Transcription
A lot of the cybersecurity concerns that we've had over the last few weeks since Mythos, that might be nothing compared to this. So, first and foremost, this is Scott Aronson. Scott Aronson is a very interesting character. He's an American computer scientist, super smart. In half the interviews, I have no idea what he's saying. He worked for Google back in the days on their quantum supremacy project.
If you're wondering what quantum supremacy means, the term refers to the use of a quantum computer to solve some well-defined set of problems that would take orders of magnitude longer to solve with any currently known algorithms running on existing classical computers. Now, does that mean that now every code is technically uncrackable? By the way, this was 2019, I believe, when this got posted. And so, the answer is no. It wasn't in 2019. It isn't right now.
However, Google and others have been making a lot of progress towards fixing these error corrections that happen with these qubits with this quantum calculations. Interestingly, AI played a pretty big role with it. It was actually an AI algorithm that seemed to be able to correct a lot of those errors. So, that was made by Google DeepMind and the name had the word qubit in it. Can you guess what the name of that AI algorithm out of Google DeepMind was? It was AlphaQubit. That's right. DeepMind basically just names everything Alpha this, Alpha that, it's all Alpha or Gemini, one or the other.
So notice the date: November 20th, 2024. So we've covered that. By the way, if you're wondering why I'm wearing glasses, ChatGPT told me I have L-Ryz and I need to start wearing sunglasses. I don't know what that means exactly, but I'll take its advice.
Now, Scott Aronson also, of course, as he puts it, moonlighted at OpenAI very early on in OpenAI's history. Now that you think about it, I wonder if he's going to get pulled into that lawsuit that's happening right now between Elon and Sam Altman and OpenAI. So here he's talking about the Super Alignment team, which he was part of. So Ilya, as I understand it, invited him to OpenAI specifically to be kind of part of the team. Not necessarily because Scott Aronson had a deep background with neural nets or anything that OpenAI was doing necessarily. I think it was literally he was seen as somebody that was just incredibly smart across the board and having yet another extremely smart person kind of look at the problem of aligning AI with fresh eyes could have been helpful. So I think Ilya campaigned to get Scott Aronson on board on the Super Alignment team at OpenAI early on. That's just my guess from what I've been hearing. So take that with a grain of salt.
But notice, you know, and Zoe, so he talked about Zoe before. He runs the "Don't Worry About the Vase" blog, which I asked in the previous video if people knew what "Don't Worry About the Vase" was a reference to. Some of you did know. It was, of course, from the classic 1999 movie, The Matrix, where the Oracle tells Neo, "Don't worry about the vase." What was great about The Matrix is that the creators knew that they'd captured lightning in a bottle. This once-in-a-lifetime movie was so good that they didn't make any sequels because they knew that it would just be a disappointment. So, The Matrix stands alone. No sequels. It's just—it's not a trilogy. I don't want to hear any contradictory facts to that. Just The Matrix, and there was nothing else.
But why am I rambling on about all this random stuff? Basically, to illustrate that Scott Aronson, this is a known person. This is a person that's highly regarded not just in the AI industry but across a lot of these high-tech science industries. This is just somebody whose opinions and words and statements carry a lot of weight. It's personally somebody that I highly respect based on some of the talks that I've listened to by him. Not really about his P equals NP stuff, which I—I don't understand, but you know, some of the other stuff about AI, I thought it was fascinating.
But I say that to say this: This is his latest blog post. So dropped sometime within the last 48 hours as of me recording sort of this video. And he's saying, "Will you heed my warnings?" Now, interesting fun fact, Scott Aronson used one of the more recent GPT models as an almost like a collaborator on one of the papers that he published. So it worked almost as a collaborator to help them with one of the proofs that he was working on. We covered it in one of the previous videos. I believe all that was happening kind of Q4 of 2025 when all these models seem to have kind of taken a noticeable shift.
So he's saying he did an "Ask Me Anything" about quantum computing and blockchain for a forum devoted to Bitcoin. Here's that forum if you want to check it out. It has some pretty nifty lightning effects if you keep scrolling. I don't know if it's like a one-time occurrence or what, but it's pretty cool. And he also collaborated with a lot of his colleagues to put out a detailed position paper about the quantum threat to cryptocurrencies and how to best respond to it. Notably, the situation evolved even while they were writing this position paper with the major recent papers from Google and Caltech/Oratomic.
So, this is where we get kind of to the meat and potatoes, I guess, is the expression. I'm sorry if I get too long to get to the point. I just want to make sure people have the right background. Here's what Scott Aronson is saying. He's saying some of the most reputable people in quantum hardware and quantum error correction, people whose judgment I trust more than my own on these topics. So this is important because it's a very high-level person saying, "I trust those people's opinions more." So these people might be a handful of the most knowledgeable people on the planet, bar none, about this stuff: quantum hardware, quantum error correction. They're now telling me that a fault-tolerant quantum computer able to break deployed crypto systems ought to be possible by around 2029.
So these things are coming in 2029, and these companies that are racing to develop them, they have no plans to slow down to give cybersecurity time to adapt or whatever. Just like there's an AI race, there's similarly this sort of quantum race as well.
Okay, so let's kind of back up a little bit and explain maybe what that means. So, first and foremost, there's a lot of encryption on the internet. Encryption is how computers lock certain information so only the right person can open it. It's like if I wanted to send you a locked box in the mail, right? So I want to make sure that only you can open it, and when it gets there, you want to know that it came from me. So, basically, both sides have the same key. It's like if two people have, you know, a key to some home, they both can open that lock. That's symmetric encryption. That allows us to encrypt large amounts of text of data very efficiently.
There's another type of cryptography that's public-key cryptography. So there's a public key and a private key. The public key can be shared with anybody, and the private key is secret. So this allows various strangers on the internet to agree on what is a secret, what isn't, to sign certain things with identity, to prove their identity, sign messages, sign software updates, etc., and also control crypto assets. So this quantum threat to encryption isn't necessarily about passwords. It's about those private and public keys.
So right now, mainly today's kind of public-key cryptography relies on math problems that regular computers, classical computers, are terrible at: on factoring large numbers, on multiplying huge prime numbers. So we have these various math problems around the internet, these little wisdom tests that computers can't effectively pass. Therefore, the things behind those math problems, they stay secure.
With a large enough quantum computer that's fault-tolerant, they can run what's called Shor's algorithm. Shor published it back in 1994, showing that certain encryptions would be quantum vulnerable. So he showed that fault-tolerant quantum computation was possible in principle in 1996. Bitcoin started over a decade later. Ethereum started in 2015, two decades later. It chose quantum-vulnerable cryptography. So did website certificates, satellite control systems, CPU microcode updates, and on and on and on.
So the point is, this Shor's algorithm breaks a lot of the encryptions that we have that power today's internet, crypto, apparently our satellites, tons of stuff. And the quantum computers, if quantum doesn't pick that lock, right? It doesn't sit there trying to pick the lock. It makes the math behind the lock stop being hard. So I kind of think of it in like, if you're trying to figure out if a number is odd or even, no matter how big the number is, right? So the bigger the number is, the more it would take to calculate stuff with it. But you can tell if it's odd or even just by looking at the last number, right? So if it ends in an even number, then it's even. So before, just adding more digits would make the problem harder. But with quantum, it doesn't matter how many digits you add to the numbers. You just look at this one thing, and it's really, really simple to figure out what's what.
So it's important to understand that quantum doesn't break everything the same way. There's a lot of other types of encryption that are more resilient to this. So who is especially vulnerable to this if and when it comes out? So Scott Aronson is saying that people familiar with the matter, people that know what they're talking about, they're saying, you know, less than three years, we're going to have this. So if that's true, who is sort of right now super exposed to it?
Well, number one: governments and intelligence targets, classified communications, diplomatic cables, military systems, any archived encrypted traffic, right? So it's no secret that a lot of the big governments—US, Russia, China, etc.—they've been saving a lot of this encrypted information, even though they couldn't read it, but they just saved it because it's out there. They just couldn't figure out what it said, but they kept saving it and saving it and saving it with the hope or the plan that one day something like this would come out and then they would be able to break that encryption and, you know, 10, 20, 30 years later, figure out what it said. So they were saving it for later in case this type of technology emerged.
Another exposed industry is banks and every financial infrastructure. So payment systems, various interbank networks, authentication systems, software updates, certificates, transactions, signing, all of that. Various big tech platforms, browsers, mobile operating systems, your phone, cloud infrastructure, identity systems, API authentication, software signing, a lot of the records like medical records, a lot of the infrastructures, power grids, satellites, we talked about a lot of that stuff. And of course, blockchains and crypto, right?
So Satoshi, the famed inventor of Bitcoin, you know, if you follow this whole thing, so his coins have never been moved. They're locked up. They're just sitting there. They're not transferred in or out. They don't move. It's this old dormant wallet that just sits there. If those encryption keys get broken, somebody could impersonate that key, that secret key, and move those coins elsewhere to use them.
So there's two kind of vectors of approach here. Now that I think about it, it's kind of like that movie Tenet, right? With the time travel, with the temporal pincer attack. Remember when they attack from from both sides of the time continuity thing? Temporal, like time, pincer, attack, like claw. They attack from both sides. So this quantum threat is kind of like that. I just connect the dots in my head. That's amazing. I'll stop patting myself on the back.
The point is that it can encrypt secrets from 20, 30 years ago, right? So it potentially could encrypt yesterday's secrets that have been secret for all this time that now kind of could be broken if and when. Again, if this is true, we don't know. But if this comes out, then all that stuff, all that data that was saved, it could be unencrypted and all of those secrets come out. We find out the truth about Area 51 and Elvis is alive, all that stuff.
But also, this quantum threat is going to impact tomorrow's trust, our ability to keep trusting stuff online. If somebody's transferring Satoshi's keys, can we trust that it is in fact Satoshi doing it, not somebody that's able to impersonate him? If your phone pops up and it's like, "Oh, hey, this is Google and there's a critical update, software update, you got to click update now or bad things will happen." Is that Google or is it somebody impersonating Google that's doing that? We're not going to know.
So my kind of first question was, you know, Google seems to be like it would be affected quite a bit by this because, keep in mind, they are one of the major players in the space. They're racing towards this quantum thing that they're building quantum fault-tolerant computers. Again, allegedly, who knows? I'm not saying any of this is true. I'm saying if this is an if-then statement, if this is true, then this thing that I'm talking about. I'm not saying it's true. I'm saying if it's true, then this. So if true, then Google seems like it would be affected by this. A lot of the infrastructure they have: Chrome web browser, they have their Android ecosystem, Google Cloud accounts, Gmail, certificates, software updates, APIs, all of it.
Here's what's interesting. Google is also preparing like crazy for this exact thing happening. So Google is setting a 2029 timeline for post-quantum cryptography migration. That can't be true. Did I just make all of it up? No. Here it is. Google is accelerating its security timeline, setting a 2029 target to migrate its internal infrastructure to post-quantum cryptography, PQC. Why would they do that? Well, it's an accelerated threat model. Google updated timeline due to faster-than-expected progress in quantum computing, reducing the estimated qubits needed to break current RSA encryption. Yeah, but who's saying that? Is it some weird blogger that's just making stuff up? No, it's Google. So, this is blog.google, google.com/innovation/AI/technology, whatever. I'll leave the links somewhere.
Notice the date: March 25, 2026. "Quantum Frontiers May Be Closer Than They Appear," just like the stuff that you see in your rear-view mirrors. They're closer than they appear. And they're setting a timeline for post-quantum cryptography migration to 2029. So, quantum computers will pose a significant threat to current cryptographic standards and specifically to encryption and digital signatures. So the threat to encryption is relevant today with store-now-decrypt-later attacks, like we've talked about, and has been working towards post-quantum crypto since 2016. Crypto, not like coin crypto, crypto like encryption. So Chrome and Cloud have their post-quantum work underway. The Android systems, they're integrating post-quantum digital signature protections.
So Google is affected because everybody's going to be affected by this. But Google's really ahead of the game. They're less exposed because they're actively preparing. But what I think everybody's missing is that if Google, who knows when the other shoe will drop, because they're—well, they're the ones that are going to be dropping the other shoe. They're polishing it up right now so they can drop it. If they're saying that the other shoe will drop in 2029, then the rest of us, you know, the other corporations, organizations, we all can't pretend that this is not going to be a problem till 2040. If Google thinks it's 2029, then it probably is 2029 or around that time. That's probably the best estimate that we're going to have.
So, how's crypto affected? Well, blockchains are public. Blockchains are permanent. Many cryptocurrencies use these encryption methods that will, you know, again, if this is true, then those encryption methods will be broken, obsolete, affected, exposed. If a wallet's public key is exposed on-chain, then a future quantum computer may be able to derive the public key and impersonate the owner and do whatever they want with the tokens, with the money, transfer it, use it, whatever, right? The coins can be stolen.
So, Bitcoin seems like it would be, you know, exposed to this. Not every Bitcoin address exposes their public key immediately. Sometimes it's when you spend it, when you transfer. So, what happens to those old abandoned coins that were thought to be lost? I think some guy early on had a little disk with a whole bunch of Bitcoin on it, and I guess he threw it away or his house cleaner threw it away, and eventually that became like tens of millions or hundreds of millions of dollars. So it could be that these famous dormant addresses, wallets, you know, from early Bitcoin era, from those addresses that they're going to be exposed, that they're going to be hackable. The coins are going to be able to be stolen out of them.
Then on the other hand, you have something like Ethereum. So Ethereum is a little bit different in that it has active governance. Mr. Vitalik Buterin, right, and gang, they're able to make decisions based on how this thing is run. So maybe that migration might be a little bit easier, although it's still incredibly complex. How to take the whole thing that was built on top of it and try to transfer it somewhere.
So with crypto, of course, everything's built on this idea of immutability, right? So if you have some old stuff that nobody can move it, can take it, or any stuff for that matter, a third party can't come in and just take your stuff. So somebody moves your coins, if some central authority figure that governs that, whatever Bitcoin or Ethereum or some other coin, like if they move it, then that could be seen as violating the rules. But of course, if they let these old vulnerable coins remain vulnerable, if they don't do anything about it, then future quantum attackers might steal them. So it's kind of like, what do you do? And this isn't just a technical problem. It's a—you know—governance. It's a constitutional problem. The question is, who gets to change the locks when we realize that the old locks aren't secure anymore? Who has that power over crypto?
And so that's what Scott Aronson is talking about. That's why he's working with Coinbase. That's—if you're wondering what does Coinbase have to do with this? Well, all those people, the colleagues that he's listed, right? So Dan Boneh is one of the world's leading cryptographers. Justin Drake is a major Ethereum researcher. And Coinbase is, of course, a major crypto institution with a direct incentive to understand whether quantum risk threatens the blockchain or not. So this isn't some random blog panic. Serious crypto and cryptography people are now actively planning what they're going to be doing about this.
So Google did have a crypto disclosure that they've posted. So Google Research recently published a post saying, "Future quantum computers may break elliptic curve cryptography that protects cryptocurrency and other systems with fewer qubits and gates than previously realized." So they did not publish an attack roadmap. They didn't publish like, "Here's the recipe to hack everybody." They're saying there is a recipe, and it's pretty simple, like it's like scrambling some eggs. It's simpler than we thought, and you know, if you make that recipe, well, you can break those encryption methods, and all sorts of bad things could ensue.
So here's the important bit to understand. So Google used a zero-knowledge proof to disclose the vulnerability responsibly without handing the bad actors a complete manual. So it's like if there's some locked vault and you need a password to get in. Google wanted to prove that they know the password, right? But instead of saying the password out loud, right, they just use this zero-knowledge proof to say, "Hey, we know what the password is. Here's proof." But they didn't, you know, let the cat out of the bag. They didn't release the password out there.
Cloudflare is very relevant to this whole thing as well. It hosts—it sits in front of a very large amount of internet traffic. Cloudflare also is targeting 2029 for full quantum security.
All right. So coming back to Scott Aronson's and heeding his warnings. One important point that he's saying here that might not be sort of out there for everybody else, not as public knowledge or understanding. This is more inside baseball maybe, but he's saying that these companies, they're not going to slow down. He's saying these things are coming. They're coming soon. And the reasoning by a lot of these labs, they're saying, "Well, isn't it better if it's done first by mostly US-based companies in the open than by, let's say, Chinese or Russian intelligence in secret?" And by the way, haven't we had tons of advanced warnings about this, right? So basically, everybody's going like, "Hey, we're going to race ahead." So this is the same exact dynamic as, you know, with AI.
So he's saying, you know, at what point are the people going to wake up? In the same way that Anthropic's Mythos model has jolted even the most ostriches about the cybersecurity risks of AI. So anyways, this whole spiel here, he's being a little bit sarcastic. He's saying, "Is the above line reasoning suspiciously self-serving and convenient, right? Isn't it the same argument that we've had for AI? Like, if we have an advantage here in the US in some US lab, shouldn't we press it? Shouldn't we develop it? Shouldn't we race towards completion as fast as possible? Isn't that what every AI company used as their argument? You know, really, you know, if you think about it, accelerating towards dangerous superintelligence is the safest course of action that we could possibly take."
So he's saying, you know, that's not his place to answer such questions. Although he's letting his viewpoints be known, I think. But he's saying this is how some of the leading, you know, quantum computing companies are now thinking about the "Shor of Damocles" that they genuinely believe now hangs over the internet. Do you see why I like this guy? Shor of Damocles, right? Like the sword of Damocles, like the threat hanging above your head, but it's not a sword. It's Shor. Like Shor's algorithm. And I'm sure if I'm mispronouncing any of those words, about a thousand of you will correct me in the comments. But whatever the case is, you see why I like this guy.
So the point here is he's sounding the alarm. As the writer of the internet's most trusted quantum computing blog since 2005, he's sounding the alarm for the rest of the world. He's saying, "If quantum computers start breaking cryptography a few years from now, don't you dare come to this blog and tell me that I failed to warn you. This post is your warning." Right? He's saying, "Start switching to quantum-resistant encryption and urge your company or organization or blockchain or standards body to do the same." And he's saying, "Heed my warning, for it comes not from some WordPress-using rando." What's wrong with WordPress? But from the inventor of—and he lists some of his sort of credentials, which again, he is the best-known quantum computer theorist in the world. He's the Schlumberger Centennial Chair of Computer Science at the University of Texas at Austin, co-founding director of UT Austin's Quantum Information Center. His research is about the capabilities and limits of quantum computers. That's what he did. That paper he published with the assistance of GPT, what, 5.3, 5.4? Before I—I don't recall. But that paper that he published, it was about quantum Merlin-Arthur, like the Merlin Knights of the Round Table, Arthur, whether they're quantum. I don't get it. If you're interested, ask ChatGPT. Obviously, it knows more about it than I do. It co-wrote the paper. He was just elected to the US National Academy of Sciences, one of the highest honors for American scientists. He's not a quantum hype man. So for years, he corrected people that would overstate, hate what quantum computers can do. This is what makes this warning even stronger. When a careful quantum skeptic says the crypto world needs to wake up, cryptography, not just cryptocurrency, I mean just crypto as in like everything encryption. When he's saying that the world needs to wake up, it carries more weight than, well, hopefully than anybody else.
So what's the sort of takeaway from all this? Scott Aronson's warning is simple. For many, many decades now, we've heard about the potential quantum threat to cryptography. We've been hearing about it. But now he's saying, "Okay, now the timeline has shifted. Now it's close. Now really get real, get serious." And of course, quantum computers, it's not like they're going to break everything, but they're going to break a lot of the current. Again, if this is true, then we'll see a lot of the current encryption technology, like the less secure versions of it, be broken, be exposed. So various things around internet identity, crypto, secure connections, software signatures. And there are a few major players that are moving to migrate everything to post-quantum cryptography. So Google, Cloudflare, Coinbase is looking into it. They have that panel. So they're of course exposed, they're dealing crypto. So they're thinking about it. Apple apparently is doing some stuff. The iMessages, they have this post-quantum cryptography. That doesn't mean that the entirety of the Apple ecosystem and infrastructure is protected.
And in case you're wondering, well, why is this an AI story? Why are we talking about it on this AI channel? Well, here's why. And I might be wrong on this. I'm probably out of my depth. I apologize. Take this with a grain of salt. Correct me if I'm wrong. But one of the biggest quantum computing challenges, one of the biggest roadblocks was these errors that would be generated by these qubits inside of quantum computers, which makes sense. They don't even know where they are. Are they here? Are they there? Maybe. Both. Point being is we couldn't figure it out. We couldn't figure out how to reduce those errors. Similar to how it was very difficult for us to figure out how the 3D structures of proteins work to give proteins their function. And then this guy, Demis Hassabis, is like, "I shall build AlphaFold, and it'll figure out how to fold the proteins." And we couldn't figure out how to solve these errors with these quantum computers. And Demis is like, "I will build AlphaQubit, and we'll figure this stuff out." And that's what they did. They created AlphaQubit, an AI-based decoder that identifies quantum computing errors with state-of-the-art accuracy. Right? If we want to make quantum computers more reliable, especially at scale, we need to accurately identify and correct these errors. This does that. Therefore, making quantum computers capable of performing long computations at scale, opening doors to scientific breakthroughs and many new areas of discovery, as well as opening the doors to your crypto wallet.
So, these neural networks figured out how to predict these quantum errors, quantum noise. And when I've said in the past videos, the last, what, three years, however many years I've been doing this, that AI will touch every aspect of our lives, that wasn't hyperbole. It really, really will. And when I say it's going to break a lot of stuff, it'll break the industry. Again, not hyperbole. It really, really, really will.
By the way, here's the Coinbase paper that they mentioned in the blog. Scott Aronson is working on quantum computing and blockchain. There's Scott Aronson. This is one of the foremost people in cryptography. Justin Drake, Ethereum Foundation, many, many other distinguished people. So, the TL;DR is: don't ignore this. It's now clearly on the horizon. So the key takeaway is PQC, post-quantum crypto, is going to be needed well, pretty much everywhere, at all blockchain layers and everywhere. Notice all these publications are very recent.
By the way, same day today, May 1st, 2026. You also might have noticed that the War Department has entered into agreements with eight of the world's leading frontier artificial intelligence companies: SpaceX, OpenAI, Google, Nvidia, Palantir, Microsoft, AWS, and Oracle. Notice one name missing: Anthropic. Because, keep in mind, one of the biggest exposed targets is the governments, including the US government. A lot of the old secrets that were out there encrypted, but saved somewhere to be potentially decrypted later. I mean, you know that they're working behind the scenes with Google to make sure that they understand what's happening and they're able to prepare for it.
Now, I wonder if Google ends up dropping this thing that breaks some of these encryptions, and everybody will quickly need to reshuffle to that PQC, post-quantum encryption or post-quantum crypto, whatever. And, you know, Google also provides the cloud services with the easiest migration path possible to transition onto to be fully quantum protected. Boy, does that seem like a really good business model.
But anyways, none of this is financial advice. But when people say, "May you live in interesting times," does that sound like a blessing or a curse? If you made it this far, thank you so much for watching. My name is Wes Roth, and I'll see you in the next.