20250227 Bits of Quantum

    6:59PM Mar 29, 2025

    Speakers:

    Rob Hirschfeld

    Keywords:

    Quantum computing

    quantum chip

    Microsoft

    Apple silicon

    AI servers

    noise reduction

    matrix multiplication

    quantum entanglement

    cryptography

    cancer research

    distributed quantum computing

    blockchain

    secure communication

    high-speed interconnect

    AI gaming desktop.

    Rob, hello. I'm Rob Hirschfeld, CEO and co founder of RackN and your host for the cloud 2030 podcast. In this episode, we dive in to all things quantum computing, starting from the idea that Microsoft managed to put a new quantum silicon chip together, but we go all over quantum, from compute to entanglement and everything in between. No you will enjoy the conversation.

    You the topic for the day, by the way, is the is the quantum chip? Oh, cool, quantum chips. Thank you.

    I'd love to see the marketing on quantum chips. Would they be like Doritos? Would they be like this, Vicky? Would they

    like, yes, yeah, they, they have the barbecues, the barbecue flavor, the, you know, the lime, lime and salt, sea salt, yeah, yeah,

    that, yeah, those are pretty good. Salt, vinegar is good,

    always good. Um, yeah, because there's um Amazon, just announced their new quantum chip, is the Microsoft chip, and then Google's been, um, kind of, in the in the same same area, doing a diffusion, a diffusion.

    I'd love to know what their what those servers are going to look like that Apple is going to build. That's going to be interesting.

    Yeah, who's tech? You know? Is it all Apple? Is it a mix? Who knows? Ah,

    I haven't heard anything about contracts being canceled with Foxconn. I haven't heard anything about TS TSMC, and I haven't heard anything about orders out of ASML,

    right? Well, I mean, the, I guess the point here is, you know, are they going to build it on Apple within Apple silicon and or is it a mix of Apple silicon and something else? What's the something else? And then we still don't know. I mean, well,

    here's my question to add to your list. How many nanometers?

    Three there's going to be some three nanometer stuff in there is my, is my projection. That is one thing that the quantum chip specifically, or no, no, we were talking about, what Apple is, is going to be building with their alpha? What is it? No, they're $500 billion they're talking about spending and and building out, you know, variety of sources and services. No, that's not quant. That's not, that's not a quantum, quantum computing investment on their part. No.

    So we were talking about what flavors they were going to make, whether they were sea salt and lime or, you know, salt and vinegar with barbecue pickle.

    Yeah, there's, that's a, yeah, I assume the apple discussion was the apple investment was, had to be aI around because export controls. But I think that's probably true. But, I mean, they don't usually sell, you know, server infrastructure. I mean, this could be desktop AI powered desktops, but, yeah, they may be pushing up for a lot of support of exactly that the AI, well, there's the desktop, but there's also, you know, the support of of real smarts in the iOS devices, based in part on, you know, the combination of on device intelligence and a, you know, a model that's not a back end, unlike what they've just announced, you know, with, you know, kind of the fall back to chat, GPT and so forth. And. Open AI, I think they are going to want to keep keep their distance from open AI, you know, at least to keep their they're likely to keep their options open there. So the question is, what are they going to be building out there? I don't think it's, I don't think it's known anymore. Yeah, well, I'm, I'm guessing there will be, there will be a lot of, probably a lot of quite distinctive and distinct kind of architectures and approaches to deployment that they're going to make, um,

    well, apparently they're producing AI servers there

    their own hardware. Yeah, I saw AI servers, but, I mean, Apple has not generally been in the market to sell AI compute, or, you know, AI serves, although, like, if you look at the new framework, Did y'all see the there were some announcements from the framework. Frameworks, a my team loves these things. They make laptops that have interchangeable parts. Very nice, yes, very nice systems. And they announced a desktop at their event yesterday. And the interesting thing on the desktop is that it has USB four interconnect, which is based, you know, basically high speed bus interconnects, and you can chain multiple machines together to act as a GPU expansion. And what's the what's the silicon that framework uses? It's open. I'm assuming they're using NVIDIA or AMD GPU pieces. That's interesting, but the inner but they would they the thing that seemed to be getting everybody's interest was not so much that they had a desktop, which are already pretty modular, but the interconnect so you could build an AI gaming desktop that had multiple GPUs chained together, or or an AI cluster, a mini AI cluster, a mini AI cluster, right, which people have been doing with the new Mac Minis, and there's a, there's a, there's this, there's a software or software and hardware combination that's using

    Thunderbolt as the way of keeping, of, of kind, of bringing them all together. And I think you have a kind of your you max out at eight of them, but, but the point is, you can actually, oh, here's, uh, here's the link for, for uh, or what, I'm sorry, oh, for the Mac mini cluster, M for mini cluster running large AI models using the Thunderbolt chained right, something called EXO. Remember what it's called, what it was, what the software was. But there's a, I think there's a box that you can use, and then there's some software, but apparently it works pretty well. Not sure what I would want with a, a Mac mini cluster, you know, probably I'm not sure what I would want with a with a big Nvidia box sitting under my desk, either? So not exactly a desktop. It's funny to build right to, yeah, you know, you're getting a whole bunch of infrastructure to basically access more GPUs, but I guess that's exactly what's going on. So, yeah, well, if you believe the scale, you know, if you believe heartily in the scaling law, yeah, that that may be, you know, the approach. I guess. The next question is, is there something else that you could do with frame that would, that would actually make better use of the distinct and different, different devices that are, that are chained together. I mean, using, using Thunderbolt, it's the as a as a kind of a high speed backbone, is not my idea of the way you want to build a build a cluster, but it's interesting that we are, you know, we have on easy reference to different, you know, multi you know, multi machine, chain. A, you know, desktop systems for AI clusters. So the trend line would indicate, you know, there's something going on there. Yeah, something going on. This is, in a weird way, an edge computing story,

    um, to me on the topic of high speed endocrine so we so framework uses used before. But I mean, last year old, all the fanfare was around Aki link, and I haven't heard anything new about that. So what happened to it?

    No idea. You're right. It was a, it got a lot of, got a lot of press, but it's still, I mean, it's, I think the press has died off. I know they're still, they're still working on those the systems. There was a whole dedicated aisle for them in Super compute, super compute for vendors trying to make this stuff work.

    So but you're right. It hasn't been as buzzy of late. They might as SXL, SOX, server interconnect. Oh, you know what probably happened? Intel,

    yes, got

    pushed to the edge of the curb.

    That would make sense. I would see it more teetering on the verge of, you know, falling flat on its face. But I yeah,

    this is actually an interesting segue. Do we want to talk about the the quantum compute chip breakthrough for i from IBM? Yeah, I think we should. I'm fascinated to hear because I have a high degree of skepticism on anything Quantum. I have a pretty high degree of skepticism on its applicability, or its actual being put into any sort of production in the near, near future. I mean, it's, I think even if you go to, you know, the the folks that are building, they are all saying, well, the more the more aggressive of them are saying, five years, and the more kind of thoughtful of them seem to be saying, you know, eight to 10 years before we're we get something really, really useful production out of quantum computing, other than the possible early use and, you know, racking open crypto. You know cryptography, which is you you know that everybody's just, you know, banging away on as hard as they can. Joan, do you have any insight into this new form of matter that they are claiming about? You know, they've they've invented as a result of their their semiconductor, their next gen transistor? Tech, do

    it, let's just say, and I'm covering my arm because that's where my blue stripes are. Asked me the same question in a year from now, and I might have a likely and reasonable answer for you until then, the word vaporware, but in hardware comes to mind. In other words, very good in the lab, really excellent in the lab. Fantastic benchmarks, but put it in the real world. Chris,

    so are you saying that it's the equivalent of a modern perpetual motion engine?

    That would be a very apt description. Close,

    I like that.

    Or the wires going behind the curtain?

    Yeah,

    I was pretty damn amazed to see it fail in a demo.

    Oh, you didn't see it failed. Yeah, I

    did. I did. I happen to have made a trip. Oh, before the, before the last blizzard. In between blizzards, there's, there's an IBM facility with a lab, and I still have some D. Deep roots in IBM, and I got invited to a very, you know, kind of invite only preview, and watched it go, yeah, splat. And, let me say, this heat, Holy Hell,

    yeah, a bad I believe it.

    I mean, literally, we walked into this room and it was, I would say, probably 50 degrees Fahrenheit. And after the demo, the thermometer on the wall, and there really was a thermometer on the wall, because they were, they were trying to see what it was, 81 degrees. Wow. And it wasn't body heat from a whole bunch of us. There were only, you know, a few people, and it was a fairly large room, and we were, you know, kind of back. So that's why I'm saying, put it in the real world. I don't think so, not quite yet. I don't think they got out over their skis. I just think that it was probably one step up from prototype Charlie. It wasn't a beta version. It was a little bit more production ready than that, but other than that, no, like I said, Ask me in a year from now. I

    mean, it's, you know, if you're, we're talking about, you know, qubits, I guess I have trouble understanding, you know, computationally, what, what problems they're going to solve, or how, I mean, I understand they're supposed to be faster certain classes of parallelization, yeah,

    sorry, in the high performance computing world, okay, it's, it's not just about how fast they transition through the numeric cycles And the stages and whatever. It's also about the interaction of the particles, shall we say, right? And they can. How do I describe this? Think about, think about looking at a part a particle screen, right? And you know how the particles bounce off and then go in different Okay, so you get close, but you don't actually touch. And they pull. They pull, from a physics perspective, they would sort of pull capability from the next one without actually touching it, which is what they're doing to reduce the heat and to go through the stages faster, right? You can be near but not touching. You can come to a certain point of interaction, but you still have distance, right? And that's for heat dissipation. Then all the other you know, sort of physics properties that go with it. But other than that, it's the it's the velocity through which these are is cycling. That's the biggest improvement.

    But, I mean, we're, when we talk about these qubits, they're not, my understanding is they're not computationally binary. They're not right. We're not using them to they're actually there. It's the way you the thing you should think about is trying to do matrix multiplication with, you know, yeah, heavily with lots of dimensionality on the matrices. Okay, doing it in parallel. So anything that's reliant on on matrix multiplication, matrix operations. That's, that's the big that's, that's the big win. The big problem in my understanding, and you can check me on this, is that, as they're expanding the number of qubits that they can manage, you know, at a time, you know, the size of these, these boxes, the the the biggest challenge is noise. And so the the issue is noise reduction and and and kind of remediation. And if you can do that, if you can kind of get that under control, then that's the win. That's when you actually get use of of the quantum computing power you get with, you know, 128 or 400 you know cubits. Whatever you know is the number of the month, number of the year,

    but the degree of the degree of noise is very unpredictable. Yeah. And that's, that's what I was trying to describe with the particulars. And maybe I came out, okay, but yeah, it's the the amount of noise. You can't go linear across it and go, okay, my little blip is going to be, always going to be two or one, or is it, you know, whatever that numeric would be. It's more like, more like a SYN graph than anything else, from a descriptor point of view. But you can't change

    the right go ahead. Isn't the big win that Microsoft is talking about with this new chip, the fact that they have made a measurably, really significant reduction in noise,

    yes, but it's still not, yeah.

    Okay, it's not, because it's not the point so if you have noise on a big, fast matrix operation, then and then the whole matrix operation goes into question. Basically, exactly, okay, in approximate, I mean, there can be value in approximate calculations, but it's, it's right, because that increases really quickly. Yeah. I mean, the whole point is to get you, well, let's put it this way. You don't spend that kind of money to then get that kind of, that kind of accuracy, or that, you know, that many floating point numbers, and then quantize it down to, you know FP 16, because you can already do fifth FP 16. So, right, right.

    I'm, I'm, I have been looking since I saw this thing. I have been looking for a word that describes in quantum terms, a hallucination. Hmm? Just like you get from Ai

    Phantasm, a

    Phantasm, that would be a good one, a phantasm. Phantasm, very good.

    So, so do they have to then, like, recalculate. If you do math right now with the with quantum system, you have to do you recalculate it to validate that it's authentic, that it doesn't have a fan.

    Because that was the other reason, when I said kind of fail, they ran four sets of calculations eight times. Oh, okay, and the percentage was only about 65% accuracy. As the number of tries went up, I would have expected it the other way, but that's just me.

    So you can, you could use the quantum system to, you know, run multiple times check to see that the results are, pick, pick the dominant result and be satisfied. Okay? And it's fast enough that that's okay, yeah, okay. And

    I also found that I have, I basically have asked for the benchmarks, but I don't think they're going to release them, because that's really proprietary stuff. But I wanted to know in the amount of noise, how you might use the machine itself to have agents running to remove the amount of noise, to make it more predictable. Yeah. And with every calculation, if you could say the noise level was going to be higher or lower, because right now, there's not a lot of ways to predict that.

    Well, there are too many. There are too many, very potential sources of noise. Is the problem when you talk about having agents in there to Well, I, you know, I suppose that.

    I don't know. Could you make it could you make it self? Adjusting, really, to get a handle on the noise. I

    I don't know enough. I don't know I don't I don't know enough about the, the the sources of noise and the, you know, what's now being categorized as noise. It could be all everything from, you know, you know, a somebody walking too close with a, you know, you know, a wrist watch on, who knows

    it's, it's a hard one to solve. And. And because I couldn't see, and this is just me, because I couldn't see, a way to make the noise less variable, more consistent, more predictable. Get rid of some of the variables to sort of, you know, slim down. These are the things you really have to work on to put it in the real world. That's why I'm like, Yeah, give me a year. You know, five guys a lab will figure this out eventually. But, but what I did find interesting, they are working on a real life problem, cancer research that was really interesting. I can't say a lot about it. It's a project that started on the world grid years ago, and it's one of the it's to try and figure out if there's a way to use quantum to start. How do I put this? Every every person who gets a particular type of cancer has a variant of it, they're not all the same, and they vary from person to person. And it's based on your DNA. It's based on, you know, your physiology, your your etymology, everything like your entire going backwards in history, your parents, your family, there's tons and tons and tons of variables. So they were running this experiment to see if they could start to try and map, figure out a way to map not your risk factors, but your treatment plans, because they know how to map your risk factors, they don't know how to map your treatment plans, because everybody you know who's on the call could be in the same situation, taking the same medication, and have absolutely, completely different outcomes, different side effects, different lifespans, the whole nine yards. That's one of the problems that they're working on with it. And to me, that was very there's

    so many variables to factor in on on those piece pieces, absolutely classic machine learning, classic machine learning, computational right? But if you can add, but if you can add in, you know, 100 more variables, or 1000 more variables on, then you, you might be able to get much better results.

    Okay, exactly,

    right, exactly. And then we're back to, can you come copy, you know, crunch the numbers Fast, fast enough to do that type of work. Well,

    they're trying to do it to proteins, amino acids, lipids, like all of the stuff in your body. So talk about a matrix that's pretty significant, because apparently even the water that we ingest changes so rapidly and in so many different ways that it's very hard to assess right? Because if you have more keratin, your muscles are larger, but they're holding more water. If you are, you know, like each each little indicator has an impact on so many others that that's where the matrix becomes very complex. But it would be amazing to see it actually work out a solution where it could take some sample from you and tell you, you know, this is your absolute optimal treatment, treatment plan, or a course of action that would be, that's no bill worthy,

    yeah, that would make a huge difference. Yes, it's, I guess I thinking on the qubit pieces, you know, I guess I've known for a long time. It's, you know, functionally better. Matrix math, faster matrix math. I guess every time I hear quantum I get you know, confused about, you know, like the, you know, scale of it, or the mechanism that it's using for it. And that seems like, you know, potentially a distraction for for what we're taught, for this, maybe not so the, every time they talk about quantum, you know, they they show the immersion, like the, you know, the big, the big gold frame with the to go in the immersion tank. And, you know, it's, there's a lot of, there's a lot of visuals in it. And I assume that there's a, you know, what was it the supercomputing people were talking about quantum entanglement, yeah. And that's a, that's a different issue. And actually quantum and. Entanglement as a as a form of of communication, data communication, right? That is almost, well, let's put it this way. It's tamper resistant, because if you tried to tamper with at all, it would just fall over. It wouldn't work, right? Right? So it's, you know, it's, it's kind of tamper evident. So the notion of using entanglement as a form of to a, yeah, yeah, multi could be multi could even the question is, can it be multi point communication? But even if it's two, two point communication. If you can get this, the speed up, that's pretty that's pretty amazing. So they are, they also demonstrating quantum entanglement in the in these pieces, okay, well, I don't know about the Microsoft stuff that I know that the there, well, two cases. I know that there have been actually results, some results published, academic publications out of the out of China, because they've got a they've got a minimal quantum computer, quantum setup for both cryptography and entanglement in a set of satellites that they've put up, and I've been aware of that they're straight distance, yeah, interesting. Okay, they don't, they don't. They don't give you all of the parameters, but, okay, they basically demonstrated the ability to stay, have some sort of a sustained or some period of time, sustained data flow using entanglement. Okay? You know, it's kind of like, okay, we've got data communications the day where everything that we're doing cryptographically can be completely blown away by quantum computing. But don't worry, because we'll use entanglement to actually have this very secure communication. It's kind of like

    that. That's point to point only. It's

    point to point. Oh, yeah, you can't securely broadcast

    or athletic more than there is no more than two points there. Yeah,

    absolutely true.

    I put a reference to a paper from irvic that I found quite intriguing and interesting. Network assisted collective operations for efficient distributed quantum computing.

    Betty, many, okay, distributed quantum computing, distributed, how or what? What's the inter protocols?

    Well, this is proposing protocols for the distribution of collective quantum operations between remote quantum processing

    and local but it's but it's conventional networking. It's not Quantum. Quantum well, but if you're saying that you can do speed of light networking, or faster than speed of light networking, between two sites and the distributed the word distributed computing changes, meaning, right? Yeah, okay.

    But it's, my point is, it's, it was a very interesting paper to read.

    In some ways. It's funny, because I think, you know, we talk about qubits all the time the quantum entanglement has is, is what would be more immediately disruptive, breaking speed of light networking for communication. Then, then, yeah, the qubits, to me, wasn't

    it tested unshown, the entanglement still followed relativistic loss. So

    a interesting question, Is it soft and light? So,

    I mean, I like faster than the light communication or transference of information is like I would have expected that to create a lot of more noise in the scientific community, because it would break a lot of physics assumptions, like causality, and

    I wasn't aware. I don't remember seeing any of that, but wasn't like I. I was doing a deep dive into it. I mean,

    the Sci Fi nerd in me would love, yes, that I like communication. It's just that I I can't see it as having been demonstrated yet, given the lack of unfair about it,

    right? Yeah, I think all that they've been able to demonstrate is, you know, for the for some value of instantaneous, they've been able to demonstrate that the whole notion of, you know, faster than light to your point, gets, you know, starts to, you know, turn your brain inside out, with regard to things like causality and you know when, when you receive the message before it was sent. You know, it does kind of tend to, you know, work, your work, your sense of reality. Here

    you have spooky action at the distance without it being immediate, immediate.

    Yes, that is true. And, and the question is, if it's, you know, if, for all intents and purposes, it is immediate that that still that gets you to some pretty interesting places.

    Make it sound number one, because it is more polite in so many ways. But what I was gonna say, actually, it was really funny. And I mean this in a totally funny way, the Back To The Future of this morning was I clicked on the meeting link for this call, and it comes up as June 22 2023 and I'm like, Well, talk about going back to

    the future, because that's how long you've been going.

    Yeah, meaning to mention that, but yes, that like the the meaning description still has some very old title, oops.

    Oh no. It was just really funny to read it, because I'm going, why is it snowing in June? Two years of my life back, you know, like

    I was, I was skipping a week, or,

    I don't know, maybe it was a Cuban in my coffee be

    an errand. Cuban,

    there you go. Better.

    You're making me think about the three body problem with the intelligent photons. But yeah, God, folded so much crazy, ridiculous tech into that, into this, that book. Yeah, you're talking the story. Yeah, the story it, I don't know. It's still a lot of fun. I enjoyed those books

    on the matter of of fictional stories the quantum back when we were talking about like reducing interference, the first thing that came to mind was something from a Neal Stephenson book, The Horizon fall of dodo where, essentially they surround each chamber with super cooled helium in order to build it right. Essentially, do do a quantum isolation of the interior of the chamber and on the on the outside. Oh, huh, of course. Then it goes into more fantasy, but Neil Simpson usually does a fair, fairly good amount of research on the lives. He does his

    homework. He does his homework. Yeah, yeah, absolutely, yeah. Those are fun.

    I can't imagine what the world is going to be like in two years from now, if they really make this thing work.

    Where, I mean, you mentioned medicine, where else do you see the most immediate application besides breaking other people's crypto, cryptography, energy cryptography, the idea of, well, there's a lot of thought. I've heard speculation, but I sure haven't figured out how exactly. But if they can have sustained computation at these speeds, then and energy. What you're talking about is controlling at just very, very fine grain, very, very fast, you know, fusion reactions, yeah. So exactly, basically, keeping the, you know, basically. The the adjustment to, you know, maintaining the the containment of Plaza, if that plasma, if that actually can be made to work on a sustained basis, that's pretty heavy duty. The new

    right, the fusion reactor. Of the sites are rather interesting in that they that, at first glance, they have a non symmetric topology. I mean, when you go into more details, like there is some symmetry in there, it's just that they look very, very weird to the human eye, on all twisted like a ribbon. Yeah? Another another place where I could see quantum entanglement Make, make a difference as well. Or is atomic transactions? I It's across jurisdictions. Like, let's say for example, like, if someone in the US and has tried to trade with the EU like, if you can tie the entanglement into a transaction, then you have to proof of that transaction happening atomically. Then you wouldn't need like escrow or anything like that.

    You You basically avoid the notion of a distributed ledger,

    right? But you, you, you, you have incorruptible proof on both sides that you're right. Okay,

    what was that you wouldn't and you wouldn't need an arbiter or a third or trusted third party of any kind? I don't understand, why? Why not? Because the, I mean, the parties are talking directly to each other here. Why? Why would you eliminate you? Still? You still need the external, yeah, well,

    the, the entangled, uh, particle is your arbiter is, yeah, like, you used entanglement to sign your transaction.

    Oh, I see what you're saying.

    Yeah, really interesting. You could get over a lot of the hurdles of blockchain, yeah,

    you wouldn't need to store the chain history, because it's implicit in the

    fact that it exists. Is, yeah, is proof? Is enough? Yeah,

    it that would be one.

    I'm not that would really be filing, how that, how that provides the proof, because if there's a dispute between the parties.

    So, so when you do a transaction on the blockchain, right the way you do it, and understood, given the street ledger is, you submit a transaction, and then you have several other parties that do a calculation or do their own sightings. And yes, I've seen this happen. It is real to me now, and once you have a sufficient threshold, then you can say yes, it's reasonable that this happened, and everyone has agreed to it. Okay. But the downside of doing this on a blockchain is that the blockchain needs to keep a transaction history of everything it's done since the beginning, in order so that you can replay it right? If you use the entanglement to sign your transaction, because the entanglement is unbreakable or or untemporable, okay, exactly. And the moment you sign the transaction, it's immediately set in stone, like you've done it is it? You don't need multiple parties to review it, because the nature of the entanglement itself means that as long as the entanglement stays, every transaction signed by it is legitimate. So

    it's absolutely equivalent on the on both, both sides of the transaction.

    Yeah. So as a result, the existence of the entangled itself is a substitute of the blockchain history. So you don't need to keep that history anymore. So it makes it a lot more resource, like it reduces the resource cost, because the biggest downside of the blockchain is that business, it becomes bigger. Yeah,

    and it's both storage and just with its size, the compute that you have to do. And transactions, transactions on blockchain, especially if they're interconnected. Blockchains, they're painful, very I mean, I see it. You trusting the synchronization the history is actually, to me, part of the value. I'm missing something in in this but, well, the history you save with the blockchain is the history of the blockchain itself. It's basically, it's all the transactions, yeah. So when, if you're, if you're looking for, for the the the lineage of that transaction, that's a question. That's actually a very good question. If you need along with the transaction, it's any dependencies, any any dependencies or or kind of history of what happened. I don't know what you do with that, with with entanglement, all that the entanglement, I'm just seeing it as a better pipe and a more trustworthy pipe. I'm not seeing the record keeping and the witnessing, which I think are, are alternate or additional pieces. If they the record

    keeping would be done. It basically means that the records don't need to be kept publicly, as long as they're signed. So you can, you can take a

    private record Okay, and verify that record

    against the entanglement to confirm that yes they are.

    So you're using, you're using the entanglement as an authentic authenticity, right? It's validating any other validator.

    It's an authenticator. It's a signer, the alleged trusted third party, or

    even just in general, like, yes, we're going back to the idea of cryptography, but entanglement as a substitution for like a TLS handshake,

    or you could use it as a what you do now with Blockchain, you depend on these you depend on oracles, right? And if there was some way in which the entanglement was the the transaction that was, you know, submitted or kind of applied by an Oracle, it would basically take away that, that little fiction that everybody seems to, you know, kind of ignore when using Blockchain, which is, yeah, I'm going to depend on oracles, and I have zero, by way of understanding the authenticity and and authentic ability of whatever the Oracle you Know, you know, jams into the network. It's just whatever the Oracle says. Of course it, trust me. Trust me. Yeah, the check is in the mail, and, and, and I need to, I need to wrap up today, because I haven't. I have a meeting. Chasing this one down did not go at all where I expected. So I appreciate the conversation.

    Do they ever Rob

    sometimes? What makes them fun? Okay? Everybody have a good weekend. See you take care of March. Bye,

    bye.

    Boy, what a lot of fun to talk through Quantum which I still remain skeptical about. But I think that the ramifications of improved matrix math, improved communication latency or zero latency can be pretty fascinating, and I hope this got you thinking about ways in which potentially, we could change how we validate and think about trust in computing systems. A lot of really heady topics in this conversation. And so what we are known for, it's what Cloud 2030 does. I hope you want to be a part of it. Come in if you've listened to this part, you are cloud 2030 curious we want to hear from you. Please just join us on one of these 30 Thursday morning discussions. Thanks. Thank you for listening to the cloud 2030 podcast. It is sponsored by RackN, where we are really working to build a community of people who are using and thinking about infrastructure differently, because that's what RackN does. We write software that helps put operators back in control of distributed infrastructure, really thinking about how things should be run, and building software that makes that possible. If this is interesting to you, please try out the software. We would love to get your opinion. And hear how you think this could transform infrastructure more broadly, or just keep enjoying the podcast and coming to the discussions and laying out your thoughts and how you see the future unfolding. It's all part of building a better infrastructure operations community. Thank you.