Latest / The Quark Side - Quantum Physics Podcast / Quantum Mechanics Is Entering a New Era of Real-World Applications

Transcript

Quantum Mechanics Is Entering a New Era of Real-World Applications · The Quark Side - Quantum Physics Podcast

• 0:01
  Speaker 1: Welcome to the Quarkside Quantum Physics Podcast, an exploration of
• 0:05
  Speaker 1: the fundamental structure of reality, where quantum laws govern matter, energy,
• 0:11
  Speaker 1: and information. Here, uncertainty is a feature, not a flaw,
• 0:16
  Speaker 1: and understanding begins at the smallest scales.
• 0:27
  Speaker 2: I want you to imagine for just a minute that
• 0:29
  Speaker 2: you've hired this wooden class Michelin star chef to cook
• 0:34
  Speaker 2: dinner right in your own kitchen.
• 0:36
  Speaker 3: Oh, that sounds amazing.
• 0:37
  Speaker 2: Right. Like we are talking about a total culinary genius here,
• 0:41
  Speaker 2: someone whose time is incredibly valuable. So you're obviously expecting
• 0:45
  Speaker 2: an absolute masterpiece.
• 0:46
  Speaker 3: Naturally.
• 0:47
  Speaker 2: But there is a catch to how this specific chef operates.
• 0:51
  Speaker 2: Out of every single hour they spend in your kitchen,
• 0:55
  Speaker 2: they spend exactly 55 minutes just like wiping down the countertops,
• 1:00
  Speaker 2: painstakingly sharpening their knives, maybe calibrating your oven temperature to
• 1:04
  Speaker 2: the thousandth of a degree.
• 1:06
  Speaker 3: Just endless prep.
• 1:07
  Speaker 2: Exactly. Perfectly aligned in the cutting board. And they only
• 1:11
  Speaker 2: spend five minutes actually cooking the food.
• 1:13
  Speaker 3: Yeah. I mean, that would be an incredibly frustrating wait
• 1:15
  Speaker 3: for dinner.
• 1:16
  Speaker 2: It really would.
• 1:16
  Speaker 3: But the thing is, that chef isn't actually being lazy
• 1:19
  Speaker 3: or difficult. What you're missing in that scenario is that
• 1:23
  Speaker 3: the chef is trying to cook inside a kitchen where say,
• 1:26
  Speaker 3: a slight change in humidity will cause the ingredients to
• 1:30
  Speaker 3: literally vaporize.
• 1:31
  Speaker 2: Oh, wow.
• 1:31
  Speaker 1: Okay.
• 1:31
  Speaker 2: All right.
• 1:32
  Speaker 3: So the overwhelming majority of their time isn't just prep.
• 1:35
  Speaker 3: It's this active, desperate battle to keep the environment from
• 1:39
  Speaker 3: completely destroying the meal before it even hits the plate.
• 1:43
  Speaker 2: Which brings us perfectly to the core of what we
• 1:45
  Speaker 2: are tackling today. Because that mind-boggling ratio of, you know,
• 1:50
  Speaker 2: preparation and environmental control to actual productive work That is
• 1:54
  Speaker 2: the exact state of cutting-edge technology right now in the
• 1:58
  Speaker 2: realm of quantum computing.
• 1:59
  Speaker 3: It really is.
• 2:00
  Speaker 2: In engineering terms, we are talking about the quantum duty cycle.
• 2:04
  Speaker 2: And it is the single biggest bottleneck standing between us
• 2:08
  Speaker 2: and this futuristic computing era we just keep hearing about.
• 2:11
  Speaker 3: Yeah, the era that's always just around the corner.
• 2:13
  Speaker 2: Exactly. So the goal today is to look under the
• 2:16
  Speaker 2: hood of these multimillion dollar machines, figure out why their
• 2:19
  Speaker 2: time management is so notoriously terrible, and look at the
• 2:23
  Speaker 2: actual physics of how scientists are attempting to fix it.
• 2:26
  Speaker 3: Because it is the defining engineering challenge of our time.
• 2:30
  Speaker 3: I mean, when people think about quantum computing, your mind
• 2:33
  Speaker 3: naturally drifts toward the theoretical potential, right? Factoring massive primes,
• 2:38
  Speaker 3: simulating molecular structures, cracking encryption.
• 2:41
  Speaker 2: All the sci-fi stuff.
• 2:43
  Speaker 3: Exactly. But none of that theoretical math actually matters if
• 2:46
  Speaker 3: the physical machine spends 99 percent of its operational time
• 2:51
  Speaker 3: just trying to keep its own components from falling apart.
• 2:54
  Speaker 2: Right.
• 2:54
  Speaker 3: So understanding the mechanics of the duty cycle, it cuts
• 2:58
  Speaker 3: through the hype. It shows you exactly where the bleeding
• 3:00
  Speaker 3: edge of the technology actually sits right now.
• 3:02
  Speaker 1: Yeah.
• 3:03
  Speaker 2: And to really appreciate the severity of this quantum duty
• 3:06
  Speaker 2: cycle problem, I think we should benchmark it against the
• 3:09
  Speaker 2: classical hardware you're likely using right now to listen to this.
• 3:12
  Speaker 3: That's a great comparison.
• 3:13
  Speaker 2: Because we understand the concept of a duty cycle in
• 3:16
  Speaker 2: everyday electronics. It's just the ratio of time a system
• 3:20
  Speaker 2: is actively performing its task versus the time it needs
• 3:23
  Speaker 2: to rest, reset, or cool down. Like an air conditioner
• 3:27
  Speaker 2: turning on and off. But a standard silicon microchip, like
• 3:30
  Speaker 2: the processor in your smartphone, has a phenomenal duty cycle.
• 3:35
  Speaker 2: I mean, it is virtually 100%. It computes constantly. It
• 3:38
  Speaker 2: doesn't need to pause to rebuild its internal logic.
• 3:41
  Speaker 3: If we connect this to the bigger picture, the reason
• 3:44
  Speaker 3: your smartphone processor can boast that near 100% duty cycle
• 3:48
  Speaker 3: basically comes down to macroscopic physics. Okay.
• 3:51
  Speaker 2: Macroscopic meaning big stuff.
• 3:52
  Speaker 1: Right.
• 3:53
  Speaker 3: A classical bit in your silicon chip is represented by
• 3:55
  Speaker 3: a voltage threshold. It takes thousands or even millions of
• 4:00
  Speaker 3: actual electrons to flip a transistor from a zero to
• 4:03
  Speaker 3: a one.
• 4:03
  Speaker 2: So it's a brute force kind of thing.
• 4:05
  Speaker 3: Exactly. Because you are dealing with such a massive cluster
• 4:08
  Speaker 3: of particles... There is this huge energy barrier protecting that
• 4:12
  Speaker 3: piece of information.
• 4:13
  Speaker 2: Which means it's safe.
• 4:14
  Speaker 3: Yes. A stray photon bouncing around or a tiny temperature
• 4:18
  Speaker 3: fluctuation doesn't have nearly enough energy to flip a million
• 4:20
  Speaker 3: electrons by accident. The classical chip is robust by its
• 4:24
  Speaker 3: very nature.
• 4:25
  Speaker 2: Okay, let's unpack this. Because this is where the paradox
• 4:29
  Speaker 2: really presents itself. We have these classical chips that cost
• 4:32
  Speaker 2: 50 bucks and work tirelessly without a single break.
• 4:35
  Speaker 3: Yeah.
• 4:35
  Speaker 2: But on the other hand, we have state-of-the-art quantum processors
• 4:39
  Speaker 2: housed in these massive dilution refrigerators, completely shielded from magnetic field.
• 4:45
  Speaker 3: And cooled down to 15 millikelvin.
• 4:47
  Speaker 2: Right, which is colder than deep space. So we have
• 4:50
  Speaker 2: supposedly removed all the thermal energy, isolated the system completely.
• 4:55
  Speaker 2: If the environment is so perfectly controlled, why is the
• 4:58
  Speaker 2: quantum processor constantly dropping the ball? I mean, why can't
• 5:01
  Speaker 2: it just calculate continuously like my laptop?
• 5:05
  Speaker 3: Well, because removing the thermal energy doesn't remove all the noise.
• 5:08
  Speaker 2: Yeah.
• 5:09
  Speaker 3: And quantum states are fragile on a level that is
• 5:11
  Speaker 3: honestly difficult to intuitively grasp.
• 5:14
  Speaker 2: They're like a house of cards, right?
• 5:15
  Speaker 3: Very much so. You mentioned the dilution refrigerators. We cool
• 5:18
  Speaker 3: these superconducting quibbets down to 15 millikelvins, sure, but they're
• 5:22
  Speaker 3: still coupled to control wires, wires that have to send
• 5:25
  Speaker 3: microwave pulses to manipulate them.
• 5:27
  Speaker 2: Oh, to actually tell them what to do.
• 5:28
  Speaker 3: Right. And those wires introduce microscopic amounts of heat. Plus,
• 5:33
  Speaker 3: there is magnetic flux noise. There are even cosmic rays
• 5:36
  Speaker 3: passing through the Earth that occasionally strike the processor.
• 5:38
  Speaker 2: Wait, cosmic rays? Seriously?
• 5:41
  Speaker 3: Seriously. And even if you could magically block all of
• 5:43
  Speaker 3: that out, you still have vacuum fluctuations, which is just
• 5:47
  Speaker 3: the inherent jitter of the quantum field itself.
• 5:51
  Speaker 2: And any one of those things causes decoherence.
• 5:54
  Speaker 3: Precisely. A qubit holding a superposition, meaning it's existing as
• 5:58
  Speaker 3: this complex probability of both 0, and 1 will collapse
• 6:01
  Speaker 3: the moment it interacts with any uncontrolled variable in the environment.
• 6:04
  Speaker 2: It's just instantly ruined.
• 6:06
  Speaker 3: Yes. We measure this fragility using what we call coherence times,
• 6:10
  Speaker 3: specifically T1 relaxation and T2 dephasing times. Okay.
• 6:14
  Speaker 2: And how long are those times?
• 6:15
  Speaker 3: In the best superconnecting qubits right now, those coherence times
• 6:19
  Speaker 3: are measured in microseconds.
• 6:20
  Speaker 2: Microseconds.
• 6:21
  Speaker 1: Yeah.
• 6:21
  Speaker 3: Maybe a few hundred microseconds if you have like a
• 6:24
  Speaker 3: world record-breaking architecture.
• 6:25
  Speaker 2: So let's do the math on that. If your quantum processor...
• 6:29
  Speaker 2: only holds its memory for, say, 200 microseconds before the
• 6:33
  Speaker 2: environment just forces it to collapse, that is your absolute
• 6:37
  Speaker 2: hard limit for doing any useful work.
• 6:39
  Speaker 3: That is your window.
• 6:40
  Speaker 1: Yep.
• 6:41
  Speaker 3: And while quantum gate operations, the actual math steps, are
• 6:45
  Speaker 3: very fast, often taking just tens of nanoseconds, the initialization
• 6:50
  Speaker 3: of the quibbets, the readout of the results, and the
• 6:52
  Speaker 3: resetting of the whole system take orders of magnitude longer.
• 6:55
  Speaker 2: So you're spending all your time setting up the shot.
• 6:58
  Speaker 3: Right. You might spend... Thousands of nanoseconds just preparing the
• 7:01
  Speaker 3: qubit to do 10 nanoseconds of actual calculation, only for
• 7:05
  Speaker 3: it to collapse immediately after anyway. That is why your
• 7:08
  Speaker 3: duty cycle just plummets.
• 7:09
  Speaker 2: It's like having an office worker who can process an
• 7:12
  Speaker 2: entire massive spreadsheet in five seconds, but they have complete
• 7:15
  Speaker 2: amnesia and you have to spend an hour reteaching them
• 7:18
  Speaker 2: how to read before every single task.
• 7:20
  Speaker 3: That's a very fair way to look at it. But
• 7:21
  Speaker 3: the time spent on initialization readout is actually just the
• 7:25
  Speaker 3: tip of the iceberg here. Oh, really? Yeah. The real
• 7:28
  Speaker 3: killer of the quantum duty cycle. The thing that absorbs
• 7:30
  Speaker 3: almost all of the machine's processing power is the attempt
• 7:34
  Speaker 3: to fight back against that decoherence. We know the errors
• 7:39
  Speaker 3: are going to happen.
• 7:39
  Speaker 2: Probably because of the cosmic rays and the heat and
• 7:41
  Speaker 2: all that.
• 7:41
  Speaker 3: Exactly. We know a stray microwave photon is going to
• 7:45
  Speaker 3: flip a quibbit when we aren't looking. So we have
• 7:48
  Speaker 3: to employ error correction.
• 7:50
  Speaker 2: Here's where it gets really interesting. Because the solution to
• 7:53
  Speaker 2: the fragility problem is... seems to create this massive administrative paradox.
• 7:57
  Speaker 3: It does.
• 7:58
  Speaker 2: In classical computing, if you want to make sure a
• 8:00
  Speaker 2: bit didn't flip by accident, you just copy it, right?
• 8:03
  Speaker 2: You make three copies, and if one flips, the other
• 8:05
  Speaker 2: two outvote it. Simple.
• 8:07
  Speaker 3: Very simple.
• 8:08
  Speaker 2: But in quantum mechanics, the no-cloning theorem says you cannot
• 8:11
  Speaker 2: perfectly copy an unknown quantum state.
• 8:13
  Speaker 3: Right.
• 8:14
  Speaker 2: And even worse, if you just look at the quibbit
• 8:16
  Speaker 2: to see if it made a mistake, The mere act
• 8:19
  Speaker 2: of measuring it destroys the superposition. It is quite literally
• 8:23
  Speaker 2: like trying to check if a bomb is diffused by
• 8:25
  Speaker 2: hitting it with a hammer.
• 8:26
  Speaker 3: Yeah, the observation itself forces the collapse. You cannot just
• 8:29
  Speaker 3: read the date equipment to see if an error occurred.
• 8:32
  Speaker 2: Exactly. So I think of it like an incredibly bloated
• 8:34
  Speaker 2: corporate office. You have one worker doing the actual math.
• 8:37
  Speaker 2: But because this worker is notoriously unreliable, you hire 10
• 8:42
  Speaker 2: managers to surround the desk and constantly check their work.
• 8:45
  Speaker 3: Okay, I like this.
• 8:46
  Speaker 2: But the catch is, the managers aren't allowed to actually
• 8:49
  Speaker 2: look at the worker's computer screen, because if they do,
• 8:51
  Speaker 2: the computer explodes.
• 8:53
  Speaker 3: That analogy captures the total absurdity of the overhead. It
• 8:56
  Speaker 3: really does. But the physical mechanism is actually much more
• 8:59
  Speaker 3: elegant than managers just guessing in the dark.
• 9:03
  Speaker 2: Okay, how does it actually work, then?
• 9:05
  Speaker 3: Instead of directly observing the data quibbit, physicists use what
• 9:09
  Speaker 3: are called ancilla quibbits. which basically just means helper quibbets.
• 9:13
  Speaker 3: Think of it less like a corporate office and more
• 9:15
  Speaker 3: like a choir singing a very specific, complex chord.
• 9:19
  Speaker 2: A choir? Wait, how does that solve the whole observation problem?
• 9:22
  Speaker 3: Well, if you are the conductor of this choir and
• 9:26
  Speaker 3: you want to know if one specific singer in the
• 9:28
  Speaker 3: alto section went slightly flat, you don't need to stop
• 9:31
  Speaker 3: the whole performance and make that one person sing their
• 9:34
  Speaker 3: note solo.
• 9:35
  Speaker 2: Right.
• 9:35
  Speaker 3: That would ruin the song. Or, in our case, destroy
• 9:38
  Speaker 3: the superposition.
• 9:38
  Speaker 2: Got it.
• 9:39
  Speaker 3: So instead, you listen to the harmonic resonance of the
• 9:42
  Speaker 3: entire section. The friction between the right notes and the
• 9:45
  Speaker 3: wrong note creates a distinct acoustic beat. It's like a
• 9:49
  Speaker 3: wobble in the sound. You are measuring the relationship between
• 9:52
  Speaker 3: the notes, not the isolated notes themselves.
• 9:55
  Speaker 2: Oh. You are measuring the parity?
• 9:58
  Speaker 1: Yes.
• 9:59
  Speaker 3: Correct. we entangle our data qubit with an ancilla qubit.
• 10:03
  Speaker 3: We don't measure the data qubit. We measure the ancilla qubit.
• 10:06
  Speaker 2: Sneaky.
• 10:06
  Speaker 1: It is.
• 10:07
  Speaker 3: The math of the entanglement allows the ancilla to absorb
• 10:10
  Speaker 3: the error signature, like a bit flip or a phase flip,
• 10:14
  Speaker 3: and then report back on the parity of the system.
• 10:16
  Speaker 3: It tells the computer, hey, an error occurred in this neighborhood.
• 10:20
  Speaker 3: without ever revealing the actual value of the data qubit,
• 10:23
  Speaker 3: thereby preserving that delicate superposition.
• 10:25
  Speaker 2: That is brilliant physics. But going back to our duty
• 10:29
  Speaker 2: cycle problem, that sounds incredibly exhausting for the machine to
• 10:32
  Speaker 2: do constantly.
• 10:33
  Speaker 3: It is entirely exhausting. To build one single, reliable, error-corrected
• 10:39
  Speaker 3: logical qubit, you might need a thousand physical qubits.
• 10:42
  Speaker 2: A thousand.
• 10:43
  Speaker 3: Yes, one physical qubit is holding the actual data together.
• 10:46
  Speaker 3: And the other 999 are functioning as this vast choir
• 10:50
  Speaker 3: of ancilla kribbits, constantly entangling, measuring parity, and reporting back
• 10:56
  Speaker 3: literally millions of times per second.
• 10:58
  Speaker 2: So the processor is working incredibly hard generating all this data,
• 11:01
  Speaker 2: but 99.9% of that effort is just internal maintenance.
• 11:06
  Speaker 3: Exactly.
• 11:06
  Speaker 2: It is the chef frantically wiping down the countertops, stabilizing
• 11:10
  Speaker 2: the room temperature, and running parity checks on the humidity...
• 11:13
  Speaker 2: just to earn a tiny fraction of a second to
• 11:16
  Speaker 2: actually cook the meal.
• 11:17
  Speaker 3: What's fascinating here is how this managerial overhead scales up.
• 11:21
  Speaker 3: As you try to run more complex algorithms, algorithms that
• 11:24
  Speaker 3: require millions of gates to execute, the probability of an
• 11:27
  Speaker 3: uncorrected error just cascades.
• 11:29
  Speaker 2: So you need even more of those helper quibbets.
• 11:32
  Speaker 3: You need more ancilla quibbets. You need faster parity checks.
• 11:35
  Speaker 3: The machine becomes this massive error-correcting engine where the actual calculation,
• 11:39
  Speaker 3: the thing you bought the machine to do in the
• 11:41
  Speaker 3: first place, is squeezed into these microscopic slivers of time
• 11:45
  Speaker 3: between endless cycles of maintenance, the useful duty cycle drops
• 11:49
  Speaker 3: to near zero.
• 11:50
  Speaker 2: Wow. So that is the roadblock. We are totally bogged
• 11:54
  Speaker 2: down in the maintenance. We are. But as we look
• 11:56
  Speaker 2: toward the future, specifically targeting those major developmental milestones set
• 12:01
  Speaker 2: for August 2026 and beyond, How does the industry break
• 12:05
  Speaker 2: out of this cycle? Because, I mean, if we just
• 12:07
  Speaker 2: keep scaling up this brute force method, we are going
• 12:10
  Speaker 2: to need a quantum computer the size of a football
• 12:12
  Speaker 2: field just to simulate a simple molecule.
• 12:15
  Speaker 3: Yeah, brute force won't work. The path out of this
• 12:17
  Speaker 3: bottleneck requires a dual approach. We need fundamentally better hardware
• 12:22
  Speaker 3: and much smarter software architectures.
• 12:24
  Speaker 2: OK, let's start with hardware.
• 12:26
  Speaker 3: On the hardware side, the goal is to increase those
• 12:28
  Speaker 3: native coherence times before error correction is even applied.
• 12:32
  Speaker 2: Making the raw materials more resilient to the environment so
• 12:35
  Speaker 2: they don't break as fast?
• 12:36
  Speaker 3: Yes. For a long time, superconducting circuits relied heavily on
• 12:40
  Speaker 3: materials like niobium. But researchers have found that shifting to
• 12:44
  Speaker 3: materials like tantalum for the capacitors dramatically reduces dielectric loss.
• 12:49
  Speaker 2: Tantalum. What makes that better?
• 12:51
  Speaker 3: Tantalum has a different microscopic structure. It doesn't harbor the
• 12:55
  Speaker 3: same kind of defects that absorb microwave photons and cause decoherence.
• 12:59
  Speaker 3: By refining the metallurgy and the fabrication techniques at the
• 13:02
  Speaker 3: atomic level, we can push coherence times from 100 microseconds
• 13:06
  Speaker 3: up into the milliseconds.
• 13:07
  Speaker 2: Oh, wow. Which means the worker can focus on the
• 13:10
  Speaker 2: spreadsheet much longer before they need their memory refreshed.
• 13:13
  Speaker 3: Exactly. But even with a tantalum upgrade, they're still going
• 13:17
  Speaker 3: to make mistakes eventually. We still need the managers. We
• 13:19
  Speaker 3: still need the choir.
• 13:20
  Speaker 2: Right, the ancillary qubits. So what's the software fix?
• 13:24
  Speaker 3: And that is where the software revolution comes in. Specifically...
• 13:28
  Speaker 3: topological error correction, like the surface code. Right now, running
• 13:32
  Speaker 3: ancilla checks in a linear brute force way is incredibly inefficient.
• 13:37
  Speaker 3: The surface code reimagines the layout of the quibbits on
• 13:39
  Speaker 3: the physical chip.
• 13:40
  Speaker 2: It's like a checkerboard pattern, right?
• 13:42
  Speaker 3: Very much like a checkerboard. Imagine the white squares are
• 13:45
  Speaker 3: your data quibbits and the black squares are your ancilla
• 13:48
  Speaker 3: helper quibbits.
• 13:49
  Speaker 2: Okay, I'm picturing it.
• 13:50
  Speaker 3: In this 2D lattice, every data quibbit is entangled with
• 13:54
  Speaker 3: multiple neighbors. And every ancillary quibbit is constantly checking the
• 13:58
  Speaker 3: parity of the data quibbits immediately around it.
• 14:00
  Speaker 2: Oh, I see. So instead of a massive top-down hierarchy
• 14:04
  Speaker 2: of checking errors, you have this localized repeating grid where
• 14:10
  Speaker 2: errors are trapped and identified instantly by just the immediate neighbor.
• 14:14
  Speaker 3: Exactly. The surface code is brilliant because it is highly
• 14:17
  Speaker 3: tolerant of local faults. An error can pop up. But
• 14:20
  Speaker 3: the grid detects it and accounts for it in the
• 14:22
  Speaker 3: software tracking without having to stop and aggressively correct the
• 14:26
  Speaker 3: physical hardware in real time.
• 14:27
  Speaker 2: Which saves a ton of time.
• 14:29
  Speaker 3: It shrinks the required overhead massively. When you combine tantalum-based
• 14:33
  Speaker 3: hardware that naturally resists decoherence with a surface code architecture
• 14:37
  Speaker 3: that manages errors more efficiently, you dramatically reduce the time
• 14:41
  Speaker 3: the machine spends fighting itself.
• 14:42
  Speaker 2: It's a two-pronged attack. Right.
• 14:45
  Speaker 3: The time spent on maintenance shrinks. The time spent on
• 14:47
  Speaker 3: gate operations grows. And your useful duty cycle finally begins
• 14:51
  Speaker 3: to climb.
• 14:52
  Speaker 2: So what does this all mean? I want to pull
• 14:54
  Speaker 2: this out of the theoretical physics lab and ground it
• 14:56
  Speaker 2: for you, the person listening right now. Because, you know,
• 14:59
  Speaker 2: it is easy to get lost in the weeds of
• 15:01
  Speaker 2: tantalum capacitors and parity checks.
• 15:04
  Speaker 3: It really is.
• 15:05
  Speaker 2: Why should you care if a quantum engineer successfully bumps
• 15:08
  Speaker 2: the duty cycle from 1% to 10% by the late 2020s?
• 15:12
  Speaker 3: Because passing that threshold is the key that unlocks fault-tolerant
• 15:16
  Speaker 3: quantum computing. Which transitions the entire technology from a laboratory
• 15:21
  Speaker 3: curiosity into a massive industrial engine.
• 15:24
  Speaker 2: We're not there yet, though. No.
• 15:26
  Speaker 3: Right now, we are in what is called the NISQ-era
• 15:29
  Speaker 3: noisy intermediate-scale quantum. The machines are just too noisy to
• 15:32
  Speaker 3: run deep, world-changing algorithms. But once that duty cycle inverts,
• 15:37
  Speaker 3: and the machine spends more time computing than resting, literally
• 15:40
  Speaker 3: everything changes.
• 15:41
  Speaker 2: Paint that picture for us. What is the first domino
• 15:43
  Speaker 2: to fall when we get a high-duty cycle, fault-tolerant machine?
• 15:47
  Speaker 3: The first major domino will be the simulation of quantum
• 15:50
  Speaker 3: mechanics itself, specifically in chemistry. Classical computers, no matter how powerful,
• 15:56
  Speaker 3: even the biggest supercomputers, cannot accurately simulate the electron interactions
• 16:01
  Speaker 3: of complex molecules.
• 16:02
  Speaker 2: It's just too much data.
• 16:04
  Speaker 3: It requires too much memory. But a fault-tolerant quantum computer
• 16:09
  Speaker 3: natively speaks that language. Take the Haber-Bosch process, for example.
• 16:13
  Speaker 2: That's the process we use to make synthetic fertilizer, right?
• 16:16
  Speaker 1: Yes.
• 16:16
  Speaker 3: It sustains global agriculture. Billions of people rely on it.
• 16:20
  Speaker 3: But it requires massive amounts of high heat and pressure,
• 16:23
  Speaker 3: consuming roughly 1 to 2 percent of the world's entire
• 16:26
  Speaker 3: energy supply.
• 16:27
  Speaker 2: Which is an insane amount of energy for one process.
• 16:30
  Speaker 3: Exactly. But nature doesn't need a massive industrial plant to
• 16:33
  Speaker 3: do this. Bacteria in the soil fix nitrogen at room
• 16:36
  Speaker 3: temperature all day long using an enzyme called nitrogenous.
• 16:40
  Speaker 2: And we just can't figure out how the bacteria do it.
• 16:42
  Speaker 3: We know what the enzyme is. But the exact quantum
• 16:45
  Speaker 3: mechanical process of how the electrons move during the reaction
• 16:48
  Speaker 3: is just too complex for classical supercomputers to model.
• 16:51
  Speaker 2: But a quantum computer could.
• 16:53
  Speaker 3: With a high-duty cycle quantum computer, we could simulate that
• 16:57
  Speaker 3: exact interaction. We could discover the catalyst required to perform
• 17:02
  Speaker 3: nitrogen fixation at room temperature, potentially saving enormous amounts of
• 17:06
  Speaker 3: global energy and revolutionizing agriculture almost overnight.
• 17:09
  Speaker 2: That's incredible.
• 17:10
  Speaker 3: We're also talking about simulating the active sites of proteins
• 17:14
  Speaker 3: for drug discovery without having to do decades of trial
• 17:17
  Speaker 3: and error in a lab. But none of this, not
• 17:20
  Speaker 3: the energy revolution, not the medical breakthroughs, can happen until
• 17:24
  Speaker 3: we solve the duty cycle bottleneck.
• 17:26
  Speaker 2: Man, it is wild to map that trajectory. I mean,
• 17:29
  Speaker 2: we started by looking at a lazy chef. And it
• 17:32
  Speaker 2: turns out the chef was just endlessly busy holding the
• 17:34
  Speaker 2: kitchen together.
• 17:35
  Speaker 3: Exactly.
• 17:36
  Speaker 2: We explored how macroscopic classical systems like a silicon chip
• 17:39
  Speaker 2: naturally resist errors, allowing for those near-perfect duty cycles. But
• 17:43
  Speaker 2: the moment you step down into the quantum realm, the
• 17:46
  Speaker 2: environment just becomes wildly hostile.
• 17:48
  Speaker 3: Very hostile.
• 17:48
  Speaker 2: And to fight that decoherence, we had to introduce the
• 17:53
  Speaker 2: heavy administrative burden of error correction, using those ancilla quibbets
• 17:57
  Speaker 2: to listen to the harmonic resonance of the system and
• 18:00
  Speaker 2: without destroying the delicate superpositions.
• 18:03
  Speaker 3: And that overhead is exactly what crashes the efficiency.
• 18:06
  Speaker 2: Right. But by pushing the boundaries of material science with
• 18:10
  Speaker 2: better superconducting materials like tantalum and deploying smarter localized software
• 18:15
  Speaker 2: grids like the surface codes, engineers are actively shrinking that
• 18:19
  Speaker 2: maintenance window.
• 18:20
  Speaker 3: They're clawing back the time.
• 18:22
  Speaker 2: They are fighting for fractions of a microsecond just to
• 18:25
  Speaker 2: buy the system enough time to run calculations that will
• 18:27
  Speaker 2: fundamentally alter our physical world. It is incredible that the
• 18:31
  Speaker 2: biggest barrier to simulating the building blocks of life essentially
• 18:35
  Speaker 2: boils down to basic time management.
• 18:37
  Speaker 3: It is a profound engineering hurdle. But, you know, this
• 18:40
  Speaker 3: raises an important question, and it is a paradox I
• 18:43
  Speaker 3: think about whenever I look at these massive, struggling, hyper-cooled
• 18:47
  Speaker 3: quantum systems. Well, we're bending over backwards, building giant dilution
• 18:51
  Speaker 3: refrigerators and inventing complex surface codes just to keep a
• 18:55
  Speaker 3: handful of quibbets coherent for a few milliseconds.
• 18:57
  Speaker 2: Yeah.
• 18:58
  Speaker 3: Yet nature operates at a fundamental quantum level, flawlessly, all
• 19:03
  Speaker 3: the time.
• 19:03
  Speaker 2: What do you mean, like the nitrogenase enzyme you mentioned?
• 19:06
  Speaker 3: Exactly. Or look at photosynthesis in plants. Or the avian
• 19:10
  Speaker 3: compass that birds use to navigate using the Earth's magnetic field.
• 19:14
  Speaker 2: Wait, birds use quantum mechanics?
• 19:16
  Speaker 3: They do. These biological systems rely on quantum coherence and
• 19:19
  Speaker 3: entanglement to function. But a leaf doesn't need to be
• 19:22
  Speaker 3: cool to 15 millikelvin. A bird's eye doesn't run a
• 19:26
  Speaker 3: 2D lattice surface code to check for parity errors.
• 19:28
  Speaker 1: Yes.
• 19:29
  Speaker 2: Yeah, that's wild.
• 19:30
  Speaker 3: Nature maintains quantum coherence in warm, wet, incredibly noisy environments.
• 19:35
  Speaker 3: It manages its own quantum duty cycle effortlessly.
• 19:38
  Speaker 2: Without a single ancillary quibbit checking the math.
• 19:41
  Speaker 3: Right. The universe is the ultimate flawlessly running heavy-duty quantum computer.
• 19:46
  Speaker 3: So the question we have to ask ourselves as we
• 19:48
  Speaker 3: engineer our way out of this bottleneck is, what fundamentally
• 19:52
  Speaker 3: limits our ability to replicate nature's efficiency?
• 19:55
  Speaker 2: Is it just our hardware?
• 19:56
  Speaker 3: Is it a limitation of our current engineering paradigms? Or
• 20:00
  Speaker 3: is there some trick to room temperature quantum coherence that
• 20:03
  Speaker 3: biology figured out billions of years ago that we're completely
• 20:06
  Speaker 3: blind to? Wow.
• 20:08
  Speaker 2: Are we just building a massive mechanical, highly inefficient version
• 20:12
  Speaker 2: of a leaf? I want to invite you to mull
• 20:14
  Speaker 2: over that thought. Look at the trees outside or the
• 20:17
  Speaker 2: birds migrating overhead and realize that the invisible mechanics they
• 20:21
  Speaker 2: are using effortlessly every single second of the day are
• 20:24
  Speaker 2: the exact same mechanics our brightest minds are spending billions
• 20:27
  Speaker 2: of dollars trying to replicate in a frozen lab. Nature
• 20:30
  Speaker 2: has already solved the duty cycle. We just have to
• 20:32
  Speaker 2: catch up, keep exploring, keep questioning, and we'll see you
• 20:36
  Speaker 2: on the next deep dive into the invisible world around us.