On The Brink with Castle Island

Prediction markets stumble during major elections, while legal and security pressures mount. Google enters institutional crypto custody, and Hyperliquid perpetuals move toward U.S. markets. Public companies push back against tokenized stocks as new share-trading models emerge. AI uncovers Bitcoin bugs, quantum advances challenge cryptography, and stablecoins find traction in machine-to-machine payments.

Could fusion reactors turn mercury into thousands of tons of gold, disrupting scarcity economics? The discussion also explores AI agents infiltrating bank accounts, cyberattacks exposing fragile internet infrastructure, and new rules for stablecoin rewards. Other topics include tokenized stocks, compute derivatives, Bitcoin demand, and Polymarket’s battle with Kalshi.

A whirlwind tour of crypto’s biggest fault lines: banking access, Bitcoin’s quantum future, and Google’s latest quantum breakthrough. The discussion also explores a Ledger supply-chain attack, OpenAI’s interest in Trail of Bits, DWF Labs’ $141 million lawsuit against BitGo, prediction-market regulation, AI infrastructure risks, and Richard Heart’s detention in Istanbul.

AI is running fantasy football teams and making startling advances in mathematics, raising questions about proof, creativity, and human expertise. The discussion turns darker with warnings that advanced AI could threaten elliptic-curve cryptography. Also covered: crypto dealmaking, tokenized stocks, prediction-market regulation, meme-coin collapses, stablecoin wallets, and the speculative frenzy around Pokémon cards.

Esther Wong, founder and CEO of Cashi, brings deep payments experience from U-Trip and Crypto.com. She discusses who uses stablecoin payments, why global users want digital dollars, and how Cashi is building spending and cashback tools. They explore tokenized securities, blockchain competition, AI-powered payments, irreversibility, regulatory challenges, and Cashi’s roadmap for rewards, licensing, yield, and tokenized stocks.

A wide-ranging look at AI’s frothy valuations, the “Fat Model” hypothesis, and whether autonomous agents could trigger lightning-fast bank runs. The discussion also covers Hester Peirce’s SEC legacy, new crypto regulations, prediction-market battles, Ethereum’s 2030 roadmap, accelerating quantum threats, Illinois’s delayed crypto tax, and California’s urban governance woes.

A wide-ranging discussion covers data-center noise, water use, and the politics of AI infrastructure. The conversation turns to dollar stablecoins, coordinated crypto regulation, Fairshake’s campaign against Sherrod Brown, tokenized portfolios, and Coinbase’s IPO access. They also examine prediction markets, BitMEX’s closure, Strategy’s latest Bitcoin purchase, a Coldcard hack rescue, Anthropic’s ambitions, and whether effective altruism has become a cult.

This week’s roundup examines the Clarity Act’s Senate collapse, crypto tax and Bitcoin reserve proposals, and Circle’s launch of a stablecoin-focused blockchain. It also covers alleged Hyperliquid insider trading, Caroline Ellison’s return to the nonprofit world, and whether AI can leak market-moving secrets. Plus: a bizarre Venezuelan oil deal involving USDT and a major federal quantum-computing challenge.

Crypto security takes center stage as a Liquid sidechain exploit drains reserves and rekindles privacy concerns. The discussion also tackles Anthropic’s warnings about AI extinction, quantum computing’s threat to Bitcoin, and whether AI systems can leak material nonpublic information. Silvergate’s collapse gets a fresh look, alongside prediction markets, Tether’s credit ambitions, and a surprising OpenAI math breakthrough.

A rogue AI agent escapes into Hugging Face, sparking debate over security, testing, and whether AI makes us less capable. Prediction markets head toward the Supreme Court, while Robinhood Chain attracts meme-coin traders and Strategy resumes its Bitcoin buying. Also covered: DeFi’s latest hack, Revolut’s U.S. bank approval, stablecoin remittances, Hyperliquid’s regulated expansion, and whether RSA-260’s factoring should cause panic.