AI Safety, Quantum Readiness and IBM’s Blockchain Beta

Hello, everybody, and welcome to Techstrong Gang for Thursday. And as usual, we have the assortment of the unbelievable. It's almost like Ripley's Believe It or Not!

out here. But let me introduce our guests. Jon Swartz, as always.

How you doing, buddy? You're in Las Vegas. I am in Las Vegas, and I'm going to apologize in advance for bulky Wi-Fi.

So if you lose me, I'm going to come back on my phone. All right. The joys of technology.

Here it is, Wi-Fi versus LTE playing out right here in person. All right. Oh, sadly, yes.

Drew Gutstein, how you doing, my friend? Good to see you, as always. Hello.

Glad to be joining. My Wi-Fi is working really well right now, so I'm happy. I'm so jealous of you, Drew.

I'm hoping not to jinx it, so I will say nothing about the Wi-Fi, because God knows what will happen. But unless you've been hiding under a rock, you would've noticed that there is a new show here in New York. It's basically called "The AI Follies," and we have all kinds of people who have come in from out of town to perform for this show, and I'm pretty sure it's going to win a Tony.

But you guys see what's happening here with all these folks. I'm going to let John summarize the events of the week, because the one thing that I cannot figure out out of all this, John, is what might actually change here? I don't think anything's going to change, Mike.

This is all obliviating and speechifying, and it's all done with things in mind. Everybody has an ulterior motive in this scenario. So yesterday, Wednesday, we had Sam Altman and Dario Amodei speaking at an emergency UN Security Council meeting or session.

And there's that promised Wi-Fi issue. All right. Well, let's try to keep this conversation going while we wait for John to come back in.

But one of the things that keeps coming back to me, and maybe Drew, help me here, but I'm like, great, even if we wanted to regulate AI, how would we regulate AI? Because I don't see any mechanisms in place that are actually going to be enforceable. So what is this conversation really about?

So I have a theory. I don't think it's actually about AI safety, or tangentially is. I think it has a lot more to do with the economics of hyperscalers building out data centers.

Over the last month, the economics of that changed drastically. Borrowing costs have come up. There are only so many investors who want to invest in AI data centers.

Even in the Middle East, there's a limit. So they might be petering out, and I think part of that, what they're facing with these increased costs, is the competitive pressure. Right?

So I think they're all kind of together taking a step back because it's starting to get too expensive. " The stuff he described is something that happens literally every day with AI. " All of the delays can be summed up like that.

And if anyone's ever used AI before, they know that that is a thing, and that's one of the biggest hurdles. " I think it's a lot more about the money. So you're saying that they knew this all along, and the fact that some 20-something-year-old kid quit a couple of weeks ago and put out a meme is not the thing that's really driving this conversation.

It's much more about maybe they're running out... Do they want the governments to kind of essentially invest in them, and that's how they're going to regulate and turn the next- I don't think that's what they want. I think they're going to be pushing for as little as possible with this, and even pushing for third parties to provide this function.

Right? So in one of the articles, OpenAI is asking for permission to create third-party reviewers and use them to test their system. It's sort of like an extended version of what was happening with the Middle East preview, so maybe it's even revenue-generating.

But yeah, I think this has a lot to do with the business model. Mm-hmm. And how that's evolving.

Do you think there might be a separation that says this class of AI model is essentially ammunition, and therefore will be controlled by the government, and then everything below this certain threshold of capability will be available for maybe commercial purposes because, well, we can at least limit the blast radius? What do you think? So I think there already are export controls on this kind of technology, especially for military or dual use applications.

So this doesn't really change that, but I think it does sort of raise the specter of kind of a jailbreak risk. So like they're really talking about context poisoning and the risks associated with that. The approaches that people have been taking to address that, so segregation, encryption, really strong access control, and very low trust, those are what the industry needs to develop and continue developing.

They don't exist yet, and I think this is an acknowledgement that we don't have the tools that we need. That's always been the case since this first started a few years back, right? So John, we're being a little cynical here, but we're saying that- Oh, a little?

You're surprising me ... this is not about safety after all. No, it isn't.

No. And in some ways, it seems to be about the money, but we're trying to figure out what the motivation here is exactly. Are they just trying to limit their liability?

Do they want the government- Yes ... to kind of step in here? What's your read?

So I think they're trying to do multiple things. And I think they are motivated by money, as they all are, from top to bottom, from the government to the tech industries, to whomever, the investors. And I think what they're trying to do is they're trying to protect themselves.

There isn't going to be federal laws of any type. They're thinking about lawsuits in the future. They're thinking about how to position themselves and try to take these responsible steps, in their opinions.

They're trying to figure out how to keep the investors happy and keep them on board and not scare them away. Because OpenAI has already, in a sense, delayed their IPO several months. They might delay it even yet again based on disclosures.

And in terms of Trump, and you probably mentioned him, he spoke on Tuesday. He says we don't need regulation of any sort. We should call a super intelligence.

And again, he has a lot of money, skin in the game, between himself, his family, and his buddies like Michael Dell, who are making a fortune right now. They don't want any regulation anywhere near this. And I think I said this yesterday, but the problem when you play this game is in the short term, you do make a hell of a lot of money.

But in the long term, it boomerangs. It comes back to bite you in the a*s, and I think it's going to happen to a number of these companies and individuals. The chickens are going to come home to roost, and right now we're kind of in a narrative of trying to be responsible about AI, when in fact that era will pass very soon probably, and they'll move on to the next thing.

I think that liability question is really important because that's completely unsettled. Yes. The way we think about internet companies is generally about content provided by users, regulation, whatever it's called.

This is not that, right? So they don't have that kind of protection that, say, Facebook has when someone posts something on their website. We're talking about the computers that are owned by these companies attacking computers that are owned by other companies.

So there is a big liability aspect to this that I think is still kind of unknown. And to that point, and here's where I go, this may turn out to be entirely crazy because essentially what they're going to conclude is the AI agents themselves are ungovernable, and they are going to do what they're going to do. So therefore, it's up to the rest of us to invest millions of dollars to better secure our data from those AI agents, and we've got to go in and put in more granular controls and all these other kinds of things around cybersecurity, and that will become the new level of regulation that will actually get installed, and all our costs will go up because, well, there's a bunch of crazy kids throwing a bunch of AI agents around the world.

So So yeah, I'm at this Okta conference called Octang, and I talked to a couple of the folks here, and they're following this really closely because they're in the security space. And I talked to a guy who is their chief product officer, who used to be the director of cybersecurity for the Obama White House administration. He also is a member of CISA and worked with Krebs and others.

And I said to him, "Absent any type of significant laws, do you see a scenario where there is a catastrophic event that occurred because of an agent? And if so, who gets sued? Who's liable?

" And he said, "You know what? That's going to happen. We just don't know when or who's going to be found liable, and that's going to be a test case.

" Mm-hmm. All right. But if that's the case, is it conceivable that, in their infinite wisdom, that somehow or other they will create a set of protection classes for the AI providers so that if something does go wrong, they're not held accountable and not liable for it?

Is that a possible outcome? I know the commerce secretary says that that's not necessarily going to happen, but there's a whole lot of other folks who are of influence here who have a vested interest in that outcome. What do you think, Drew?

It's a big ball of yarn right now. I don't think there's a clear strategy that's going to win here, especially once we start thinking about, sort of move away from frontier models for a second and think about maybe the cheaper ones that people want to run on-prem and things like that. So, there's a big difference between an attacker using their subscription to Claude to attack people because they're using Claude's infrastructure that's part of that attack, and Claude should be the one who's responsible for stopping it.

Right? That's the sort of the only area where you can assign any sort of accountability. This is all up in the air.

None of it's in legal code, so we'll have to have some cases to see what happens. Can I add? There's something that I find really ironic about all this.

So we have OpenAI and Anthropic that keep talking about how we have to pump the brakes. We have to slow down development. Didn't earlier this week they both announce new models?

I mean- ... this happens every month, and yet they're talking out of both sides of their mouth. And tonight, to add another layer to it all, there's going to be a state dinner where the tech bros are going to be there with Trump and President Xi, and they're all going to probably mollify one another and, at least on the US side they will.

So I just don't see anything really changing. I mean, they're telling us one thing and they do something quite different. So let me ask you another question about the United Nations itself, because I always feel like it's the court of last resort.

It's the place where people show up when they can't solve it any other way. It's usually generally hopeless anyway. But right now we have a situation where one to two members of the Security Council have more AI capabilities than the world combined, and they're basically essentially dictating terms to other countries.

At least one of them is for sure. The other one is kind of trying to be a little more subtle about it, but terms they are. " Because it's not just a nuclear treaty.

This impacts your everyday life and your country, your economy, and everything that goes with it. What do you think, Joe? Yeah.

Armaday, in a sense, said this is a threat to humanity, and again, this is like a last resort stage, and I think it's every country for itself. I think the Western countries are much more concerned about this than, say, others maybe in Asia. I almost think of the UN as like the Senate.

The Senate, nothing ever gets accomplished, and that's by design. That was part of the charter of the United States is that the Senate is for speechifying and for filibustering and for blocking. It's friction and resistance.

And I think this whole thing, the follies is a good way, it's a charade, I believe. And I think everyone has an ulterior motive. They're all speaking from their carefully worded, legally vetted scripts.

And they're going to just plow on until there's a major event. We had HuggingFace really didn't gain traction among beyond the industry. I don't think the general public has any idea what that is or cares.

Mm-hmm. The FBI, that's terrifying what's happening with the FBI. But again, within this crazy world where everything is being thrown against the wall, this is just another minor event, UN events, that will come and go, and then next week, within two or three weeks, we're going to have new models from OpenAI and Anthropic that will be even faster and even cheaper and even more, I don't know, salivating for the investors when these companies go public.

Mm-hmm. My take is there will be coalitions of countries that form partnerships around AI, and they will make collective investments together, and they'll have some collective policy that they implement together. And they will do that to try to counterbalance what's happening here from the US and China.

And they may not have the resources to do that, but from their perspective, they don't always need the latest and greatest model. They can go with something that is just good enough. So Drew, are we going to break up into a lot of AI city-states?

What do you think? So I think this is a fascinating conversation, and I'll reveal a little bit of how nerdy I am. Have you guys been to a tour of the UN?

Yeah. Yeah. Okay, so great.

We're all dorks. I'd love that tour. And if anyone goes on that, they'll learn very quickly that the UN is for nuclear arms.

It was created to stop the world from destroying itself with nukes. And since the nuclear arms race has sort of subsided, it's been, I think, a little bit adrift and maybe climate change being sort of the next existential threat. But to the extent that the UN is built for existential threats, I think there's a role.

On the other hand, I completely agree on market sovereignty and the concerns that European countries have and other countries that are sort of caught in the middle. And the industry's trying to be in there. These are huge markets.

So the hyperscalers care a lot about being in there. It's a question of whether they can provide the controls that different sovereign nations want. Yeah.

The whole situation to me is just, it kind of defies explanation, what's going on. And the irony, I think I mentioned this in the green room earlier, was we have this whole kind of ban of the media from the White House or from these events. So tonight we're going to have this event with the tech bros talking to the presidents of China and the US in this AI arms race, and it's going to be covered by the Chinese media but not the American media.

There's not going to be a press pool. And I think that kind of sums up... And the other thing too is, I wonder, Mike, we usually kind of look down on these alliances, but Okta told me about something called this Blueprint Alliance that they're part of.

There are 15 members and they turned away some folks. They actually are convinced that that will actually be effective in terms of safeguards or the dreaded guardrails that we talk about. I don't know if you share that opinion, but the gentleman who used to work at the Obama White House, Eli, was really sold on this.

I really applaud the effort. My problem is, there's 15 other ones that I can't distinguish any more from the other ones. That's what I mentioned to him.

" I mean, he really didn't address it because he doesn't want to throw shade at those other organizations. But he was convinced given his... His experience is valuable, too, so I respect what he has done and a lot of the work he did.

But I don't know. Even he said this is just Wild West time. It's just- All right.

And- ... grab whatever you can and move as fast as you can. I'm going to end this here because I think you're spot on.

And the lesson from all of this is despite all the speechifying, if you're a business or an IT leader in an organization, you should consider yourself to be on your own because there's no help coming, promise. Oh, wait, one other thing, Mike, I was going to say, I forgot to say. So the fastest adopters right now, according to VMware, the fastest adopters of AI are heavily regulated industries like healthcare and the financial services I think that's terrifying.

Yeah, even it isn't, because there's also the places where there's the most mundane routine, paper processing- That's true ... soul-choking workflow stuff. I can see that, but it's also the- I also think about the data, though, and what could possibly go wrong with- Yeah ...

the attacks that are going on. And I think of HuggingFace, I think of how fast these attacks hypothesize and how effective they are anyway. If we just gave away all the data in the first place, we wouldn't be sensitive, we wouldn't have to pay to all secure it, and all these problems would just go away.

Just make the data free and readily available to anyone. All this secrecy stuff is the problem. Anyway, I digress.

But we are going to shift the topic to something that is equally, well, shall we say, potentially terrifying, and it's quantum computing. And Alan Shimel has an article up on Security Boulevard talking about how the need to actually start thinking about securing our data is getting very real because, well, the latest and greatest quantum computers, there's been some advancements, and nobody's quite saying this out loud yet, but originally they were saying that we would have Q day, the day where these quantum machines could break our encryption would be sometime in 2029. But as I look at these advances and these new machines that are coming down the pipe with a little help from AI, I got a sneaking suspicion that that date is going to move up to 2028.

So Drew, I'm going to toss this to you a little bit, but have we reached some sort of Rubicon here with the quantum stuff? Because last time I checked, it took a lot of effort to replace all these encryption schemes, and so therefore I've got maybe two years if I'm lucky, and that's not enough time to do that, correct? Yeah, and it was important here to really consider the threat vector, right?

So, the article talks a lot about encrypted backups, and the problem with encrypted backups is the sort of capture and store technique that a lot of adversaries use. So, if they're sniffing internet traffic and they see a bunch of encrypted stuff, they'll just store it, and then later on, try to decrypt it, maybe in a few years when computers get faster. This has been a cycle that's been going on since TLS 1.

So every few years, we reach a point where computing outpaces the algorithm, and you have to increase the key size and roll it all out before everyone has that kind of computer, right? And this is actually the exact same effect. It's scarier because it has the word quantum in it, but I don't think it's really anything different than when you need to retire some encryption and introduce another form of encryption, something the industry has done repeatedly.

Yeah. Now, that said, Mike, your point about when is Q day, I think it is coming closer. People are making real advances using AI in this space, but it's not like an AI quantum computer.

So there's a lot of fear about what these things will do, and for the most part, they're not going to be able to do a lot, at least not for a long time. So yeah, they might start cracking encryption that was captured decades ago, right? And that will happen, and that's going to happen regardless.

So, really, this is really about maintaining hygiene. When this always happens, we have to do a full inventory of where's all our encryption and where's all this stuff. Yeah.

And the nice thing is you don't need a quantum computer to stop it. So there are algorithms that the quantum computer can't solve- Yeah ... and we just use those instead.

Yeah, and no doubt, and I think the issue is just that the time and effort to go do that is large, and the time window keeps shrinking, and nobody's doing anything about it yet. So we're going to try to have a mad dash to the goal line at the end here, and as we all know, that's not going to work because, well, let's think about your average enterprise. How long would it take them to replace and upgrade all these encryption schemes?

At least two years. So at some point now I've got to decide which data I'm going to try to protect and upgrade, and which data I'm just going to suck it up and maybe it's going to get lost or seen by somebody, and hopefully, it's not particularly relevant. Is that where I am?

I think that's spot on. And also, there's a solution here. Don't back up so much stuff.

I think organizations tend to back up way too much data. Obviously, there are requirements for regulated domains, but even those, the data that's being preserved is just not useful for that. And so I think we end up with a lot more sensitive data in the logs that are being backed up than we really need to have.

So I think, A, rolling over your encryption algorithms, great. If you're going to do that every few years, every time computers get faster, and now we're going to do it maybe every one year because quantum computing or whatever, so we've got to get good at that. And the other thing, look at what you're backing up.

You don't need to back up all that stuff. Maybe you can eliminate some risk by eliminating backups. Do you know what's weird is, when we were talking about, you talk about this accelerated timeline, Mike, it's almost like deja vu now.

We're talking about things moving too fast, and I remember the Okta executives kept bringing up the word velocity, and that's their greatest concern on the security element. And when I think about what this is about with quantum versus AI, I think of security, and we've always talked about this as the last consideration, and until something happens, it's addressed. And I wonder if we're just going through the same exercise right now with quantum that we're doing with AI, where things are moving too quickly, and we don't think about the consequences.

And with quantum, to me, it's more terrifying. But again, I see this pattern developing. It's just like it all comes down to money again, as everyone's in this mad, frantic dash to cash in while they can before some larger event takes place that they can't control, and I'm thinking about an economic downturn.

I think- I think that the conflation of those two technologies, like in the materials we're looking at, the AI hysteria or quantum hysteria. People just are thinking that there's some quantum AI. That's not what's happening.

There's no quantum AI, it's just computers, and they're expensive, and they're really slow. Right. And so one day- Quantum AI is essentially a natural language interface to launch a quantum process, but that's not it.

It's just as far as- Yeah, that's right. And to the extent that AI is accelerating development everywhere, right? Yeah.

That's definitely happening. But the other side of this too is, people talk about Q day like it's going to be Valentine's Day. " It's probably going to be broken for a significant period of time before anybody realizes, and maybe a couple of researchers will have gotten around to making some sort of public statement.

" So Drew, are we preparing for something that may come, but we don't know exactly when it will? Yeah. Like everything else, all these technological advances, right?

So, okay, again, I think these kinds of tellings generate a lot of clicks and sound very dramatic. We are in the boring, slow world of technology still. It's not all that.

So yeah, it's something for sure. If you're responsible for the encryption that your company is deploying, start looking into this, definitely. Mm-hmm.

But for consumers or people worried about getting hacked by a quantum AI or something like that, that's all science fiction. That's not a thing. All right.

" Yeah. All right. True in technology, there's always the unknown we look forward to, like what's this going to lead to?

Now it's almost like a sense of dread over it recently, right? What about the fear of the unknown? Yeah, that's because we finally figured out that this sword actually cuts both ways, right?

Exactly. Yeah. Hey, there's a negative and a positive- I'm sure quantum computing will solve some of life's mysteries, and that will help us in a lot of places, especially medicine and areas like that.

So I don't think we're saying on any level that we shouldn't be pursuing this stuff. I'm just saying that you should buckle up, because there're going to be a lot of unintended consequences, which is the current theme. So let's move on to our next topic.

We have IBM today talking about... And it was just your standard boring IBM announcement as usual. But basically, they said that they have this mechanism for managing your digital assets, wallets, and all the stuff that goes with cryptocurrency, and they're hooking that up to their mainframes.

And those mainframes are the things that process all our transactions that are out there, and I think it's still in the neighborhood of 80 to 85% of all transactions are on a mainframe, and that's called the real economy. And then we have this emerging economy around cryptocurrencies. And what IBM is saying, and also so are a lot of the banking networks now, that we're going to connect these digital currencies, AKA stable coins, and we're going to process those transactions and keep track of them using the same infrastructure that we use to pay our other bills.

And so as I look at all this, Drew, it's like have we finally come to the point where the IT mechanisms that are going to allow this transition to digital currencies are finally falling into place? I think so, actually. And this is interesting because this has been a long time coming.

Many years ago, I worked in banks, and the SWIFT network has for a very long time been incredibly insecure. There's some great hacks that you guys are going to look at. Look up what happened to the SWIFT network over the last 20 years or so.

So there's always, since the beginning of crypto, was the sense that, oh, blockchains are going to solve the financial problem. We're finally going to be able to get rid of the middleman. Because when you look at what all these mainframes are doing, they're just matching up the buyers and sellers and creating the report.

It's not intelligent in that way. It's fairly mundane, but incredibly sticky. So where I worked, we had a permanent group of gray-haired people in the corner who were responsible for the mainframe that was still responsible for a huge amount of the business.

That is still the case, and IBM's made a great business out of charging a lot for supporting old hardware. All these banks that are stuck on mainframes are paying out the nose. Now they're sending that into crypto, and I think that's actually really fun.

The goal of that endeavor has always been transparency, and perhaps there's a way with a blockchain-based approach to moving money around banks. Maybe we'll be able to fight fraud a little better. That would be really nice.

Right. We'll have to see. Yeah.

We can do that on a private network where we know all the end users, but I think there was a piece in "The New York Times" today talking about how Russia has been using public Bitcoin networks to evade sanctions and launder money, and that's kind of a whole separate conversation. But John- Yeah ... let me add this little wrinkle for you.

So now we have stable coins. Stable coins are issued by companies or organizations that are processing something around a particular class of transactions you might have. I don't know, I'll make something up, stable coins for the diamond industry or something.

And so they are using those coins as their currency, and then they'll exchange them at the end for real dollars. But what happens when one of those stable coins collapses? Because now it's really tied into the real economy.

It's not some isolated little thing. So, is there a ripple effect now because all these "stable coins" are going to be interconnected and interdependencies between industries? Yeah, that's a terrifying notion, and thanks for passing that on.

If I'm not paranoid enough after this conversation. Yeah. I guess maybe hence this is why we're having, and I haven't followed it closely enough with this crypto bill- Mm-hmm ...

that was trying to make its way through Congress, and it's not going to, I guess. It all comes back to speed and monetization and convenience with unintended consequences, and you think of the ripple effect that that could evoke. It gives me pause.

All of this right now, we're kind of in this state right now of uncertainty and kind of reckless enthusiasm. That kind of ties together all these segments. I wouldn't say no thought has been put into it.

Plenty of thought's been put into it, but it's also about spinning your yarn to fit your purposes and not scare away your potential investors. And in a sense, it just gives me even more of a cynical viewpoint of the government. And I think a lot of this, I think this kind of ties into polling that shows this incredible distrust of the public, more so than ever, in government.

And I think a lot of it is tied into this government's embrace of data centers, of super intelligence, of crypto. It's almost like a- Mm-hmm ... Ponzi scheme in a weird sense.

Well, there's massive amounts of money from the crypto people pouring into these political campaigns around this particular issue. But can I scare you a little bit more for fun? Oh, God, no, please.

Okay, go ahead. Consider this possibility. Most of the transactions around these stable coins and whatever else today are pegged to the US dollar, and the US dollar has always been kind of the economic engine, which we can say the bat that we pull everybody in the line into, because if we mess with the value of the dollar, we can destroy whole economies.

Now, imagine this. I can take my little stable coin and decide not to peg it to the US dollar anymore, and I can peg it to a different set of currencies. And suddenly I have a lot more economic freedom that is not dependent upon whether or not there's somebody in the White House has indigestion.

So, we could be... Or insomnia, depending on which case it is. Yeah, that's messed up.

That's accurate. But so my point being here is, we in the United States could be ceding in our zeal for driving cryptocurrencies and the potential money around that, but it could be leading to something that is not necessarily in our current self-interest. What do you think, Drew?

I think there's a massive macroeconomic effect that's in play on these crypto markets. I've seen some analysis, but I'd love to see a research paper on this. What happens is when you put money into, say, like a stable coin, you actually have to post that amount to the exchange.

So you have to actually, it's not like a futures market where you only have to cough up the margin, the small difference. You actually have to put all that money on the exchange, and that prevents it from being used for other things. Now, what happens is that money in the exchange gets sort of recycled into other cryptocurrencies, and so the money is all churning around, but it's not reaching the regular economy.

So it's almost like money kind of gets stuck in these crypto- Kind of sounds like those investments in AI companies I keep reading about. Yes. So, okay, I think there's just been so much abuse of blockchain technology for so long.

Everybody understands that it's being used for really horrible criminal money laundering, all kinds of different things. That may be why it was created, but that's not what stable coins are for. Stable coins are meant to hedge your investment.

So if I put money into this Bitcoin, I might want to use a stable coin to maybe offset some of that risk, because the crypto markets are so unpredictable. So when you talk about a collapse of a stable coin, it has a ripple effect in the larger economy around crypto. So that money that is now kind of gone when the wallet is stolen or whatnot, causes just this huge proliferation of just noise within the blockchain.

It's really ugly. So for fun, think about this. John, NVIDIA could start issuing its own stablecoins with pictures of Jensen on it and start handing these out to various companies that it's investing in and cross-investing in.

And then, of course, Elon will have his, and Larry Ellison will have his, and then we'll just go crazy with this. What do you say? Well, you're making me more and more alarmed as we speak right now.

But yeah, I actually could see that. Normal. And given the power these guys have and the access they have, especially the access they have, they probably could just whisper this tonight maybe.

At the state dinner, they could whisper this idea or concept to a certain individual and have their way and play chicken with the economy or play Russian roulette with the economy and blow a hole in not just ours, but the rest of the world. So you think they're going to be watching the Techstrong gang tonight too? Oh, yeah, I do.

So, Mike, I'm going to blame you. I'm going to blame you if there's a calamitous world event because in a sense you've given these James Bond villains an insidious idea that I think will lead to all of our extinctions. So you heard it here first.

Techstrong gang has precipitated a world collapse. Yes. Yeah.

Well, and Drew, correct me if I'm wrong, but the Federal Reserve has influence over the US dollar, but it may not have any ability to exercise any control over stablecoins that are issued internationally by companies that are essentially being their own bank. Yeah. It's a really mixed bag also in the US.

So there are some coins that are regulated under the same regulator that regulates futures trading, but not all of them. And internationally, really what happens is everyone just trades in Hong Kong, where there's really no rules. So creating rules in the United States might not actually have the desired effect.

Mm-hmm. But, I think that we're going to see this connection between blockchain technologies and AI, and NVIDIA is right in the middle. So they have a real strategic advantage for this market.

What do Bitcoin miners and AI people both want a lot of? Mm-hmm. Graphics cards.

And guess who has all the graphics cards? So yeah, it's a good play for NVIDIA. I think they have so much control over this market.

It's kind of like taking candy from a baby. Right. So what is your ultimate confidence in the ability to secure these networks?

Because to your point about Swift, Swift's been a number one target for cybercriminals for as long as anybody remembers, and still probably one of the most widely used networks out there. But ultimately, are we just going to see cybercriminals stealing millions of dollars worth of stablecoins and no one's going to report it because, well, that would be too embarrassing? Yeah.

There's a lot there. So, I think it will be a net positive, right? So using a blockchain allows every participant to validate, right?

And it makes it much harder to hide transactions, especially with the types of analysis that we can do nowadays on the transaction logs. So I think it's like everything. People are going to use it for bad, some people are going to use it for good.

But I think we're not done with blockchains yet. We're going to see a lot more of this sort of regulatory usage, I think. All right.

I think, folks, I'm going to leave this here for now because it's pretty clear that we are experimenting with things at a level of scale that's probably never been considered, and the implications thereof are mighty. So I'm not saying you should lose any sleep over any of this stuff tonight, but in the next couple of months or so, you might want to start thinking a little bit harder about the implications involved. Maybe the UN has some ideas here.

Well, these are like nuclear weapons. We're talking about nuclear weaponization of a sort. So maybe you're onto something, Drew.

Maybe they can save us. No, I- Maybe not ... I think we're going to have to bring back the League of Nations.

I don't think the UN is the thing anymore. Wow. Nice.

We'll have to go back old school because the fundamental problem with the UN is unless you're one of those five permanent members of the Security Council, it's a stacked game. It's a stacked game no matter what. And I think a lot of countries are going to sit there and go, "You know what?

" That's just me. All right. " Please stay tuned for the rest of the Techstrong TV replay for the day, and we'll see you all tomorrow.