Privacy is a fundamental human right. Yet no network delivers permissionless, programmable, self custodial privacy, with formal verification and post quantum security. Most projects are compromises dressed up as solutions: private payments without programmability, privacy layers hostage to someone else's governance, "privacy" chains with invite lists and emergency decryption keys. All compromises which lead to dead ends. This matters more than ever. AI will supercharge surveillance and widen the divide between those with access to frontier intelligence and those without. Financial sovereignty has never been more important. what does a solution look like? A permissionless privacy network from day zero: programmable, formally verified, and built for the post quantum post AI era. And most important of all, lean. Lean from an engineering standpoint, lean from a bureaucracy standpoint, and lean from a Lean4 standpoint. Any serious attempt must be uncompromisingly cypherpunk. Lean, fast moving, zero bureaucracy and resistant to political capture. Its principles cannot be sacrificed for funding, convenience, short term adoption, or the whims of some VC. when it comes to privacy, censorship resistance, and sovereignty, even a minor compromise is a complete compromise. When did the blockchain industry become so afraid to build cypherpunk technology? So afraid to go against the grain and build something genuinely bold? Somewhere along the way, compromises became the norm, and what could have been great became mediocre.

Really looking forward to giving a workshop on the math behind the Ethereum proximity prize with @aiv768 at @EFDevcon! We will explain the prize problem from the basics, then explain latest research results related to proximity gaps!

seeing the P2IDE (pay to ID extended) note in action with real funds is super cool! Reclaimable, time lockable public or private UTXO. Sending USD privately on Miden! I remember writing this exact UTXO script in masm more than a year ago, now all that low level Miden assembly is all wrapped up beautifully by the @joinbread wallet! The level of engineering brainpower to make this all work end to end is craaazy. Cryptography team, VM team, compiler team, node team, protocol team, client team, wallet team, devrel team, business team… Need to write a separate tweet about each part of the Miden stack… 🤯 I’ve been testing this all on testnet for years at this point, but seeing it with real assets finally (albeit super small amounts) is just incredible. Btw the wallet itself is a private multisig!

I haven’t been as excited about a programming language since when I first started learning Rust. Lean4 is an extremely powerful language. However, one thing it’s missing is good tutorials and examples! So I built an online Lean4 learning tool! I predict Lean becoming one of the most popular languages in the coming years amongst AI labs, blockchain researchers, and agents. It’s what powers verifiable auto-research. Lean4 increases the velocity at which humanity can make mathematical discoveries. To be ahead of the curve, learn some lean today! Open to feedback and suggestions! 😅

Woohoo! I am at the top of the leaderboard! 🎉🥳 Big thanks to i34-9 for acknowledging my earlier queued submission was valid! Really appreciate it! The PR which got merged: The approved PR had 76 of its 81 submission files byte-for-byte identical to my earlier queued PR, which got stuck bc of Yukon's flaky infra: So what did I do to get this cutting edge result? I still did manual PR review and manual simplification of the Lean written by the agent. Had to tell the agent a lot, “this is way too verbose, this can be simplified.” Honestly, just basic PR cleanliness & basic CS principles helped push the frontier. Overall, is a cool project. Glad I am on the leaderboard, but submissions should stay private until verification finishes. Otherwise, others can resubmit your work while you’re still waiting for CI. If i34-9 hadn’t added me as a co-author, I could have ended up with 0 leaderboard credit. The CI infrastructure was having problems, and my earlier submission failed with this message: “The submission was never checked — the verifier could not reach a verdict. This is an infrastructure fault, not a judgement on the proof.” Also, if anyone knows who i34-9 is, I’d really like to DM them to thank them personally. Anyways, I am very happy. It’s a team effort after all!

Mitchell Amador is a true visionary in our industry. DeFi owes him one big thank you for preventing billions lost in hacks. Insane speaking skills too 😅 @immunefi @consensus2026