In this episode, we sit down with Dr. Thomas Baker, Canada Research Chair in Quantum Computing for Modeling of Molecules and Materials at the University of Victoria.

Together, we dive into the fundamental differences that set quantum computers apart, the interdisciplinary challenges and breakthroughs in the field, and the real-world hurdles facing quantum’s transition from theory to practicality. Dr. Baker shares how creativity, flexible thinking, and collaboration across physics, chemistry, and engineering are vital to progress in quantum information science—and why learning skills like programming and public speaking still give students an edge. Whether you’re a quantum enthusiast or simply curious about the future of technology, this episode offers accessible insights, advice for newcomers, and candid reflections on where this exciting discipline is headed.

Links

• Thomas Baker on LinkedIn - https://www.linkedin.com/in/drbakerte/
• Watch on YouTube - https://youtu.be/hH8L6zKOn0c

Time Stamps

00:00 Explaining Quantum Computing Basics

03:19 Getting into quantum computing

08:19 Quantum vs Classical Algorithm Testing

11:46 Quantum error correction challenges

13:54 Discussing quantum computing and error correction

20:23 Challenges in interdisciplinary quantum fields

24:05 Comparing qubit types to fuel sources

25:34 Discussing quantum computing concepts

30:57 Challenges of Quantum Information PR

32:05 Adapting talks to different audiences

38:11 Science Meets Parliament experience

38:59 Importance of Funding Quantum Science

45:29 Using Julia for student projects

47:09 Using Julia for easy programming

50:22 Importance of typing and coding skills

53:19 Discussing the Quantum Podcast

Have kind of contact with mathematics when they're in school. And so if

you say that you, like, do math, and they're like, oh, yeah, I add numbers

together, too. I know what's going on there. When you say that you're in physics,

though, that's already a huge barrier. And then when you say that you're in quantum

information or quantum computing, that's already. You've pretty much lost

everybody by the time you get there. Welcome

to Impact Quantum

Podcast. Turn it up fast. Candace and Frank, blowing my

mind at last. Quantum Podcast. They're breaking the.

Hello, and welcome back to Impact Quantum. The podcast. We

explore the emerging field of quantum computing and what that

means for the job market at large. You know, you don't need to have

a PhD, don't need to be a researcher. You just need to be a little

bit curious. And with me is the most quantum curious person I

know, Candice Kahouli. How's it going, Candace? It's

great. I want to say Happy New Year to you because you still get to

say it, I think. That's right. We're still in the first week of January.

You still get to say it. Exactly. So I'm really excited because

today we have Thomas Baker, and he is

Canada Research Chair in Quantum

Computing for modeling of molecules and

materials from the University of

Victoria. So how are you,

Thomas? I'm doing well. Happy New Year to you guys, too. And, yeah,

thank you. I'm glad to be here. Awesome.

So that is quite a mouthful. So

it's a tough roster. Well, yeah.

Like, you know, when presumably you were at cocktail parties or holiday

parties, like, what do you tell people you do? Oh,

first of all, I try and allay their fears. I think a lot of people

have kind of contact with mathematics when they're in school. And so

if you say that you, like, do math and they're like, oh, yeah, I had

numbers together, too. I know what. I know what's going on there. When you say

that you're in physics, though, that's already a huge barrier. And then when you say

that you're in quantum information or quantum computing, that's already. You've pretty

much lost everybody by the time you get there.

Really what I do is I'm trying to figure out how to use and how

to make quantum computers, which is a fundamentally new type of

computer. So everybody watching this is

probably using some sort of computer either in their cell phone or in their

laptop. And quantum computers are able to do

computations with fundamental laws of physics. And the idea or the

hope is that we can derive different results that are more efficient than what

you can get with classical computers. Oh, I like

that. That was really straightforward. I was really

fantastic. I wish you would have been with me when my son was

here over Christmas, and I could have given that answer, you

know, And I'm talking about the infinite possibilities

between 0 and 1, and I'm like, no, no, no, I've totally lost

him. I've totally lost him. You know, spooky action at a

distance, but that was really, really interesting. So let me

ask you this. Okay. So what

first drew you to quantum computing?

Was there a specific moment, a problem, a

curiosity? Yeah, great, great question.

So I was floating around in another field of

physics, condensed matter, before jumping into quantum

computing. And with a little bit of quantum chemistry, it turns out that everything that

I was working on falls under to what I do now. So right now my

position is this big synthesis of everything I've done in my life.

But the.

Point that I really got into quantum computing was actually through one

guy at the University of Sherbrooke, where I did a postdoctoral

position for three years, and his name was David Ploulon.

Unfortunately, he passed away right at the beginning of the pandemic.

But before that, even before I'd met him, he put up this paper basically saying

that the current way that we were thinking to generate the

initial state on the quantum computer was not all that efficient.

And this had just made. It had ricocheted through the community. So I actually heard

someone talk to me about this paper. And then they said, okay, quantum computers might

not be that good. And then when I went there and I realized that it

was him and that they were offering me this incredible position

to go work with these incredible people, I just had to say yes. And

so I wound up going. And I remember he. He kidnapped me

into his office in the first day that I was there and held me for

three hours while he explained everything that he knew about quantum computing. And I

picked up maybe just a little bit of it, but it was enough to get

me thinking, like, okay, let's try and do this. And then he threw a really

hard problem at me that we kind of

solved there at the end, but it was just a good

training experience. I found that it was testing everything that I'd ever done before.

I didn't have to leave anything behind. And I just thought, this is the thing

that I need to dive into and figure out. Very

cool, very cool.

What's exciting. I read your bio, which is both

in English And French. Or at least your hello, bonjour.

So Candice actually lives in Montreal. My dad's side of

the family is from Montreal, so I

appreciate the francophone outreach.

So you're researching about how to use

quantum information to find new ways to work with

chemistry, computer science and physics. That's interesting because most

people we talk to tend to isolate in like one of those silos.

You're really the first person I've seen that

does. You know, you've. I wouldn't say you're a silo smasher. Maybe you

are, I don't know. But like you, you, you kind of work across those

three different verticals. I think that's interesting. Yeah,

I. It's hard, though. I've met a couple of other people that have

been kind of shoehorned into this interdisciplinary thing.

And you probably see a lot more people that are in one silo,

or however you want to call it, just because it's easier. If

you get to know a group of people, then you can send your papers out

to them and there's some general familiarity with the work that you've done before.

I have to, you know, hats off to the people in my group. Oftentimes we're

writing papers in between several different fields. And if

you get the referee that has been working in one of them and you approached

it in a different way, then this can be really

a calamity that is waiting to happen. I would say that

the point where quantum information and quantum computing is, is at

this point, you need the people who are interdisciplinary. You need

to be thinking in different ways. A lot of the proofs of principle

come from really heroic papers coming out of what I would say is

largely computer science. So computer science, like modern

computer science, is like, here's like software engineering and

here's how we, like, I don't know, make the cache size different on a

computer. But like old school theoretical computer science where

they're like, analyzing stability of algorithms or updating

matrix operations on a computer. Some really brave people decided

to try and figure out how quantum computers could work. And so a lot of

our first demonstrations of how quantum computers could work are

more computer science. But that original dream of, oh, we want to

actually model quantum systems, we actually want to do things in

quantum chemistry. We want to go beyond what we can do with our best

supercomputers on the quantum computer. These are all things that

require the interdisciplinary knowledge. And so what you've seen is

you've seen a lot of teams of people. So a lot of like, huge author

lines where they get somebody to try and go between all those little

subgroups in the team. I think one of the advantages in our group is that

we're kind of, we're throwing everything against the wall, seeing what

sticks and trying to publish what we can.

Interesting. So by working on all

these papers, you know, you must come across really

exciting breakthroughs in quantum algorithms and

and error correction. Is there anything that

really stands out? Yeah, I mean,

well, so the recent stuff, there's been like a subtle algorithm

revolution, I would say, that's been going on both

classically and in quantum computing. Because people who have

worked on sort of one area, like maybe a tensor

network, whenever there's a quantum algorithm that comes out, it better be tested

against the top of the line, the state of the art, to see if the

quantum computer can actually beat what's on the classical computer. Because we're talking about an

investment of millions of dollars. You better, you better

test it quite rigorously and see, see what can happen.

The, the point where quantum computing is right now

though is people are starting to use actual properties of quantum

physics, including things that aren't necessarily proven

mathematically. We just, in the physics training, you kind of, you kind

of do what works and you're supposed to move in, move fast and break things

a little bit. You're supposed to be careful and, and do the right things. But

you know, if you're using some 30 year old hypothesis to base your quantum

algorithms on that isn't necessarily proven, I can see how people in

other fields wouldn't necessarily stumble onto that. So I think what's really been

exciting recently has been how do we actually use

even untested things in quantum mechanics in order to

actually try and find solutions that might actually work for

what's going on. That's interesting.

You have to go through and kind of prove that stuff out.

I mean, if you start from a contradiction, you can wind up with anything.

Right? So like there's that old argument from Bertrand

Russell, the logician, and let's see if I get it right. It's like

if you're standing next to the Pope in a room,

you start from a contradiction that just like one equals two, then

someone might reasonably come up with an argument to say that you're the Pope. Because

if there's two people in there, that really means that there's one person. And so

since you and the Pope are in the room, you must be the Pope. So

trying to base science off of something that's obviously contradictory,

this doesn't work, but basing something that we might have like 30 years of evidence

for or that we might, you know, believe, because

it must limit to this. Even if we don't have the formal mathematical argument,

these are all things that are very important to

investigate to make sure that we understand how to use quantum computers in the best

way.

So in your view, what are the biggest

scientific or technological hurdles

slowing the transition from theoretical

quantum computing to practical, widely

used systems? Yeah, good question.

I mean, yeah, this is sort of a preoccupation with the field right now.

One is that long term we

don't know how to make the quantum computer noise free,

like, you know, whatever computer you're using on a daily basis, or cell

phone. A lot of years of development of quantum error

correction or I'm sorry, of classical error correction. Classical error

correction has happened so that you can make sure that like if you had one

of like the original cell phones, you'd have like, like your, your voice

sometimes like echo, like right in the middle of the conversation. I remember people

being like, oh, you physicists, you should just. Yeah, right, you should

just, you physicists, you should just figure out how to make a better cell phone.

But it turns out that that was like a big engineering question for a long

time in electrical engineering. It was like, how do

we get basically a better protocol to

make sure that we actually received the message that was sent? And so this took

a long time and you probably saw through successive generations of cell phones that they

things got better. One of the big things that we have to do in

quantum computing right now is we have to do exactly that. We have

to build in ways to make sure that the

natural quantum fluctuations on the quantum computer are not

there when we actually use it to get some useful quantity

out of it at the end. So this is kind of a

major technological hurdle that I think is on the roadmaps of most major

quantum computer companies. But I would

say that it's not the farthest thing away because we've already had some rudimentary

demonstrations of the most well known versions of quantum error

correction. Things like a toric code, which is very

complicated. But other things

like finding the point at which they can

reasonably sustain a noiseless state on the quantum computer.

Some logical qubit that can be expressed there, these things, I think they're forecast

to come by 2029. So that's not, maybe that's not so far away anymore.

And I've been kind of interested in the progress in this and how

fast it's come on. It used to Be that we would say that those things

would come in 50 years or something. Now it seems like it's around the

corner. They're also branching out into two dimensional architectures, which

would really help with this. And I assume that three dimensional

systems, networks of qubits that are in three dimensions would be right around the

corner. But I would say that, you know, even with all this

technological advancement and the Nobel prizes that are coming for

the coherent control of quantum states, and for

all of that, we still need good methods that are

reliable on the quantum computer and good compilation

and transpilation so that we can actually make those algorithms something that you can

actually run on the quantum computer. And these, I would say, are some of the

big challenges that, that we're lacking at and that the community is looking

at as well. No, I think it's a good point,

like, because like, you know, where we are with quantum

computers is probably where we were in the, maybe the 50s or 60s,

right? Because when I was in school for computer science, error

correction was taught. But even one of the professors I had was

like, you're probably never really going to encounter this

because this is an almost solved. Easily. It's almost, it's an

almost solved problem. And we have this debate in the last couple

of years was, do we keep this in the curriculum? Because yeah,

doesn't really matter. I don't know, like it became like a thing.

And error correction is something I think everyone within the

sound of my voice can understand, right? Because like one of the

things. So one of my first jobs out of college is like my second or

third job I was at [email protected] before they were

barnesandnoble.com and one of the store systems, if you look at

any barcode, the last digit there is actually an error

correcting code. So it doesn't have to be complicated,

but it is something that is in everyday life and it's largely a

solved problem, at least in conventional or classical computers.

It's just fascinating. Kids today who go through

a traditional comp sci program, they're probably. What do you mean, error

correction? I had a younger colleague ask me about why do you need to do

error correction? And I was like, well, truth is, we've.

Error correction has been a thing. It's just been solved.

Although interestingly, for space systems, again, you're the physicist,

so you know, for space systems, because cosmic

rays or there's some kind of radiation that will

occasionally mess with electronic systems,

space systems have to have like a triple redundancy.

So that way they check the math just in case. Right. So

the chances of one of them being impacted is.

Is. Is very real. But the chances of all three of them being. Or two

of the three being impacted the same way,

isn't it, you know, not going to happen, realistically? Yeah,

that's exactly right. Yeah. The. There's a book on my shelf about

basics of information, Information Theory. And it starts

off with this guy, Claude Shannon, that was right after

World War II, and he said, how do we. What's like the best way

to make sure that the message that you send is

received by the receiver? And so the start of this is always

just that. You write down three or four axioms, things that you assume to be

true. It's like finding someone in New York is harder than finding someone in a

small village, or that the probability of error should be

continuous, or the other one that

it should add like a logarithm, which is something that

you can derive. But basically, there's a

theorem that comes out of this analysis by Claude Shannon,

which is exactly what you're talking about, where they say

that if you send the message multiple times, then you can

take a majority vote of the answers and assume that that is

the one, that that was actually meant to be sent. If you do

this slow enough and with some other caveats, then you can

come to a conclusion that, yes, I did receive the right

message and everything is as it should be. Yeah, I think when

you held up the barcode, I think that's the Reed-Solomon

code that is there. Sounds familiar.

I think that's the right one because my first. Yeah,

go ahead. No, my first job there was making

basically tying into barcode scanners. And I. There was a

whole book about this history of the barcode and like the technical

specifications, how much of the barcode can be missing before it

will just give up. And there's a whole science to it. And it's like everyday

tech. We don't think about. I'm sorry, I cut you off. No, no, no, no,

no, no, no. And that's actually a perfect example. I had a much more complicated

example in mind. This idea that, you know, if you miss some of the

information, you can still recover the message. In other words, extrapolate to the code word.

That's exactly this idea behind information theory. So, yeah, it's all

coming out of this. And, you know, whether you use a Reed Solomon code or

a polar code or something, these are all the same ideas that we want to

implement on the quantum computer. I would say, to your point,

about the computer scientists and whether like error correction is actually something to

study like one. You're absolutely right. Quantum computing is

very much a 50s 60s thing right now. Even like the algorithms that we

compare in quantum chemistry, they're best thought of in comparison

with algorithms from the 50s or the 60s. So not like you're not going to

use these to make your cell phone, you're going to get very rough

results. But then the other point that you're making about

like computer science students, like, yeah, I mean for both this field

and artificial intelligence, there's a lot more to prove. And what I hear

from sort of the modern computer science students is like, oh, whoa,

should we actually get into this? So it's important to make advances in these

fields, right? But there's opportunity in those advances.

Like this is brand new untouched land,

so to speak. I was joking with a previous show.

They're still naming algorithms after people, Right.

And that's a good indicator that this is new territory. Right.

So it's, it's. Yeah, I mean it's not going to be a well worn path,

but that's where the opportunity always is. Right. You know, eventually those

streets will be named after the first people who kind of,

you know, walk through those woods, so to speak. Right? Yeah,

exactly. But it's, it's scary. So it's like, it's not, it is very scary. Like

it's not for the, not for the timid, but you know, it's like

that's the fundamental training that you get in physics is. It's like somebody goes to

a physicist to hire them when they want to solve something that they don't know

how to solve and then they assume that the physicist has all this knowledge that

is there. But really you're just, you're just trying to put things together based on

what you know and yeah, it's exactly. This process of being creative is scary,

right? Well, it's always that tension between, you know, physicists,

physics, physics and kind of engineering. Right. Like

physics is about the fundamental understanding the fundamental laws, but

engineering is, you know, what trade offs do you need? Do you make to

comply with the laws of physics as well as comply with

whatever it is you're trying to do. Right. There's always engineering

is, is a science of trade offs, so to speak.

But I think to your point, like it's very difficult to draw the line

for some fields at this point. And this is a constant conversation that happens

inside of physics departments everywhere. It's like, is this field

at the point where it's more applied. And so like a

chemistry department should be hiring these people or so mathematical

that a math department should have these, these professors. Or should an

engineering department be optimizing the photonics or

circuitry that is required there? This is a really tough place to

find this line. And I think it's even more difficult because for quantum

information and quantum algorithms in particular, if you come up with

an algorithm that is complete in a computer science sense,

and what I mean by complete is if it solves one kind of problem, it

solves every other problem in that, in that,

in that class of problems, then it's like I show one

algorithm to my students, but I might be sending them off into biophysics

and then maybe the next project is going to be on, I don't know,

quantum Internet or the next problem is going to be on like optimizing like

train schedules in the uk. So from one algorithm in

this fields you are going to intersect with a lot of theory

that is going to come out of this. And I think that's the real, like

the, the challenge of being so interdisciplinary. I wonder

how computers used by everybody. Yeah, I was going to say, like, I wonder like

how, you know, what were these conversations had in the 50s

where you had the physics department, you had the mathematicians, you probably

had electrical engineers in the room all debating about like, who

should, who should own what, right. You know, and then, you know,

eventually kind of spun out to its own thing. I wonder if we're going to,

you know, kind of have a very similar conversations now and over the next,

easily the next five years. Right. Like probably

the, probably the equivalent conversation right now is what Qubit platform

is going to win out in what, over what timescale?

To take your example from the 50s, like. So in the 50s, I think the

Nobel Prize went to. It was definitely Bardeen and two

other, I think two other people for the transistor. Inventing the transistor,

which it actually turned out that a Canadian person had the

patent for the Shockley, I think had the patent for this. Oh really? Earlier.

But then they took the idea from this and ginned it up and then they

won the Nobel Prize at Bell Labs. So

before that I suppose you would have had maybe vacuum tubes or something. I don't

know. It was in those original computers. But I'm sure that

there were people that were like, no, no, no, no, no, no, you don't want

to trust the solid state transistor. You want to go with the vacuum tubes.

And I know that there were people that were Saying, well, the computers are just

going to get bigger and bigger and bigger. But then it turns out that everything,

you know, the right technology won out through the scientific conversation.

I would say right now our big questions that we have to sort out

is what Qubit platform is actually the best one to use? And I would

say that the other one is what? How do we actually use the

quantum computer so that it's not just for defense purposes or

for, you know, whatever else we need it to be for.

These are all the things that I think are probably the modern

iterations of those. Those are those age old arguments that are still going to happen.

Nothing ever gets solved. You just find new things to argue about. You find new

things to argue about. That is very true. No, I think it's, it's,

it's, it's. There was also a science fiction movie I read. Not

a science fiction movie I read science fiction book I met, I read where it

was. There was about a society who built computers, but out of basically

pipes and valves. So it was a fluidic based computer.

And you know, part of the plot was just how massive the thing was

and how many people they needed to like pull the levers to do the

computation and things like that. So like, but, but you know,

to your point, like obviously a transistor, vacuum tubes are more efficient than that

and transitions transistors are more efficient than that.

But to your point, like do you really think we're going to settle on one

type of qubit? Because that's one of the debates that we've had is like,

I think it'll be kind of like fuel sources for

vehicles, right? You'll have electric, you'll have, you know,

diesel, you'll have gasoline and there'll be some, you

know, I think, you know, I

think that we'll end up having the majority will be one and then

they'll still be like sizable minorities of other ones. Kind of like we have with

fuel, right? Like with automobiles, like most of it is gasoline,

you know, with an increasing ev. But also diesel also

plays a pretty big role too. Like. Yeah, yeah. And

continuing your analogy, there's also been proposals for organic qubits.

So there's also this like, oh, what's the sustainability angle and how

do we regular chemistry system and tune the energy level so that

we can manipulate the qubit states in that. So

yeah, no, this is well taken. I think it's too early to say, like as

I was about to say, like oh well, single qubit readout is hard on this

one. So Maybe it's this one. I think it's too early to say if there's

some like, breakthrough on one of these that maybe that'll work out. I,

I will say that I've been really impressed by

photonics and spin cubits. I just, I think that the ideas behind

that are really interesting and it is kind of a, like if I was going

to teach a course on it, I'd probably start with a photonics based

setup. Just because it's like you talk about Alice and Bob,

you talk about Cubans being or, you know, states being

entangled with each other and it, it's just a very obvious

manifestation of all of those things. But the leaders right now, I would

say are still pretty much superconducting qubits. Just because

they have, you know, it's scalable, you need to keep it at low temperature,

but it allows you to have a

nonlinear response that is still stable. And so

this is kind of the, I think the direction that people

are going in right now at least. Yeah, I think photonics is probably

going to win out. It'll be the gasoline, so to speak, of the future.

Just because the room temperature thing I think goes a long way.

Yeah. To making it practical. Right.

You can scale, you know, super cooled systems, but it,

the cost of that is going to be astronomical. Oh yeah,

that's especially, especially as we think about, you know,

sustainability and whatnot. I live just across the river from

Loudoun County, Virginia, which is apparently somebody told me this

data center alley they call it, but apparently Loudoun County, Virginia has more

data centers than China, like the entire country.

And it's crazy. I don't know if that's true, but

driving around there, you could believe it. Like, it is literally

I can't tell you. We were in the middle of shopping for a car

and we were at a big car dealership there. And like

every building was an office building without windows. Like they were clearly

painted to look like an office building. But one of the big

complaints is, is the energy usage and the noise and things

like that. So I can't imagine, you know, those are just

cooled to, you know, room temperature. I can't imagine

if you have to get them down to like what, one or two kelvin, right.

Like what, how much more pollution and how much noise.

You know, it's amazing that we even have that much data in the first place.

Like if I think about how much data, like I've never seen like average

amounts of data that people have, but it's like my

total data footprint must be less than 100 gigabytes,

maybe with some videos or something floating around. I'm sure that this is,

you know, maybe a balloon or something, but having

only a couple of gigabytes per person, it's like, wow, okay,

so there's a lot of people using this. Plus there's probably like medical data

that's saved or your browser tracking data, your

mouse data. There's a whole lot more data. There's data

you know about. It's probably like an iceberg. Right. There's what you know about

and then there's what was really there. Right. From government records to

all that sort of thing. Yeah, yeah. But we're sort of waiting to see how

much better quantum computers can be, because whatever that number is, take a

logarithm of it, and that is how many qubits you need to represent that. And

it like, you know, it's sort of an order of magnitude game that you're playing.

So if it's a gigabyte, then it's, you know,

I don't know, a few dozens of qubits in

order to represent all that. So even if it, even if you need to cool

the quantum computer down, there's still, oh, you would still get used

to get a savings. That is a good point. I hadn't thought of that. And

you mentioned logarithm a few times, and it's been way too long.

My, my, my teenager's taking AP math. Sure.

And I don't need to give him another reason to, you know, think his old

man doesn't know anything. But I also forgot. So logarithm is

a way to, to break down a number into a

smaller number. That's right. Okay. So, yeah.

Without busting out a slide rule. Right, right, right, right, right,

right. So if you're picturing the number in like scientific

notation, or if you're just picturing any number to the

power of some other number, the magical property of the

logarithm is that whatever was not the exponent stay

like the, the logarithm is going to produce some other number. The, the key point

though is, is that the exponent that comes down in front.

And so when I say like, if I say like a million,

think like 10 to the 6. If I take a logarithm of that base 10,

then I get 6. So,

okay, so 2 to the whole idea of 32, you would get 6.

32. 32. Sorry, yeah, base

2. Yeah, yeah. So logarithm, base 2 on 32

would be 6. Yeah. So the the base of the logarithm matters.

You're thinking qubits, which is the right one for this. This podcast. I'm thinking like

10, 10 level q dits. I get you. Okay,

okay, okay. No, that helps, that helps. And the last time I had

to deal with logarithms on a day to day basis, I think Kurt Cobain was

still alive. So. Been a while.

Smells like mathematical rigor. There you go.

There's the title for the episode. Okay, well, yeah.

So, you know, the more we talk about this, you know, quantum

technology is notoriously hard to communicate.

What strategies do you use when teaching and

explaining these concepts to students and to non

experts? Well, you've seen a couple of my pitches.

The first thing that I'll say about quantum information, and one of the reasons that

I like it, it does not have a good PR department. I think we can

just admit that, right? Like astrophysics, they can show you a video of like two

stars colliding. Ooh. Or like the amazing

images that come out of the James Webb telescope. These are like, these are, these

are fun to look at. If I'm giving a talk, it's like, here's a very

abstract thing that I'm going to talk very abstractly about. And then if I'm

successful, you're going to take it for granted. Just like you take your

wall clock for granted after it's been engineered. Just like you take your

computer for granted after it's been well engineered. You're not constantly thinking about

like, what's going on underneath. So getting people to engage with

the minutiae of this is very, very difficult. And

I think pretty much everybody in this field has an experience where

you've been put up for a three minute thesis or some sort

of short public talk about it. Oh, very hard.

If you can't educate your audience over an appreciable amount

of time, you're relying on the prior knowledge. And this

is way off the deep end. So it sort of depends

on what I'm talking about when I, when I give a talk. You

sort of saw my pitches already in our conversation because when I was

talking about error correction, I went back to classical error correction.

And so a lot of students have come to me and said, can I please

learn more about quantum error correction? I say, sure, let's go study classical error

correction. Because all the ideas poured over and I try and reason by analogy

between the two and then the other. Like

if somebody comes to me and says like a quantum algorithm, I'm just like,

okay, let's let's do a very simple example and work through something in a paper

in order to do this. But yeah, I mean, it very much depends on

your audience, depends on what their background knowledge is, depends on what they want to

use it for. Like politicians, you're gonna gear more towards

like a, you know, your, your encryption might be broken by a quantum

computer. If it's somebody working in medicine, you might be, well,

we might be able to model a better molecule. And, and you basically reason from

what they know to. And then, and then somewhere in there, I'll say,

okay, this is the magic that I have. This is the thing that I think

I can do that you think maybe is impossible or haven't

thought about doing before. But this is the part where I think the quantum computer

can sit and let me go try to solve

that for you, and we'll see how it goes.

Okay, so what are the essential skills or

mindset you think new students should develop to

contribute meaningfully to quantum information

science? That is such a deadly good question. That's so

good. Yeah. I mean, the actual

skill set that you would need in order to be successful in this area, I'd

say one is like, flexible thinking.

There are deep divisions in physics that exist.

There's preferences. There's even entire articles that have been written

in like a monthly column format, bashing other areas of

physics over the 20th century, like, you know, let's not use computers.

Let's reason everything out with mathematics. And I think,

you know, having students that, that come in fresh

is such a valuable asset to this field because sometimes they

see things through a completely different lens or they pick up a little computational

skill that is going to, to help you. Like, I worked with

three really good undergraduates last summer, and I had a project that I thought was

going to take a graduate student. But those three guys, they, they just

worked together, asking my other excellent

graduate student some questions, and she was able to answer,

but they were able to essentially put everything together just because

they didn't have any preconceptions. They were even solving things that we had

an external colleague and he did, you know, he's already been thinking about it for

a little while. So he went down the practiced way, but they just went in

a particularly new direction. I would say that I'm a little bit

partial to physicists and people who have a physics background. No

offense to everyone else. Like, I definitely need

chemists in my group because they know about things that I don't. Like

electronic structure. I need engineers to talk to because

they know about how to actually put these things into experiment. And what

I should expect if I actually make a claim that this will be

implementable on a quantum computer, I definitely need

mathematicians and computer scientists for the rigor

of what happens. But in terms of coming from a physics background and trying to

innovate and trying to be creative in the field, I would say that if you

just have some math background and if you're willing to touch a

computer and to, you know, program a little bit,

I think if you just come in relaxed and creative, I think this

is the best set of skills for anybody coming in, in the academic environment,

I should say. Yeah. And I think that's an interesting segue

into this is a field where it is

going to be commercialized over the next three to five

years, if not already. Oh, yeah. I mean, it's, it's, it's

definitely like what? I, I definitely think that there

are different cultures amongst academia and kind of

the commercial startup crowd, which is probably an understatement.

But what, how do you

see that, that being mitigated, like, in terms

of conflict? Yeah. Somebody

was telling me about condensed matter physics and they had seen it kind

of rise and then they decided to switch to quantum information. And I'm not sure

that this quote is quite accurate to condensed matter physics, but I like the quote,

so I'm gonna preserve it here. Everybody

was really collaborative when they were successful, and then as

things sort of got, you know, more

static and things were less successful, people started to not

be so collaborative. So I think it's a little

bit of luck and I think it's a little bit of, you know,

we need to stop trying to put our names

on things and it's very important to actually get the thing solved.

Right. Like, I would say that AI has this problem

quite a bit right now. AI, it

gets things wrong. There haven't been an explosion of

materials come out of this that could be that

we're going to find ways to do it in the future. Right. But

this, this additional pressure that modern science has

of the companies need to make their quarterly profits

and they need to do fundraising. This is something that is very,

very new in the grand context of science

and publicly funded science in particular. So I think it's very important that,

like, if you're a student or a postdoctoral researcher or a new faculty

or something very important to just kind of play your game and

see, see how things are going to turn out. And even if some idea is

not popular now, maybe it will be soon. Right. So

what Signals should industry and

policymakers look for to

know when quantum is moving from

experimental to operational?

Sure, yeah. For policymakers, this is, this is,

I think it's a question of what your motivations are. I

was really lucky and I was able. There's a program called Science Meets Parliament that

is up here in Canada. And so I, as one of the research chairs,

I was allowed to go to the, the Houses of Parliament.

It was a really great experience. But one of the things that we were talking

about in the lead up to this was how you would message for what

I know because I'm from the US and how you would talk to

a US politician versus a Canadian politician.

I think sort of the consensus was is that US politicians have a long

history of looking at their constituencies and identifying

how technologies have brought jobs and broad economic development and

how they've, they've helped things in Canada. We're a little bit newer to that

idea and there's been some recent headlines about this and in the

newspapers about why it's important to fund quantum computing and

quantum mechanics. So if I was talking to a politician about this, I

think I'd probably start off with like, like what is your motivation? Like

do you want to just engineer on what there already is? If

so that will change how you fund things. If you actually want to innovate and

you actually want to push technology to the limit. History has shown

us that funding and science has been so lucrative

it has a huge return on investment. This is why

a lot of immigration programs are set up to bring in high knowledge

people into a country is because it just has huge benefits.

I think kind of what I've been seeing with the recent

activities and things is that idea of innovation

from the end of World War II, in particular of publicly funded science.

I think people are a little, have sort of lost the,

the image of like what science can provide. Like huge technological

changes come from this stuff.

And but we're, we're at a point now where we're promising that

it's going to happen in a couple of years, which is the standard scientific pitch.

But then if we don't deliver on it, I think people just get an idea

into their head that maybe it's not coming, maybe the money isn't worth it, but

it always is, it always leads to good things. And

there's been a huge raft of funding that's come out in Canada that I'm very

grateful for and that they seem to be very serious about

funding this area. And I think places that invested in are

going to reap the benefits as they go forward. Yeah. I

mean, the next wave of economic innovation is always in some weird

corner of science that hasn't been explored yet. Right.

I mean, you know, Claude Shannon, when he was working through his stuff,

was even when you go back to Babich and

Ada Lovelace. Right. Like, that was weird. That was just

some weird science project that, you know, no

one else was interested in. Ada Lovelace saw the potential, but no one

else did. Right. It wasn't until World War II where,

hey, we need to crack these codes or we need to figure out what

those physicists are doing because they just built a bomb that can level. Yes, exactly.

We better keep track of these guys. Right, Right, right, right.

Yeah. But you know, social scientists have figured out

that traditionally the best way to do it is to give a lot of people

at least a little bit of funding and see and let the winners happen naturally.

Right? Yeah, definitely. Definitely. You're using good examples there of

historical situations.

Sorry, Candace, I cut you off. No, no, no. I'm just, I'm just thinking.

I'm just kind of taking in everything.

So let me ask you this. What role do open source

tools and academic collaboration

play in accelerating the progress

compared to proprietary or closed approaches?

This is such a weird time for open source software.

It's been amazing, even just over my comparatively short scientific

career. It used to be like, everything open source is good

and I still encourage my students to use a lot of open source

software and things. Weirdly, I've been hearing, even just with this

disruption from AI that some computer science professors are

saying, don't put everything online because it just gets scraped by

large language model. And then I'm not, I haven't heard the full argument

about that. I just heard sort of the, the top thesis statement,

even, even little things like, like a driver to

interface with general printers. Like I think some big

company bought up the, the, the, the driver that most,

that a lot of computers use to interface with printers. And

the transition away from like open source standards with

GPUs towards the more proprietary versions of the

software, it's like there even just in my limited

scientific career, there's been a transition towards closed source

technologies. I don't know what's going to win

out. That's kind of a major question for me right now in my research. And

what, what will be, you know, doing. And as we deal more and more with

research security, it's like we have to play these games of making sure that our

computers are secure and that they don't cross certain borders and things like this.

But I would say that the, like the battle

between open source and closed source, if you're making your own

stuff, I think you have to go with open source. So

most of the people in my group will use a Unix based operating

system rather than something that is more

proprietary, something more general. And it's

really, really important that if you're making your own, like if you're making innovations on

computer stuff, you better write your own code. Trying to rely on somebody's else,

somebody else's code to be super general enough to make new discoveries, this can be

difficult. So I don't have a perfect answer for this. It's just for

research. Open source is still good. Yeah, well, it's.

The cost of admission is low too. Right.

And I saw on your profile you linked to and this is going to get

a little nerdy, Candace, so I apologize in advance. Saw you have at least two

projects on GitHub. Yes. At least

they're both in Julia. And I find that interesting

because Julia was supposed to be this, you know, I

started in AI way before it was cool. So there was Python and R,

but Julia was, was meant to be a third contender. Now what

I've seen in industry is R

is still used by companies. Python is definitely the winner.

You don't hear much about Julia. So, you know, you're,

you're probably the first like Julia programmer I've ever had on a

podcast. So I have to, I have to ask,

where is it used? I don't, I don't, I don't mean that. I don't want

to sound like a smug Python coder. Right. Because I have no right to be

a smug. Because I used to write Silver Lake code and

Windows Phone apps. Right. So like smugness is not something

I've learned my lesson about being smug. No, I'm just curious. Like, is like, you

know, and, and for people that want to, you know, knock on Julia as a

language jupyter notebooks. The J part is

Julia, you know. Yeah. So sorry, go

ahead. No, no, no, no, no. So that was

the reason that I've been coding in Julia at this point. The reason that I

maintain those within the group is pure

pragmatic reasons. Students coming into my group from a

physics background don't necessarily have all the high level coding skills that they would need

in order to quickly sort through a C or

a Fortran code or something lower level. Even something like Rust,

I would say would be just a little bit more than they would get. So

Given a master student who has nominally two years to

complete their degree and they take courses for maybe the first year and

then they get a summer, the fall, and then they need to start writing up

their thesis. I have to find ways to lower the amount of time

for people to get into the research. Now the question is, could I have accomplished

the same thing with Python? The answer is yes, except that

there's this statement that floats around Julia that is

very contentious that Julia's for loops are optimized in a way

that you can get performances like lower level languages.

So the original reason that I started in Julia was I wanted to figure this

out. It turns out that it's a more complicated

process than the help pages would have you think. But those

codes are, they do run at the lower level language

standard for speed. In some cases they're a little bit faster. And I've

actually that One repository starred 50 times,

whatever that means. But I've noticed that some of the tricks have been filtering into

other codes. Maybe people are finding them independently.

So very much like a developer's code that is listed there,

but purely for pragmatic reasons. If I had to go back and start it

again, if I would have switched to something like Rust, I would have lost

the ability to do multiple dispatch in Julia. The idea that you can have

one function that takes many inputs, you don't have to explicitly

make all the functions in Julia and this makes programming much

easier. But there are, you know, there's, there's trade offs,

there's, there's things about it that, that I'm like why did,

why did I even touch a computer in the first place? Why didn't I just

stick with theory? But that it's, at this point it's just a pragmatic

reason for. So that the students find it a little bit easier. And the,

you know, I won't comment on the name of the

language. It probably, probably could have had a different name, but it's okay. No,

I was just curious, you know, because like I, it, you know, it was one

of those things where it was going to be the next big thing and it

kind of, it definitely has traction but it Denver really

I think lived up to its potential in terms of what it could do.

Now a lot of people filtered in made packages and they're deprecated

now. So like I think pie plot into Julia is

not working. And the pie plot I think still looks better than the regular

Julia standard. So I do think they do suffer from a lot of deprecation

and things But I would say scientific software, maybe

to your point, There's a network of computer supercomputers in Canada.

Turns out we're one of the only groups that uses Julia. But our

new. Our new computers that are coming in, it's like, these are going to network

really well with Julia. And I don't know, it's for ease of use at this

point because so much work has gone into it already. All right, that makes sense

again then, Like, I don't have a favorite language anymore. Right. I think Python, you

know, I learned the hard way. I used to be a C

developer.

Well, before that I was Java. It really had more to do with

commercial availability of work, really was.

Now I go with Python. Python's not the best

language I've ever worked with, and it's certainly not the worst.

It gets the job done and it can get

a lot of different jobs done. Yeah. Like Julius Python.

But you don't have to worry about the indents and you can optimize the for

loops. Allegedly. Yeah. And

I don't find too much difference. I think, honestly, if there's like a student

out there, like, listening to this and it's like, oh, what do I do? Just

learn any computer programming language. The skills and the tools that

you'll pick up learning just one, they'll translate to other languages.

And if you start at Julia and you don't explicitly set types for

variables, you'll pick it up through time when you move to the other thing and

the compiler will tell you. Yes. Compilers

aren't shy. Well, I started with BASIC on the

Commodore 64. Right. I did, too. Yeah. Awesome.

BASIC was the first one I touched. Yeah. So BASIC is, you know,

and then, you know, that was

arguably whether or not it's compiled. And I think it depends on which implementation. BASIC

and all that. But I mean, like, again, you're right. It's. It's the thinking in

structures, thinking in process, thinking that that's the

skill, that how you do that is secondary

and, you know. Yeah, yeah, yeah,

exactly. So, yeah. But I think as

we sort of transition more to, like, TikTok and stuff, where you don't have to

punch things into a keyboard, very important that we maintain those coding

skills. Like, absolutely. Maybe I'm fighting uphill because it's like, it's like. It's like

arguing that someone should continue to learn, like, car mechanics even as we

transition to, like, you know, cars being more computerized, but I still think

that, like, knowing how to type quickly, knowing maybe, maybe that's A good answer to

Candace's question. Like, if you know how to type quickly, that can just pole

vault you over people who might be smarter than you like, but can't type as

fast. So, I mean, little soft skills like this, like computer

programming, like, I don't know what's. What's

another good one. A little bit of public speaking. A little bit. Oh, yeah. Public

speaking will put you to the top 10%. Yeah.

Even. Even if you're not good at it, even if you're okay with it and

you're comfortable with it, you're in the top 25% right there.

Because most people will say, like, what's. What's this? There's some

ridiculous statistic where, like, you know, people are less afraid of jumping out

of a plane, something like that. But, like,

I don't know if that's true. But just. Even that. That sounds plausible is the

problem. Right? Yeah, yeah. No, I.

I think all these. These points are well taken, but, like, how

to recommend success? At the end of the day, you kind of have to be

lucky. Like, you can have all the skills in the world and just. You picked.

You picked. You went left instead of right. And so you do need a bit

of luck at some point, but you can also kind of make your own luck.

So. I don't know. Make your

own luck. I like that. So. So that

was fantastic. Fantastic. That was really great. Thanks

for. Thanks for coming on the show. I really enjoyed it and. Any final

thoughts, Candace? Honestly, there's so much more to

talk about, but I really enjoyed it, and I want to work on those

soft skills that are needed for better PR and

better communications and better marketing. That's it.

Yeah, that's where I am. Awesome. And where can folks find out more about you?

Thomas? I've got a research website. If you go to the

Department of Physics and Astronomy or the Department of Chemistry websites, it links to my

own website. We've got some really exciting work that's coming out of the

group on quantum algorithms and classical algorithms and things with quantum computers

and. Yeah, feel free to reach out on my

email on those pages, as you like. Awesome.

With that, we'll play the outro music.

The multiverse is skanking Skanking in time Black holes are

wailing in a horn line so fine From Planck scales to planets they're

connecting the dots Candace and Frank, they're the cosmic

hot shot.

Quantum podcast, turn it up fast Candace and Frank,

blowing my mind at last Quantum podcast, they're breaking

the mold. Science, science has got beats it's bold

and it's gold.