Quantum Computers: Truth, Facts And Hype

Mark Fielding (00:33)
Bob Sutor, welcome to Thinking on Paper. Thank you for thinking on paper with us today.

Bob Sutor (00:38)
Thank you. Happy to be here with you guys.

Mark Fielding (00:40)
In the past twelve months, in all your world travels in the quantum realm, what is the most impressive experiment, use case, or quantum computing

Nugget that's stopped you in your tracks more than anything else. What stopped you in your global travel and God? ⁓ my god.

Bob Sutor (01:02)
honestly, to to answer slightly negative at first, ⁓ sometimes I think people are aiming more for the history books than anything else. You know, here here's the quantum supremacy du jour, here's this week's quantum advantage, and things like this. ⁓ for me, I you know, I think it comes down to basic engineering.

Right. So people who are are really doing very high quality ⁓ work on this innovation. Right. So science can be messy. S science you know, you you can't guarantee scientific results. I was mathematician, you know, at some point my advisor said, On Monday your dissertation will be finished and this was Thursday. And I was like, really? Okay. Amazingly it was.

Jeremy Gilbertson (01:44)
Yeah.

Mark Fielding (01:47)
Those constraints

Bob Sutor (01:48)
But now seeing the ⁓ the clean processes that people are using to advance a lot of what's happening in quantum. So bringing real engineering, so fabrication, for example, ⁓ trying out different vendors, ⁓ A B testing on on on different elements of what goes into an actual quantum system. So we are a long, long way from ⁓ from not having any more science. In fact, that will never end.

But it's engineering discipline that I see when I travel that really gives me the hope that we're moving towards something solid.

Jeremy Gilbertson (02:24)
you talk us through you mean by, you know, you reference science and then what you mean by engineering as it relates to what's going on in quantum?

Bob Sutor (02:32)
Sure. And and there's the third ⁓ element actually, which is the mathematics. So for a long, long time, mathematics and physicists have ⁓ usually happily joked with each other as to which is more important. I'm a mathematician, so it's pretty clear which is more important. But there's another ⁓ interplay which is between scientists or physicists and engineers.

Jeremy Gilbertson (02:49)
Yeah.

Bob Sutor (02:57)
And the physicists will say, well, this is impossible, or this will take you know a long time until we figure this out. And engineers say, look at what I just did, it works. Right? So there's this natural tension between theory and implementation. And we see that all the time. We see that in some of the recent papers about the reduction of the number of qubits one needs to do Shor's algorithm and things like this. These are all theoretical, right?

Under the most perfect conditions of this and that and whatever. Yes, we believe that we can do this, but by the way, none of this actually works today. ⁓ but then the headlines scream, it's done, it's happening Monday, right? And so that is what l is lost on on people, is is not understanding saying ⁓ the theory, this is how the pieces fit together. but again, to repeat this, in theory, right.

Mark Fielding (03:35)
Ha ha.

Bob Sutor (03:52)
But the actual building of it. And I I think we can see this today because whatever the number of qubits anybody needs to do anything, you know, it's thousands, it's tens of thousands, hundreds, millions. And people proudly say, look at my twelve qubits. We can extrapolate from here to a billion. It's like, you've got to be kidding me, It it's like they're on one side of the Grand Canyon and say, ⁓ I've got ten qubits, you know, like or you know, I can jump across that whenever I want to.

Yeah, I've taken two tiny little steps here. So that that's the ⁓ that's where a lot of the hype comes from as well, is people not distinguishing what the honest innovation from the best people on the science for quantum and then understanding what people

Mark Fielding (04:33)
Does that apply to IBM

and their recent Cleveland clinic trials with the with the protein simulation?

Bob Sutor (04:39)
well that's that's solid work, right? ⁓ because part of it, the the science part is understanding how to actually represent the information about the chemistry in the quantum computer so you can compute with it, you know, I I I would challenge most people listening to this, saying, You got a big protein, okay, put it in a quantum computer. What the heck does that mean?

Mark Fielding (05:03)
What does that mean?

Bob Sutor (05:05)
No, and so even we start talking about qubits and superposition and entanglement and this and that and whatever and Hamiltonians and you know, all all this this fancy stuff. It's just like, yeah, yeah, I stuck a protein inside the computer, it was amazing. It was great, and and things like this. So a lot of those things, like, yeah, it it it it's a very nice achievement, but we're we're learning about how to do this. ⁓ I'll also say we are in the prehistory of quantum. Okay.

Someday in the future, and I hope to be around. In fact, this is what I'm trying to do with everything I do is to speed things up. We will be in, let's say, the golden age of quantum computing, where the systems are big enough and you know, high enough fidelities, that meaning they do what they're supposed to do without losing information. It's gonna be fantastic. All these promises, right? And we can do this, that, and whatever, and you won't even know. We're not close to that. We are in this prehistory, this very noisy period.

I think a lot of the things that people are doing today that they are lauding, yeah, they're milestones, but they're gonna be completely forgotten and ignored and thrown away. They're important to the innovation, the understanding of the science, the early stages of the engineering, but it's like gonna be the ancient days. It's gonna be like looking at a black and white television in nineteen fifty compared to this screen we're looking at right now, you know, and saying in nineteen fifty, we're so close. We're so close. Right.

In in in one year, two years and three years we're gonna have flat panel displays with, you know, four K and this and it's like Yeah.

Mark Fielding (06:39)
But some

Jeremy Gilbertson (06:40)
are two different versions of coming to life right now. There are people harnessing what quantum can do right now and building businesses around that that are getting funded and that are doing pretty interesting things. But then they're also like a whole nother sector that are pointing to the future with fault tolerance. What happens to the companies that are building businesses around today once we hit fault tolerance?

Bob Sutor (07:02)
Well, they plan to evolve, I would hope, right? Because the the systems today are are are the the raw metal, if you will. They're they're really low level, right, at the at the hardware level. ⁓ I I I gave a talk, I mentioned I was in Slovakia ⁓ several weeks ago, and ⁓ I started my talk by saying the the three most important things in the quantum industry today are in this priority order money.

Sovereignty and then yeah, technology.

Money is a huge bit. So if you're a startup, you need money. You have to survive, right? So, you know, ⁓ you can ask this very question of a company, what are you doing merely to survive, to keep paying the bills, to keep buying whatever hardware you need, right? things like this, paying your people. we're seeing an awful lot of investments going on. We're seeing a few specs, ⁓ people getting pretty good size series A, series Bs, and things like this.

lots of bets, FOMO, fear of missing out, that seems to be flipping on and off month by month. Right? It's like, ⁓ everybody who is go going to invest has invested. And then suddenly there's this great flurry and things like this. So for a lot of those little companies, right, they need to have something that works as soon as possible. They need to get paying customers so they can survive as companies.

Now, unfortunately, a lot of the paying customers are universities. ⁓ they're really kind of the esoteric research parts of enterprises. You know, there it's not mainstream. It's not the part of the the enterprise that does like the databases and the security. It's like the research ends. And that's not going to be the mainstream.

revenue. I mean, it's just like data centers and and systems today. Eventually it will be real use of a lot of these systems. And so another area of hype related to this are the huge market projections. Quantum will be worth X billion trillion dollars in so many years. BS. Yeah, I'm sorry. You know, you take a lot of money, a lot of numbers which you've made up, multiply them by other numbers that you've made up and you add them together.

And you confidently say this is the market size. Yeah, right. You want a bigger number? I can find someone. You want a smaller number? I can find someone. Right. So so money is driving, you know, is behind survival ⁓ for a lot of this. It's an investment. We saw the US last week ⁓ say that they were going to do a lot. We saw France. You're saying smaller things in in Finland, Romania actually is going to spend a hundred million dollars on.

So money is so so dominating everything that's happening here. And I think it's hiding a little bit of who are the best and who maybe won't survive or maybe who shouldn't survive because they just don't have a hope of competing.

Mark Fielding (10:06)
You you mentioned the second pillar being sovereignty after money. Could you just explain what you meant by that first?

Bob Sutor (10:13)
Well, so why did the United States, for example, say that they were going to invest two billion dollars, right, in nine quantum companies? Well, two two of the nine IBM and global foundries, ⁓ which get one point three seven five billion, so the vast majority of this is for foundry work related to this. And so, you know, we can look at the past of what's happened with semiconductors and foundries and overseas ⁓

you know requirements on the technology, things like this. And then they were to smaller companies, ⁓ all of which were mostly American, and then there was Dirac, which was Australian. ⁓ but that's a special case because it's Silicon Spin the least advanced of of the modalities. France then turned around almost immediately and said, Hey, we are giving, you know, this much money to French companies. You've got Procure in the UK, which is

yet another billion, which is a competition, ⁓ which many people have pointed to as the right way of doing this. Why are government officials picking winners and losers as opposed to going through rigid competitions of who has the best technology? ⁓ I be ⁓ the the US has done that in the the DARPA benchmarking initiative and things like this. ⁓ Finland

spends an awful lot of time talking about how wonderful they are in quantum and things like this. countries don't want to be left behind in whatever this huge market is going to be with quantum. It's also a national security thing. So if you are a let me just say for the sake of, you know, a a major country who can afford a billion dollar quantum computer, where are you going to get that from? Do you want it to be homegrown? Do you want to buy it?

From US, from China, from somewhere else? ⁓ is that going to restrict what you can do? We talk about these great things like chemistry. You know, you mentioned IBM, a Cleveland Clinic. all this bit about national intelligence and cryptography and things like this. Do you think intelligence agencies aren't paying attention? Do you think they don't want large quantum computers to do fun things with? Of course they do. So it's economics.

Which is what they always say, workforce, which is great, and then it's national security issues and that's the sovereignty angle. And damn it, we're not gonna have to pay somebody else to give us a quantum computer.

Mark Fielding (12:45)
What's happening with Chinese quantum is since we're speaking about national security.

Bob Sutor (12:51)
I have tried to answer that question various ways through the years. And no matter how I answer it, somebody has a problem with it. No, no, I'm going to I'm going to answer it, but I'm just telling you right ahead, and I'm warning your listeners, right? So ⁓ so what I said the last time I gave a talk was from what we understand from public sources.

Mark Fielding (13:00)
We have no problem.

Give us the most controversial take.

Bob Sutor (13:20)
And what we can imagine regarding the quality of science around the world, they're probably roughly speaking where we are. You know, that is the West, the United States, ⁓ Europe, and so forth and things like this. Okay, well, that's the controversy. No, Bob, they're further ahead. No, Bob, they're faking it. They're not further ahead, right? You know, so you know, it's like we're well

Mark Fielding (13:44)
Yeah.

They're wallowing in our dust.

Bob Sutor (13:49)
Well, all of you get on one side of the room and the other get on the other side of the room and and and then we'll deal with it. But the fact of the matter is we d just don't know. And I I was taught by a former CIA agent to always insert this from public sources. Because, you know, we don't know. What now what I have observed is occasionally there's an announcement of this computer, whatever, and the information is almost impossible to find. Right. So it shows up in some newspaper or news outlet somewhere.

And maybe there's one professor talking to somebody. So is this really a company? Is it manufactured? Is it really just research and things like this? So although I said there were ninety-five companies, you know how many of those are really serious? I don't know. But I think it's a mistake to say that they are behind where we are. Let's put it that way. That's probably a reasonable thing. We don't know if they are truly ahead, but we have some really, really

Mark Fielding (14:48)
We we need a some Chinese

quantum dissidents to to leave the country and come and share their secrets. Is has that happened? Is there any any history of that?

Bob Sutor (14:54)
Well you know

Well, they haven't talked to me. I don't know. well we do know is I mean and and this is true in in general, it's not just Chinese. You know, people come to the United States, they go to the best universities here in Europe, elsewhere, and then sometimes they go back to whatever countries they happen to be from. Right. China did bring a whole lot of expats back a few years ago. There were announcements of somebody from Australia, for example, who went back to China to lead. ⁓

Mark Fielding (14:59)
Ha.

Bob Sutor (15:27)
But there's always movement and this is what I'm saying, it's hard to really say. And are they making the progress? I I I I will tell you a story. I gave a talk in Beijing must have been seven years or so ago. And ⁓ it was back before quantum had this kind of international geopolitical thing. It was all just nice, interesting science. And I I

I gave a talk and then I was sitting at dinner with a a Chinese scientist and he he leaned over to me and said, Every time you people at IBM make an announcement, the government calls me up and says, When are we gonna do that? Right? So maybe that's not the biggest indication, but they are certainly paying attention and we are paying attention as well. But ⁓ I will say that we whatever you wanna call us, the West, whoever,

Boy, we can't help telling the world every little tiny thing that we do. Right? So if there were a country who had a a great quantum development effort and was really doing superior work and decided we're just not gonna tell the rest of the world because why should we tell the world we're close to having a system that can break encryption? Whereas the United States is out there, yeah, we just did this. We increased it by one qubit. You know, our science is like this. This is the modality. Here's a paper on exactly how we yeah.

So it's it's a real strange contrast. We just don't know.

Jeremy Gilbertson (16:55)
are humans going to be the bottleneck to the adoption of quantum when when you think about our ability to ask new questions that quantum is capable of answering. It's like a whole new lens that opens up. Could we be the bottleneck to that?

Bob Sutor (17:08)
Some people are concerned about the actual breadth of algorithms that people are or may use with with quantum. That is i in classical computing, you know, we have tens, hundreds of classes of algorithms and then thousands of different different you know, implementations. Quantum the number is much, much smaller. ⁓ if if if your listeners want to go and look at them.

⁓ there's a website called the Quantum Zoo, which lists the algorithms and the variations and things like this. this is one area as we spoke about before that I think some of the methods that we use now on these noisy computers, the variational methods, those will largely, I believe, be thrown away once we have fault tolerance. quantum i is not, you know, as we've been saying for ten years, quantum is not a replacement for classical computing.

You know, to control the bits and the and and the pixels on your phone, we don't need a quantum computer, right, to have a crystal clear display, right? But for certain types of problems that are large enough, we think quantum will win. So if it's not one of those problems, or if it's a small problem we can do otherwise, don't waste your time doing quantum, right? So there's this threshold, right, where quantum will make sense for certain types of problems.

Otherwise the classical methods we do now are just fine.

Mark Fielding (18:38)
you mentioned that there are hundreds of classes of classical algorithms? could there be one day hundreds of classes of quantum algorithms or is quantum limited by nature?

Bob Sutor (18:49)
well it is nature. You know.

Mark Fielding (18:51)
Well th well that's what I mean. B the fact that it is nature,

is it limited down so it doesn't need all these many algorithms because it could just have a few because nature is more perfect.

Bob Sutor (19:03)
I s I suspect quantum is going to be a little bit more like AI, just on the algorithmic front. So you know, if you if you th so think of all that you know, the chat GPTs, the anthropics, the this is and the whatever. first of all, people are far more expert than I am saying, you know, in ten years we won't be doing anything the way we do it now in AI, right? We'll be smart. But nevertheless, it's all about the models, it's all about the training, it's about the parameters. The the

While we keep improving the raw infrastructure of the implementation, it's the numbers that feed into these algorithms that adjust the outcomes. Right? So it's it it's like you have a framework that you then pump in a lot of numbers and depending on what those numbers are, your answer will come out. ⁓ I think quantum has some of those characteristics. And that's where I think quantum and I AI will intersect. Well, one will help the other one. That is

How how how do you set up the context for in which the algorithm can run and deliver your eventual result? Those are the parameters of the numbers that I'm talking about. ⁓ but so far, ⁓ you know, I I mean dancing with qubits, which you know my book, which you mentioned, you know, it has I don't I don't know, a dozen standard algorithms. And you look at other quantum books and they're all the same twelve algorithms, right? And these don't go back that far. Whereas you pick up

a a book on classical algorithms. It's seven, eight hundred pages think and goes on and on and on, this variation, that variation, and things like this. So I think it's a I think it's a different class, let's put it this way, of kind of algorithms. but always reminding us all that computing is computing and get the job done, use whatever algorithm makes sense on that type of processor if you're using a GPU.

You darn well better be using GPU appropriate algorithms and not, you know, single threaded CPU algorithms and things like that.

Mark Fielding (21:06)
in one of your recent reports, you mentioned sixteen real world applications of of of computational categories. We've spoken about Cleveland Clinic and IBM. That's the most famous, it's the well the most well known. Could you what are your favorite from those sixteen, some of the lesser known?

domains, some lesser known research and experiments in which which are happening

Bob Sutor (21:28)
one of the three general areas. so let me first start by by saying these are the three main areas where we think of applying quantum computing and then I'll I'll spread them out a little bit more. So the first area we say, ⁓ well, artificial intelligence. why do we say that? Well, from the earliest discussion, there's a lot of money in them there, artificial intelligence. You know, you better be working there, right?

But really, down deep in a lot of AI and machine learning, it's a part of mathematics called linear algebra. Quantum computers are essentially linear algebraic supercomputers. Okay, so therefore the type of math that quantum computers naturally do very well applies itself very well to machine learning. Well, what else? Well, computational fluid dynamics, CFD, which I think is a great phrase.

And what this means is a liquid or air. Imagine you're driving your car and and the air is flowing over your car, and you want it to be, quote, aerodynamic. Well, you design the car in a certain way, it's not random, right? That science of understanding how the air flows over it is computational fluid dynamics. It turns out it's very complicated because air gets turbulent and you know it hits one surface and bounces onto another, and things like this. Well,

Mark Fielding (22:24)
Agree.

Bob Sutor (22:52)
So a lot of people are looking at quantum computing to use CFD, and now we bring in the automotive companies, we bring in the aeronautic companies. Aha. Okay. So we we we've gone from looks like AI to a lot of other areas of water around ships, ⁓ maritime. Right. So you can kind of see how this is growing out from the from the core application area. we mentioned chemistry a little bit. This is ⁓

Apples to apples. So nature is quantum. You are quantum. You are a quantum app, if you will, right? ⁓ so so and there is an area called quantum chemistry. Right. So therefore we're gonna start with chemistry. The great big hope is drug development. you know, personalized medicine. ⁓ we were talking about the history of computing. Every single big disruptive thing that has come along in the last thirty years is going to give us that personalized medicine.

Right? We know it. And quantum's the next one to do it. Okay. but the idea is to saying can we actually do the chemistry in the quantum computer as opposed to in the lab? Okay, well, so drug development material science, new alloys. how about batteries? You know, lithium sulfide. Well, lithium and sulfur are both pretty small elements in terms of the molecule in terms of the atoms and and working together.

Therefore, that should be amenable. You know, small project, small things will will work faster. batteries, ⁓ energy industry, ⁓ saving energy, electric cars. ⁓ okay, gee. So we've gone from from our our quantum chemistry, and now we've enlarged once again into the various right energy networks. Where are you storing this information? ⁓ and then

Mark Fielding (24:41)
You're just describing

it like the last ten episodes of Thinking on Paper, Bob.

Bob Sutor (24:45)
Well well, you know, I mean the this is how you know you start at the core and you start saying, Well, where do we use this? ⁓ and the f final area ⁓ which I'll highlight, ⁓ there are a couple of smaller ones, is ⁓ just optimization types of problems. So the sorts of problems, combinatorial optimization, these are the things where you say ⁓ you can sm start with a relatively small amount of data, but as you start considering this, ⁓ the problem just blows up and blows up.

⁓ the traveling salesperson example. ⁓ people tend to know, you know, I give you fifty cities, you have to compute the shortest path going through each city once and every once once once ⁓ only once and then return to where you came from. well that's a great example. Quantum doesn't actually help you at all, but it's a great example. Anyway, ⁓

But there are lots of these types of combinatorial problems. And in fact, many of them are in finance. ⁓ Many of them are related to like pricing ⁓ financial derivatives. Risk assessment is another one, right? You want to build a factory, right? You'd like to have it built in three years. What are all the possible risk factors? Well, you have to weigh them all. There's a 10% chance of this happening, a 20% chance.

Right. You have to have an environment stud environmental study. What's the risk of that slowing down? So many interacting factors, and they're not independent, right? One can affect another one. And and just alone risk risk assessment is like, well, that's every business in the world. Because everything you do is a risk, right? And things like this. So it's in these areas. But once again, I do want to emphasize they're only problems that are truly big enough that are quite

that our classical methods can't just, you know, solve them anyway. kind of a

Mark Fielding (26:39)
Is is anybody working

on climate change specifically?

Bob Sutor (26:43)
they say they are, you know, and I don't know. ⁓ on honestly, it's really hard to say what it really means. a reason why I'm being a little hesitant about this is because seven or eight years ago people were claiming they were doing it. And as far as I could tell, they were doing absolutely nothing. It just sounded really good, right, as an application area. ⁓ the fourth area, I said there were three, but here's the secret. Fourth one is differential equations.

Mark Fielding (26:45)
I know, but are they?

Bob Sutor (27:11)
Which of course all your readers, you know, you remember from college, your your course in partial differential equations. Okay. Well, you know, things that change over time. And here you want to think about the weather, right? And pr weather prediction. And again, we have pretty good classical methods supplemented by a lot of sensors. Is it raining over there? What's the wind? You know, ten miles over here, we can kind of predict. Differential.

Mark Fielding (27:13)
Ha ha.

Jeremy is partial to one of those.

Bob Sutor (27:37)
Equations actually, for many, many things in in the universe, describe how they evolve. And so how the weather evolves and therefore how ultimately maybe the climate can evolve. That is an area that I would say within the last year has gotten rejuvenated in quantum and people are showing some interesting results. So yeah, it might work out that way. ⁓

But I would be careful about kind of the incredibly broad climate change as opposed to dropping down two or three levels on what might be causing it, right? And asking if quantum can tell you anything there.

Jeremy Gilbertson (28:10)
So

dive into one of those particular examples. And let's just say, a battery company comes to you and they're seeing all of this stuff in the market, like talking ba basically they're hearing people shouting from the mountaintops about tr the new equivalent of transistors and how they work and how electrons flow transistors and it'll help them make better batteries down the road. how do you help someone like that balance

The noche of what they're hearing in the market and the application of, man, how do I take the first step to take advantage of something to make a better battery?

Bob Sutor (28:44)
So the people who would know about this are the chemists and the chemists who are working on batteries today. You know, Odd as it may seam right? you occasionally do have the CEO or somebody or CTO saying, Hey, we should be working on this. We have had automobile companies working on this in the past, Volkswagen, I remember just for one several years ago, because they want to have the best ultimately the best electronic vehicles and they would love to have their own proprietary battery technology that will last longer and things like this.

So you you have to get the attention of the scientists who may be working non-quantumly, to think if you know that is not with quantum computers. You have to kind of pull them in. And now more and more of them over time are actually seeing the work that's being done. So they are enlarging their scientific and to some extent their engineering breadth of knowledge about what to do then. Then

Happens, you know, in any particular ⁓ company. I mentioned I mentioned Volkswagen, but in any automotive company, how do you convince the people up high that they should be investing in the research on this? Right? And so you have to take it from there. You have to prove the hard numbers of what you think you can improve, when you can improve it, do you build it yourself? Do you partner and all the usual types of things?

Jeremy Gilbertson (30:03)
So when when someone thinks about a new technology, a new computing technology, you would think that that would the proper place for that to reside would be under the technology org org, the CIO, the CTO, that sort of thing. It would seem to me as we're talking through this, quantum is a resource. If you were to stand up a quantum center of excellence to use a traditional bureaucratic business term, that would almost seem likely to live under like the science and research.

arm in Org not even mess with IT until it gets to a point where it becomes data, right?

Bob Sutor (30:36)
Yeah, I I think that's absolutely right. I I mean, while people in IT might be experimenting, and here I'll get in trouble again. ⁓ people aren't solving really big, really important problems with quantum yet. I'll get in trouble because some people say, yes, we are. Yes, we are. We're we're doing okay. Except those those people who are listening who who are. But for everyone else who's not actually doing anything, yeah, it's research, and that's why in banks.

A lot of the people are the quants and the quants and the research, the same people who started investigating the out using ⁓ well, years ago improving the algorithms for quantitative trading, right? ⁓ then they started doing AI and now these are the same people, right, in fintech who are looking at where quantum may be applicable. If you're in day to day operations in a bank or in trading, you may be curious about quantum, but you're not using it and you're not not investigating it.

⁓ again, I'm sure there's one or two people who are, but as a general statement.

Mark Fielding (31:38)
Well let's let's ⁓

c let's call them out. B the the Parato, the eighty twenty principle, who are the twenty percent who are doing the big heavy lifting in quantum? Which are the companies which are really behind doors pushing it onwards?

Bob Sutor (31:53)
So most of the pharmaceutical companies are doing some serious work and they're starting to engage. ⁓ and it varies. ⁓ I have ⁓ one of the reports I I do end users, so I actually list them which pharmaceuticals and things like this, ⁓ which automotive companies and things like this. many, many banks are looking at them. So ⁓ JP Morgan was an early pioneer in in in doing an awful lot of this.

and of course I'm going to immediately blank on on on the names of them. ⁓ but those that you would imagine having the ri the richest research ⁓ departments. ⁓ HSBC ⁓ is another one that has done some good work.

Mark Fielding (32:37)
Whose

hardware and software are they using?

Bob Sutor (32:40)
⁓ the recent announcement they did with ⁓ IBM. But that's actually that's a key question ⁓ because more than the end users, it's the big company. So if you're an IBM, if you're an IonQ if you're an Infleqtion whomever, you need users. okay, so this this is your bonus cat for the I I warned you. ⁓ that's Chester.

Mark Fielding (33:01)
Yes.

What's the you can answer the question, what

is a cat qubit later?

Bob Sutor (33:09)
What is a cat qubit? Yeah. if I if I make quantum computers, I need users. I need users who are making breakthroughs. So I am going to go and cut some sort of deal with the car companies, with pharmaceutical companies, whatever, to please, please, please try to do something with this. Please write some papers with us. so that others will want to use the quantum computers. But there is again a researcher scientific angles that

The the the people in the industries are the domain experts. And the people in the quantum computing companies, while some of them may come from a particular industry, ⁓ they know the computing technology and you need to mix them well. ⁓ the situations that by the way that don't work well are, for example, I'm a quantum computing software company and I have an algorithm and that's all I know. And now I start going, This is my hammer, and I try to hit every industry over the head with my hammer, like

I can solve your problem. I know nothing about what you do, but I can solve your problem, right? So the proper blending is the experts on the hardware and software working with the domain experts in the particular industries.

Jeremy Gilbertson (34:20)
is it frustrating for domain experts like scientists, chemists, whomever, to now have to cause they have to get in deep on quantum computing, I think, too, in order to make use you almost have to yeah, you almost have to understand it enough to build it, then to figure out how to use it. Are they gonna have to be experts in two things now?

Bob Sutor (34:44)
People who would do it are already computational experts. Right? It's not like they never use a computer for anything. ⁓ and so they know the best algorithms on the best high performance computing systems to get as far as they can. So this is the next natural step. And I would say that for those type of people, it is intellectually fascinating to be able to do this.

because it is so much closer to the actual chemistry, the science itself. Right. And so ⁓ people like that always get excited. there's something new to learn. That's great. I can think about new things. ⁓ and that's why they went into in the first place.

Mark Fielding (35:30)
Question for you on that.

I kind of feel sorry for the chemists and the biologists, but also very envious because the quantum computing industry is flirting for their attention. The AI industry wants their attention. We spoke last week about microgravity. The space industry wants their attention. Everybody wants them because they are the the piece that makes everything work in any particular domain.

How do they choose? Like the yes, they love the curiosity, they love pushing themselves, but they can't do everything.

Bob Sutor (36:02)
well that's what the CTO does, right? I mean the th that's why Yeah, to choose. That's why you have to work. If if if I am a ⁓ I mean, I'll call it rank and file chemist, just meaning that ⁓ I'm actually doing the chemistry. Maybe I'm a manager, I lead a group or something like this, but I'm actually doing the chemistry. I'm not doing it randomly because it's interesting. I have goals, I have projects I'm working on, and those are determined at least in commercial settings.

Mark Fielding (36:05)
I choose.

Bob Sutor (36:30)
By people higher up to saying this is our business, this is what we're trying to do, this is our strategy, and things like this. ⁓ if they are being pulled by their management in lots of different directions, then that's a management problem. I have seen that happen as well. Like every new exciting thing that floats by the senior office, let's do that too. You know, a little discipline is a good idea.

Jeremy Gilbertson (36:54)
Well, and then a lot of these

investigators are are working under grants and the and they're working at universities that have specific obligations to them. So the university might say, Well, here's your tool set. We want you to use this that they've negotiated with one of these, you know, system vendors. And then the grant is specific to a specific technology. And then it gets, yeah, you almost need an orchestrator or someone in the middle. And I think you said like the CTO or VP provost of research or or whatever, trying to figure out those tool sets.

Bob Sutor (37:22)
Well, in in academia, ⁓ there usually isn't a CTO per se, right? ⁓ and and grants rule, right, you gotta get somebody to give you a whole lot of money to fund your lab, to fund your your your students, your graduate students, your postdocs, and things like this. the good news is there's more money going into this, although let me just simply say there have been some

unusual things in one or two governments cutting scientific funding. Leave it at that. so well well, you know a country is not going to accelerate and advance if you kill basic research. As should be clearly obvious to everybody. So so

Mark Fielding (37:59)
You don't have to leave it at that. Jeremy won't.

Bob Sutor (38:19)
There is in certain parts of the world, including in the United States, there there's less money sometimes, right, from the government. So you have to find alternative sources. ⁓ and they they could come from, let's say, the pharmaceutical companies or motor that is, they could be grants from industry to do these sorts of things. ⁓ but on the other hand, the competition is picking up more and more and more. I I have this quantum daily update that we publish. ⁓ I say five days a week, but it was seven days last week because there's so much news.

I started it myself about a year ago just so I would stay abreast and now I just it I've automated so much of it. ⁓ and you know, it used to be it's like here are the f here are the five quantum news items today and I'll skip tomorrow. You know, two days ago I there were thirty-two really good high quality links of things happening around the world related to quantum in in applications and algorithms and error correction and sensing and communications and networking and this and that.

and and and whatever. So so ⁓ there are a lot of people who are now looking at this. ⁓ there's more competition, there are more people looking at because maybe there is more money doing this. ⁓ I certainly ⁓ would encourage by the way if you are a university ⁓ researcher to go out to these quantum companies to these small quantum companies and and see if you can work out some sorts of arrangements to

To do research together, with the caveat that some universities are extremely difficult to work with regarding intellectual property.

Jeremy Gilbertson (39:56)
Yeah, TTO's technology transfer office. It gets complicated.

Bob Sutor (39:56)
And they're

There is many a startup in lots of areas that totally messed up because of ⁓ the cost and the difficulty of of getting intellectual property rights from universities when they started. So that's not a problem you can ignore.

Jeremy Gilbertson (40:18)
say you got, say you have like 16 different, let's call them forces, And you have different subcurrents under those forces, you have currents, right? Which are more of a ⁓ specifically applied version of that.

And they're all intersecting in this giant soup. Could quantum computing help model or help visualize the complexity of those interactions?

Bob Sutor (40:42)
Well, here's the definite maybe. in in the sense of we have to ask some technical questions about your your forces, describing them. Are you describing them discreetly? Are you describing them continuously? How much math is there behind there? How much is there for a lot of these sorts of things, you have to actually map it to physics in some ways. So a lot of these optimization sorts of problems.

Because they evolve over time, right? And so even when we look at some of the financial services applications, may start in the language of finance, but by the time you're running it on a quantum computer, it's an honest-to-goodness physics object that you are computing and you're looking at how it changes over time. And ideally, when you stop looking at it, you get the answer and things like this.

So it's the complexity of how you would imagine modeling it. Just as if, you know, forgetting quantum, let's say you wanted to do this using AI.

How what would that model look like? What is the data? What are you observing? and again, is it continuous? Is is it, you know, points in time, things like this. ⁓ a lot of people have been trying to do this. a researcher a after IBM won Jeopardy, if you remember it seems like ancient history, ⁓ one of the researchers left to join, I think, a hedge fund because he was going to model the entire world, you know, with what we knew then of AI years ago.

And and that's the that is always the hope, you know, to see everything and it's usually by people who want to make a lot of money too.

Mark Fielding (42:24)
just reminded of the Nexus by Julio Ottino and the modelling of it was like three square meters of Pacific Ocean off Canada and just how impossibly chaotic and complex that three square meter piece of ocean was. And the the hubris, the arrogance to say I'm gonna model the earth and everything in it is just, I don't know if it's

It's very ambitious, but is it crazy or is it genius? I don't know.

Bob Sutor (42:51)
Well we have to keep doing it. We have to keep understanding and we have to understand what the simplifications are. Right. So what can we safely ignore? ⁓ so let's say we're looking at the current flow through that stretch of water. ⁓ does the salinity of the water affect the current? I suspect it does a little bit somehow, some way, but is it such a small thing that we can safely ignore it? So just like in AI, right, you imagine the feature set.

You know, which of the inputs are are high value, which are lower value, which can you statistically reduce and things like this to get a good enough answer. ⁓ I I think people should realize and and you know, they see it every day ⁓ using like Chat GPT or Cloud or whatever. You know, they're looking for answers that are good enough. They're not looking for exactly perfect answers. Maybe in some computations you would, but good enough.

And in the same way, a lot of what happens in quantum computing is approximations. And is it a good enough approximation that you can do something useful with? It's not necessarily this is the perfect answer, right? Particularly when you're modeling, as you described. Can you get close enough so that you can infer something interesting and then act on it in some way?

Mark Fielding (44:11)
I love that idea of what can we ignore. I think we should hold on to that because that that that's pretty powerful stuff.

How many quantum companies did you say there were in the world?

Bob Sutor (44:22)
By my count there are ninety-five quantum computing hardware companies. And so these are the people who somehow harness qubits, right? So they may manufacture them, as is the case with IBM and Google, several other ones, Alice and Bob, or capture them, meaning neutral atoms or ions, the inflections of the world, the continuums of the world. ⁓ and they're about ninety-five. And I'm sure I'm sure there are a few more little ones, maybe. I mean

New one seems to pop up every two weeks and some of these maybe aren't that real. But that gives you a a ballpark.

Mark Fielding (44:57)
Last question and it's over to you, Jeremy. I c I can't have a quantum conversation anymore without speaking about Helium-3. Bob. You've spoken to all of these companies. You've been around the world speaking to and about quantum. How many times are you coming across Helium-3 as a bottleneck? Is it being used? How is it being used? Is it a real thing? Or should we forget about asking that question?

Bob Sutor (45:10)
Okay.

If they are worried about it, they're not telling me about it. And they're not planning moon exposit ⁓ expeditions either. Right. ⁓ there are only cer there are only certain types of modalities, as we call them. There there are nine ways of making quantum computers. and you know, some of them advertise as being room temperature, except when they're not room temperature. It's like is my refrigerator room temperature? Well, sort of, right.

Mark Fielding (45:29)
No expeditions to mine the moon.

Bob Sutor (45:50)
The big systems like IBM, ⁓ the superconducting systems, the silicon spin systems, yeah, they need sort of really super cooling down. ⁓ other systems ⁓ like neutral atoms will eventually need a little cooling, but not nearly that much. So ⁓ you you shouldn't you shouldn't look at an IBM and say all 95 companies will need this much helium-3 or or liquid nitrogen or whatever and and the hardware to do it.

some of the systems like the photonic systems, ⁓ they're they're gonna need a lot of it, and we will see. and here I'm going to, by the way, unasked, insert. We don't know who the winner is gonna be. People always want to know what's you know, what is the one. Right now there are four with a bonus fifth that seem to be

⁓ the best contenders and they will vary a little over time. We saw this huge surge in neutral atoms. February inflection was the first company that went public to do that. We also saw Google, which did superconducting, say, hey, we're doing neutral atoms too. And they had previously invested in neutral atoms. So it's like, you know, every week it's like, there are two more ion companies and things like this. So we don't know. The best way of thinking about it though, is for all the applications we've talked about, is there will be

Some things that some of these computers will be better for, even though the fundamental principles are better. There are also going to be certain types of constraints. So you might say, hey, you know, I want the most accurate answer possible from the quantum computer I use. Or I want a good enough answer at the lowest possible price, or that uses the least amount of energy, right? And things like this. So they're going to be dials you turn.

if I'm a hedge fund, I might say, I want that answer in ten minutes. If I'm a chemistry company, I might say, you know, I'm willing to save money and get the answer in two days. So just thinking there's going to be one type that just blasts out for every type of problem that's going to be a deal. That that isn't going to happen.

Jeremy Gilbertson (48:01)
so do you think there's like there's a place for a quantum solution aggregator that can help select almost to the same point of like here's my question, and like in like in chat bot models, there's like, hey, use this model, use this model, use this model, and they're all good for different things. Is there room for aggregation to help people figure out what to use?

Bob Sutor (48:21)
There are, I mean, in the field of optimization, there are algorithms that sit on top of optimizers that decide the best optimizer to use for certain types of problems. ⁓ there is serverless cloud computing, which means, hey, I I don't know where this thing is gonna run, and maybe I don't even know on what type of processor. It could be a GPU, it could be a CPU, could be something else. That's your job. This is the problem, I'm gonna state it. Here's the data, you go off.

figure out where to run it, right? Problems aren't mo the solutions to problems are not monolithic. Here's my problem. Here, quantum computer. Give me the answer. No, it it it's like, well first of all, you might have used a laptop to submit it. You use the network to get over there. You've got cloud computers, you've got things that convert digital to analog s signals. I mean these systems are very complicated and there's workflow and there could be AI here, there. So it it's

We we should be talking more about the workflow that involves quantum computing, isn't simply quantum computing. and it may determine, in fact, hey, this problem is too small, don't waste your money running it on a quantum computer, we can do it classically better, right? But it's crossed a certain threshold in terms of complexity. Now we're gonna run it on that. So yes, they're gonna be layers, just as there are now in classical computers.

It's a fun area, by the way, if I haven't given that. It's it's a microcosm of all of computing that at least we can keep in our heads all at the same time still, although that's time seems to be passing pretty quickly.

Mark Fielding (50:00)
a shout out to anyone who's listening that we have interviews with many of the companies that Bob has mentioned Infleqtion, D Wave, IonQ IBM, NVIDIA, Horizon. Check out thinkingonpaper.xyz. Our paths first crossed because at least in my world, you're known as debunking the quantum hype. So to finish the show, let's let's debunk together.

Anyone listening to this, the next eighteen months, they're gonna be bombarded with hype and noise. And amongst that, there's going to be real signal, real value, real thinking on paper episodes.

how should people think about that? What what filter should they be using to judge quantum announcements, let's say, over the next eighteen months?

Bob Sutor (50:49)
So let's let's fix ideas and say you're looking at LinkedIn and whatever other social media outlet, you know, it's similar. So you see something and it's about quantum. If it's obviously AI generated, and you can usually tell by the beautiful graphic, right? And five word sentences sep separated by blank lines, right? You know it when you see it. AI slop. Stop reading. Just don't read it. Just just just pass it by or block.

Or say you're not interested, just ignore all that because they almost all have egregious errors in them. Whatever the hype, right? They have egregious errors. They often praise results. There's this amazing one that keeps popping up about this this stunning result in China that only happened four years ago, right? And I think we know a little bit about that. So so throw away all that. Now, when you see a press release by somebody.

What I suggest is irreverently you chuckle to yourself and say, Yeah, right. Okay. That is don't be cynical. Well, maybe that is cynical, but say you've now got to prove to me that you've really done something useful and you just haven't wordsmithed this thing so much that it sounds amazing. So beware quantum supremacy, quantum advantage, and things like this, right? ⁓

Any sort of result which is going to change the future of computing, don't immediately repost, give it a few days. Right? Let some other people sort of weigh weigh in on this. ⁓ in fact, there's there was a result that's going back and forth between D-Wave and the Flatiron Institute from a D-Wave paper last year, and Flat Iron responded saying, No, we can do this on a laptop. And D-Wave said, No, you can't do all of it on a laptop. So there's this ping-ponging back and forth. That's that's interesting to say. ⁓

Just don't, you know, consider it a as steady progress. Don't get overexcited. You know, don't do these things like the future of computing is now, you know, we've moved from theory to reality. That's such

You know, we we've moved it's only an engineering problem. The science is finished. That's wrong. Right. So it's so I I'm trying to give a sense of, you know, if you if it's just smacks of being silly and and over just ignore it, right? There's plenty of other content that is out there. And always go to the original source, if you will. So even if you're reading an article and they mention, here's the press release or here's the paper, go and glance at that, see what they really said.

All of this said, there's wonderful work being done, right? The hype comes in for so many reasons. People, you know, want to get attention to themselves, people want to be social media influencers. You know, ⁓ find a few trusted sources. Starting ⁓ in fact, this last January on the my Substack, I started doing once a week quantum follow Fridays. And I did a couple of after that. And I listed people.

With their LinkedIn saying, follow these people. These are real people. They say, I think by the time I was done, I did a dozen each week. So I I I did about 60 and I'm due for another one. But those are at least good, smart, thoughtful experts or people who analyze what's going on and and start with those. But please, please, please don't just repost AI slide.

Jeremy Gilbertson (54:19)
good advice just in general. I think that's that's great. Bob, thanks, thanks so much for the context there. One thing too, I want to remind people of is you had a great quote a few years ago, just setting the stage for for all of this stuff and just the complexity of what quantum computing is trying to do and the complexity of trying to make that into something useful that can help humans. Like we got to the moon faster than

Quantum computing has come together, right? I think you said it was like eight years it took us to to get from from this rock to that rock. Like if you put it in that perspective, ⁓ we're in a we're in a good spot. So thanks for thanks for that. I thought that was I thought that was well played.

Bob Sutor (54:59)
Yeah, it it's I was saying people have to stop using the phrase moonshot because we and now we're ten years on, you know, at least, right? And it was only eight. ⁓ so ⁓ it's quantum's gonna be with us. another quote I I I I have so there's a book called Quantum Supremacy, ⁓ by Michio Cacao. I I always mispronounce his name. And

The book has its critics, but I like it because I'm quoted on the bottom of the first page. And and what I say is quantum computing is going to be the most compute ⁓ the most important computing technology of this century. And I still stand by that. I think that's right. You know, AI will change, it's important, yeah, that's not questioned, but as a fundamentally different thing that most of us have not seen in our lifetime, quantum is gonna be it.

Mark Fielding (55:52)
You you're a much better guest than Michu

Kaku. We did a an eight series book club breakdown, chapter by chapter breakdown of quantum supremacy, and he never returned our emails.

Bob Sutor (56:04)
Yeah, well it's I don't know him. I d I I don't know him. ⁓ it it it covers a lot of topics, you know, as I said I'm quoted and he lists my book in the back for further reading. So you know, my mom always used to say if you know, if they spell your name right, don't complain. So

Mark Fielding (56:06)
Do you know him?

Good book.

Jeremy Gilbertson (56:24)
There you have it. Thanks for joining us today. This is this has been a pleasure. Great insight, great peek into the quantum world. Would love to stay in touch as you continue to investigate and uncover some stuff. We're we're doing the same. So ⁓ yeah, let's let's keep it going. Appreciate you being here today. Mark, closing thoughts.

Bob Sutor (56:28)
Pleasure.

Mark Fielding (56:40)
Thank you. Bob Sutor, thank you for thinking on paper with us today.

Thinking on paper at XYZ for all our quantum episodes, our book club episodes at AI. Jeremy, I'm gonna leave you with a final thought and it's about the the traveling salesman. Have you heard about the new traveling salesman

Jeremy Gilbertson (57:00)
I know the old one. I didn't know there was a revision. Tell me about it.

Mark Fielding (57:03)
Well, there's a new one that people are brandishing around and it links to our guest in the next couple of weeks from ⁓ AstroForge. And it's about essentially the traveling salesman, but for asteroids, to find the most optimal path for mining asteroids. So you go out into the Kuiper belt and you have to find the optimal path to go jump to each asteroid to check, measure, see what's available to mine it. And that's

Jeremy Gilbertson (57:29)
Don't forget to pack your quantum computer.

Yeah.

Mark Fielding (57:31)
Well, you need a

Bob Sutor (57:32)
Right.

Mark Fielding (57:32)
quantum

computer. ⁓ no, well no, because Bob said the quantum computer wasn't very good at that. So maybe you just need a classical computer to to solve the asteroid mining salesman problem. But yeah, that in a few weeks we'll be speaking to AstroForge about that. until then, be disruptive, stay curious.

Jeremy Gilbertson (57:48)
Keep thinking on paper.