In this episode, we dug into a rough stretch for Bitcoin infrastructure and security, starting with the Coldcard entropy exploit, the messy hotfix fallout, and related concerns around BTC Pay Server and Lightning vulnerabilities. From there, we explored how AI-assisted security research is rapidly changing the landscape: open-weight frontier models and tools like Ghidra and Wireshark are making it dramatically easier to inspect code, reverse engineer binaries, and uncover weaknesses in both open and closed-source Bitcoin software. We talked through the tradeoffs between closed proprietary AI providers and more permissive open-weight models, the hardware and cost constraints of running these systems, and why this new era of AI-driven auditing is likely to expose a lot of long-hidden assumptions across the Bitcoin stack.
We also spent a lot of time on Bitcoin mining: how AI-based reverse engineering is being applied to proprietary mining firmware, what that could reveal about dev fees, remote controls, and operational risk, and why open-source mining infrastructure matters more than ever. We covered the BIP 110/Ocean controversy, miner and node power dynamics, and the urgent need for better decentralized pooling options and more home miners running their own nodes. We wrapped with updates from the 256 Foundation ecosystem, including progress on Bitaxe Proto and Bonanza support, Magina ports, ideas for cheaper home mining nodes, and upcoming hands-on workshops and events for builders who want to get deeper into mining and open-source infrastructure.