March 17, 2026 • 39 mins
In Episode 136, Patrick and Ciprian dive into the last five years of Entangled Things and explore potential of quantum computing over the next five years, focusing on the critical signals that indicate breakthrough moments. They discuss the parallels between quantum and AI advancements, highlighting how sudden leaps in technology can redefine industries. The conversation covers the evolution of quantum sensors, the synergy between classical and quantum computing, and the importance of error correction and qubit stability.
Listen
Watch
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_00 (00:30):
Welcome to Entangle Things.
Hosted by Patrick.
SPEAKER_04 (00:40):
Hey Tiprine, how are you doing?
Hey, Patrick.
I'm doing great.
Looking forward for anotherepisode of Entangle Things.
SPEAKER_03 (00:47):
Well, so we're a little late, but but we're we
want to talk a little bit of aretrospective after five years.
February 8th was our fifthanniversary of making the show.
Can't believe it.
It's really been crazy.
SPEAKER_01 (00:59):
It's really unbelievable.
SPEAKER_03 (01:01):
And that means you and I have been talking for
about 10 years about this topicand attracting crowds at uh at
different places.
I I don't know if I ever toldyou the story, but I was I was
down in Hawaii on a businesstrip and I was uh walking
through a park with a couple ofpeople who work with me and I
was talking to them aboutquantum and we sat down at a
bench and a uh uh gentlemanstarted listening in and joined
(01:24):
in the conversation.
Uh and so it kind of reminded meof the old days when we'd be at
a conference and people wouldgather around.
Um so quantum's much betterunderstood now.
It's it's it's something thatpeople have heard of.
Um back five years ago when westarted this, that wasn't the
case.
It was quantum computing.
(01:44):
What are you talking about?
You know, is that you know, isthat for physics?
And the answer was yes and no,and you know, everything in
between.
So so it sounds like it's a goodtime to talk about this.
But what I really want to getto, and your expertise uh and my
experience, I think, will beinteresting here is what can we
learn about what the next fiveyears might look like given what
(02:06):
we've seen AI's boom be?
My take is that the AI boom isnow.
Okay, I can't imagine the hypegetting much bigger.
Um, and then in a few years,probably by the end of this
decade, we're gonna see aquantum break-free in the same
way.
Now, maybe I'm wrong.
Maybe it's a decade away, maybeit'll never happen.
(02:26):
I I don't think that's the case.
But I'm I'm looking for likelessons learned, what to look
for, how people equipthemselves, things like that.
I think that would be a good wayto frame a retrospective on five
weeks.
What do you think?
SPEAKER_01 (02:39):
Yeah, yeah, absolutely.
I mean, um I am a big fan oflearning things from the history
of computer science because uhthings have this interesting
tendency of repeatingthemselves.
SPEAKER_03 (02:53):
Yeah, so I think Mark Twain said that history
doesn't repeat, but it rhymes.
SPEAKER_01 (02:58):
Yeah.
SPEAKER_03 (02:58):
Yeah.
And I think uh a friend RichardCampbell's that's one of his
favorite quotes.
SPEAKER_01 (03:03):
And I think there are similarities in how AI
evolved and evolves and quantum,but there are also differences.
Um I think the most strikingsimilarity for me is the fact
that development is essentiallynot a continuous slope, right?
It happens in in incremental butsignificant steps forward that
(03:28):
step function.
It's literally a step function.
SPEAKER_03 (03:31):
Yeah, that's how I a lot of people and that's and
that's the problem with makingpredictions of well, it took us
18 months to get here, so in 18months we'll be there.
Yeah, that might or might not betrue.
SPEAKER_01 (03:43):
The perfect example is right, the stars of today,
the large language models, theyhave been around uh like well
before 2022 when we had theso-called Chat GPT moment,
right?
But that was when things reallyuh exploded, and then obviously
we have other uh examples.
And I think in the five yearsthat we've been privileged to
(04:07):
have uh amazing guests on thison this show, we've seen how
those the developments are alsohappening in quantum where
things seem to be essentially uhblocked or capped uh in a
certain way.
And then all of a sudden thereis an announcement that someone
did this and that.
SPEAKER_03 (04:26):
So this is this is key.
And prognostication is adangerous art uh or or science.
Maybe it's a science.
I think it's more of an art.
Um I'd like to talk briefly, orat least for a little bit, about
what we think is are thepotential, and we're probably
gonna get this wrong, chat GPTmoments for quantum.
Because right now, when I talkto people about quantum, they're
(04:48):
like, wow, what's that allabout?
And there's, you know, I can'tsay, well, you know, you can do
this and your business will getX number of profits next year
because of this.
You know, there's some of thatwith the cubos and you know,
with the things that D-Wave'sdoing with optimization, and
there's some scientificdiscoveries.
But I've always harped on theShores algorithm.
(05:09):
When when Shores breaks RSA 2048in in a day or an hour, then I
think that would definitely be amoment.
But I think there's somethingbefore that's gonna come.
I think there's gonna besomething that captures the
attention and makes industrysay, oh, this is actually a
thing.
This is actually gonna happen.
(05:30):
It's it's kind of like the daythat you know fusion opens a
power plant and powers a city.
Um people will be like, oh, Ithought this was just 20, I
thought this was still 20 yearsaway.
Yeah, as it used to be.
What do you think those otherpotentials could be other than
shores?
SPEAKER_01 (05:46):
Uh before getting to that, I just want to finish my
idea in terms of uh what is notsimilar between AI and quantum,
right?
I I think there is a majordifference between how these two
things evolve in that we reallyhaven't seen, and it's unlikely
that we will see a quantumwinter as opposed to AI, which
(06:09):
had its fair share.
And I think the big differenceis because there is this threat,
you already mentioned it, right?
Quote unquote threat, uh posedby the Shore's algorithm, which
would have massive impact in alllevels, from private
organizations all the way togovernments.
And this is why I think thefunding for the development of
(06:32):
quantum computing uh is lesslikely or much less likely uh uh
to dry out.
It's assured almost.
Uh yeah, it's it's essentiallyassured because it becomes
something that is critical forthe well-being of states, uh,
which was not the case with AIuh in the past.
SPEAKER_03 (06:52):
So this is one important difference where now
when you say a winter, it's aloss of enthusiasm, it's the uh
bottom of a hype cycle, it'sit's a combination.
SPEAKER_01 (07:01):
It's a combination of all, right?
It's funding dries up, uh, lotsof companies are essentially
either shifting focus orcompletely getting out of that
particular business.
Um, and there is obviously asteep decrease in interest from
end customers, uh,organizations, uh, and and and
(07:21):
so forth.
SPEAKER_03 (07:22):
Last count, there was like 88 major companies in
the space, and that might justbe North America.
Um I haven't heard of I've heardof companies like morphing or
being absorbed.
Uh I've heard of you knowresearch teams changing topics
and going and doing somethingelse, but I haven't heard of
anybody just outright going outof business.
SPEAKER_01 (07:44):
No.
No, that's so it doesn't.
And that's because the money isthere, right?
SPEAKER_03 (07:48):
Right.
Because it's like, well, youknow, we're gonna keep that bet.
And DARPA's shown interest inthis space.
Yeah, which means they thinkthat it's going to um disrupt
things.
SPEAKER_01 (07:59):
And and I think there is there is significant
significantly more government uhinvolvement, whether it's the US
or Europe or China, right?
Significantly more governmentinvolvement than it used to be
with AI in its early days.
This is also a significantdifference, right?
Yeah uh that we that we see outthere.
(08:21):
Now, getting back to your toyour question, um I think at
least what I expect is to belike the next step in this uh
stepping function, uh will be anon-trivial uh improvement in
error correction and qubitstability that will not be
(08:42):
enough to run efficiently uh orto solve the shore problem,
right?
The shore algorithm, but it willopen the door for, let's say,
smaller problems with the codessmaller that would have a direct
impact in things likeoptimization or material science
or or things like that.
(09:04):
I think before shore, and Ifully agree with what you said,
Patrick, before we will have aquantum computer that will just
be able to get an RSA uhencryption and and break it,
before that we will see uhsmaller scale um improvements
that will start to have uh veryclear practical uh applications,
(09:27):
whether it's as I said, materialscience or or or anything else.
SPEAKER_03 (09:31):
Do you think that's well the material science one's
interesting?
Do you think that's enough tocapture because as you said,
LLMs were around for a while andnobody cared?
I I think there has to be somelike breakout where like you you
talked about the Fubinakomolecule.
Um, yeah.
Yeah, and that you know, that'swe we use two percent of one to
(09:54):
two percent of you of globalenergy to make fertilizers, and
if we can figure that moleculeout, that that could be rapidly
and dramatically reduced.
I think that could be somethingthat captures the imagination.
If if we had a breakthrough ofmaterial science that that
changed people's lives on aday-to-day basis, if we got
superconductors that worked,that were easy to make, cheap
(10:15):
materials, and that we couldreplace all of our wiring with
that worked at at roomtemperature or at ambient
temperatures, those are the kindof breakthroughs that may or may
not uh ever occur.
And but I think that's what itmight take for it to really kept
capture the zeitgeist.
SPEAKER_01 (10:31):
And I think there are other potential areas as
well uh which uh do notnecessarily need to be
exponential improvements.
Like just imagine if we couldcut down uh either the time or
the cost or both of traininglarge language models to half or
(10:54):
a quarter, right?
It it doesn't need to beexponentially like two at the
power of whatever, right?
But even if we can come up withsomething via quantum computing,
right, um addressing a part ofthe massive optimization
problem, which is training adeep neural network, right?
Uh, what if we can come up withsomething that will cut costs in
(11:17):
two or in four or in eight,right?
So I don't think it needs to beuh as spectacular as we kind of
all expect it to be.
Um, but if it has practical umapplications, and I think we
have enough problems today, andwe are starting to get better at
(11:39):
how to frame those problems forquantum systems, that the next
breakthrough will actually behey, it's not gonna cost us a
hundred days to train a model,but we can do it in 40 days, uh,
or things like that.
This is just another anotherexample.
So, because uh you know what wesee more and more and more is
(12:02):
there is no such thing as a purequantum thing, or uh quantum
computing does not live in abubble that is completely
disconnected from whatever it'saround it.
On the contrary, all theapplications that we know of
today, even the ones that areonly theoretical, like shore,
they involve a combination ofquantum computing and classical
(12:25):
computing.
Right.
And I think the synergy is uhalso going to help the
advancement uh quite a lot.
SPEAKER_03 (12:33):
Interesting point.
So when we talked to Murray Tomfrom D-Wave, he pointed out that
there were optimization problemsthat they could solve that could
be solved on a regular computer,but they could solve them at
lower energy costs.
And and and with the you know,the talking about energy with
with AI is, you know, we'retalking about percentages of our
(12:57):
energy infrastructure beingdedicated to that.
So there might be so I could seewhere if we had some companies
like Anthropic and OpenAIsaying, well, we've been able to
reduce our uh our our datacenter energy cost by 50%
because we're using quantum umoptimizations instead of
classical optimizations for thissmall.
(13:19):
Again, Shores, the the lessonlearned from Shores is you don't
have to solve the whole problem.
You just gotta solve the hardestpart.
You just gotta solve the partthat nobody thought you could
solve.
And if we can do that withoptimization, now I know there's
you know lots of tensorcalculations and there's there's
so many things going on in in AIthese days.
(13:42):
But it would be interesting tosee if you think that that's a
likely scenario.
Is it likely that we're gonnafind things that are necessary
in inference engines and in coin training quantum computing,
uh not quantum computing, in inAI where that kind of
optimization would be a abreakthrough?
SPEAKER_01 (14:01):
I think it's likely.
SPEAKER_03 (14:03):
Okay.
SPEAKER_01 (14:03):
I think it's it's it's it's likely.
Um and that's because the scaleof these problems keeps uh going
up, right?
Yeah uh the more powerful modelswe need and the more powerful
models we want to train, themore problematic becomes the
optimization uh uh part.
But again, it might not be AI atall, right?
(14:27):
It might be a different fieldwhere having a large enough
number of qubits, which is stillnot in the hundreds of
thousands, right?
Or tens of thousands, but maybelet's say in the thousands,
having like a couple of thousandstable qubits with with
universal quantum computing, Ithink will will open the door to
(14:49):
some unbelievable advancements.
SPEAKER_03 (14:51):
So I have a theory, and again, this is you know,
prognostication's a delicateart.
Um, I believe that the boom thatwe're seeing with AI will
facilitate the next boom, whichI don't think is going to be
quantum.
I think there's gonna be anintervening one in robotics.
So as I think as of right now,last I heard, there's about two
million industrial robots in usein China, and there's another
two million in the West (15:14):
Japan, Australia, or Europe, the United
States each have about half amillion, give or take.
So there's about four millionindustrial robots being used
right now.
And there's a big push for humanhumanoid robots.
I'm not sure if that's gonna gowhere people think it's gonna
go.
But but the robotics revolutionseems to be the thing that's
(15:35):
enabled.
And I'll I'll do one more seguehere, one more.
There's a company called TALAS,T-A-A-L-A-S, that took that made
a chip and they put the wholeLLM, 8 billion parameter LLM on
that chip.
And if you go to their homepage,they have a link to the chatbot
they've powered with this chipcalled Chat Jimmy, and it serves
(15:58):
up 15,000 tokens a second, whichis mind-boggling more than it's
way, way more than most otherengines can can accommodate over
the web.
Um, and the idea is if I can puta an LLM in a chip, I can get
rid of some of that delay, andthat enables robotic solutions,
(16:19):
real world solutions.
And I could see where sensors,quantum sensors, start taking an
outside role, outsized role inenabling the robotics
revolution.
Because you know, we've talkedto several people about sensors,
not as many as we'd like, but Icould see where we'd have
fundamental problems in therobotics space, where you need a
(16:43):
fast model.
Well, putting the putting awhole model on a chip, that
seems to be solved.
That's not a quantum problem.
But having the sensing so that arobot can go into a mine and do
the mining instead of a personhaving to go down in the mine
and do the mining.
Remember, we talked to uh acompany that does the magnetic
sensoring, and they talked aboutartificial diamonds with you
(17:04):
know vacancies.
I could see something like thatcapturing the imagination to un
for people to understand, wow,this wouldn't, these robotics
revolution wouldn't be possiblewithout quantum.
And so I think it might come inone of these side angles where
it captures the the uh you knowGPS.
A lot of people didn't, theymaybe don't realize that that
(17:26):
you know quantum mechanics isone of the things that makes GPS
possible.
Um but I'm wondering if that'show it would seep into the into
the uh the the consciousness ofthe business community and
people, uh other people.
It's definitely gonna have animpact.
And I think it's gonna be AIrevolution, robotic revolution,
probably ongoing at the sametime, and then quantum
revolution in sometime in thenext decade.
SPEAKER_01 (17:49):
Yeah, I think it's very likely that that things
will will evolve, will evolvethis way.
Um definitely with robotics, nowwe are we're starting to see
massive advancements in thetechnologies used to help the
robots learn.
That's that that was one of thethings, right?
(18:09):
I know, for example, Tesla hasthis project where they are
planning to put several hundredrobots in in factories.
And the interesting thing isactually these robots are
matched so that whatever onerobot learns to do, which
essentially amounts to traininga bunch of deep neural networks
at the end of the day, right?
(18:30):
It will instantly share with allthe others.
So think of it like one needs tolearn how to pick up uh
something with its uh left hand,the other one has to learn how
to pick up something with itsright hand, and they can share,
right?
So that will cut down approacheslike this, will significantly
cut down the time needed forthese machines to uh learn right
(18:53):
there their their things.
SPEAKER_03 (18:54):
So that's that was Waymo's kind of narrative when
they said, well, for every everyone of our cars gets an hour of
driving time, they all benefitfrom the learning.
And there's also some people umyou know getting into the uh I
don't want to go stray too muchinto AI, but but one of the
things um I also wanted tohighlight is have you heard of
Eagle Eye from Andorrill, U.S.
government contractor?
(19:15):
Lucky Palmer?
Lucky Palmer, the guy that cameup with um the uh Oculus.
Oh, yeah, he's one of thefounders.
They have this goggle set for USArmy uh soldiers where if you
can see a bad guy, then I canhave it rendered on my vision in
(19:37):
AR, augmented reality, so I knowwhere they are behind the
building, even though I can'tsee them.
SPEAKER_01 (19:44):
Yeah.
SPEAKER_03 (19:45):
Because someone else has eyes on them, or some drone
has eyes on them.
So it's that shared experience.
So I could see you know a sensorbeing deployed in that regard so
that you could kind of figureout what a building is made of.
So you could decide, you know,what's safe to do there, what's
(20:06):
not safe to do there.
I mean, the the the comp thepossibilities are endless.
And and unfortunately, I thinkthe military might be the place
where a lot of thesetechnologies get their start.
Just you know, NASA pushed thethe computer chip, uh, it's
usually government efforts.
SPEAKER_01 (20:20):
And the PowerPoint pushed the internet.
SPEAKER_03 (20:22):
Exactly, yeah, yeah, and autonomous vehicles and
everything else.
Um so I think sensors issomeplace that I'd like to you
know understand better who'splaying there, what they're
doing, uh, because I think thatcould have an outside role uh in
getting us to that moment.
Now, once we get to that moment,one of the things that's gonna
be a scramble is figuring outhow quantum will disrupt each
(20:47):
and every market and and facetof life.
And right now, I don't thinkwe're prepared for that.
Most people, again, haven't eventhought about what this
technology is or or why itexists.
But AI is now disrupting partsof the economy that we didn't
imagine even two years ago.
And so I think it's yeah.
SPEAKER_01 (21:07):
That's something that if you look historically,
right, that has happened withwith every major breakthrough
technology, right?
Yeah, you had some types ofdisruptions that you were
expecting, and then others thatyou were literally not
expecting.
Yeah.
Um, I'm just thinking about anegative one, which was the
(21:27):
dot-com bubble with theinternet, right?
SPEAKER_03 (21:30):
But still, the internet didn't go away after
that, even though it was.
SPEAKER_01 (21:32):
Of course.
Yeah.
Yeah, but it was an interestingside effect, right, of the the
growing popularity and theenthusiasm.
SPEAKER_03 (21:39):
Yeah.
And unfortunately, things likecorner bookstores have have
fallen into uh single digit umuh numbers as opposed to where
they were before.
SPEAKER_01 (21:48):
So I I I and let's also not forget about quantum
communications, right?
Because typically when we talkabout quantum, we need to think
about the computing, we need totalk, we need to think about The
communications part, we need tothink about the sensing part,
right?
Because all these are kind ofdifferent aspects uh uh around
(22:08):
the the development of the ofthe technologies.
So and we see advancements onall fronts.
That's that's the interestingthing from from from my
perspective.
And then the other thing that Iwant to to highlight is uh it
hasn't necessarily has itdoesn't necessarily have to be a
(22:29):
uh major step in terms of say uhwe've increased 10 times the
number of qubits, right?
And and and things like that.
It also can be uh a majoradvancement in terms of an idea,
right?
And if you ask me, uh one of themost important ideas that uh
(22:50):
occurred in the past few yearswas uh kind of thinking outside
the box and implementing thisidea of moving qubits instead of
trying to use static topologiesfor building out gates,
especially two-qubit gates,right?
The the core idea to move qubitstogether to create uh uh
(23:12):
two-qubit gates, right?
That in itself, for me, is agood example of a major step in
that stepping function, becausethat idea enabled in a
relatively short amount of timesignificant improvements in
terms of error correction andeven uh the the way that that
more complex gates liketwo-qubit gates can be
(23:35):
implemented.
Uh, and it also I think openeduh the the road towards um
increasing the temperature at atwhich some of these chips have
to have to run, right?
Um allowing us to move out ofthose crazy uh 30, 20, 30, 40
(23:57):
millikelvin temperatures tothings that are let's say at
least more more levels that areat least more manageable.
SPEAKER_04 (24:05):
Yeah.
SPEAKER_01 (24:05):
So that's that's the other thing that that and
remember there were times at thebeginning where we were like,
hey, we we don't see uh many newalgorithms, like we don't see a
new shore, we don't see a newgrover, and and so forth.
Well, looking back to these fiveyears, right, for me it's uh
(24:27):
interesting that it wasn't analgorithm, but it was an idea,
yeah, right, in in how toaddress a potential uh uh
physical problem, which is howdo we build the two qubit gigs.
SPEAKER_03 (24:39):
Well, we were focused in those early days,
early, early year, actually.
Um I was focused more onsuperconducting and photonics.
I those were the modalities thatI had heard of that I knew of.
SPEAKER_01 (24:52):
And trap ions, trap ions because of their legacy in
atomic clocks and and other kindof previous developments.
SPEAKER_03 (25:00):
I don't think the players had really fleshed
themselves out in that space.
So I I they looked likeinteresting potentials at that
time.
And then D-Wave with the um withtheir uh Adiabatic app.
Um now D Wave is still activeand and doing solving problems.
(25:21):
They've got the Cubo system andthey're they're actively trying
to help companies figure out howto use this optimization.
Um but they're also looking atdoing uh universal quantum
computing, from what I've heard.
And maybe they're seeing youknow the Yeah.
So the yeah, the the algorithmthing didn't really go the way
(25:42):
we thought.
I I really thought that we well,there'd be we had Shores and we
had drovers, and we'll have youknow, Cyprians and Patrick's,
you know, there'll be other uhsystems.
But we also kind of explored thefact that there might not be as
many really fundamentally hardproblems to tackle, right?
(26:05):
That we that was also the thingthat was kind of surprising is
we we're looking around, well,what kind of problems could be
solved?
Well, what kind of problems arethere right that in the in that
world?
And again, I think AI androbotics are introducing a whole
new set of problems wherequantum might be um you know a
valuable uh thing to do, avaluable place.
SPEAKER_01 (26:27):
And and to that point, Patrick, I think one
other potential kind of uhmassive leap forward can be
exactly in this area, uh, abetter understanding of how to
model existing problems into thequantum world and into the
quantum space.
Because this is one of thethings that that are still
(26:50):
relatively difficult because youdon't really have the same
primitives with quantumcomputing like you have with
classical.
You can't really store a lot ofinformation in the classical uh
uh kind of sense of the uh ofthe word, right?
Uh you also don't really havelong term, and really when we
say long-term in the world ofquantum, we're uh uh really
(27:13):
referring to seconds.
That's essentially long-term inin quantum, right?
So there is a a non-trivial umlevel of difficulty in taking a
problem uh that is wellunderstood in the practical
world and is well defined, andessentially morphing it into a
(27:33):
representation that could bevalid or uh can can work with
quantum computers.
I think there's a lot to improvethere, and this could be another
area where we could seeimportant breakthroughs uh in in
the coming years.
SPEAKER_03 (27:48):
Yeah.
I I'm not predicting um thatquantum computing becomes like
let's the fi world on fire andmainstream before the end of
this decade.
I I don't I don't see that as asin the cards.
I may be wrong.
Maybe maybe I'm completelywrong.
SPEAKER_01 (28:03):
I share I I share your view, yeah.
SPEAKER_03 (28:05):
Yeah, I think but I think we're making consistent
steps forward and filling inthese gaps.
Um, you know, the wild card forme on chores, just to pick one,
because I think that one is aguaranteed.
If if if suddenly RSA 2048 isendangered by quantum computers,
that's gonna be something that'sgonna get noticed in in
(28:27):
instantly.
Um when talking to these some ofthese companies that are making
efforts to chain together to doa distributed computing approach
to quantum computing.
Well, now we don't need a a5,000 qubit quantum computer.
We need a hundred fifty qubitquantum computers.
And that's a much easier thingto imagine us cobbling together
(28:51):
quickly than waiting for thereto be a 5,000 logical qubit
quantum computer.
I don't know if that's gonna bethe way it plays out, but that
feels like where we're gonna seea sudden step function occur.
SPEAKER_01 (29:04):
Yeah, yeah.
I mean, think about the earlydays of our podcast, right?
Um chips with five qubits orchips with with eight qubits
were really exciting, yeah,right?
Yeah uh uh back then.
And now we have chips with what,48, 50.
SPEAKER_03 (29:22):
Well, we were talking about this before there
were any qubits where they werejust theoretical.
SPEAKER_01 (29:28):
Of course, yeah, yeah.
So uh things are are improving,right?
Things are are are improving.
Um and uh again, um becausethere is this continuity in
terms of funding, right?
This is why I believe um, atleast in this kind of first part
of its history, like the firstuh I don't know, 20, 30 years of
(29:51):
its history, quantum computingwill have a smoother ride uh uh
than uh AI had or or or theother the other technologies.
Um now I also think that thatanother um um important aspect
is going to also be the impactof this almost uh unexpected
(30:19):
success of AI, right?
Because on one hand, I think ithelps, but on the other hand,
now with AI, we can alsocontinue the advancements in
classical computing, right?
We can essentially design andimplement much uh better data
(30:41):
centers, uh much better uhcomputing grids that are based
on GPUs and and other and otherthings.
So I believe there is also uh asteep development in classical
computing, which I think isstarting to also chip into some
of the problems, especiallyaround optimization problems
(31:05):
that were kind of deemed to bethe exclusive realm of quantum
quantum computing.
Right, yeah, it's it's the neverendless a never-ending
discussion about quantumsupremacy where the quantum side
says, okay, here's somethingthat we figured out, right?
We calculated, and then a fewmonths later there's a classical
(31:26):
computing team coming, sure.
Look, we use this supercomputer.
Hold my computer, we uhsignificantly optimized and
improved this and that and that,and now look, you you don't
really prove quantum supremacyyet.
So it's an interesting kind ofrace because classical
computing, right, even if itsuffers, let's say, from an
(31:49):
exponential uh disadvantagecompared to quantum, but
classical is also marchingahead.
And uh Well, I think that's abenefit.
SPEAKER_03 (32:00):
I think I I remember back in the old days when there
was when there's more than oneweb browser that is a
reasonable, optimal, you know,even optimal choice, they both
get better faster.
Right?
When Internet Explorer had realcompetition, we saw better
browsers overall for everyonebecause there was there was
(32:22):
push.
Um and but when we have amonoculture, everybody sits back
and says, well, you know, thingsare as good as they're gonna
get.
And and to that, I do believethat supercomputing is being
pushed by quantum and quantum isbeing pushed by supercomputing.
Eventually, I do believe thatquantum will find things that it
can do and do better, but Ithink it's helping quantum as
(32:43):
well.
It's putting a realism on it,it's preventing a quantum hype
bubble, which is dangerousbecause we're gonna have a hype
bubble with AI.
Some would argue that we'realready like dangerously in it
and it's actually close tobreaking.
Others would say, you know, wehave plenty of space to go.
I think the robotics um pushwill actually give it a lot more
(33:04):
runway than people are giving itcredit because I think AI is the
driver of robotics in this case.
With the quantum, the longer wepush off that hype bubble, then
the the further we'll get beforewe get a winter.
We'll get a winter with quantumeventually.
But it may be 10 years from now.
It may be after it it becomeslike you know a household name
(33:27):
and everybody's trying to becomea quantum engineer, and
hopefully everybody's listeningto old entangled thanks
episodes.
And um, but I honestly I thinkthat's a good thing because
moving the goalposts just makesyou run faster.
SPEAKER_01 (33:43):
Yeah, yeah.
And then there's another thing,Patrick.
Um, since we're speaking aboutboth AI and quantum, I think the
advancements in AI, especiallyaround large language models and
even more especially aroundcoding agents, are helping
tremendously the learning curvefor quantum computing related
(34:06):
things, whether it's quantumcomputing related pro uh
languages or uh approaches tooptimize designs of quantum
computing gates or whatever,right?
It's now significantly easier tolearn about this stuff.
And I think that's gonnacontinue for at least the next
(34:29):
uh the next few few few years.
Because uh if you talk, I don'tknow, let's say a random
example, Q sharp, right?
The uh programming languagedeveloped by by Microsoft for
for quantum computing, right?
Today it's really trivial to goask a large language model to
give you an example of like howto do, I don't know, a quantum
(34:50):
Fourier transformation or Idon't know, uh the uh an i uh an
ising optimization, you name it,right?
Yeah, and it's it's a resourcethat especially the younger
generation has, which makes theentry level significantly lower
(35:12):
than it used to be even likefive years ago or or ten years
ago.
And remember, we always talkabout the fact that the real
kind of explosive developmentwill probably happen when the
what we call the quantum-borngeneration is going to be the
one driving the field, where youwill have people that were not
(35:35):
uh impacted either in a positiveor negative way, that's up to
debate, but they were notimpacted by previous uh uh
learning of classical computing.
SPEAKER_03 (35:47):
So those of us that are old enough to have gone to
high school with Newton is that.
SPEAKER_01 (35:51):
Right?
Paradigm and uh who are going tobe people that the only
computing that they ever learnedor knew, right, was the the the
quantum computing uh uhapproach.
And then for the rest of us,obviously, uh lowering the the
uh kind of the the friction inentering the the this particular
(36:15):
field, making it easier to learnthings, making it easier to get
examples of things, making iteasier to get explanations of
various types of things.
Uh it's gonna help us well uh ina very, very significant way, if
you ask me.
SPEAKER_03 (36:32):
Yeah, I think I I mean we're over we're we're
getting over time here.
I usually I think, oh, it's justme and Cyprian talking, it'll be
quick.
Nope, never.
Uh so uh, you know, there'sstill a lot to talk about here,
um, but I think we need to drawto a close soon.
That said, I think what you'resaying is is very valid.
(36:52):
And one of the things a lot ofpeople are searching for is
well, how do I stay relevant ina world where AI can write code,
it can, you know, writesymphonies, it can do all sorts
of things.
Well, the the the there's stilla a border of knowledge.
And while I think AI can come upwith some novel solutions to
some problems, invent drugs, dosame things in material science,
(37:14):
we still need people to figureout how to do these novel
approaches because a largelanguage model is only as good
as the data you put into it.
SPEAKER_01 (37:23):
Exactly.
SPEAKER_03 (37:23):
And so this is the these frontier sciences, these
frontier you know things.
Um some of the things I look at,I'm just amazed at.
Well, somebody thought to use adiamond.
I keep coming back to thatexample of a flawed diamond for
detecting magnetic.
SPEAKER_01 (37:39):
Isn't that great?
SPEAKER_03 (37:40):
Like it is, it is.
It was one of my favoriteexamples.
We need to talk to more of theseuh these sensor people.
And when the when I'll admit,when when they first started
talking about quantum sensors,I'm like, what are you talking
about?
What what's and now I get it.
I get the fact that you know weuh they're sensitive and they're
small and they're and they'regonna have an impact.
So um I'm very excited for thenext five years.
(38:01):
Um I hope we can keep uh keepeveryone entertained and
informed.
And I think we're still on thesame mission is is we think this
is an important technology forscience, humanity, and
everything else.
And we want to make sure otherpeople are uh uh come along for
the ride.
SPEAKER_01 (38:19):
Absolutely.
Absolutely.
That's it's uh gonna be anexciting uh uh five years to
follow, for sure.
SPEAKER_02 (38:27):
Imagine saying that after this last two years.
That's amazing.
All right.
So thanks, everyone.
SPEAKER_00 (38:34):
Thank you.
Thanks, everyone.
Bye.
Bye.
Cybercrime is one of the biggestthreats to businesses of all
sizes and industries.
With almost half a million opencyber positions, the problem is
compounded by the lack ofavailable talent in the
marketplace.
At Pulsar Security, our eliteteam of highly credentialed
experts collaborate with you toassess your current defenses and
(38:56):
develop solutions tailored toyour specific needs.
With services ranging fromcybersecurity education to
advanced penetration testing andred teaming, you can start
reducing your risks today.
Visit pulsarsecurity.com andlet's secure your digital future
together.
Welcome to Entangle Things.
Hosted by Patrick.
SPEAKER_04 (00:40):
Hey Tiprine, how are you doing?
Hey, Patrick.
I'm doing great.
Looking forward for anotherepisode of Entangle Things.
SPEAKER_03 (00:47):
Well, so we're a little late, but but we're we
want to talk a little bit of aretrospective after five years.
February 8th was our fifthanniversary of making the show.
Can't believe it.
It's really been crazy.
SPEAKER_01 (00:59):
It's really unbelievable.
SPEAKER_03 (01:01):
And that means you and I have been talking for
about 10 years about this topicand attracting crowds at uh at
different places.
I I don't know if I ever toldyou the story, but I was I was
down in Hawaii on a businesstrip and I was uh walking
through a park with a couple ofpeople who work with me and I
was talking to them aboutquantum and we sat down at a
bench and a uh uh gentlemanstarted listening in and joined
(01:24):
in the conversation.
Uh and so it kind of reminded meof the old days when we'd be at
a conference and people wouldgather around.
Um so quantum's much betterunderstood now.
It's it's it's something thatpeople have heard of.
Um back five years ago when westarted this, that wasn't the
case.
It was quantum computing.
(01:44):
What are you talking about?
You know, is that you know, isthat for physics?
And the answer was yes and no,and you know, everything in
between.
So so it sounds like it's a goodtime to talk about this.
But what I really want to getto, and your expertise uh and my
experience, I think, will beinteresting here is what can we
learn about what the next fiveyears might look like given what
(02:06):
we've seen AI's boom be?
My take is that the AI boom isnow.
Okay, I can't imagine the hypegetting much bigger.
Um, and then in a few years,probably by the end of this
decade, we're gonna see aquantum break-free in the same
way.
Now, maybe I'm wrong.
Maybe it's a decade away, maybeit'll never happen.
(02:26):
I I don't think that's the case.
But I'm I'm looking for likelessons learned, what to look
for, how people equipthemselves, things like that.
I think that would be a good wayto frame a retrospective on five
weeks.
What do you think?
SPEAKER_01 (02:39):
Yeah, yeah, absolutely.
I mean, um I am a big fan oflearning things from the history
of computer science because uhthings have this interesting
tendency of repeatingthemselves.
SPEAKER_03 (02:53):
Yeah, so I think Mark Twain said that history
doesn't repeat, but it rhymes.
SPEAKER_01 (02:58):
Yeah.
SPEAKER_03 (02:58):
Yeah.
And I think uh a friend RichardCampbell's that's one of his
favorite quotes.
SPEAKER_01 (03:03):
And I think there are similarities in how AI
evolved and evolves and quantum,but there are also differences.
Um I think the most strikingsimilarity for me is the fact
that development is essentiallynot a continuous slope, right?
It happens in in incremental butsignificant steps forward that
(03:28):
step function.
It's literally a step function.
SPEAKER_03 (03:31):
Yeah, that's how I a lot of people and that's and
that's the problem with makingpredictions of well, it took us
18 months to get here, so in 18months we'll be there.
Yeah, that might or might not betrue.
SPEAKER_01 (03:43):
The perfect example is right, the stars of today,
the large language models, theyhave been around uh like well
before 2022 when we had theso-called Chat GPT moment,
right?
But that was when things reallyuh exploded, and then obviously
we have other uh examples.
And I think in the five yearsthat we've been privileged to
(04:07):
have uh amazing guests on thison this show, we've seen how
those the developments are alsohappening in quantum where
things seem to be essentially uhblocked or capped uh in a
certain way.
And then all of a sudden thereis an announcement that someone
did this and that.
SPEAKER_03 (04:26):
So this is this is key.
And prognostication is adangerous art uh or or science.
Maybe it's a science.
I think it's more of an art.
Um I'd like to talk briefly, orat least for a little bit, about
what we think is are thepotential, and we're probably
gonna get this wrong, chat GPTmoments for quantum.
Because right now, when I talkto people about quantum, they're
(04:48):
like, wow, what's that allabout?
And there's, you know, I can'tsay, well, you know, you can do
this and your business will getX number of profits next year
because of this.
You know, there's some of thatwith the cubos and you know,
with the things that D-Wave'sdoing with optimization, and
there's some scientificdiscoveries.
But I've always harped on theShores algorithm.
(05:09):
When when Shores breaks RSA 2048in in a day or an hour, then I
think that would definitely be amoment.
But I think there's somethingbefore that's gonna come.
I think there's gonna besomething that captures the
attention and makes industrysay, oh, this is actually a
thing.
This is actually gonna happen.
(05:30):
It's it's kind of like the daythat you know fusion opens a
power plant and powers a city.
Um people will be like, oh, Ithought this was just 20, I
thought this was still 20 yearsaway.
Yeah, as it used to be.
What do you think those otherpotentials could be other than
shores?
SPEAKER_01 (05:46):
Uh before getting to that, I just want to finish my
idea in terms of uh what is notsimilar between AI and quantum,
right?
I I think there is a majordifference between how these two
things evolve in that we reallyhaven't seen, and it's unlikely
that we will see a quantumwinter as opposed to AI, which
(06:09):
had its fair share.
And I think the big differenceis because there is this threat,
you already mentioned it, right?
Quote unquote threat, uh posedby the Shore's algorithm, which
would have massive impact in alllevels, from private
organizations all the way togovernments.
And this is why I think thefunding for the development of
(06:32):
quantum computing uh is lesslikely or much less likely uh uh
to dry out.
It's assured almost.
Uh yeah, it's it's essentiallyassured because it becomes
something that is critical forthe well-being of states, uh,
which was not the case with AIuh in the past.
SPEAKER_03 (06:52):
So this is one important difference where now
when you say a winter, it's aloss of enthusiasm, it's the uh
bottom of a hype cycle, it'sit's a combination.
SPEAKER_01 (07:01):
It's a combination of all, right?
It's funding dries up, uh, lotsof companies are essentially
either shifting focus orcompletely getting out of that
particular business.
Um, and there is obviously asteep decrease in interest from
end customers, uh,organizations, uh, and and and
(07:21):
so forth.
SPEAKER_03 (07:22):
Last count, there was like 88 major companies in
the space, and that might justbe North America.
Um I haven't heard of I've heardof companies like morphing or
being absorbed.
Uh I've heard of you knowresearch teams changing topics
and going and doing somethingelse, but I haven't heard of
anybody just outright going outof business.
SPEAKER_01 (07:44):
No.
No, that's so it doesn't.
And that's because the money isthere, right?
SPEAKER_03 (07:48):
Right.
Because it's like, well, youknow, we're gonna keep that bet.
And DARPA's shown interest inthis space.
Yeah, which means they thinkthat it's going to um disrupt
things.
SPEAKER_01 (07:59):
And and I think there is there is significant
significantly more government uhinvolvement, whether it's the US
or Europe or China, right?
Significantly more governmentinvolvement than it used to be
with AI in its early days.
This is also a significantdifference, right?
Yeah uh that we that we see outthere.
(08:21):
Now, getting back to your toyour question, um I think at
least what I expect is to belike the next step in this uh
stepping function, uh will be anon-trivial uh improvement in
error correction and qubitstability that will not be
(08:42):
enough to run efficiently uh orto solve the shore problem,
right?
The shore algorithm, but it willopen the door for, let's say,
smaller problems with the codessmaller that would have a direct
impact in things likeoptimization or material science
or or things like that.
(09:04):
I think before shore, and Ifully agree with what you said,
Patrick, before we will have aquantum computer that will just
be able to get an RSA uhencryption and and break it,
before that we will see uhsmaller scale um improvements
that will start to have uh veryclear practical uh applications,
(09:27):
whether it's as I said, materialscience or or or anything else.
SPEAKER_03 (09:31):
Do you think that's well the material science one's
interesting?
Do you think that's enough tocapture because as you said,
LLMs were around for a while andnobody cared?
I I think there has to be somelike breakout where like you you
talked about the Fubinakomolecule.
Um, yeah.
Yeah, and that you know, that'swe we use two percent of one to
(09:54):
two percent of you of globalenergy to make fertilizers, and
if we can figure that moleculeout, that that could be rapidly
and dramatically reduced.
I think that could be somethingthat captures the imagination.
If if we had a breakthrough ofmaterial science that that
changed people's lives on aday-to-day basis, if we got
superconductors that worked,that were easy to make, cheap
(10:15):
materials, and that we couldreplace all of our wiring with
that worked at at roomtemperature or at ambient
temperatures, those are the kindof breakthroughs that may or may
not uh ever occur.
And but I think that's what itmight take for it to really kept
capture the zeitgeist.
SPEAKER_01 (10:31):
And I think there are other potential areas as
well uh which uh do notnecessarily need to be
exponential improvements.
Like just imagine if we couldcut down uh either the time or
the cost or both of traininglarge language models to half or
(10:54):
a quarter, right?
It it doesn't need to beexponentially like two at the
power of whatever, right?
But even if we can come up withsomething via quantum computing,
right, um addressing a part ofthe massive optimization
problem, which is training adeep neural network, right?
Uh, what if we can come up withsomething that will cut costs in
(11:17):
two or in four or in eight,right?
So I don't think it needs to beuh as spectacular as we kind of
all expect it to be.
Um, but if it has practical umapplications, and I think we
have enough problems today, andwe are starting to get better at
(11:39):
how to frame those problems forquantum systems, that the next
breakthrough will actually behey, it's not gonna cost us a
hundred days to train a model,but we can do it in 40 days, uh,
or things like that.
This is just another anotherexample.
So, because uh you know what wesee more and more and more is
(12:02):
there is no such thing as a purequantum thing, or uh quantum
computing does not live in abubble that is completely
disconnected from whatever it'saround it.
On the contrary, all theapplications that we know of
today, even the ones that areonly theoretical, like shore,
they involve a combination ofquantum computing and classical
(12:25):
computing.
Right.
And I think the synergy is uhalso going to help the
advancement uh quite a lot.
SPEAKER_03 (12:33):
Interesting point.
So when we talked to Murray Tomfrom D-Wave, he pointed out that
there were optimization problemsthat they could solve that could
be solved on a regular computer,but they could solve them at
lower energy costs.
And and and with the you know,the talking about energy with
with AI is, you know, we'retalking about percentages of our
(12:57):
energy infrastructure beingdedicated to that.
So there might be so I could seewhere if we had some companies
like Anthropic and OpenAIsaying, well, we've been able to
reduce our uh our our datacenter energy cost by 50%
because we're using quantum umoptimizations instead of
classical optimizations for thissmall.
(13:19):
Again, Shores, the the lessonlearned from Shores is you don't
have to solve the whole problem.
You just gotta solve the hardestpart.
You just gotta solve the partthat nobody thought you could
solve.
And if we can do that withoptimization, now I know there's
you know lots of tensorcalculations and there's there's
so many things going on in in AIthese days.
(13:42):
But it would be interesting tosee if you think that that's a
likely scenario.
Is it likely that we're gonnafind things that are necessary
in inference engines and in coin training quantum computing,
uh not quantum computing, in inAI where that kind of
optimization would be a abreakthrough?
SPEAKER_01 (14:01):
I think it's likely.
SPEAKER_03 (14:03):
Okay.
SPEAKER_01 (14:03):
I think it's it's it's it's likely.
Um and that's because the scaleof these problems keeps uh going
up, right?
Yeah uh the more powerful modelswe need and the more powerful
models we want to train, themore problematic becomes the
optimization uh uh part.
But again, it might not be AI atall, right?
(14:27):
It might be a different fieldwhere having a large enough
number of qubits, which is stillnot in the hundreds of
thousands, right?
Or tens of thousands, but maybelet's say in the thousands,
having like a couple of thousandstable qubits with with
universal quantum computing, Ithink will will open the door to
(14:49):
some unbelievable advancements.
SPEAKER_03 (14:51):
So I have a theory, and again, this is you know,
prognostication's a delicateart.
Um, I believe that the boom thatwe're seeing with AI will
facilitate the next boom, whichI don't think is going to be
quantum.
I think there's gonna be anintervening one in robotics.
So as I think as of right now,last I heard, there's about two
million industrial robots in usein China, and there's another
two million in the West (15:14):
Japan, Australia, or Europe, the United
States each have about half amillion, give or take.
So there's about four millionindustrial robots being used
right now.
And there's a big push for humanhumanoid robots.
I'm not sure if that's gonna gowhere people think it's gonna
go.
But but the robotics revolutionseems to be the thing that's
(15:35):
enabled.
And I'll I'll do one more seguehere, one more.
There's a company called TALAS,T-A-A-L-A-S, that took that made
a chip and they put the wholeLLM, 8 billion parameter LLM on
that chip.
And if you go to their homepage,they have a link to the chatbot
they've powered with this chipcalled Chat Jimmy, and it serves
(15:58):
up 15,000 tokens a second, whichis mind-boggling more than it's
way, way more than most otherengines can can accommodate over
the web.
Um, and the idea is if I can puta an LLM in a chip, I can get
rid of some of that delay, andthat enables robotic solutions,
(16:19):
real world solutions.
And I could see where sensors,quantum sensors, start taking an
outside role, outsized role inenabling the robotics
revolution.
Because you know, we've talkedto several people about sensors,
not as many as we'd like, but Icould see where we'd have
fundamental problems in therobotics space, where you need a
(16:43):
fast model.
Well, putting the putting awhole model on a chip, that
seems to be solved.
That's not a quantum problem.
But having the sensing so that arobot can go into a mine and do
the mining instead of a personhaving to go down in the mine
and do the mining.
Remember, we talked to uh acompany that does the magnetic
sensoring, and they talked aboutartificial diamonds with you
(17:04):
know vacancies.
I could see something like thatcapturing the imagination to un
for people to understand, wow,this wouldn't, these robotics
revolution wouldn't be possiblewithout quantum.
And so I think it might come inone of these side angles where
it captures the the uh you knowGPS.
A lot of people didn't, theymaybe don't realize that that
(17:26):
you know quantum mechanics isone of the things that makes GPS
possible.
Um but I'm wondering if that'show it would seep into the into
the uh the the consciousness ofthe business community and
people, uh other people.
It's definitely gonna have animpact.
And I think it's gonna be AIrevolution, robotic revolution,
probably ongoing at the sametime, and then quantum
revolution in sometime in thenext decade.
SPEAKER_01 (17:49):
Yeah, I think it's very likely that that things
will will evolve, will evolvethis way.
Um definitely with robotics, nowwe are we're starting to see
massive advancements in thetechnologies used to help the
robots learn.
That's that that was one of thethings, right?
(18:09):
I know, for example, Tesla hasthis project where they are
planning to put several hundredrobots in in factories.
And the interesting thing isactually these robots are
matched so that whatever onerobot learns to do, which
essentially amounts to traininga bunch of deep neural networks
at the end of the day, right?
(18:30):
It will instantly share with allthe others.
So think of it like one needs tolearn how to pick up uh
something with its uh left hand,the other one has to learn how
to pick up something with itsright hand, and they can share,
right?
So that will cut down approacheslike this, will significantly
cut down the time needed forthese machines to uh learn right
(18:53):
there their their things.
SPEAKER_03 (18:54):
So that's that was Waymo's kind of narrative when
they said, well, for every everyone of our cars gets an hour of
driving time, they all benefitfrom the learning.
And there's also some people umyou know getting into the uh I
don't want to go stray too muchinto AI, but but one of the
things um I also wanted tohighlight is have you heard of
Eagle Eye from Andorrill, U.S.
government contractor?
(19:15):
Lucky Palmer?
Lucky Palmer, the guy that cameup with um the uh Oculus.
Oh, yeah, he's one of thefounders.
They have this goggle set for USArmy uh soldiers where if you
can see a bad guy, then I canhave it rendered on my vision in
(19:37):
AR, augmented reality, so I knowwhere they are behind the
building, even though I can'tsee them.
SPEAKER_01 (19:44):
Yeah.
SPEAKER_03 (19:45):
Because someone else has eyes on them, or some drone
has eyes on them.
So it's that shared experience.
So I could see you know a sensorbeing deployed in that regard so
that you could kind of figureout what a building is made of.
So you could decide, you know,what's safe to do there, what's
(20:06):
not safe to do there.
I mean, the the the comp thepossibilities are endless.
And and unfortunately, I thinkthe military might be the place
where a lot of thesetechnologies get their start.
Just you know, NASA pushed thethe computer chip, uh, it's
usually government efforts.
SPEAKER_01 (20:20):
And the PowerPoint pushed the internet.
SPEAKER_03 (20:22):
Exactly, yeah, yeah, and autonomous vehicles and
everything else.
Um so I think sensors issomeplace that I'd like to you
know understand better who'splaying there, what they're
doing, uh, because I think thatcould have an outside role uh in
getting us to that moment.
Now, once we get to that moment,one of the things that's gonna
be a scramble is figuring outhow quantum will disrupt each
(20:47):
and every market and and facetof life.
And right now, I don't thinkwe're prepared for that.
Most people, again, haven't eventhought about what this
technology is or or why itexists.
But AI is now disrupting partsof the economy that we didn't
imagine even two years ago.
And so I think it's yeah.
SPEAKER_01 (21:07):
That's something that if you look historically,
right, that has happened withwith every major breakthrough
technology, right?
Yeah, you had some types ofdisruptions that you were
expecting, and then others thatyou were literally not
expecting.
Yeah.
Um, I'm just thinking about anegative one, which was the
(21:27):
dot-com bubble with theinternet, right?
SPEAKER_03 (21:30):
But still, the internet didn't go away after
that, even though it was.
SPEAKER_01 (21:32):
Of course.
Yeah.
Yeah, but it was an interestingside effect, right, of the the
growing popularity and theenthusiasm.
SPEAKER_03 (21:39):
Yeah.
And unfortunately, things likecorner bookstores have have
fallen into uh single digit umuh numbers as opposed to where
they were before.
SPEAKER_01 (21:48):
So I I I and let's also not forget about quantum
communications, right?
Because typically when we talkabout quantum, we need to think
about the computing, we need totalk, we need to think about The
communications part, we need tothink about the sensing part,
right?
Because all these are kind ofdifferent aspects uh uh around
(22:08):
the the development of the ofthe technologies.
So and we see advancements onall fronts.
That's that's the interestingthing from from from my
perspective.
And then the other thing that Iwant to to highlight is uh it
hasn't necessarily has itdoesn't necessarily have to be a
(22:29):
uh major step in terms of say uhwe've increased 10 times the
number of qubits, right?
And and and things like that.
It also can be uh a majoradvancement in terms of an idea,
right?
And if you ask me, uh one of themost important ideas that uh
(22:50):
occurred in the past few yearswas uh kind of thinking outside
the box and implementing thisidea of moving qubits instead of
trying to use static topologiesfor building out gates,
especially two-qubit gates,right?
The the core idea to move qubitstogether to create uh uh
(23:12):
two-qubit gates, right?
That in itself, for me, is agood example of a major step in
that stepping function, becausethat idea enabled in a
relatively short amount of timesignificant improvements in
terms of error correction andeven uh the the way that that
more complex gates liketwo-qubit gates can be
(23:35):
implemented.
Uh, and it also I think openeduh the the road towards um
increasing the temperature at atwhich some of these chips have
to have to run, right?
Um allowing us to move out ofthose crazy uh 30, 20, 30, 40
(23:57):
millikelvin temperatures tothings that are let's say at
least more more levels that areat least more manageable.
SPEAKER_04 (24:05):
Yeah.
SPEAKER_01 (24:05):
So that's that's the other thing that that and
remember there were times at thebeginning where we were like,
hey, we we don't see uh many newalgorithms, like we don't see a
new shore, we don't see a newgrover, and and so forth.
Well, looking back to these fiveyears, right, for me it's uh
(24:27):
interesting that it wasn't analgorithm, but it was an idea,
yeah, right, in in how toaddress a potential uh uh
physical problem, which is howdo we build the two qubit gigs.
SPEAKER_03 (24:39):
Well, we were focused in those early days,
early, early year, actually.
Um I was focused more onsuperconducting and photonics.
I those were the modalities thatI had heard of that I knew of.
SPEAKER_01 (24:52):
And trap ions, trap ions because of their legacy in
atomic clocks and and other kindof previous developments.
SPEAKER_03 (25:00):
I don't think the players had really fleshed
themselves out in that space.
So I I they looked likeinteresting potentials at that
time.
And then D-Wave with the um withtheir uh Adiabatic app.
Um now D Wave is still activeand and doing solving problems.
(25:21):
They've got the Cubo system andthey're they're actively trying
to help companies figure out howto use this optimization.
Um but they're also looking atdoing uh universal quantum
computing, from what I've heard.
And maybe they're seeing youknow the Yeah.
So the yeah, the the algorithmthing didn't really go the way
(25:42):
we thought.
I I really thought that we well,there'd be we had Shores and we
had drovers, and we'll have youknow, Cyprians and Patrick's,
you know, there'll be other uhsystems.
But we also kind of explored thefact that there might not be as
many really fundamentally hardproblems to tackle, right?
(26:05):
That we that was also the thingthat was kind of surprising is
we we're looking around, well,what kind of problems could be
solved?
Well, what kind of problems arethere right that in the in that
world?
And again, I think AI androbotics are introducing a whole
new set of problems wherequantum might be um you know a
valuable uh thing to do, avaluable place.
SPEAKER_01 (26:27):
And and to that point, Patrick, I think one
other potential kind of uhmassive leap forward can be
exactly in this area, uh, abetter understanding of how to
model existing problems into thequantum world and into the
quantum space.
Because this is one of thethings that that are still
(26:50):
relatively difficult because youdon't really have the same
primitives with quantumcomputing like you have with
classical.
You can't really store a lot ofinformation in the classical uh
uh kind of sense of the uh ofthe word, right?
Uh you also don't really havelong term, and really when we
say long-term in the world ofquantum, we're uh uh really
(27:13):
referring to seconds.
That's essentially long-term inin quantum, right?
So there is a a non-trivial umlevel of difficulty in taking a
problem uh that is wellunderstood in the practical
world and is well defined, andessentially morphing it into a
(27:33):
representation that could bevalid or uh can can work with
quantum computers.
I think there's a lot to improvethere, and this could be another
area where we could seeimportant breakthroughs uh in in
the coming years.
SPEAKER_03 (27:48):
Yeah.
I I'm not predicting um thatquantum computing becomes like
let's the fi world on fire andmainstream before the end of
this decade.
I I don't I don't see that as asin the cards.
I may be wrong.
Maybe maybe I'm completelywrong.
SPEAKER_01 (28:03):
I share I I share your view, yeah.
SPEAKER_03 (28:05):
Yeah, I think but I think we're making consistent
steps forward and filling inthese gaps.
Um, you know, the wild card forme on chores, just to pick one,
because I think that one is aguaranteed.
If if if suddenly RSA 2048 isendangered by quantum computers,
that's gonna be something that'sgonna get noticed in in
(28:27):
instantly.
Um when talking to these some ofthese companies that are making
efforts to chain together to doa distributed computing approach
to quantum computing.
Well, now we don't need a a5,000 qubit quantum computer.
We need a hundred fifty qubitquantum computers.
And that's a much easier thingto imagine us cobbling together
(28:51):
quickly than waiting for thereto be a 5,000 logical qubit
quantum computer.
I don't know if that's gonna bethe way it plays out, but that
feels like where we're gonna seea sudden step function occur.
SPEAKER_01 (29:04):
Yeah, yeah.
I mean, think about the earlydays of our podcast, right?
Um chips with five qubits orchips with with eight qubits
were really exciting, yeah,right?
Yeah uh uh back then.
And now we have chips with what,48, 50.
SPEAKER_03 (29:22):
Well, we were talking about this before there
were any qubits where they werejust theoretical.
SPEAKER_01 (29:28):
Of course, yeah, yeah.
So uh things are are improving,right?
Things are are are improving.
Um and uh again, um becausethere is this continuity in
terms of funding, right?
This is why I believe um, atleast in this kind of first part
of its history, like the firstuh I don't know, 20, 30 years of
(29:51):
its history, quantum computingwill have a smoother ride uh uh
than uh AI had or or or theother the other technologies.
Um now I also think that thatanother um um important aspect
is going to also be the impactof this almost uh unexpected
(30:19):
success of AI, right?
Because on one hand, I think ithelps, but on the other hand,
now with AI, we can alsocontinue the advancements in
classical computing, right?
We can essentially design andimplement much uh better data
(30:41):
centers, uh much better uhcomputing grids that are based
on GPUs and and other and otherthings.
So I believe there is also uh asteep development in classical
computing, which I think isstarting to also chip into some
of the problems, especiallyaround optimization problems
(31:05):
that were kind of deemed to bethe exclusive realm of quantum
quantum computing.
Right, yeah, it's it's the neverendless a never-ending
discussion about quantumsupremacy where the quantum side
says, okay, here's somethingthat we figured out, right?
We calculated, and then a fewmonths later there's a classical
(31:26):
computing team coming, sure.
Look, we use this supercomputer.
Hold my computer, we uhsignificantly optimized and
improved this and that and that,and now look, you you don't
really prove quantum supremacyyet.
So it's an interesting kind ofrace because classical
computing, right, even if itsuffers, let's say, from an
(31:49):
exponential uh disadvantagecompared to quantum, but
classical is also marchingahead.
And uh Well, I think that's abenefit.
SPEAKER_03 (32:00):
I think I I remember back in the old days when there
was when there's more than oneweb browser that is a
reasonable, optimal, you know,even optimal choice, they both
get better faster.
Right?
When Internet Explorer had realcompetition, we saw better
browsers overall for everyonebecause there was there was
(32:22):
push.
Um and but when we have amonoculture, everybody sits back
and says, well, you know, thingsare as good as they're gonna
get.
And and to that, I do believethat supercomputing is being
pushed by quantum and quantum isbeing pushed by supercomputing.
Eventually, I do believe thatquantum will find things that it
can do and do better, but Ithink it's helping quantum as
(32:43):
well.
It's putting a realism on it,it's preventing a quantum hype
bubble, which is dangerousbecause we're gonna have a hype
bubble with AI.
Some would argue that we'realready like dangerously in it
and it's actually close tobreaking.
Others would say, you know, wehave plenty of space to go.
I think the robotics um pushwill actually give it a lot more
(33:04):
runway than people are giving itcredit because I think AI is the
driver of robotics in this case.
With the quantum, the longer wepush off that hype bubble, then
the the further we'll get beforewe get a winter.
We'll get a winter with quantumeventually.
But it may be 10 years from now.
It may be after it it becomeslike you know a household name
(33:27):
and everybody's trying to becomea quantum engineer, and
hopefully everybody's listeningto old entangled thanks
episodes.
And um, but I honestly I thinkthat's a good thing because
moving the goalposts just makesyou run faster.
SPEAKER_01 (33:43):
Yeah, yeah.
And then there's another thing,Patrick.
Um, since we're speaking aboutboth AI and quantum, I think the
advancements in AI, especiallyaround large language models and
even more especially aroundcoding agents, are helping
tremendously the learning curvefor quantum computing related
(34:06):
things, whether it's quantumcomputing related pro uh
languages or uh approaches tooptimize designs of quantum
computing gates or whatever,right?
It's now significantly easier tolearn about this stuff.
And I think that's gonnacontinue for at least the next
(34:29):
uh the next few few few years.
Because uh if you talk, I don'tknow, let's say a random
example, Q sharp, right?
The uh programming languagedeveloped by by Microsoft for
for quantum computing, right?
Today it's really trivial to goask a large language model to
give you an example of like howto do, I don't know, a quantum
(34:50):
Fourier transformation or Idon't know, uh the uh an i uh an
ising optimization, you name it,right?
Yeah, and it's it's a resourcethat especially the younger
generation has, which makes theentry level significantly lower
(35:12):
than it used to be even likefive years ago or or ten years
ago.
And remember, we always talkabout the fact that the real
kind of explosive developmentwill probably happen when the
what we call the quantum-borngeneration is going to be the
one driving the field, where youwill have people that were not
(35:35):
uh impacted either in a positiveor negative way, that's up to
debate, but they were notimpacted by previous uh uh
learning of classical computing.
SPEAKER_03 (35:47):
So those of us that are old enough to have gone to
high school with Newton is that.
SPEAKER_01 (35:51):
Right?
Paradigm and uh who are going tobe people that the only
computing that they ever learnedor knew, right, was the the the
quantum computing uh uhapproach.
And then for the rest of us,obviously, uh lowering the the
uh kind of the the friction inentering the the this particular
(36:15):
field, making it easier to learnthings, making it easier to get
examples of things, making iteasier to get explanations of
various types of things.
Uh it's gonna help us well uh ina very, very significant way, if
you ask me.
SPEAKER_03 (36:32):
Yeah, I think I I mean we're over we're we're
getting over time here.
I usually I think, oh, it's justme and Cyprian talking, it'll be
quick.
Nope, never.
Uh so uh, you know, there'sstill a lot to talk about here,
um, but I think we need to drawto a close soon.
That said, I think what you'resaying is is very valid.
(36:52):
And one of the things a lot ofpeople are searching for is
well, how do I stay relevant ina world where AI can write code,
it can, you know, writesymphonies, it can do all sorts
of things.
Well, the the the there's stilla a border of knowledge.
And while I think AI can come upwith some novel solutions to
some problems, invent drugs, dosame things in material science,
(37:14):
we still need people to figureout how to do these novel
approaches because a largelanguage model is only as good
as the data you put into it.
SPEAKER_01 (37:23):
Exactly.
SPEAKER_03 (37:23):
And so this is the these frontier sciences, these
frontier you know things.
Um some of the things I look at,I'm just amazed at.
Well, somebody thought to use adiamond.
I keep coming back to thatexample of a flawed diamond for
detecting magnetic.
SPEAKER_01 (37:39):
Isn't that great?
SPEAKER_03 (37:40):
Like it is, it is.
It was one of my favoriteexamples.
We need to talk to more of theseuh these sensor people.
And when the when I'll admit,when when they first started
talking about quantum sensors,I'm like, what are you talking
about?
What what's and now I get it.
I get the fact that you know weuh they're sensitive and they're
small and they're and they'regonna have an impact.
So um I'm very excited for thenext five years.
(38:01):
Um I hope we can keep uh keepeveryone entertained and
informed.
And I think we're still on thesame mission is is we think this
is an important technology forscience, humanity, and
everything else.
And we want to make sure otherpeople are uh uh come along for
the ride.
SPEAKER_01 (38:19):
Absolutely.
Absolutely.
That's it's uh gonna be anexciting uh uh five years to
follow, for sure.
SPEAKER_02 (38:27):
Imagine saying that after this last two years.
That's amazing.
All right.
So thanks, everyone.
SPEAKER_00 (38:34):
Thank you.
Thanks, everyone.
Bye.
Bye.
Cybercrime is one of the biggestthreats to businesses of all
sizes and industries.
With almost half a million opencyber positions, the problem is
compounded by the lack ofavailable talent in the
marketplace.
At Pulsar Security, our eliteteam of highly credentialed
experts collaborate with you toassess your current defenses and
(38:56):
develop solutions tailored toyour specific needs.
With services ranging fromcybersecurity education to
advanced penetration testing andred teaming, you can start
reducing your risks today.
Visit pulsarsecurity.com andlet's secure your digital future
together.
Popular Podcasts
Betrayal Weekly
Betrayal Weekly is back for a new season. Every Thursday, Betrayal Weekly shares first-hand accounts of broken trust, shocking deceptions, and the trail of destruction they leave behind. Hosted by Andrea Gunning, this weekly ongoing series digs into real-life stories of betrayal and the aftermath. From stories of double lives to dark discoveries, these are cautionary tales and accounts of resilience against all odds. From the producers of the critically acclaimed Betrayal series, Betrayal Weekly drops new episodes every Thursday. If you would like to share your story, you can reach out to the Betrayal Team by emailing them at betrayalpod@gmail.com and follow us on Instagram at @betrayalpod and @glasspodcasts. Please join our Substack for additional exclusive content, curated book recommendations, and community discussions. Sign up FREE by clicking this link Beyond Betrayal Substack. Join our community dedicated to truth, resilience, and healing. Your voice matters! Be a part of our Betrayal journey on Substack.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by Audiochuck Media Company.
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.