A lot happened this week across Cannes, quantum research, and the broader privacy ecosystem. Here’s my read on it. @EthCC drew a smaller crowd than previous years, and honestly, that’s fine. The traveling crypto circus is leaner now and only the people who need to be there show up. What struck me most wasn’t the conference itself but the contrast it made clear: tradfi attendees aren’t talking price action. They’re tracking the CLARITY Act, studying RWA tokenization structures, and thinking in decades. The crypto-native crowd is still fixated on the next bull run. That gap is widening, not narrowing. On the quantum front, things got more concrete this week. Google’s research now estimates cracking Bitcoin’s ECDSA could require fewer than 500,000 physical qubits, a dramatic reduction from previous consensus figures. An attack could run in roughly 9 minutes at a 41% success rate, faster than a Bitcoin block confirmation. Google’s own internal migration deadline is 2029. The U.S. government has mandated a full transition away from non-PQ-secure schemes by 2035. @alexanderlee314 made the point clearly at @EthCC: the window to migrate away from elliptic curve cryptography is closing. What most commentary is missing is what this means specifically for ZK systems. Almost every privacy-enabled chain deployed today, Aleo, Aztec, Zcash, uses SNARKs built on elliptic curve pairings. PLONK, Halo 2, the whole stack. Shor’s algorithm breaks these directly. The proofs that give you privacy and compression are built on assumptions a large enough quantum computer would destroy. As @bobbinth laid out in detail, migrating a SNARK-based protocol to post-quantum primitives isn’t a patch job. In some cases it means rebuilding from scratch. STARK-based systems like @0xMiden have a different profile. STARKs rely on collision-resistant hash functions, not elliptic curves. That’s a structural advantage, not a rebrand. The proof core survives a post-quantum world in a way that curve-based systems simply don’t. There’s still work to do around key management, network encryption, and identity layers, but that’s renovation, not demolition. On the architecture side, @0xMiden also published a clean breakdown of how the protocol is actually structured this week. Three building blocks: Accounts, Notes, Transactions. Every account is a smart contract, your assets live inside it, not in a shared global contract. Accounts communicate through Notes: programmable messages that carry assets and logic. A transaction is what happens when an account produces or consumes a note. Simple, composable, and built for privacy from the ground up. @alexanderlee314 described Miden Notes as programmable UTXOs, powerful enough to represent CLOB orders, option contracts, HTLCs, and more. Separately, Miden engineer Paul Schoenfelder walked through how @0xMiden compiles Rust for MidenVM at @EthCC, explaining why the team routed through WebAssembly rather than building a custom LLVM backend or writing a compiler from scratch. Not a flashy announcement, but the kind of foundational engineering decision that compounds quietly over time. Zooming out, the broader privacy infrastructure story is accelerating. Nasdaq-listed SOL Strategies agreed to acquire Darklake Labs this week, folding their ZK team in to build private onchain finance for institutions. China’s tax and financial regulators jointly directed banks to deploy blockchain and privacy-preserving computation for lending, the specific use case being credit verification without raw tax data leaving government systems. Two very different contexts, same underlying signal: privacy is no longer a niche feature. It’s becoming regulatory infrastructure. The teams building now with hash-based proof systems and privacy-first architecture are in a structurally better position than they’ll get credit for until it’s obvious. @0xMiden is one of them.