AllianceDAO's Wang Qiao Says Crypto Can Outprepare TradFi for AI and Quantum. The Signature Layer Disagrees.

Wang Qiao, co-founder of the crypto accelerator AllianceDAO, has a thesis that fits inside a single sentence: within two years, the world will realize that crypto — attacked relentlessly, publicly, for fifteen straight years — is far better prepared for AI and quantum threats than the internet or traditional finance it is so often measured against. No whitepaper. No metric. No signature-migration roadmap. Just the sentence, plus a framing that converts a trillion-dollar attack surface into a badge of honor.

That framing is rhetorically excellent, and I want to be precise about where it is worth something. It is worth a lot in the narrative layer, where positioning gets priced in weeks. It is worth considerably less in the cryptographic layer, where I have spent enough time reading audit findings to know that being shot at is not the same as wearing armor. The whale didn't blink. The wallets didn't move. A sentence moved, and in this market a sentence is the cheapest form of capital there is.

AllianceDAO is not a random handle in a Telegram group. It runs one of the more consequential early-stage funnels in crypto — an accelerator program that takes equity and token warrants in pre-launch teams, hands them a demo-day runway, and exits into whatever liquidity the cycle happens to offer. Wang Qiao's own background, built through research-side roles in the data and analytics layer of this industry, gives his statements a gravitational pull an anonymous account could never manufacture. That matters more than people admit. When someone with a research pedigree says something optimistic, desks like mine quote it verbatim; when an anonymous account says the identical thing, it dies in the timeline by lunchtime.

The timing is not accidental either. The two live narrative buckets in this cycle are AI-adjacent compute and post-quantum readiness. Both are long-duration themes, both are hard to falsify inside a quarter, and both let any project — or any incubator — claim strategic relevance without shipping a single line of code. I have watched this pattern through three cycles. The claim arrives, the vocabulary gets defined next, and the capital gets allocated after that. The allocation is what the claim is for.

Two assertions are bundled into that sentence, and they do not share an evidentiary basis. Separate them, and the confidence splits cleanly.

The adversarial-security claim is the strong one. Crypto genuinely runs the most aggressive offensive-security culture in financial technology. Bug bounties are denominated in seven figures. Red teams ship against live mainnet rather than a staging environment nobody uses. Formal verification, once confined to aerospace and academic cryptography, is now a standing line item in DeFi audits. I have sat in rooms where a protocol's core team burned nine days arguing over a single reentrancy path no attacker had ever found. That culture is real, and it is measurably more hostile to sloppiness than the average bank's quarterly penetration test, which typically ends in a PDF and a promise.

Now the ledger. Industry forensics desks put annual crypto theft in the billions every year — roughly $3.8 billion in 2022, roughly $1.7 billion in 2023, back above $2 billion in 2024 — with cross-chain bridges and private-key compromises absorbing the bulk of it. Set those figures against the sector's total value and you do not get hardening. You get a system where the cost of a successful breach remains low relative to the prize, and where the median exploit is not a clever cryptographic break but a leaked key, a compromised signing environment, or a phone number ported at 2 a.m. Three bridges alone accounted for a staggering share of one year's losses, and not one of them failed because of a mathematics problem. Being attacked constantly tells you the attack surface is enormous. It does not tell you the defense matured. Adversarial pressure selects for survivors, not for best practices — and the two are most frequently confused by the party selling the narrative.

The quantum claim is where the argument inverts, and this is the part that deserves the least charity.

Bitcoin and Ethereum secure ownership with ECDSA over the secp256k1 curve. That construction's hardness rests on the elliptic-curve discrete logarithm problem. Shor's algorithm solves that problem efficiently on a sufficiently large fault-tolerant quantum computer. Symmetric primitives — hashes, block ciphers, HMAC — do not face the same cliff; they degrade gracefully and get restored by lengthening keys and outputs. So the structural asymmetry runs in the opposite direction from the claim. An industry whose entire asset-ownership layer sits on discrete-log signatures is more exposed to quantum migration risk than a payments incumbent whose core settlement rails remain largely symmetric, centralized, and perimeter-defended. Traditional finance has quantum problems too — TLS handshakes, HSM firmware, correspondent-bank signing — but those are edge problems with upgrade paths. They are not the definition of who owns the money.

Exposure is also not uniformly distributed, and it is not future-dated. Any output whose public key is already revealed on-chain is a standing target for a harvest-now, decrypt-later adversary: reused addresses, the early pay-to-public-key outputs from Bitcoin's first years, exchange hot wallets that have signed outbound transactions hundreds of thousands of times. Estimates in the low millions of BTC sit in exposed or reused form. Those coins cannot be rescued by a node upgrade. They can only be moved by their owners, and many of those owners have not touched a private key since 2013.

The migration math is the least discussed and most decisive part. NIST finalized its post-quantum signature standards in 2024: ML-DSA and SLH-DSA. Neither is a drop-in replacement for ECDSA at comparable cost. A Dilithium-class signature runs about 2.4 kilobytes. The hash-based option runs from roughly 8 kilobytes to tens of kilobytes. ECDSA signatures are 64 bytes. Run that arithmetic against block space and a standard one-input, two-output transaction inflates roughly tenfold — before anyone touches a hardware wallet's secure element or a consensus change measured in thousands of lines with a multi-year activation path. Signature aggregation softens the blow. It is not free, not chain-neutral, and not deployed.

I have audited teams that ran formal verification against their own contracts and still could not tell me which of their dependencies would survive a curve swap. That is not a knock on their competence. It is a description of the distance between a security posture and a cryptographic migration program — different disciplines, different budgets, different timelines, routinely mistaken for one another in marketing decks.

Here is the angle nobody is publishing. The strongest evidence for the thesis is also the strongest evidence against it, depending on which reading you prefer. A trillion-dollar ledger of irreversible transfers is an extraordinary honeypot: public, persistent, and impossible to claw back. That property trains defenders fast. It also guarantees that the return on a successful attack is unusually high, which keeps attack volume high, which produces the loss figures above. You can read one dataset as either hardening or chronic bleeding, and the reading you choose usually depends on whether you are raising a fund or writing a postmortem.

There is a structural interest worth naming plainly. An accelerator's portfolio is priced off the perceived durability of the sector it incubates. A claim that crypto is the most battle-tested security environment on earth is not a neutral observation — it is a valuation input. I remember the summer of 2020, when I published a piece arguing that Compound's governance distribution had concentrated voting power among a handful of early addresses. DeFi purists called it FUD. The distribution data did not care then, and it does not care now. Governance is a silent coup, not a vote — and the same rule governs narratives. Whoever defines the frame collects the value.

Credit where it is earned, though: the AI half of the claim holds up far better than the quantum half. Chain-level AI defense is genuinely advantaged. Ledgers are public, immutable, and traceable — a forensic environment that fraud teams inside any bank would trade a decade of budget to inhabit. Anomaly detection over transparent state transitions works, and it works at machine speed. Offense scales too, especially social engineering and automated vulnerability discovery, but the on-chain asymmetry favors the defender more often than not. Alpha is not given; it is seized in the noise — and the AI dimension is precisely where the noise is currently mispriced in the optimists' favor.

Watch the migration, not the sentiment. A public roadmap from a top-five chain for signature-scheme transition. A NIST-standardized scheme shipping in a production wallet with real user volume. Quarterly loss data that finally inflects downward for two consecutive quarters. Any of those validates the thesis with evidence rather than attitude. Until then, price it as what it is: a well-timed confidence statement from someone whose portfolio benefits if you believe it. The chart lies; the ledger does not blink. Neither major network has announced a curve swap yet, and that silence is the only datum in this entire conversation that will still matter in 2030.