There is a sentence that keeps returning to me, the way a particular line of code returns to you at three in the morning when sleep refuses to come. Move your assets. It sounds like advice. It sounds like a plan. Haseeb Qureshi, managing partner at Dragonfly, recently argued the opposite โ that in a scenario where AI models break cryptographic signatures, migrating your coins to a "safer" chain is theater. A fire drill inside a building that is already burning. The phrasing has been paraphrased through a dozen aggregators by now, which is itself a small tragedy of how this industry transmits its most important ideas. But the claim, stripped to its bones, is this: the migration is not a defense. And the uncomfortable part โ the part that made me close my laptop and stare at the ceiling โ is that he is technically right, and almost nobody in this bull market wants to hear why.
I have spent sixteen years watching this space, and I have learned that the most dangerous vulnerabilities are never the ones with a CVE number. They are the ones we agree not to discuss because discussing them is bad for the narrative. A $100M raise can buy a lot of marketing. It cannot buy a new mathematical foundation. So let me trace the code back to the conscience behind it, because what Qureshi is really pointing at is not a bug in anyone's protocol. It is a bug in how we think about security itself.
Context: What We Actually Depend On
To understand why asset migration fails as a defense, you have to understand what a blockchain signature actually is, and how fragile the assumption beneath it has always been. Every transaction you have ever signed on Bitcoin or Ethereum rests on ECDSA over the secp256k1 curve. Solana and Cardano rest on EdDSA over Ed25519. Different curves, same foundation. Both depend on the hardness of the Elliptic Curve Discrete Logarithm Problem โ the ECDLP. The security of your entire portfolio is a bet that a specific class of math problem stays hard for the foreseeable future.
That bet has been remarkably good. It has held for decades. But it was never a promise; it was a probability. And two forces are now pressing against that probability at once.
The first is quantum computing. Shor's algorithm, in theory, solves both the discrete logarithm problem and integer factorization in polynomial time. This is not a secret. It is not fringe. It is the reason NIST spent years standardizing post-quantum cryptography and finally published ML-KEM, ML-DSA, and SLH-DSA in 2024. The cryptographers saw this coming a long time ago. What they did not see coming was the second force: artificial intelligence.
Qureshi frames the threat through AI models rather than quantum machines. That framing choice matters, and I will return to it. For now, hold both possibilities in your mind simultaneously. The threat is not one thing. It is a composite: AI as an accelerant to cryptanalysis, AI as a tool that lowers the barrier to breaking things that used to require a nation-state, and โ in the most speculative reading โ general intelligence applied directly to the problem of defeating encryption.
Here is where the industry's response has been almost entirely wrong. The reflexive answer, the one repeated in every bear-market thread and every exchange blog post, is that we will simply migrate. We will move to quantum-resistant chains. We will upgrade our wallets. We will rotate our keys. Problem solved, thesis intact, buy the dip. It is a comforting story. It is also, in the specific scenario Qureshi describes, a story that cannot be executed.
Core: The Migration Paradox
Let me walk through why, because this is the part the aggregators dropped, and it is the part that contains the actual insight. The migration paradox has three layers, and each one is worse than the last.
Layer one: migration requires a signature. To move your assets from an old address to a new, post-quantum address, you must sign a transaction with your old key. If that old key is already compromised โ if the primitive has already fallen โ then the attacker does not wait for you to migrate. They front-run you. They sign first. You do not get to escape a burning building by walking calmly out the front door when the arsonist is holding the door. This is not a theoretical edge case; it is the default outcome. The act of migration is itself the attack surface.
Layer two: harvest now, decrypt later. This is the layer that keeps me up at night, and it is the one the market understands least. An adversary does not need to break your key today. They can collect your public keys and your historical on-chain transactions today โ all of which are permanently public โ and decrypt them later, when the capability arrives. Your exposed public keys are a time capsule of future vulnerability. Every address you have ever reused is a promissory note to a future attacker. You cannot un-publish a blockchain. You cannot retroactively hide a public key that has been broadcast to ten thousand nodes. The data is already harvested. The clock is already running.
Layer three: the chicken-and-egg of the upgrade. Switching to post-quantum algorithms requires โ at the protocol level โ a signature authorizing that switch. But the entire premise of the threat is that signatures are compromised. So the authorization for the defense depends on the integrity of the thing being defended against. This is the migration paradox in its purest form: the escape hatch is locked by the very lock that has failed.
When I audited the first generation of ERC-20 token standards in Cape Town back in 2017, I found reentrancy vulnerabilities in two projects that later collapsed. I saved investors roughly $45,000 by documenting those flaws publicly. I learned something then that I have carried ever since: technical precision is a form of social protection. The reentrancy bug was dangerous not because it was clever, but because people assumed the contract would behave the way they expected. The migration paradox is the same disease at planetary scale. It is dangerous not because it is exotic, but because everyone assumes migration is a solution when it is, structurally, a trap.
So what does an actual defense look like? Not migration. Renovation. The answer has to happen at the protocol layer, before the threat materializes, and it has to happen across the entire stack simultaneously.
On the cryptography side, the primitives exist. NIST has standardized them. ML-DSA โ formerly Dilithium โ and SLH-DSA โ formerly SPHINCS+ โ are real, reviewed, and deployable. The Bitcoin community has discussed quantum-resistant address proposals like BIP-360 and P2QRH. Ethereum's direction points toward account abstraction via ERC-4337, EIP-7702, and โ critically โ key rotation. Key rotation is the quiet hero of this entire conversation. If your account can rotate its signing key without moving its funds, then a compromise is recoverable. You do not need to escape the burning building; you need the ability to change the locks while you are still inside.
On the hardware side, the implications are brutal and almost nobody is pricing them. A hardware wallet is only as quantum-resistant as its secure element. Most secure elements in circulation today do not support post-quantum signature schemes. That means the device on your desk โ the one you trust with your seed phrase โ may be obsolete in a way no firmware update can fix. This is not a software problem. It is silicon. And silicon takes years to redesign, re-certify, and ship.
On the custody side, the picture is equally uncomfortable. Institutional custodians carry the strictest security requirements, which means they will likely migrate first โ and their migration will be a massive, coordinated operational event. History teaches us that exchange cold-wallet migrations are among the most dangerous windows in this industry. Every large migration is a moment when funds are in motion, keys are in use, and mistakes compound. Now multiply that by every custodian, every exchange, every DeFi contract account on Earth, all trying to move at once. The coordination problem alone is a systemic risk.
And this is why the threat is genuinely systemic rather than project-specific. Signature security is a public good. It is the shared floor beneath every chain, every wallet, every bridge. When the floor moves, everything standing on it moves together. There is no island of safety. A single chain that upgrades early might earn a security premium, but it cannot protect a user whose other assets live elsewhere. We build bridges, not just blocks, between people โ and right now we are building those bridges on a foundation we have not audited for the flood.
The Contrarian Angle: The Threat Is Real; The Hype Is Manufactured
Now let me say the thing that will make some people in this bull market uncomfortable, because a guardian who only tells you what you want to hear is not a guardian at all.
Qureshi's framing is powerful. It is also, in a specific and important sense, a reframing. The quantum threat to cryptography is old news โ it has been discussed, dismissed, and re-discussed for over a decade. What is new is the word "AI." The old narrative wore out its welcome because the timeline was always twenty years away, and twenty years away is a place where nothing gets funded. "AI" is the hottest word in every room right now. Repackaging a decade-old existential risk in the language of the moment is a masterclass in agenda-setting โ and Dragonfly, as a first-tier fund, is very good at agenda-setting. That is not a criticism. It is an observation about how capital shapes the conversation.
The real danger is not that the signature break arrives tomorrow. The real danger is that the narrative arrives before the technology, and the market does what it always does with a good story: it builds a speculative vehicle around it. We have watched this movie. The "quantum-resistant" token sector has flickered into life several times over the years โ small caps, high volatility, pure narrative โ and it has never sustained a real trend because the fundamentals never arrived. This time, wrapped in AI, it might burn hotter and faster. And when it does, the coins that pump will not be the projects with working post-quantum implementations. They will be the projects with the best marketing.
This is a pattern I have learned to distrust on principle. Liquidity fragmentation is not a real problem โ it is a manufactured narrative that funds use to justify launching yet another product. The same instinct is at work here. Scarcity of a genuine solution is being converted into a scarcity narrative that can be monetized. I watched exchange launchpad returns decay from a hundredfold to a tenth of that, and the lesson was not that the products got worse. It was that the narrative got exhausted while the monetization machinery kept running. When a story outlives its fundamentals, the fundamentals are the last thing anyone checks.
And then there is the regulatory layer, which everyone forgets until it bites. Europe's MiCA was sold as clarity. What it actually delivered was a set of stablecoin reserve requirements and CASP compliance costs calibrated for institutions, not for the small teams doing the actual cryptography work. The projects most likely to pioneer post-quantum key rotation are exactly the ones least able to afford the compliance overhead. Regulation written for safety can quietly strangle the safety research it claims to want. This is the recurring irony of this industry: the rules designed to protect users often eliminate the builders who would have protected them best.
So here is my contrarian read, stated plainly. The technical threat Qureshi describes is real and under-priced. The narrative built around it is inflated and will be abused. Both things are true at once, and the discipline required is to hold them together without collapsing into either denial or hype. The people who will be hurt are not the VCs. They are the retail users who buy the wrong token, and the artists and small creators who migrate their work to a chain that markets safety without shipping it. Artists own their pixels; we just hold the keys โ and if those keys are sold to them on a lie, the failure is ours, not theirs.
Takeaway: Education Is the Only True Defense
So where does that leave us? Not in the bunker, and not on the bandwagon. Somewhere harder: in the work.
The migration paradox tells us that passive defense is not defense. You cannot wait, and you cannot run. The only durable path is the unglamorous one โ protocol-level post-quantum readiness, key rotation as a first-class primitive, hardware that can actually hold a new signature scheme, and coordination across chains, wallets, and custodians before the threat crystallizes rather than after. None of this is exciting. None of it will trend. That is precisely why it matters.
And the deepest layer of the defense is not cryptographic at all. It is human. The reason the migration paradox exists is that people assume the obvious solution is the real one. The reason the narrative will be abused is that people trust stories over structures. The only thing that breaks that pattern is education โ patient, technical, honest education that teaches people to read the underlying primitive instead of the headline. Education is the only true decentralized currency. It is the one asset that no attacker can front-run, no custodian can freeze, and no regulator can recall.
I keep thinking about the bear market of 2022, when portfolios fell eighty percent and I sat with developers one-on-one, helping them hold themselves together long enough to keep building. What got us through was not optimism. It was the willingness to look at hard truths and keep working anyway. This is that moment again, dressed in better clothes. The question is not whether the signature will hold forever. The question is whether we will do the work while it still does. Every line of code is a hand extended in trust โ and the honest ones are the only hands worth taking.