Sneak peek 👀 Our web version is coming. Fully self-custodial BTC → XMR atomic swaps, right in your browser. Secure swaps with no intermediary, no arbitrators, but complete mathematical soundness. No downloads. No installs. Works in Tor Browser for full privacy. Swaps up to 5 BTC. Low fees. Soon.

Stablecoin settlement never sleeps. Bank liquidity rules do. Tuongvy Le, General Counsel at @veda_labs and a former GC at a federally regulated bank, walked through what 24/7 programmable money does to safety and soundness on the latest episode of Tokenized with hosts @sytaylor and @cuysheffield. Wholesale deposits are hot money. Consumer deposits are sticky. Silicon Valley Bank showed what happens once the hot money can move instantly. One bank modelled the hit to its liquidity coverage ratio from offering weekend stablecoin settlement to institutional clients. The result was some deposit flight and a real balance sheet cost. "You can offset that by charging fees for people to make payments over a weekend that they couldn't before." 🎙️ Listen to the latest episode on

CLAUDE OPUS 4.8 JUST BROKE A $9 BILLION PRIVACY COIN. 🚨 A security researcher used Anthropic's new model to find a 4-year-old soundness bug in Zcash's Orchard pool. The flaw could create unlimited, undetectable counterfeit $ZEC. Discovered May 29. Patched June 3. Could have been exploited for over 4 years. The team has no answer.

lowercase snarks Words like laser, scuba, radar began uppercase. LASER — Light Amplification by Stimulated Emission of Radiation SCUBA — Self-Contained Underwater Breathing Apparatus RADAR — RAdio Detection And Ranging When a technology matures and becomes reliable, trusted, commoditised, it earns the lowercase. Lean Ethereum is a bet on snarks, not Succinct Non-interactive ARguments of Knowledge. Post-quantum security. Provable soundness. End-to-end formal verification. Deep cryptanalysis. Real-time proving. zkVM programmability. Simplicity and elegance. All essential for the lowercase. All inevitable. Ethereum L1 has 10y uptime and $1T secured with hashes and signatures, our cryptographic workhorses. I believe in 100y uptime and $1Q secured with snarks, our cryptographic jet engines. * L1 scale — 10K TPS gigagas scale with real-time zkEVMs * L1 security — post-quantum security with snarked signatures * L1 privacy — Zcash-grade stealth with wormholes (eg EIP-7503) Shipping snarks is a cryptographic Manhattan Project, one the EF is investing tens of millions into: * verified-zkevm[.]org — formal verification * poseidon-initiative[.]info — deep cryptanalysis * ethproofs[.]org — real-time proving * proximityprize[.]org — provable soundness * zkevm.ethereum[.]foundation — enshrinement * pse[.]dev — privacy Step by step, the EF is evolving into a snark-first org: * cryptography team — driving soundness and cryptanalysis * snarkification team — driving formal verification * zkEVM team — driving protocol integration * Ethproofs team — driving real-time proving * PSE team — driving privacy * PQ consensus team — soon™ Believe in something magical. Believe in lowercase snarks.

I just had a fascinating call with Eyad Alkassar (@AlkassarEyad) about how ChatGPT 5.5 Pro helped move the frontier of an open problem in computer science. The problem sounds abstract, but it is surprisingly intuitive: Imagine four people dividing nine objects. Each person may value them differently. The goal is to find a division so fair that, after any single object is removed from someone else’s pile, nobody would prefer what remains to their own pile. This criterion is called envy-freeness up to any good, or EFX. Whether complete EFX allocations always exist for four or more people remains open. For four people, the previous general guarantee stopped at seven objects. A new preprint by Eyad Alkassar, Mahmoud Fouz and renowned computer scientist Kurt Mehlhorn now presents a proof for up to nine. The story behind it is remarkable. Over dinner, Eyad challenged Mehlhorn to give AI one of the harder open problems in his field. Eyad and Mahmoud are both computer science PhDs who had spent years building startups and were newcomers to fair division. They built a research workflow around several AI systems. At one point, Fable had questions about one of Mehlhorn’s papers and repeatedly urged Eyad to contact him. After Eyad refused twice, the AI pointed out that they were only around 30 kilometers apart and suggested that a short drive would be worth it. According to Eyad, the situation became even stranger once the proof was finished. Fable then advised him against sending the result to Mehlhorn, warning that he might be a competitor and that an incorrect proof could be embarrassing. Eyad contacted him anyway. Mehlhorn reviewed and verified the proof architecture, refined the arguments and joined the paper as a co-author. According to the paper, Claude Fable performed the analysis and lemma proofs, authored the trusted verification layer and audited its soundness. ChatGPT 5.5 Pro proposed most of the candidate attacks and optimizations. Eyad and Mahmoud selected the systems, assigned their roles, coordinated their communication and provided the computing infrastructure. The proof divided the space into 36,152 canonical cases after symmetry reduction. The closing run verified all 122,553 certificates using Z3 and completed 141,878,161 per-clause soundness checks with zero reported failures. The general theorem for an unlimited number of goods remains open. This result moves the known frontier from seven goods to nine. The authors describe the process as “AI led, human assisted and verified.” For me, this is one of the clearest examples yet of current AI systems contributing substantial work to a new mathematical proof. I found the story too remarkable not to share. Paper: