AI companies at the time. Were the latter really worth about 50 times as much? The discussion does not resolve that valuation question; it uses the inaccessible comparison to explain why investors might want distribution beyond their domestic brokerage universe.
The second opportunity is companies raising capital without following the conventional IPO route. Badreg estimates that a serious US listing is unlikely to cost less than roughly $30 million and can require 1.5–2 years of organizational change. He qualifies that estimate as based on numbers he has seen, rather than personal IPO experience. A good company earning tens of millions might still find that process economically unattractive.
Banks corroborates the burden through his own SPAC experience with B. Riley, including the need to raise a PIPE, and says taking a company private is also expensive. Ansem connects tokenization to Hyperliquid’s distribution of value to users and team members. Badreg sees convergence with broader capital formation, citing a16z’s expansion into frontier businesses: crypto rails could serve operating companies whose businesses are not themselves crypto.
8. MemeFi exposes the cultural component already present in stock markets
Badreg says he initially thought pairing memes with stocks made little sense, then changed his mind after watching the activity closely. Traditional finance, he argues, understates how much assets trade on “narratives, community, culture, momentum.” Tesla, GameStop, and Nvidia illustrate his point: they are companies with financial statements, but also communities with identities. Tokenization makes that cultural layer more directly accessible to new trading behavior.
Eli Ben-Sasson interprets the experiment as isolating attention and narrative from the conventional asset, while Tulip King supplies the uncomfortable counterexample: the meme could perform well while its paired stock tanks. Badreg accepts that the structure can deliver “the best of both worlds” or “the worst of both worlds.” His enthusiasm is for experimentation and participation, not a claim that attaching an equity automatically improves an investment’s economics.
Tulip places the experiment in crypto’s longer history. Many assets traded on Bittrex and Cryptopia in 2017, or during 2020–2021, failed; he thinks those failures helped reveal which mechanisms worked. “They didn’t call them meme coins back in the day. They just called them altcoins.” Badreg also sees tokenization forcing public-company CEOs to take positions, with resistance and eventual acceptance resembling earlier battles around stablecoins.
9. Sunrise’s product is the market around the token
Badreg identifies a strong market-maker network as the first requirement for usable tokenized equities. Coming from the trading world, he emphasizes understanding how liquidity providers actually operate rather than assuming they appear after issuance. The goal is an asset that can trade reliably across an ecosystem, not a token whose existence technically proves that an off-chain security has been represented on a blockchain.
Mint-and-redeem infrastructure provides the second requirement. If a tokenized stock trades above or below its reference market price, authorized participants need a practical way to arbitrage the discrepancy. That mechanism is what can pull prices back toward one another. Badreg treats it as fundamental plumbing: without working issuance and redemption, an apparent listing may not become an efficient market for ordinary users.
Distribution is the third requirement, encompassing centralized exchanges, wallets, DeFi applications, aggregators, routers, and DEXs. Sunrise tries to combine the components into something that looks simple externally but is difficult internally. Badreg’s claim is conditional: solve enough of these connected problems and a team can launch 100 assets on Solana while maintaining functioning markets, then expand from there.
Asset selection also requires judgment. The initial heuristic was to bring whatever the community wanted, but Badreg describes that community as highly opinionated and difficult to satisfy consistently. His criticism of indiscriminate catalogs is memorable: “We have 8,000 stocks,” but “you have zero markets.” He prefers listings that capture attention, provoke research, and make a trader ask why this particular asset is arriving now.
10. Programmable equities and social participation create different sources of demand
Badreg rejects a simple contest between spot tokenized stocks and stock perpetuals. Spot markets support ownership and composability; perps support hedging and more complicated strategies. “You’re trying to make Apple stock programmable.” The objective is not merely replicating a brokerage position on another screen, but allowing assets to participate in financial operations that other protocols, bots, and agents can execute together.
His example is moving collateral through several operations that conventionally might require 2–3 days, 4–5 companies, and repeated reconciliation. On-chain infrastructure could let a user or agent coordinate those steps programmatically. The challenge he leaves to Solana developers is concrete: what can someone do with a tokenized stock that they cannot already do through traditional finance or the applications on their phone?
A different source of growth comes from younger people’s willingness to trade and share publicly. Badreg and Tulip King speculate about participation moving from below 1%, perhaps 0.1%, among earlier 18-year-old cohorts toward much larger current numbers; Tulip suggests 10–20%. They do not establish those figures as measured statistics. Their underlying observation is a cultural shift toward active investing, visible portfolios, and earning from capital rather than labor.
Ansem offers the darker version: AI could concentrate economic power in a few companies while displacing jobs, leaving ownership of those companies increasingly important. Badreg adds reports of banks restricting AI-agent access and cites the interpretation that bots were moving deposits toward better yields. He frames that interpretation as incumbent resistance to competition; the conversation does not establish it as the banks’ documented reason for restricting access.
11. Starknet’s L1 discussion reflects a changed problem, not an abandoned scaling capability
Eli Ben-Sasson introduces himself through academic cryptography, his realization at a 2013 Bitcoin conference that zero-knowledge techniques could help scalability as well as privacy, and subsequent work co-founding Zcash and StarkWare. His description of StarkWare is “the skunk works or the Bell Labs of crypto”: a research organization turning inventions into usable systems. That is his positioning, rather than a neutral ranking of competing technologies.
He says a Starknet community vote concerning becoming an L1 concluded that day with substantial support. The next step is to investigate and build toward that direction, rather than claim the migration is already complete. The intended combination includes post-quantum security, programmable privacy, and improved usability. He describes the prospective network as a “safe haven” or high ground if a quantum threat materializes.
The strategic change has two parts. Starknet successfully built Ethereum-scaling technology, but Ethereum currently has multiple scaling options and insufficient demand for that capability alone. Separately, Starknet’s security as an L2 depends on Ethereum’s migration speed. Ben-Sasson characterizes the community’s message as: “We’re not confident that Ethereum’s gonna move fast enough for our safety, so we’re gonna move ourselves.”
12. The “quantum tsunami” is a preparedness thesis with an uncertain clock
Asked when a quantum threat might arrive, Ben-Sasson first says, “I don’t have the crystal ball. I don’t know.” Pressed for a guess, he identifies “the next two years” as the period when he would be most worried. His reasoning is accelerating innovation across multiple quantum-computing teams and hardware approaches. He does not provide a probability estimate or assert that an attack is certain within that window.
The metaphor carries his prioritization: “This is the damn ark that you can board if the flood comes.” Living on an ark is not necessarily the most convenient experience, but convenience becomes secondary if a house and garden sit beneath 100 feet of water. He therefore frames the prospective L1 around safety under a particular threat, rather than primarily as another competition over transaction-speed marketing.
Ben-Sasson nevertheless claims Starknet already offers superior efficiency, cost, safety, and usability, and highlights account abstraction and privacy that can compose with DeFi. The interview does not supply comparative benchmarks for those categorical claims. His point is that these capabilities already exist in the technology; the new strategic argument is to make security less dependent on how quickly another chain changes its underlying cryptography.
AI enters the same discussion from both sides. StarkWare is exploring hyperscale transactions and micropayments for agent commerce, without announcing a finished product. Ben-Sasson also worries AI could accelerate quantum development or weaken existing cryptographic assumptions. He argues that zk-STARKs’ reliance on hash functions provides a safer foundation against those threats: infrastructure for agents, and infrastructure designed to withstand advances in the tools agents use.
13. A zk-STARK lets an untrusted operator prove that the work was correct
Ben-Sasson’s explanation begins with Darth Vader processing a day of Bitcoin transactions and reporting the new system state. Without a proof, the concern is obvious: Vader might simply assign the bitcoin to himself. A zk-STARK changes the acceptance rule. The operator can do the computation, but the result is accepted only after a verifier checks a short cryptographic proof that the required rules were followed.
In the example, the proof is smaller than a Bitcoin block, while standing behind the much larger set of transaction updates. Ben-Sasson calls it a kind of cryptographic “cement” enforcing integrity, reliability, and correct execution even when the operator is not being watched. He believes Satoshi would have embraced usable ZK, noting that early discussion recognized its privacy potential before practical implementations could deliver the broader scaling capability.
The eventual user experience he wants resembles handing over a $20 bill: a transaction between counterparties without an intermediary observing and controlling everything. He extends that intimacy to lending, borrowing, investing, and perpetuals, with usability comparable to Robinhood. The important addition is self-custody and privacy rather than a central application holding funds and seeing every action: familiar interface quality applied to decentralized financial relationships.
14. Long adoption timelines do not excuse waiting passively for developers
Ben-Sasson accepts Ansem’s observation that Zcash spent many years without strong market recognition. He dates the underlying paper he co-authored with six other cryptographers to 2014 and Zcash’s implementation to 2016; meaningful attention took roughly nine years after launch. “It was painful,” he acknowledges. The admission preserves a distinction between believing a technology is important and expecting markets to recognize it quickly.
His comparison is monetary adoption: paper money took centuries to become accepted, while ATMs and credit cards required decades of trust-building. On that scale, digital money taking years to establish which properties matter is unsurprising. He applies the same patience to Starknet. Privacy, scarcity, and security may each become salient at different times; his stated commitment is to keep building the properties he believes usable money needs.
He is less forgiving about the original ecosystem-development strategy. “Build and they will come” did not produce a great return on the funding deployed. Some strong DeFi and gaming projects emerged, but too few funded teams became durable, self-sustaining businesses. That assessment is an explicit change in approach: excellent infrastructure and grants alone did not reliably generate the applications and users the network needed.
StarkWare is consequently building some applications itself while continuing to support external teams. Ben-Sasson names SuperVega, at supervega.app, as a retail options product with a focus on good UX and exposure to assets including Bitcoin, Zcash, and STRK. He hopes internally developed applications can bring users and attention more directly, rather than leaving the entire demand problem to independent developers.
15. Tulip’s bullishness is measured in decades and contingent winners
Tulip King enters with the opposite emotional register from Ansem: “I’m, like, so fucking bullish.” He thinks the scale described by Sunrise is directionally right—trillions in assets and potentially quadrillions in trading—and asks who genuinely expects less activity on crypto rails two years from now. Repeated over successive periods, he thinks that migration adds multiple orders of magnitude to today’s markets.
His HYPE “100 bagger” statement has a stated premise: “Just presume that Hyperliquid wins.” He accepts that another perpetuals venue could become the better allocation and require a switch. Against the upside he imagines, hypothetical HYPE prices of $60 or $40 do not alter his long-term enthusiasm. “I’ll sell my kidney” is the deliberately extreme expression of that conviction, rather than an articulated risk-management rule.
He extends the opportunity to options, lending, spot markets, and other DeFi businesses, with buybacks supplying the link to tokenholders. The horizon is the next decade, not the next week. “This is my retirement cycle.” He says his own portfolio is at highs and criticizes treating a disappointing broad market as proof that no attractive assets exist: the distinction is which projects investors actually own.
16. Options preserve financial exposures that perpetuals cannot replace
Tulip’s Derive thesis begins with the payoff structure. A perpetual is a leveraged linear exposure with path risk, which is why liquidation exists to protect protocol solvency. Options can offer convex outcomes. In a market with extreme tails such as crypto, he regards that difference as economically valuable rather than an inconvenient product feature that superior perpetuals interfaces will simply eliminate.
Options also permit hedges, spreads, and positions expressing a view on volatility rather than the asset’s direction. Tulip allows that perpetuals might compete with retail zero-day options activity, but says the rest of the options market remains largely untapped. His argument is therefore not that one interface wins all speculative trading; it is that distinct exposures support a separate market as crypto’s financial structure matures.
He quotes Derive as having around 90% of options premiums and sees that leadership as a reason not to overcomplicate the selection. His estimates of remaining volume growth are deliberately rough: perhaps 10,000X, perhaps 100,000X, “however many zeros it is,” to reach traditional-finance scale. Those are his order-of-magnitude expressions of the opportunity, with explicit uncertainty about the calculation, rather than a detailed valuation model.
Competition does not invalidate owning the present leader. He would consider Lighter or Variational once they show 20–30% share, and names Hyperliquid and Paradex as other potential challengers. If another venue genuinely overtakes Derive, “you’ll see it coming.” He thinks buying after that evidence can still leave substantial upside, because the investment premise is expansion through another order of magnitude or two.
17. The barbell favors distinct exposures over multiplying correlated positions
Asked for his largest holdings, Tulip opens with “my entire net worth is Eli Lilly,” then describes Bitcoin, HYPE, and Zcash as roughly similarly sized positions whose relative weights change with prices. He also says early purchases of Pearl and Derive have grown into significant holdings. The conversation conveys strong concentration and several meaningful exposures, rather than a precise, internally reconciled portfolio allocation.
His governing categories are “coins that are money” and “coins that make money.” The first basket includes Bitcoin, Zcash, and Pearl, with possible comparisons such as Zcash versus Monero or Pearl versus other AI-linked assets. The second contains operating protocols whose economic activity benefits their tokens. He wants investors to articulate which role an asset serves, rather than rely on a generic expectation that crypto rises.
Cards and Venice appeal partly because they occupy different lanes from Hyperliquid and Derive. “I don’t think you should be chasing beta. I think you should be chasing exposures.” His proposed monitoring process is to identify each market’s leader, own it, and use a Grok bot for daily updates on competitors. Reallocation should follow durable competitive gains, not a single strong day or week.
Pump fits his operating-business framework through execution: PumpSwap, its application, and its bridge extend the business vertically. He also resists treating wallets and social trading apps as purely zero-sum rivals. A Pump coin traded through FOMO can generate income for both, while Phantom provides cash and spending functionality. He thinks the businesses benefit from one another more than their competitive public positioning acknowledges.
18. Stablecoin adoption challenges the claim that every successful chain needs native money
Tulip says he once liked the monetary thesis for Ethereum: growing blockchain economies would naturally use ETH as money or collateral because it lacked an external counterparty. The same logic could apply to SOL. His objection is empirical rather than a denial that the theory once sounded attractive: “It just doesn’t show in the numbers.” He says the assets most widely used as collateral increasingly are stablecoins.
Solana’s swap pairs provide the distinction. SOL still participates heavily because of meme trading, but tokenized stocks are being paired against USDC. Someone bullish on stock tokenization is therefore also bullish on USDC becoming the denomination for that activity. Tulip sees similar behavior in Ethereum’s lending and trading markets, weakening the assumption that ecosystem growth automatically expands native-token monetary use.
The resulting problem is counterintuitive: “The more mature Solana and Ethereum get and the more robust they become, the less and less the native asset is actually being used like money.” For Tulip, successful adoption can undermine one component of the token thesis rather than confirm it. His argument depends on following what users choose as collateral and trading currency, not merely counting applications or transactions.
He consequently thinks ETH and SOL need to be evaluated more as assets that make money. That is a categorical shift in his own framework, and he connects it to their disappointing performance around earlier highs. It is not a claim that the networks have no useful activity. The contested question is how much of that usefulness reaches the native asset through monetary demand or captured revenue.
19. Scaling blockspace can help users while weakening the scarcity premium
Tulip’s second objection is that transaction growth need not produce fee growth. He says Solana transactions have increased sharply while fees have fallen because the network is doing what it should: scaling. His analogy is NFTs. An individual collection can be scarce, while NFTs as an asset class are not. Likewise, existing blockspace can feel scarce before networks expand capacity or users adopt alternatives.
Ethereum’s expensive swaps helped make competing chains attractive. Tulip asks how much Solana could raise fees before Aptos, Sui, or another network captured marginal demand. He is not denying technological differences between chains; the issue is whether a network can preserve enough pricing power to support its valuation while continuing to deliver the cheap execution that attracts users in the first place.
He states the counterargument himself: a critical mass of applications could make Solana’s shared environment uniquely valuable. Users might pay for access to Pump, DePIN networks, and other interoperating applications rather than interchangeable execution capacity. “If I’m wrong, it will be for that reason.” He thinks investors can wait for evidence of stronger economics, because sustained fee generation is unlikely to appear invisibly overnight.
Tulip cites roughly $300 billion for Ethereum and $90 billion for Solana when arguing that a quarter-trillion dollars or more is misallocated. He explicitly rejects dollar-for-dollar rotation: $100 billion lost from one asset might redirect only $10–$50 billion elsewhere. His provocative comparison is investors accepting SOL at “20,000X earnings” while rejecting HYPE at 100–200X; those are his framing multiples, not independently calculated figures supplied by the show.
20. Ansem disputes how quickly holders abandon ecosystem and monetary valuations
Ansem understands Tulip’s mechanism but disagrees with the proposed capital rotation. He does not think ETH and SOL holders have historically treated the assets simply as businesses. They also represent money and exposure to applications that will be built later. Reduced native-token use will not necessarily make those holders decide, all at once, that “it’s not money.” Existing valuation habits can persist despite changed usage.
He grants more of the argument for new, traditional allocators. Someone asking “What am I buying here?” may struggle to value a base-layer token whose economic capture is unclear. Ethereum’s 2020–2021 fees once supplied a more straightforward numerical case, and the loss of that picture matters. Ansem acknowledges being an early ETH bear while also admitting strong personal bias toward SOL and its ecosystem.
Their clearest agreement concerns applications’ optionality. Ansem thinks a successful business token can become more money-like more easily than a chain can capture substantially more revenue. Pump using custom pairs illustrates the former route. His remaining SOL defense is exposure to applications that do not yet exist; Tulip calls that “an extremely reasonable take” and also recognizes the involved, talented Solana team as an advantage.
21. The dotcom analogy favors finding durable businesses after infrastructure scarcity fades
Tulip compares Ethereum’s earlier position with Cisco during the dotcom boom: investors faced an enormous new addressable market and bought the provider of scarce connectivity. The internet thesis could be right while the scarcity premium attached to its infrastructure supplier eventually became excessive. He thinks Ethereum occupied a similar position when useful blockspace was difficult to obtain, before broader scaling changed the supply picture.
His preferred point of comparison is roughly 2003. The broader technology index took many years to recover—he loosely recalls “ten or sixteen years”—while selected companies such as Apple, Amazon, and Microsoft became powerful long-term investments much earlier. His claim is about dispersion: investors did not need the old index leaders or the whole market to regain highs before identifying businesses with substantial future compounding potential.
The response he advocates is labor rather than prediction theater. Tulip says he has about 120 slides assembled and intends another multi-hour discussion of worthwhile projects. Four hours produced a list; roughly two weeks of investigation followed. He thinks a dedicated investor might be one to three weeks from developing much better theses. That is his encouragement to research, not evidence that those eventual selections are guaranteed winners.
22. Pearl connects Bitcoin’s security-budget problem to the AI infrastructure buildout
Tulip’s categorical Zcash call is “Zcash is going to $1 trillion,” immediately qualified on timing: “I couldn’t tell you when.” His broader monetary theory is zero-sum competition for liquidity and store-of-value demand. Bitcoin gained room because the dollar had weaknesses; alternative crypto monies need weaknesses in Bitcoin to create their own opening. In his formulation, “good” money succeeds by filling a vacuum left by worse money.
Privacy, quantum exposure, and the security budget are longstanding Bitcoin questions, but he thinks scale and proximity now make people care. Zcash reopening the market for alternative money is the signal he watches. He nevertheless calls quantum a “pretty bad narrative,” without developing the promised explanation. That remains a substantive difference from Ben-Sasson’s urgency, rather than a resolved debate between their technical positions.
Pearl addresses the security-budget branch of the thesis. Tulip points to public bitcoin miners moving toward AI because converting power into AI computation can pay more than mining bitcoin. He also qualifies the evidence: Bitcoin has experienced hash-rate downturns before, price weakness affects mining, and conditions are recovering somewhat. He explicitly does not want to over-project a temporary hash-rate decline into a permanent conclusion.
His simplified description of Pearl is: take Bitcoin, remove SHA-256, and insert matrix multiplication. Instead of competing independently for energy, mining could connect with the vast effort to convert energy into intelligence. He likes preserving a Bitcoin-like monetary design rather than adding broad programmability, and says proof of stake is “bad money” in his framework. The attraction is a differentiated monetary exposure, not merely another AI-themed application token.
23. Research begins by finding upstream thinkers and testing what seems stupid
Tulip says he lacks a rigid research template because markets occupy most of his waking attention. He deliberately seeks “off-path thinkers”: fiction, lawyers, new AI tools, and unfamiliar protocols all contribute to how he forms connections. When a rising project looks stupid, he buys a little and tries it. The discipline is exposure to evidence before dismissal, rather than treating the first reaction as a finished judgment.
Before Twitter likes became private, he traced good traders backward through what they liked before posting their own ideas. Following those chains could lead to an upstream thinker with 400 followers or fewer than 5,000. He applies the same method to podcast guests, books, and newsletters: someone with good taste may point to valuable sources outside an investor’s familiar subject area.
One discovery is a fast-rising chart associated with a Russian-ruble stablecoin, whose name is unclear in the captions. Tulip links interest to capital flight during the Ukraine war and conversion into Tether. Another route is inspecting GitHub integrations and commits around projects discussed by Vitalik, including privacy-related work. His point is that product dependencies and actual development can reveal connections that headline market commentary misses.
He also finds revenue-producing, buyback-oriented projects between roughly $5 million and $100 million in value, but identifies a moral quandary. Publicizing a legitimate undervaluation could attract followers who overbid it, creating an exit for him. He says he has not promoted those discoveries in that manner: “Integrity’s the most important thing in crypto.” The research opportunity and the audience-incentive problem appear together, rather than as separate subjects.
24. A dashboard multiple starts the investigation; persistence of revenue finishes it
Tulip gives GMX as an explicitly unfinished example and tells listeners not to rush into the low-cap token. Looking at DeFi Llama, he sees a valuation around 1.2 times revenue alongside several months of growth and asks whether it might be cheap. The multiple does not answer the question. He wants to investigate the actual markets and activity responsible for producing that revenue.
AI helps retrieve underlying information: “Claude, like, go pull, like, the actual holders on this.” His technical background makes it easier to check whether growth comes from something temporary or a mechanism he thinks will persist. The investment case is not merely that revenue rose. It is understanding why it rose, then connecting that explanation to a broader view of how the market develops.
Ansem emphasizes the time horizon required for information to become reflected in prices. Three or four months of weakness is not necessarily long against a thesis lasting six months, a year, or more. He recalls repeatedly wondering why an asset was not higher, only to recognize that other investors had not reached the same conclusion. His accompanying warning is that excessive leverage or short deadlines can prevent waiting.
Tulip adds an uncomfortable effect of daily media work: discussing the market every day made him more short-term oriented, and he had to relearn longer-term thinking. Sometimes the right content is explaining that there is no good trade. Ansem says producing the show has deepened his research without changing his trade frequency. Their experiences differ; constant market exposure does not automatically improve every part of execution.
25. Trading entertainment rewards communication as well as trading skill
Tulip’s route into the business runs through electronics and firmware work associated with CERN and a Beijing particle-physics lab, radar jammers, and four years in engineering at Messari. Crypto occupied his free time until he took a substantial pay cut to work in it. A public Aster trade brought him onto ThreadGuy’s stream, then into Counterparty’s creative work, partnerships, and distribution operations.
The career decision was partly about enjoyment: “Am I about to go from being a successful engineer to working for a Twitch streamer?” He chose the job because it looked exceptionally fun. The commercial thesis is more specific than simply liking streaming. He and the Counterparty team believe active investing is becoming entertainment, and that people capable of explaining trades compellingly can build valuable audiences around it.
His chess analogy distinguishes the best player from the biggest creator: Magnus Carlsen can be outstanding without being the largest entertainment channel. Banks extends the parallel to gaming, where elite decision-making and entertaining delivery are separate strengths. Trading content therefore need not be reserved for someone with a nine- or ten-figure P&L. It needs enough genuine ability and an audience that receives something worth returning for.
Kimchi supplies Counterparty’s observable distribution example: an offhand discussion became its best TikTok, then the next day’s Kimchi discussion did so again. Banks explains the feedback loop more bluntly, acknowledging that he once announced he would farm a big clip and told a “fabricated story” and an “exaggerated story” about Kimchi. The account is evidence of deliberate attention engineering, not corroboration of the lore being amplified.
26. The media business compounds through reach and distribution efficiency
Tulip asks listeners to imagine a stream with 100 live viewers generating a million clip impressions. The commercial audience is much larger than the live counter suggests. He focuses on two measures: live viewership, which seeds engagement, and total impressions, which reflect the distribution operation around the show. Clips are not a peripheral byproduct; converting a stream into widely viewed moments is part of the product.
Growth can occur in both dimensions. More live viewers expand the starting audience, while better moment selection, clipping, and distribution increase impressions per viewer or per stream. Tulip says Counterparty has grown both and has not yet found a ceiling. He calls the combination an “exponential business”: the practical claim is that improvements in audience size and distribution efficiency can multiply rather than merely add.
The operating questions are concrete: what sponsors pay per impression, what clippers cost, whether the impression base can grow, and how efficiently output is produced. Skilled clippers are difficult to find and need to be retained and treated well. Tulip describes that distribution backend as a major part of his back-office responsibility, rather than presenting the entire company as a host speaking spontaneously into a camera.
Ansem adds a trading benefit to working as a research group: discussion exposes missing assumptions. His example is a friend, Jeremy, being bullish on Serum, which prompted Ansem to investigate both Serum and Solana and conclude, “I don’t think it’s Serum. I think it’s Sol.” He thinks writing and speaking with people who disagree can create insights that reading another isolated market post would not.
27. Audience wealth and public records are proposed checks on bad incentives
Tulip argues that sustained financial-media economics favor an audience retaining and growing capital. Content is free and paid for by sponsors, who want customers with money and trading volume. A viewer who blows up provides little future activity. The claimed alignment is therefore lifetime customer value: “Your audience getting richer” is the win condition he describes, while acknowledging that saying something on stream cannot itself guarantee that outcome.
He contrasts that model with making a quick exit at the audience’s expense. The long-term business requires repeatedly providing useful information and preserving relationships. Banks agrees, pointing to trading volume as central to sponsor economics. Their claim is about incentives and intent, not an audited demonstration that every sponsor, recommendation, or trading decision benefits the viewer; the episode supplies the operating logic rather than quantified audience returns.
Tulip thinks public leaderboards can gradually reward better behavior. On FOMO, someone displaying large unrealized gains can fall down the rankings after losing them, while traders who preserved gains rise. He pairs that mechanism with public on-chain records and investigators such as ZachXBT. His expectation is that durable positive operators become distinguishable over time, reducing how much depends on a persuasive narrative at a single moment.
On copy trading, Banks argues that endlessly dumping on followers would eventually exhaust the follower capital available. Tulip suggests some rapid copying may come from sophisticated small funds or bots responding to publicly profitable traders such as Frank. Those are their explanations, not demonstrated identities behind particular transactions. Both think social interfaces make previously specialist wallet-following information accessible to less experienced participants.
28. Crypto’s reputation problem divides the hosts from Tulip’s adoption-first view
Banks sees crypto’s poor branding as an impediment to adoption, especially when people collapse different events into the same accusation. His example is a coin being rugged versus an unrelated community coin losing attention to another version: charts can look similar while responsibility differs. Eli Ben-Sasson adds that CT attacks even wholesome material, recalling hostile responses about his family after his following grew; changing the tone of content alone is no cure.
Tulip asks whether fixing the overall brand is necessary. People say they hate the internet, social media, and AI while using all three constantly. If useful crypto products keep growing, reputational dislike may not stop adoption. His personal answer is to run his own operation well rather than solve every public narrative: “It will resolve itself, or it won’t resolve itself,” but he can still try to behave responsibly.
Banks’s pushback is that finance demands deeper trust than using a search engine or homework tool because users commit capital. Tulip accepts that distinction as “a good point.” Banks expects the categories eventually to blend into ordinary finance and trading, perhaps without people saying “crypto” at all. Until then, he sees media and credible examples as important bridges; Ansem names Hyperliquid’s Jeff as an especially strong builder example.
29. Stock-market narratives can finance the reality they initially exaggerate
Ansem finds AI-company communications unsettling: relatively small stock declines seem to trigger reassuring headlines, making market prices appear almost as important as technical progress. He also worries about eventual OpenAI and Anthropic issuance after extraordinarily fast rises toward valuations he loosely describes as $1.5–$2 trillion. New models could improve the economics again, he concedes, so he expects substantial changes of mind rather than a simple one-way market.
Tulip disagrees that energetic promotion is itself strong evidence of weakness. He calls his view “hyperstitious”: acting around an imagined future can help construct that future. A higher stock price can improve financing terms, support employee compensation, and give a company resources to grow. He thinks corporate leaders have learned something crypto discovered earlier, when less constrained communications made reflexive price-and-adoption loops easier to observe.
Tesla during COVID is his main example. In his telling, Elon promoted a valuation the existing business could not yet support, issued substantial stock, financed gigafactories, and then expanded production toward the implied future. “The financial illusion that Elon created back-constructed the reality moving forward.” The argument is causal: an elevated price can become an input into operating success, rather than merely an inaccurate statement about current fundamentals.
He extends the loop to crypto. Publicity around a major successful trade might attract users to one exchange rather than another, creating the activity that later justifies enthusiasm. HYPE used as collateral supplies a more direct, though he says relatively small, connection between token price and trading capacity. Ansem accepts that companies understand their market impact, while the broader concern about late-cycle equity weakness remains unresolved.
30. Tulip would rather own a differentiated opportunity than price an AI utopia
Tulip shares concern about the AI complex unless government intervention changes the picture, which he says is definitely possible. He is not interested in buying dips there. Eli Lilly appeals partly as equity exposure less entangled with that trade. The episode does not develop a drug-level investment case; his stated preference concerns a separate opportunity set alongside crypto assets where he believes substantial upside remains.
His personal-game-theory argument starts from trillion-dollar AI-company valuations. If those companies outperform assets he thinks can rise 10–100X over a decade, the world may change so radically that his original portfolio choice matters much less. “I’ll either be dead or euphoric.” The nuclear-war comparison makes the same point: he sees limited value in optimizing a portfolio around an outcome that overwhelms the ordinary purpose of investing.
Fiction provides another way he explores extreme outcomes. Both speakers have read Accelerando; Tulip is reading Daemon and Neuromancer and praises Red Rising. Ansem worries about creative thought converging into an internet hive mind. Tulip prefers a variance framework: the internet and AI can produce much more low-quality consumption while also letting exceptional creators reach scales and niches that previously were unavailable.
The practical implication he draws is that original thought may stand out more amid automated replies and generic articles. He connects his own writing, following, and job opportunity to offering reasonably clear, distinctive ideas in that environment. The broader principle echoes the earlier discussion of intelligence becoming commoditized: commitment, consistency, agency, and the ability to produce something people actually value can become increasingly consequential differentiators.
31. Mature financial rails may let old crypto-social ideas work on a second attempt
Ansem asks why relatively few projects seem to emerge from nothing into substantial new businesses. Tulip admits he expected FWB to do much better and remains interested in Paper Trade, Cards, and Venice. His explanation is that large incoming flows require mature protocols capable of absorbing them. A slower pace of financial invention may be part of the same maturation that makes institutional-scale activity possible, rather than simply evidence of stagnation.
Beyond that initial concentration, Tulip hopes “everything old becomes new again.” ENS, IPFS, decentralized web infrastructure, and Lens-like shared social graphs could be reconsidered once scaled networks have real users. He gives a creator’s concrete frustration: Counterparty cannot simply carry ThreadGuy’s Twitter followers across every other distribution platform. A shared graph with different interfaces and algorithms could address a genuine operational problem rather than exist solely as a token story.
Ansem’s version is trading as “the world’s greatest MMORPG,” whose characters and histories are still scattered across disconnected profiles. Wallets can prove early protocol use, profitable activity, and experience, while ZK tools might selectively conceal details. If trillions of dollars and their associated users move on chain, he thinks systems that surface reputation and activity could become more useful. Tulip agrees, but says buying the old social tokens now would be premature.
Friend.tech keeps the optimism grounded in losses. Both loved the product; Tulip traded through spreadsheets and recalls high volume, large fees, gains, and eventual damage. Ansem says he may have had the largest airdrop, round-tripped it entirely, and “didn’t sell a single dollar.” Neither treats that experience as proof the category is finished. They expect to keep experimenting, while the exchange preserves how an exciting social product can still be a disastrous holding.