Ideas
Bullish on AI crypto creators.
Mike remains very bullish on AI creators in crypto. He mentions AIXBT but explicitly says it is not necessarily that ticker or project; the broader AI-crypto creator theme is what he finds super interesting.
Agentic payments will be big.
Mike believes onchain agentic payments are still a ways out, but they are inevitable and will be big; whoever facilitates them will make a lot of money. This is a long-term thematic bet on agent payment infrastructure.
RWAs are real and working.
RWAs are very real and are one of the areas actually working in crypto, driven by institutional and market demand. Mike treats them as a validated use case even while other AI-crypto themes remain nebulous.
Stablecoin demand is enterprise-heavy.
Stablecoin issuers are inundated with demand from institutions, chains, and enterprises. Myles sees this as one of the areas with real market pull and enterprise-heavy demand, even though much of the value may accrue to companies or equities rather than tokens.
Policy spend beats HYPE buybacks.
Hyperliquid's $28 million Policy Center led by Jake Travinsky is an incredible use of funds. Regulatory and policy spending can expand Hyperliquid beyond crypto-native users, help it compete with Binance, and produce upside worth many multiples of the amount spent; he would prefer this to a token buyback.
ETH and SOL retain network effects.
Parts of the fat protocol thesis still stand: Ethereum and Solana are the biggest winners, have durable developer ecosystems and liquidity, capture value from applications, and are not going away soon. Their EVM and SVM network effects make it difficult for other chains to compete.
HYPE airdrop built native distribution.
Hyperliquid's distribution borrowed the infrastructure playbook: a great product plus a large airdrop made stakeholders rich, which drove social-media sharing and gave Hyperliquid native distribution. Xavier sees this as the closest app analogue to early Ethereum and Solana and doubts the playbook can be copied successfully.
Polymarket media distribution reaches mainstream.
Polymarket gained distribution through traditional media because its stats and odds were treated as better than polls, exposing its data to non-crypto audiences. Xavier highlights this as a novel distribution strategy that cracked mainstream media, which most crypto apps cannot do.
Ethena combined DeFi whales and exchanges.
Ethena built a trusted brand early, aligned crypto-native DeFi whales with capital, marketed well, and used exchange distribution including Binance; it is now moving toward banks. Xavier sees its distribution strategy as unique and hard to copy.
FOMO uses social trading distribution.
FOMO is a social trading app that borrowed Web2 and TradFi distribution tactics: easy onboarding, strong UX, trade sharing, and commenting. Xavier sees that social product design as a novel crypto distribution strategy and a source of virality.
Axiom referrals turn users into distributors.
Axiom has executed cashback and referrals well, incentivizing its community to spread the product. Referrals had not worked well in crypto before Axiom, but Axiom turned users into distributors.
Telegram bots enable viral distribution.
Trojan and Banana Gun leveraged Telegram-native distribution. The model is simple, easy to copy, share, use, and add people to, making it easy to create virality, which Xavier frames as an effective crypto distribution channel.
Prediction markets have real demand.
Prediction markets are among the few crypto areas with clear market pull and demand over the past two years, alongside stablecoins and RWAs. Myles uses this to validate prediction markets as one of the working use cases.
General-purpose infra no longer works.
Myles argues that crypto infrastructure cannot simply launch a general-purpose framework and promise a huge design space. Projects now need opinionated use cases with clear market pull today while showing investors how the wedge expands over time.
Crypto data could explode next year.
Xavier sees crypto data as underexplored. APIs are too expensive and hard to access, while agents want cheap, fast, relevant, contextual data at the right time. He expects the data area in crypto to explode over the next 12 months using crypto rails.
Crypto API payment rails opportunity.
Xavier argues private data becomes more valuable in an agent-driven world, and the business model for how agents pay for APIs is unsolved. Building crypto rails for API and data payments is an opportunity because agents will rationally seek the best data but need efficient access and payment coordination.
Open-source AI beats closed models.
Xavier thinks open-source AI will ultimately beat closed systems. Incumbents may acquire open projects or shut APIs, but that pressure will make open-source AI shine, favoring open, multi-model, multi-agent architectures.
This Bell Curve video, published February 20, 2026,
features Mike Ippolito, Myles O'Neil, Xavier
discussing AI-SECTOR, Onchain agentic payments, RWAs, STABLECOINS, HYPE, ETH, SOL, POLYMARKET, ENA, FOMO, AXIOM, Trojan, BANANA, PREDICTION MARKETS, General-purpose crypto infrastructure, Crypto data, Crypto API payment rails, Open-source AI.
17 trade ideas extracted by AI with direction and confidence scoring.
Speakers:
Mike Ippolito,
Myles O'Neil,
Xavier
· Tickers:
AI-SECTOR,
Onchain agentic payments,
RWAs,
STABLECOINS,
HYPE,
ETH,
SOL,
POLYMARKET,
ENA,
FOMO,
AXIOM,
Trojan,
BANANA,
PREDICTION MARKETS,
General-purpose crypto infrastructure,
Crypto data,
Crypto API payment rails,
Open-source AI
