December 20, 2013 • 28 mins
How does quantum computing work? From superposition to quantum entanglement, we tackle the concepts behind quantum computers and explain what they'd be used to do.
Learn more about your ad-choices at https://www.iheartpodcastnetwork.com
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Brought to you by Toyota. Let's go places. Welcome to
Forward Thinking. Hey there, everyone, and welcome to Forward Thinking,
the podcast that look at the future and says I'm
making a note here huge success. I'm Jonathan Stricklandon, I'm
Lauren Bock Obama, and I'm Joe McCormick. And guys, you know,
(00:24):
I have this problem. I was using my computer the
other day to work on a non trivial math problem,
and it gave me an estimated time of completing that
particular task as four hundred and seventy eight years, and
I've got plans that day. I'm actually washing my hair.
So I was hoping that maybe we could talk about
(00:46):
classical computers, how they process information, and maybe why you're
still washing your hair and your cyborg body and years
years his cyborg body has hair. Yeah, it actually it
takes to wash only my body has hair. But yes,
it's uh, it's you know, let's not I don't want
to go too far into my personal details here, so
(01:09):
we all understand classical computers are reading machine language, binary language,
zeros and ones. Technically there's actually something electronic going on there,
but we're specifically saying that you know, it interprets information
in sequences of zeros and ones that represent other things. Yeah,
any piece of software running on your computer is actually
(01:30):
a really long, I'm generally really long sequence of on
and off switches of ones and zeros, right, and it's
following specific rules. Yeah, and so there are two options
as to work with. It has on and off their
own one or one and zero whatever. On and off
would be one in zero um. And everything that it
(01:52):
does has to be translated down to that level before
the hardware can make use of it, right, because the
hardware doesn't understand things like run Assassin's Creed four. It
actually has to break it down into understands one zero
one one. Right. But if you tell it that run
equals one zero zero zero one one one, right, Yeah,
(02:14):
if you establish what the rules are, then it can
it can interpret it through multiple levels of translating from
whatever language you're using into machine language that that binary system.
So these sort of this sort of approach is really
good for certain types of operations. For example, if you
want something just really simple, like you want to multiply
(02:36):
one number by another number, and you input then too
the computer and it understands what the operation is understands
how to multiply then, and by understand I mean it
has the instructions to do so, not that it actually comprehends. Yeah,
you can listen to our artificial intelligence episode and hear
that whole argument again. But the the idea of being
(02:57):
that it does that very well. It can do that
operation quickly, you get it. You know, if you use
a calculator, you're going to get that information very quickly.
That's first computer, more quickly than a human being, certainly
than most most human beings. Yes. Yeah, your classic computer
is very good at math, and it's very good at
doing lots and lots of math problems way faster than
you can. But if it needs to do a whole
(03:18):
series of those in order to get the one results
you want, because the question you're asking is actually going
to require many, many, many operations before you can get
an answer, that's where the classical computers start to slow down.
Even when we're talking about modern classical computers, which often
have multi cores, right, you know, you'll hear about these
multi core processors. They can break down certain types of
(03:41):
problems into parallel chunks, and each core of the processor
can work on a chunk of that problem, which cuts
down on the amount of time it takes to solve
the overall question that you've asked. But even then it's
you know, you're talking like like maybe sixteen cores. That's
still nothing if you're talking about a truly difficult problem.
(04:02):
And that's that's traditionally where we run into a barrier
with classical computers. There are certain types of non trivial
problems that classical computers have trouble solving. And here's an
example we've talked about before, the traveling salesman problem. You know,
the idea that you've got a traveling salesman who needs
to visit let's say ten cities, and your job is
to try and find the most efficient route for that
(04:24):
traveling salesman to go through to spend the least amount
of energy on this trip. And there are a lot
of different potential wants to choose from, you know, choosing
city A first and then going to city D and
then going back to CITYB, or going straight through, and
really you don't know what the right answer is until
you've compared all those options. And as you add cities
(04:45):
to this this question, it becomes more and more difficult
to answer, and a classical computer essentially what it has
to do is go through and run every single possibility
and then at the end of running all of them,
compare all the results to other to come up with
your answer, which can take a long time. So that
(05:06):
means that we need to look at maybe an alternative
to classical computing if we want to be able to
solve those types of problems in a more efficient, faster manner.
But how do we get more efficient than than a bit?
I mean, on or off seems like a pretty pretty
simplified What if you could be both on and off
at the same time? That's crazy talk. It is crazy talk, Jonathan. Yes,
(05:28):
I have never seen a light switch that was both
off and on at the same time. No wonder you
haven't seen it. If you had observed it, it it would
either be off or on. Yeah, can't. You can't see
it and see it that it's both off and on
at the same time, then you've observed it. Well. The
problem is because the light switch is actually a really
huge thing. Yeah, it's a macro level thing. It's on
(05:51):
our scales, um, so it's position tends to be pretty stable, right, Yeah,
But if we're looking at things on say a quantum
ski ill that's sub atomic, tiny world where things just
don't make sense an electronic Yeah, we're talking like tiny
little particles. They exhibit behaviors that if if that same
behavior were to suddenly exhibit itself on the on the
(06:16):
on the macro scale, on our scale, we would all
just think that we had been sucked into some sort
of David Lynch weird al. Yeah, would make sense. No,
nothing would make sense, because the quantum world and the
classical world are are very different in the way they behave. Well, yeah,
our our intuitions are evolved to deal with you know,
(06:36):
like trees and rocks and animals, nothings, superpositions and in
further more, trees and rocks and animals that don't suddenly
shift three miles to the right for no apparent reason. Okay,
so what what are we talking about? How can something
be in two positions at once? It is called superposition,
And asking me how is the wrong way to go
(06:57):
about a joe, because I certainly could not tell you
how I can tell you what's going on here. So
super position is a concept within the field of quantum engineering,
quantum mechanics, where a sub atomic particle is able to
coexist in multiple states at the same time. And they
don't mean states like the United States. I'm talking about
(07:18):
actual states of being. So, for example, with electron, we
often talk about spin, Like let's say that it could
either have a spin that's up or spin that's down,
and that's going to determine its magnetic field as well. Right,
the direction of its magnetic field is that will be
determined by this electron spin. Now, on the quantum level, technically,
(07:38):
an electron can be in superposition, meaning that can inhabit
both an upspin and a downspin at the same time.
It has a probability of being in either one or
the other at any given moment, were you to observe
that electron and therefore lock it into that one one
state or the exactly, So, if you were to observe
the electron, the electron would then suddenly in have but
(08:00):
just one of those two states, and that would be
determined by the probability of which state it was most
likely to be in. Sometimes it's going to be in
the less likely state. That's why they're probabilities. Right, it
might be a chance that will be a spin down
and a six chance of spin up and you observe
it and it's spin down. That can still happen because
as long as the probability exists, that's how things can
(08:23):
sometimes shake out. Now, if you were to do that
same sort of experiment over a really long run, then
the probabilities would start to manifest themselves, assuming that everything
else was identical, which would never happen. But anyway, superposition
is that crazy idea that a sub atomic particle can
exist in both of these states at the same time,
(08:45):
at least until you observe them, at which point, once
you observe them, you would say that the system decoheres.
That's it's an idea where a quantum system is a
very delicate thing and if you interfere with it in
any way, if you try to interact with it in
various ways, it will decohere and become a classical system
where things behave more the way we would expect them
(09:06):
to based upon our own experiences. In terms of thought experiments.
This is kind of going back to if you've heard
about it Stronger's cat, It's it's it's you know, poking
the system and seeing, you know, trying to identify what
a particle is doing is going to make the cat either. Yeah.
So the basic Strodinger's cat thought experiment is that you've
got a box with uh, some sort of castor in
(09:29):
it that is going to that could release a poisonous
gas anytime after thirty minutes have passed. So you've got
a cat inside this box with the cast of poisonous gas,
and you wait for thirty one minutes to pass. And
at that at thirty one minutes, there's a fifty percent
chance that the castors released the gas and a fifty
percent chance that it hasn't, which means at that moment,
(09:49):
before you open up the box and observe it technically
from a quantum level, the cat is fifty percent alive
and fifty dead. And then once you observe it it
those those uh that quantum state deco here's and it
forms a classical state. Right. Of course, the idea of
Schrodinger's cat was first introduced a sort of like a
reductio out of serdum. The idea was like, this is
(10:10):
so ridiculous, exactly, yeah, but it turns out like works.
Quantum physicists were like, well that's tough, you know, Yeah,
on the On the macro scale, of course, it's ridiculous.
You know, you would never say that the cat is
both alive and dead at the same time. It's either
one or the other. And because you open up the
box doesn't change that at all. But on the quantum
scale it certainly does matter. Okay, But so if if
(10:32):
this is how does this relate to quantum computing? Are
we giving the cat a bunch of buttons to push?
The cats and quantum computing? Lauren, I don't know where
you got your notes, but let's just without cats. I
think is a really sad. I think computers are made
entirely of cats. Internet is made of cats, not computers.
Uh no, no, no, we're going to We're going to
(10:52):
back off the cats and the Internet and the quantum
computing for just a second. There's one other there's one
other concept of quantum that we have to cover, which
is entanglement. Yeah. This is the idea of where you
have multiple sub atomic particles that are entangled in some
quantum way. So remember when I was talking about the
spin of the electron being either up or down. If
(11:14):
you have two entangled electrons, those electrons are going to
be kind of opposite but mirror images of one another,
and that if you know the behavior of one, you
know what the behavior of the other one was at
that moment when you observed the first one. Knowing that
from that moment forward, you can't really predict anything, but
(11:34):
being that if if two electrons are entangled and one
is spinning up, the other one would be spinning down.
For that, for that particular set of features, that's not
just limited to spin. There are other things we have
to take into consideration, but that goes well beyond just
the basic idea of entanglement. Entanglements very important with this
when it gets to quantum computing, the concepts of superposition, entanglement,
(11:58):
and coherence are all really really important. So you asked,
you know about quantum computing. That's that's the where we're
getting at. The idea of quantum computing is being able
to harness these features of the quantum world in a
way that can do compute computational work for us and UH.
And the the base unit of that. You know, if
(12:19):
you were to say the base unit of a computer
is the bit, either a zero or a one. The
base unit for a quantum computer is the cubit, which
I had to be reminded, is not a little orange
guy who bounces around the pyramid. Okay, so instead of
a silicon microprocessors, say, you would have computational activities being
(12:41):
done by something like a photon or an electron or
preferably lots of them. We need at least two. But um,
although I guess I know you need at least two.
So what the cubits are? Uh? Interesting in that if
a bit a bit can only be a zero or
a one, it's one or the other. It gives you
a single value. Yes, cubits are both zeros and ones
(13:02):
at the same time superposition uh, and technically all values
in between, although that's not really that important. So the
interesting thing about cubits is that the the relationship between
cubits and computing power is exponential. You take two to
the power of however many number of cubits you have,
and that is the equivalent of your quantum computer's power,
(13:25):
meaning that with with you know, two um cubit's you
have four potential values. There. Uh. You have to look
at the things as like, uh, two zeros zero, one
one zero or one one um and they're all the
same things, and there are all of those at the
same time, right, But if you were to add another
(13:46):
cupid in there, then you're talking about two to the
third power. So you're talking about eight pieces of information
from three cubits, which is different from the way it
would be if it were just bits. And as you
add cubits it becomes exponentially more powerful. By definition, you're
talking about actual exponent here the inn in that two
to the nth power. So the interesting thing here is
(14:10):
that all of these different cubits could uh and inhabit
these two values of zero and one at the same time.
If you have enough of them, then you could, in theory,
run a very complex problem through a quantum computer and
it could solve for all aspects of that problem simultaneously
in parallel because it's essentially doing all of those calculations
(14:32):
at once because all of the cubits are all possible values.
So it's kind of uh great for very specific types
of difficult problems. Okay, so this sounds like, though, um,
it's not going to be a replacement for the kinds
of computers we use now, not at all. Because while
(14:53):
it's great for certain complex problems like the traveling salesman
problem where it could solve for all of those different
variations simultaneously. It's not necessarily going to be any faster,
and in fact, might even be slower than a classical
computer for your basic computing functions that that regular schmos
like like myself, like like I do. If it's not
(15:14):
gonna show YouTube super faster, you're not going to be
able to run the latest video game even faster, like
I love the idea of running a video game on
a quantum computer and all possible outcomes of the video
gameplayouts simultaneously, like I was both good and evil and
everything in between. But that's that's not exactly what would happen. Okay,
So you're saying that a machine like this might have
(15:36):
really incredible powers in some kind of specialized way, very specialized,
like cryptography breaking. If if we had a working quantum
computer of sufficient power, cryptography as it exists now would
be meaningless. And the reason for that is that basically
the way cryptography tends to work is take two really
(15:58):
large prime numbers, like really really large. We're talking digits
that are hundreds of digits long, like it's it's an
enormous number, and then you find another enormous prime number,
and you multiply the two of them together and you
get that product, and then your your encryption is based
upon a party that's authorized having one of those two
(16:18):
large prime numbers, and as long as it's one of
the two right to correct numbers, they can get access
to that uh, that particular information or site or whatever.
I'm oversimplifying for the purpose of this podcast. Now, publicly,
all you can see is the product. So you see
this huge product. I mean, it's a no enormous number.
(16:39):
I remember, it's the product of two big prime numbers.
And if you don't have the information already, trying to
figure out which two prime numbers made this even bigger
number is really hard to do. In a way a
classical computer would do it, is it would start by
dividing by prime numbers and then run through all of
the prime numbers that possibly could be. So, if you're talking,
(17:02):
if you pick a large enough prime number, that alone
is going to guarantee that any computer working on trying
to force this cryptography, this brute force stile attack is
going to take longer than it would be you know,
feasible to run. So you would you know, most people
would not ever bother to try, because to do so
successfully would take forever. But if you had a quantum
(17:24):
computer that could solve for all potential prime numbers at
the same time, you could crack that roll relatively quickly.
In fact, yeah, it could. It can make the most
advanced encryption tools useless. It could also usher in a
new era of quantum cryptography, which would be even more
(17:45):
difficult to crack. But you know, it's it's one of
those things already where that's just one application. Obviously, there
are lots of other applications for quantum computing. Yeah, that
was one of the early applications. Actually, in the in
the early nineties, Peter Shore of Bell Labs developed a
quantum algorithm that that was a method of of entangling
cubits and using superposition to um to find prime factors
(18:07):
of an integer. Although that's not to say that that
we can run that on all of our fancy current
quantum computers. It was really like a proof of concepts
saying that once we are able to do this, it's
going to change our world, and it's good to know
about it now rather than three months after the world's
fastest quantum computer is made, and then we all realize
all of our stuff is public. It's better to know
(18:28):
it now, so we can say, huh, that's a problem.
How do we fix this? So I mean, but yeah,
it's a great example, and a lot of work has
been done on quantum computers since just that that that
algorithm was really for a hypothetical quantum computer. But now
we've got people who have actually built at least preliminary
quantum computers. Yeah, what's out there? I think I think
(18:51):
the fanciest one that we've got was built in two
thousand and seven. It's it's called the d Wave and
it's a sixteen cubit quantum computer. I know they were
working on one that would have been five and twenty
eight cubits, which would have been phenomenal, and that that
would have been unveiled within the last year or two,
But I honestly don't know if they ever uh successfully
(19:13):
demonstrated that one. But the fact that we've had people
demonstrate this at all is pretty phenomenal because keep in mind,
you have to be really careful with the way you
operate one of these things, because just by observing it,
by by trying to interpret the results you could cause decoherence,
and then you end up with a very primitive classical
computer that can't do much of anything, Like can you
(19:34):
imagine going from a sixteen cubit computer to a sixteen
bit computer? Would not be great? Right, And they've been
working on these kinds of problems for UM since the
since the sixties and seventies and eighties when all of
this was sort of starting to come together, and it
required a lot of work in UM and first of all,
bringing reversible logic gates into the computing world, which which
(19:57):
allows you if you have a one way gate, then
you're going to experience a lot of data and therefore
heat loss in your computer system. UM having a reversible
gate lets you UM basically not burn out your processor
every time you turn it on. Essentially, UM and and
quantum electrodynamics, which is just starting to hum look at
(20:18):
how electrons and photons interact with each other, so that
we can start creating these little quantum pieces of of
electrical information, right right, yeah, I mean, how do you
harness this stuff? It's it's really tricky. Are you have
to be able to create entangled particles? That's already kind
of tricky there's certain minty materials that are being used
right now in an experimental way that that are potentially
(20:42):
a source of entangled photons. That's pretty exciting stuff. You
have to figure out how to create a system that
these can work in that is not going to allow
it to go into decoherence. You have to figure out
how to program for it so again that you can
take advantage of it using it for the right sort
of problems, and you have to figure out how to
get the solution out of it again without disturbing the system,
(21:06):
which is pretty tricky stuff. And even then, you're talking
about probabilistic results, right, you mean you're getting results that
are assigned certain probabilities of being correct versus incorrect. And
sometimes the probabilities you get are so high that you
might as well say it's a certainty. I mean, you
can't really say that statistically speaking, there's always some room
(21:29):
for uncertainty, but you know from human experience and be like, well,
you know, times out of a hundred it's right, but
there's still a chance that could be wrong and not
all and of course some results may end up being
like we're sure this is the right answer, So it's
a it's a very uh specific, kind of niche oriented
(21:51):
version of computing. It's not something you're not gonna go
and get your uh your your your laptop. That's gonna
have you know, the cotum version of the last model
you owned, right, sure, and it is it is. Yeah,
like you said, just slow going. I mean it wasn't
until two thousand and one that there was a successful
demonstration of shorts algorithm, and that was with a seven
(22:12):
cubit computer that found the factors off. Yeah, they have
factors of It's still it's still something. I mean, hey, no, no, no,
I mean that's impressive. I mean I think that the
very first one added one in one and everyone was
so excited. I mean I mean understandably so. But nonetheless,
and you know, and there's there's a few different ways
that you can work on this coherence problem. I think
(22:32):
that that is the largest issue that we're talking about. Um, yeah,
that's that's I mean, none of these are trivial, but
I would say that's the hardest or from what I understand,
keeping in mind, I'm not a quantum physicist, right because
I mean you can either I mean, you can like
trap ions and super cool them until they're in a
quantum state so that you can work with them, or
um you can use liquid to to kind of wrangle
(22:54):
cubits so that you can spread a single cube it
across a few different molecules, which will decrease your your
rate of coherence. Right, But all of this is you know,
I mean, we're talking about very tricky particle physics. Yeah.
And while we're talking about you know, kind of humble beginnings,
keep in mind that other things that we depend upon
today very heavily had very humble beginnings. For example, the Internet.
(23:17):
You know, if you remember back when the Arpanet was
first being put into into action, you know, that was
a predecessor to the Internet. Some people call it kind
of like the grandfather to the Internet, but it was
it was a network of networks, was the idea, and
it was this um. The one of the earliest messages sent,
actually the first message set UH ended up crashing midway
(23:41):
through the message. It was a one word message. So
you know, you've got, you know, something where you point
to that and like look at that. If we had
just assumed that everything that followed that that failure was
also going to be a failure, we would not have
the Internet right now. So while we talk about these
kind of tiny exam balls, you know, when you're thinking
(24:01):
about proving a concept to be viable, that's huge, right,
even if it seems relatively tiny and you're talking about, oh,
you found the factors of fifteen. Wow, that's something that
I learned in third grade. But you know, for something
that until that point was hypothetical, that's amazing. So yeah,
I mean, these quantum computers have a lot of interesting potential,
(24:23):
not just in decryption or even the traveling salesman problem. Joe,
you had, you saw something, right, Yeah, they might be
able to help us actually study some things in the
real world that are really difficult to study with the
computers we have today, or might be really expensive, say
to simulate physically. I've got a press release from last
year from the n I s T Information Technology Lab UM,
(24:44):
and what this talks about is that quantum computers might
be able to simulate particle collisions. So the kind of
work that we have to do now with like the
particle accelerator, like the large Hadron collider and something it's
miles across and die amateur, Yeah, it could could actually
be simulated just in software. And of course, like with
(25:05):
a regular computer, it's really hard to do this because
digital you know, the computers we have today can't determine
all these quantum states. There's just like too much information
to keep track of um. But this press release talks
about a team that came up with an algorithm that
could basically run on any quantum computer, regardless of what
(25:26):
it's quantum hardware was um, and it would simulate all
of the different ways that two different types of particles
could collide and interact. Interesting. Yeah, see, this is kind
of fascinating stuff. The idea that we can uh add
in all these known factors that we are aware of
and create a simulation that could potentially create stuff we
(25:49):
aren't aware of. It's kind of to me, that's that
that's almost like magic at that at that stage where
you're like, all right, I know what this does, and
I know what this does, I don't know necessarily all
the things that can happen when the two collide, and
you create a simulation that can actually show you that.
To me, it boggles my mind. You know, it's just
such a phenomenal thing. And yeah, again, the quantum computer
(26:11):
would be a use in that situation, but not so
much if you want to play mind Sweeper. Well though,
we're not saying necessarily that a quantum computer could never
run a game or a video player or something like that.
We're just saying like, there's really no reason you'd need
one when a classical computer can do the same thing
(26:32):
for a long time. Because, I mean, while while your
power of your quantum computer does, uh does expand exponentially
with the number of cubits you add, you still would
have to add a lot for it to really uh
for for classical applications, for it to be better than
a classical computer. Now quantum applications, it would just leave
(26:52):
everyone else in the Yeah, but you would you would
need you know, a whole gallon of quantum, right, Yeah,
you gotta get that. Gotta go over to the quantum
refueling station and fill up your quantum tank and bring
it back just to just to play like like Super
Mario Brothers. Oh yeah, I don't know about you, guys,
but I'm ready to go and play some video games.
(27:13):
So I'm gonna wrap this up. Guys. If you enjoy
our show, please go check f w thinking dot com.
That's our home web page where We've got all the
blog posts and videos and podcasts, cool stuff that you
don't want to miss. Go check that out. I remember
you can interact with us on Google Plus, Facebook, and Twitter.
Just look for f W thinking that's us. We're happy
(27:35):
to talk to you, and we will talk to you
again really soon. For more on this topic in the
future of technology, visit forward Thinking dot Com, brought to
(27:59):
you by Toyota Let's Go Places
Brought to you by Toyota. Let's go places. Welcome to
Forward Thinking. Hey there, everyone, and welcome to Forward Thinking,
the podcast that look at the future and says I'm
making a note here huge success. I'm Jonathan Stricklandon, I'm
Lauren Bock Obama, and I'm Joe McCormick. And guys, you know,
(00:24):
I have this problem. I was using my computer the
other day to work on a non trivial math problem,
and it gave me an estimated time of completing that
particular task as four hundred and seventy eight years, and
I've got plans that day. I'm actually washing my hair.
So I was hoping that maybe we could talk about
(00:46):
classical computers, how they process information, and maybe why you're
still washing your hair and your cyborg body and years
years his cyborg body has hair. Yeah, it actually it
takes to wash only my body has hair. But yes,
it's uh, it's you know, let's not I don't want
to go too far into my personal details here, so
(01:09):
we all understand classical computers are reading machine language, binary language,
zeros and ones. Technically there's actually something electronic going on there,
but we're specifically saying that you know, it interprets information
in sequences of zeros and ones that represent other things. Yeah,
any piece of software running on your computer is actually
(01:30):
a really long, I'm generally really long sequence of on
and off switches of ones and zeros, right, and it's
following specific rules. Yeah, and so there are two options
as to work with. It has on and off their
own one or one and zero whatever. On and off
would be one in zero um. And everything that it
(01:52):
does has to be translated down to that level before
the hardware can make use of it, right, because the
hardware doesn't understand things like run Assassin's Creed four. It
actually has to break it down into understands one zero
one one. Right. But if you tell it that run
equals one zero zero zero one one one, right, Yeah,
(02:14):
if you establish what the rules are, then it can
it can interpret it through multiple levels of translating from
whatever language you're using into machine language that that binary system.
So these sort of this sort of approach is really
good for certain types of operations. For example, if you
want something just really simple, like you want to multiply
(02:36):
one number by another number, and you input then too
the computer and it understands what the operation is understands
how to multiply then, and by understand I mean it
has the instructions to do so, not that it actually comprehends. Yeah,
you can listen to our artificial intelligence episode and hear
that whole argument again. But the the idea of being
(02:57):
that it does that very well. It can do that
operation quickly, you get it. You know, if you use
a calculator, you're going to get that information very quickly.
That's first computer, more quickly than a human being, certainly
than most most human beings. Yes. Yeah, your classic computer
is very good at math, and it's very good at
doing lots and lots of math problems way faster than
you can. But if it needs to do a whole
(03:18):
series of those in order to get the one results
you want, because the question you're asking is actually going
to require many, many, many operations before you can get
an answer, that's where the classical computers start to slow down.
Even when we're talking about modern classical computers, which often
have multi cores, right, you know, you'll hear about these
multi core processors. They can break down certain types of
(03:41):
problems into parallel chunks, and each core of the processor
can work on a chunk of that problem, which cuts
down on the amount of time it takes to solve
the overall question that you've asked. But even then it's
you know, you're talking like like maybe sixteen cores. That's
still nothing if you're talking about a truly difficult problem.
(04:02):
And that's that's traditionally where we run into a barrier
with classical computers. There are certain types of non trivial
problems that classical computers have trouble solving. And here's an
example we've talked about before, the traveling salesman problem. You know,
the idea that you've got a traveling salesman who needs
to visit let's say ten cities, and your job is
to try and find the most efficient route for that
(04:24):
traveling salesman to go through to spend the least amount
of energy on this trip. And there are a lot
of different potential wants to choose from, you know, choosing
city A first and then going to city D and
then going back to CITYB, or going straight through, and
really you don't know what the right answer is until
you've compared all those options. And as you add cities
(04:45):
to this this question, it becomes more and more difficult
to answer, and a classical computer essentially what it has
to do is go through and run every single possibility
and then at the end of running all of them,
compare all the results to other to come up with
your answer, which can take a long time. So that
(05:06):
means that we need to look at maybe an alternative
to classical computing if we want to be able to
solve those types of problems in a more efficient, faster manner.
But how do we get more efficient than than a bit?
I mean, on or off seems like a pretty pretty
simplified What if you could be both on and off
at the same time? That's crazy talk. It is crazy talk, Jonathan. Yes,
(05:28):
I have never seen a light switch that was both
off and on at the same time. No wonder you
haven't seen it. If you had observed it, it it would
either be off or on. Yeah, can't. You can't see
it and see it that it's both off and on
at the same time, then you've observed it. Well. The
problem is because the light switch is actually a really
huge thing. Yeah, it's a macro level thing. It's on
(05:51):
our scales, um, so it's position tends to be pretty stable, right, Yeah,
But if we're looking at things on say a quantum
ski ill that's sub atomic, tiny world where things just
don't make sense an electronic Yeah, we're talking like tiny
little particles. They exhibit behaviors that if if that same
behavior were to suddenly exhibit itself on the on the
(06:16):
on the macro scale, on our scale, we would all
just think that we had been sucked into some sort
of David Lynch weird al. Yeah, would make sense. No,
nothing would make sense, because the quantum world and the
classical world are are very different in the way they behave. Well, yeah,
our our intuitions are evolved to deal with you know,
(06:36):
like trees and rocks and animals, nothings, superpositions and in
further more, trees and rocks and animals that don't suddenly
shift three miles to the right for no apparent reason. Okay,
so what what are we talking about? How can something
be in two positions at once? It is called superposition,
And asking me how is the wrong way to go
(06:57):
about a joe, because I certainly could not tell you
how I can tell you what's going on here. So
super position is a concept within the field of quantum engineering,
quantum mechanics, where a sub atomic particle is able to
coexist in multiple states at the same time. And they
don't mean states like the United States. I'm talking about
(07:18):
actual states of being. So, for example, with electron, we
often talk about spin, Like let's say that it could
either have a spin that's up or spin that's down,
and that's going to determine its magnetic field as well. Right,
the direction of its magnetic field is that will be
determined by this electron spin. Now, on the quantum level, technically,
(07:38):
an electron can be in superposition, meaning that can inhabit
both an upspin and a downspin at the same time.
It has a probability of being in either one or
the other at any given moment, were you to observe
that electron and therefore lock it into that one one
state or the exactly, So, if you were to observe
the electron, the electron would then suddenly in have but
(08:00):
just one of those two states, and that would be
determined by the probability of which state it was most
likely to be in. Sometimes it's going to be in
the less likely state. That's why they're probabilities. Right, it
might be a chance that will be a spin down
and a six chance of spin up and you observe
it and it's spin down. That can still happen because
as long as the probability exists, that's how things can
(08:23):
sometimes shake out. Now, if you were to do that
same sort of experiment over a really long run, then
the probabilities would start to manifest themselves, assuming that everything
else was identical, which would never happen. But anyway, superposition
is that crazy idea that a sub atomic particle can
exist in both of these states at the same time,
(08:45):
at least until you observe them, at which point, once
you observe them, you would say that the system decoheres.
That's it's an idea where a quantum system is a
very delicate thing and if you interfere with it in
any way, if you try to interact with it in
various ways, it will decohere and become a classical system
where things behave more the way we would expect them
(09:06):
to based upon our own experiences. In terms of thought experiments.
This is kind of going back to if you've heard
about it Stronger's cat, It's it's it's you know, poking
the system and seeing, you know, trying to identify what
a particle is doing is going to make the cat either. Yeah.
So the basic Strodinger's cat thought experiment is that you've
got a box with uh, some sort of castor in
(09:29):
it that is going to that could release a poisonous
gas anytime after thirty minutes have passed. So you've got
a cat inside this box with the cast of poisonous gas,
and you wait for thirty one minutes to pass. And
at that at thirty one minutes, there's a fifty percent
chance that the castors released the gas and a fifty
percent chance that it hasn't, which means at that moment,
(09:49):
before you open up the box and observe it technically
from a quantum level, the cat is fifty percent alive
and fifty dead. And then once you observe it it
those those uh that quantum state deco here's and it
forms a classical state. Right. Of course, the idea of
Schrodinger's cat was first introduced a sort of like a
reductio out of serdum. The idea was like, this is
(10:10):
so ridiculous, exactly, yeah, but it turns out like works.
Quantum physicists were like, well that's tough, you know, Yeah,
on the On the macro scale, of course, it's ridiculous.
You know, you would never say that the cat is
both alive and dead at the same time. It's either
one or the other. And because you open up the
box doesn't change that at all. But on the quantum
scale it certainly does matter. Okay, But so if if
(10:32):
this is how does this relate to quantum computing? Are
we giving the cat a bunch of buttons to push?
The cats and quantum computing? Lauren, I don't know where
you got your notes, but let's just without cats. I
think is a really sad. I think computers are made
entirely of cats. Internet is made of cats, not computers.
Uh no, no, no, we're going to We're going to
(10:52):
back off the cats and the Internet and the quantum
computing for just a second. There's one other there's one
other concept of quantum that we have to cover, which
is entanglement. Yeah. This is the idea of where you
have multiple sub atomic particles that are entangled in some
quantum way. So remember when I was talking about the
spin of the electron being either up or down. If
(11:14):
you have two entangled electrons, those electrons are going to
be kind of opposite but mirror images of one another,
and that if you know the behavior of one, you
know what the behavior of the other one was at
that moment when you observed the first one. Knowing that
from that moment forward, you can't really predict anything, but
(11:34):
being that if if two electrons are entangled and one
is spinning up, the other one would be spinning down.
For that, for that particular set of features, that's not
just limited to spin. There are other things we have
to take into consideration, but that goes well beyond just
the basic idea of entanglement. Entanglements very important with this
when it gets to quantum computing, the concepts of superposition, entanglement,
(11:58):
and coherence are all really really important. So you asked,
you know about quantum computing. That's that's the where we're
getting at. The idea of quantum computing is being able
to harness these features of the quantum world in a
way that can do compute computational work for us and UH.
And the the base unit of that. You know, if
(12:19):
you were to say the base unit of a computer
is the bit, either a zero or a one. The
base unit for a quantum computer is the cubit, which
I had to be reminded, is not a little orange
guy who bounces around the pyramid. Okay, so instead of
a silicon microprocessors, say, you would have computational activities being
(12:41):
done by something like a photon or an electron or
preferably lots of them. We need at least two. But um,
although I guess I know you need at least two.
So what the cubits are? Uh? Interesting in that if
a bit a bit can only be a zero or
a one, it's one or the other. It gives you
a single value. Yes, cubits are both zeros and ones
(13:02):
at the same time superposition uh, and technically all values
in between, although that's not really that important. So the
interesting thing about cubits is that the the relationship between
cubits and computing power is exponential. You take two to
the power of however many number of cubits you have,
and that is the equivalent of your quantum computer's power,
(13:25):
meaning that with with you know, two um cubit's you
have four potential values. There. Uh. You have to look
at the things as like, uh, two zeros zero, one
one zero or one one um and they're all the
same things, and there are all of those at the
same time, right, But if you were to add another
(13:46):
cupid in there, then you're talking about two to the
third power. So you're talking about eight pieces of information
from three cubits, which is different from the way it
would be if it were just bits. And as you
add cubits it becomes exponentially more powerful. By definition, you're
talking about actual exponent here the inn in that two
to the nth power. So the interesting thing here is
(14:10):
that all of these different cubits could uh and inhabit
these two values of zero and one at the same time.
If you have enough of them, then you could, in theory,
run a very complex problem through a quantum computer and
it could solve for all aspects of that problem simultaneously
in parallel because it's essentially doing all of those calculations
(14:32):
at once because all of the cubits are all possible values.
So it's kind of uh great for very specific types
of difficult problems. Okay, so this sounds like, though, um,
it's not going to be a replacement for the kinds
of computers we use now, not at all. Because while
(14:53):
it's great for certain complex problems like the traveling salesman
problem where it could solve for all of those different
variations simultaneously. It's not necessarily going to be any faster,
and in fact, might even be slower than a classical
computer for your basic computing functions that that regular schmos
like like myself, like like I do. If it's not
(15:14):
gonna show YouTube super faster, you're not going to be
able to run the latest video game even faster, like
I love the idea of running a video game on
a quantum computer and all possible outcomes of the video
gameplayouts simultaneously, like I was both good and evil and
everything in between. But that's that's not exactly what would happen. Okay,
So you're saying that a machine like this might have
(15:36):
really incredible powers in some kind of specialized way, very specialized,
like cryptography breaking. If if we had a working quantum
computer of sufficient power, cryptography as it exists now would
be meaningless. And the reason for that is that basically
the way cryptography tends to work is take two really
(15:58):
large prime numbers, like really really large. We're talking digits
that are hundreds of digits long, like it's it's an
enormous number, and then you find another enormous prime number,
and you multiply the two of them together and you
get that product, and then your your encryption is based
upon a party that's authorized having one of those two
(16:18):
large prime numbers, and as long as it's one of
the two right to correct numbers, they can get access
to that uh, that particular information or site or whatever.
I'm oversimplifying for the purpose of this podcast. Now, publicly,
all you can see is the product. So you see
this huge product. I mean, it's a no enormous number.
(16:39):
I remember, it's the product of two big prime numbers.
And if you don't have the information already, trying to
figure out which two prime numbers made this even bigger
number is really hard to do. In a way a
classical computer would do it, is it would start by
dividing by prime numbers and then run through all of
the prime numbers that possibly could be. So, if you're talking,
(17:02):
if you pick a large enough prime number, that alone
is going to guarantee that any computer working on trying
to force this cryptography, this brute force stile attack is
going to take longer than it would be you know,
feasible to run. So you would you know, most people
would not ever bother to try, because to do so
successfully would take forever. But if you had a quantum
(17:24):
computer that could solve for all potential prime numbers at
the same time, you could crack that roll relatively quickly.
In fact, yeah, it could. It can make the most
advanced encryption tools useless. It could also usher in a
new era of quantum cryptography, which would be even more
(17:45):
difficult to crack. But you know, it's it's one of
those things already where that's just one application. Obviously, there
are lots of other applications for quantum computing. Yeah, that
was one of the early applications. Actually, in the in
the early nineties, Peter Shore of Bell Labs developed a
quantum algorithm that that was a method of of entangling
cubits and using superposition to um to find prime factors
(18:07):
of an integer. Although that's not to say that that
we can run that on all of our fancy current
quantum computers. It was really like a proof of concepts
saying that once we are able to do this, it's
going to change our world, and it's good to know
about it now rather than three months after the world's
fastest quantum computer is made, and then we all realize
all of our stuff is public. It's better to know
(18:28):
it now, so we can say, huh, that's a problem.
How do we fix this? So I mean, but yeah,
it's a great example, and a lot of work has
been done on quantum computers since just that that that
algorithm was really for a hypothetical quantum computer. But now
we've got people who have actually built at least preliminary
quantum computers. Yeah, what's out there? I think I think
(18:51):
the fanciest one that we've got was built in two
thousand and seven. It's it's called the d Wave and
it's a sixteen cubit quantum computer. I know they were
working on one that would have been five and twenty
eight cubits, which would have been phenomenal, and that that
would have been unveiled within the last year or two,
But I honestly don't know if they ever uh successfully
(19:13):
demonstrated that one. But the fact that we've had people
demonstrate this at all is pretty phenomenal because keep in mind,
you have to be really careful with the way you
operate one of these things, because just by observing it,
by by trying to interpret the results you could cause decoherence,
and then you end up with a very primitive classical
computer that can't do much of anything, Like can you
(19:34):
imagine going from a sixteen cubit computer to a sixteen
bit computer? Would not be great? Right, And they've been
working on these kinds of problems for UM since the
since the sixties and seventies and eighties when all of
this was sort of starting to come together, and it
required a lot of work in UM and first of all,
bringing reversible logic gates into the computing world, which which
(19:57):
allows you if you have a one way gate, then
you're going to experience a lot of data and therefore
heat loss in your computer system. UM having a reversible
gate lets you UM basically not burn out your processor
every time you turn it on. Essentially, UM and and
quantum electrodynamics, which is just starting to hum look at
(20:18):
how electrons and photons interact with each other, so that
we can start creating these little quantum pieces of of
electrical information, right right, yeah, I mean, how do you
harness this stuff? It's it's really tricky. Are you have
to be able to create entangled particles? That's already kind
of tricky there's certain minty materials that are being used
right now in an experimental way that that are potentially
(20:42):
a source of entangled photons. That's pretty exciting stuff. You
have to figure out how to create a system that
these can work in that is not going to allow
it to go into decoherence. You have to figure out
how to program for it so again that you can
take advantage of it using it for the right sort
of problems, and you have to figure out how to
get the solution out of it again without disturbing the system,
(21:06):
which is pretty tricky stuff. And even then, you're talking
about probabilistic results, right, you mean you're getting results that
are assigned certain probabilities of being correct versus incorrect. And
sometimes the probabilities you get are so high that you
might as well say it's a certainty. I mean, you
can't really say that statistically speaking, there's always some room
(21:29):
for uncertainty, but you know from human experience and be like, well,
you know, times out of a hundred it's right, but
there's still a chance that could be wrong and not
all and of course some results may end up being
like we're sure this is the right answer, So it's
a it's a very uh specific, kind of niche oriented
(21:51):
version of computing. It's not something you're not gonna go
and get your uh your your your laptop. That's gonna
have you know, the cotum version of the last model
you owned, right, sure, and it is it is. Yeah,
like you said, just slow going. I mean it wasn't
until two thousand and one that there was a successful
demonstration of shorts algorithm, and that was with a seven
(22:12):
cubit computer that found the factors off. Yeah, they have
factors of It's still it's still something. I mean, hey, no, no, no,
I mean that's impressive. I mean I think that the
very first one added one in one and everyone was
so excited. I mean I mean understandably so. But nonetheless,
and you know, and there's there's a few different ways
that you can work on this coherence problem. I think
(22:32):
that that is the largest issue that we're talking about. Um, yeah,
that's that's I mean, none of these are trivial, but
I would say that's the hardest or from what I understand,
keeping in mind, I'm not a quantum physicist, right because
I mean you can either I mean, you can like
trap ions and super cool them until they're in a
quantum state so that you can work with them, or
um you can use liquid to to kind of wrangle
(22:54):
cubits so that you can spread a single cube it
across a few different molecules, which will decrease your your
rate of coherence. Right, But all of this is you know,
I mean, we're talking about very tricky particle physics. Yeah.
And while we're talking about you know, kind of humble beginnings,
keep in mind that other things that we depend upon
today very heavily had very humble beginnings. For example, the Internet.
(23:17):
You know, if you remember back when the Arpanet was
first being put into into action, you know, that was
a predecessor to the Internet. Some people call it kind
of like the grandfather to the Internet, but it was
it was a network of networks, was the idea, and
it was this um. The one of the earliest messages sent,
actually the first message set UH ended up crashing midway
(23:41):
through the message. It was a one word message. So
you know, you've got, you know, something where you point
to that and like look at that. If we had
just assumed that everything that followed that that failure was
also going to be a failure, we would not have
the Internet right now. So while we talk about these
kind of tiny exam balls, you know, when you're thinking
(24:01):
about proving a concept to be viable, that's huge, right,
even if it seems relatively tiny and you're talking about, oh,
you found the factors of fifteen. Wow, that's something that
I learned in third grade. But you know, for something
that until that point was hypothetical, that's amazing. So yeah,
I mean, these quantum computers have a lot of interesting potential,
(24:23):
not just in decryption or even the traveling salesman problem. Joe,
you had, you saw something, right, Yeah, they might be
able to help us actually study some things in the
real world that are really difficult to study with the
computers we have today, or might be really expensive, say
to simulate physically. I've got a press release from last
year from the n I s T Information Technology Lab UM,
(24:44):
and what this talks about is that quantum computers might
be able to simulate particle collisions. So the kind of
work that we have to do now with like the
particle accelerator, like the large Hadron collider and something it's
miles across and die amateur, Yeah, it could could actually
be simulated just in software. And of course, like with
(25:05):
a regular computer, it's really hard to do this because
digital you know, the computers we have today can't determine
all these quantum states. There's just like too much information
to keep track of um. But this press release talks
about a team that came up with an algorithm that
could basically run on any quantum computer, regardless of what
(25:26):
it's quantum hardware was um, and it would simulate all
of the different ways that two different types of particles
could collide and interact. Interesting. Yeah, see, this is kind
of fascinating stuff. The idea that we can uh add
in all these known factors that we are aware of
and create a simulation that could potentially create stuff we
(25:49):
aren't aware of. It's kind of to me, that's that
that's almost like magic at that at that stage where
you're like, all right, I know what this does, and
I know what this does, I don't know necessarily all
the things that can happen when the two collide, and
you create a simulation that can actually show you that.
To me, it boggles my mind. You know, it's just
such a phenomenal thing. And yeah, again, the quantum computer
(26:11):
would be a use in that situation, but not so
much if you want to play mind Sweeper. Well though,
we're not saying necessarily that a quantum computer could never
run a game or a video player or something like that.
We're just saying like, there's really no reason you'd need
one when a classical computer can do the same thing
(26:32):
for a long time. Because, I mean, while while your
power of your quantum computer does, uh does expand exponentially
with the number of cubits you add, you still would
have to add a lot for it to really uh
for for classical applications, for it to be better than
a classical computer. Now quantum applications, it would just leave
(26:52):
everyone else in the Yeah, but you would you would
need you know, a whole gallon of quantum, right, Yeah,
you gotta get that. Gotta go over to the quantum
refueling station and fill up your quantum tank and bring
it back just to just to play like like Super
Mario Brothers. Oh yeah, I don't know about you, guys,
but I'm ready to go and play some video games.
(27:13):
So I'm gonna wrap this up. Guys. If you enjoy
our show, please go check f w thinking dot com.
That's our home web page where We've got all the
blog posts and videos and podcasts, cool stuff that you
don't want to miss. Go check that out. I remember
you can interact with us on Google Plus, Facebook, and Twitter.
Just look for f W thinking that's us. We're happy
(27:35):
to talk to you, and we will talk to you
again really soon. For more on this topic in the
future of technology, visit forward Thinking dot Com, brought to
(27:59):
you by Toyota Let's Go Places
Fw:Thinking News
Hosts And Creators
Jonathan Strickland
Joe McCormick
Lauren Vogelbaum
Popular Podcasts
United States of Kennedy
United States of Kennedy is a podcast about our cultural fascination with the Kennedy dynasty. Every week, hosts Lyra Smith and George Civeris go into one aspect of the Kennedy story.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com