MARKET BUBBLE SEARCH · episodes ←Every episode Ask ThreadGuy anything

ThreadGuy · October 14, 2026 · 1 hr 18 min

Can AI Break Crypto? Mert, Illia Polosukhin and ZK Dragon

ThreadGuy hosts Mert, Illia Polosukhin, and cryptographer ZK Dragon (Dev) to unpack Justin Drake's quantum/AI decryption scare and whether AI can realistically break crypto's cryptography. The panel largely concludes the fear is overblown given massive compute gaps, while discussing how Near, Zcash, and Solana are built or upgrading to handle post-quantum risk and AI-powered smart contract exploits.

Watch the recording

Topics

ThreadGuy opens explaining his own understanding of the two threats to crypto: quantum computing cracking public keys and AI accelerating brute-force decryption, using a Naruto sand-shield analogy.

ZK Dragon (Dev) explains cryptography as hard math problems like RSA factoring and elliptic curve discrete log, and says OpenAI's 700+ new math results showed no meaningful progress on these core problems.

Dev states that even with all Bitcoin mining hardware repurposed, it would take a trillion times more compute to break elliptic curve cryptography based on current knowledge.

Illia argues blockchains need to stay adaptable like Chrome did when it quietly added post-quantum cryptography support, rather than ossifying like Bitcoin and Ethereum tend to.

Mert explains why Zcash is well positioned for quantum risk due to its narrow scope (store of value only) and its cryptographer-heavy dev team, while Solana has proven it can handle major rewrites from its 2022 outages.

Illia describes how NEAR detached cryptography from the account model, allowing multiple key types (ECDSA, Edwards curve, and this year added lattice-based post-quantum ML-DSA) and multi-signature protection for large holdings.

Mert argues Justin Drake's bunker-mode post was exaggerated, says losing money by panicking is far more likely than this risk occurring, and notes Zcash and Near already have contingency plans outlined.

Illia details NEAR's push on cheap formal verification (lemma-proving 200x cheaper than GPT/Claude) and building a fully formalized VM so developers can prove contract properties, addressing AI-powered hacks at the protocol level, which he says is the more real daily risk than quantum.

Dev and Mert discuss who could realistically achieve a cryptographic break, concluding it would almost certainly be a nation-state like the US or China since the hardware/power requirements (tens of gigawatts for a million years) are beyond any private actor, and quantum computers would likely be used to decrypt state communications rather than attack blockchains.

Dev and Illia discuss an end-state for 'perfect' cryptography, describing multi-signature setups requiring multiple cryptographic assumptions to break, while Illia notes NEAR upgrades every six weeks to stay agile against evolving hardware and algorithms.

Mert closes by saying smart contract hacks are a far bigger practical risk than cryptography breaking, and that core developers are already prepared, so bunker mode is unwarranted.

This summary was written from the recording's transcript by a language model. For the exact words, ask the ThreadGuy archive: every answer there is quoted word for word and linked to the second it was said. Market Bubble Search is independent and not affiliated with the show. Nothing here is advice.