Latest / TFTC: A Bitcoin Podcast / #733: The Truth About The Quantum Threat with Brandon Black

Transcript

#733: The Truth About The Quantum Threat with Brandon Black · TFTC: A Bitcoin Podcast

0:00
Brandon, welcome to the show, sir.
0:02
Thanks. Good to see you, man.
0:05
Good to see you. I'm very happy to get you on to talk about quantum, the risk, the perceived risk, the perceived reality.
0:14
What is the reality? Maybe that's where we'll start.
0:18
And I'll reiterate what I was saying right before I hit record, which is it's hard as somebody who's not a quantum physicist, is not a cryptographer to really get a grasp of what the true reality with advances at quantum computing are and specifically how they relate to Bitcoin's security in the short, medium and long term.
0:41
and I've watched you on the front lines of sort of debating the merits of the advancements in
0:49
quantum physics over the last year. I actually listened to your podcast that you did with
0:53
Gwerdy last year. I thought that was really good and that helped me get a better understanding of
1:00
your perspective. But I think starting with this question of how can you be confident
1:06
on one side of the other?
1:08
How can you be confident that it's not as
1:10
big of a risk
1:12
quantum computing as to Bitcoin today?
1:15
And then conversely,
1:16
why do you think
1:17
people are saying we need to rush
1:20
to become quantum resistant
1:22
or as confident as they are right now?
1:26
Yeah, I mean, I think
1:26
the thing that gives me
1:28
confidence, and I said on Guardian,
1:30
I think I still stand by it.
1:33
You know, I can't, I have an emotional
1:34
confidence that we're
1:36
50 to 100 to forever years from a quantum computer breaking a meaningful cryptographic system.
1:41
But that's not scientific. That's a gut feeling.
1:44
Looking more kind of in a scientific lens at it,
1:48
we live in a world constrained by the reality of physically building things.
1:54
And it's hard to physically build things.
1:57
That doesn't make any particular physical building challenge impossible.
2:00
But I think the quantum field is full of these, you know, call it if this is possible, then all we have to do is build it kind of perspectives.
2:13
And they totally just deny the realities of the difficulty in building physical things that manipulate tiny subatomic particles.
2:22
Or in the case of neutral atoms, you know, individual atoms being manipulated by laser tweezers, I think they're called, or optical tweezers.
2:28
like these are incredibly difficult things to build and especially difficult to scale right
2:33
let's say you can build the first optical tweezer and manipulate one atom okay now you want to build
2:39
a grid of thousands of atoms and manipulate all of them with different tweezers at the same time
2:44
think about how that that difficulty scales and and so maybe it's possible which is the thing i
2:50
have to acknowledge when i'm being reasonable as opposed to emotional about this but the
2:54
difficulties are dramatically underplayed. And that's what the evidence shows. I think that's
2:58
what people kind of don't acknowledge as the reality. When they look at quantum, they're like,
3:04
oh, all we have to do now is scale it up. But we've been working on quantum as a
3:09
colony species for 40 years or something, as long as I've been alive. And every time they go to scale
3:16
it up, they hit new difficulties that they weren't expecting. That's the evidence we have,
3:21
is that it's been over and over and over and over for decades
3:24
that scaling up is hard.
3:26
That doesn't mean it's impossible,
3:28
but the hardness is what gives me that confidence,
3:30
is that based on the evidence we have seen for decades,
3:34
repeatedly and consistently,
3:36
they may find a way to scale it up,
3:38
but it's going to be clawed out
3:40
by individual small improvements
3:42
over years and years and years.
3:45
And at some point,
3:46
they might hit on a technology that will work.
3:48
And when they do,
3:49
we'll see that technology go from five logical qubits in coherence for a while to 10 in coherence
3:57
for a similar length of time to 20 to 40, et cetera. We'll see doublings, like we saw with
4:02
transistors, of the same technology progressing through a scaling roadmap. And then we'll be able
4:07
to say, oh, now it's doubled three times. It took 2.5 years each time to double the number of logical
4:15
qubits in this technology. Okay. So at that rate, we'll get to a cryptographically relevant computer
4:20
in Y years. But until we see a roadmap like that, that's the successful scaling in one technology,
4:27
the evidence is it's going to be just hard fought, tiny wins, new technologies.
4:32
And there's just no evidence that it's going to come anytime in the next decade or really any
4:36
time in the next 20 years. So that's where my confidence comes from. On the flip side,
4:41
can you ask the other side of the question too? I think the other side gets their confidence from
4:45
the, let's say the impressiveness of the wins they do get. These are incredibly difficult,
4:54
mathematically challenging algorithms and designs. And the folks working on them are some of the
5:01
most brilliant people in the world. And they go through the fundamental physics of our world.
5:07
They go down to the very smallest subatomic particles.
5:10
They figure out how things work.
5:12
And then they say, oh, my God, I figured out that we can do this in this easier way or in this faster way or this new method of manipulating subatomic particles.
5:25
And those discoveries are absolutely amazing.
5:28
And they're the cutting edge of humans understanding our own world.
5:32
Like there's nothing more impressive than that.
5:35
And so it's a reality that these are the most brilliant people doing incredible work to understand our world.
5:40
And people get really excited about that.
5:44
And so that's where they get their confidence is that the most brilliant people are working on that problem.
5:48
But that doesn't change.
5:50
Like there's still this separation between it's probably literally possible, which is their perspective.
5:57
I don't think it is.
5:58
But in their world, it's probably possible.
6:01
and these fundamental understandings of the universe give them confidence that it's going
6:05
to be solved in the real world in physical devices. And that's the, so the way I understand it,
6:14
let me know if I'm wrong. Cause I did see you retweet my understanding of it the other day,
6:20
which is like, Hey, I think it's very obvious that in the theoretical realm of like, what is
6:26
possible. They are making advancements. And the biggest chasm that exists in reality,
6:33
which I think you just described, but trying to distill it down for an ELI-5 for people listening,
6:39
is that yes, we're making these theoretical advancements of what we can do.
6:43
However, there is a chasm between what we can do and the physical reality of
6:47
building machines that can actually sustain an uptime and persistence to make that theoretical
6:54
advancement a applicable reality yeah exactly and i think that in quantum there's this like extra
7:02
well there's extra detail to it which is um it's often possible to build call it the toy example
7:11
of something where you know that we've seen super connecting qubits we've seen neutral atom qubits
7:16
we've seen uh shooting the other technology that's kind of commonly pushed right now but there's a
7:20
bunch of different technologies that where they've built a built a few useful qubits and done
7:25
something interesting with it um but again the the specific thing that happens is is that taking that
7:31
and going to the next step and building twice as many qubits or four times as many qubits and and
7:36
they keep running into problems there um so so that i think that's also where they get their
7:41
confidence is oh we built that thing the physics described it we built it but then they tried to
7:48
build a slightly bigger one and they hit a brick wall. And we even saw like the way these things
7:52
are published, the Majorana paper from Microsoft kind of glossed over that. And when you dig deeper,
7:59
the physical device that they were publishing about had a single, I think a single physical
8:03
qubit or something of that, like tiny, tiny nature. It's like, oh, they built the first
8:08
qubit of this type. That's the tech I wanted to mention, using a Majorana particle. Okay,
8:13
Super cool.
8:14
You got one bit of that.
8:16
Now, what happens when you try to scale up?
8:18
And why has it now been almost a year since that result with no one building a two-Majorana qubit device?
8:24
Like, what's the deal?
8:27
Well, I think this is an important detail to dive into, too, is the difference between a logical qubit and a physical qubit and the relationship between the two.
8:38
Because from what I understand, logical comes down to the math and what you can do.
8:43
And then physical is like, all right, how do you organize all that and make it a computer, an operating software, whatever it may be?
8:50
Yeah, you're right.
8:53
And it's a complicated thing where depending what your physical underlying technology is, that affects the relationship between the physical and the logical qubit.
9:03
So there's like different avenues of research happening in quantum.
9:08
There's the mathematical research, which is assuming we have logical qubits mathematically, how do we apply that to solving real world problems?
9:16
That's what Google published about this week, right?
9:18
They published a paper that says, here's some new mathematics we could apply to running Shor's algorithm on logical qubits.
9:25
that makes it take fewer logical qubits to execute Shor's algorithm.
9:30
And in this case, also fewer steps of computation.
9:34
That's the math side, but that all runs on logical qubits.
9:37
And then the question is, how do you organize some physical device
9:41
into logical qubits on which you can run math?
9:44
And that's where the physical qubits come in.
9:47
So the other paper that came out this week on neutral atoms
9:50
was a new way of error correcting physical qubits
9:55
to make useful mathematical logical qubits.
9:58
They, within the context of this neutral atoms architecture
10:03
that uses the optical tweezers to move atoms around,
10:06
because they can physically move the qubits around,
10:08
they can grab a physical atom that's a qubit
10:11
and has been entangled in some way with some other qubits
10:13
and they can move it to another part of the chip or the device.
10:17
That's the cool thing about it.
10:20
Because of that, they can do what's called non-local error correction.
10:23
And the theory described in this paper is that because they can do non-local error correction,
10:28
because the physical qubits can be moved by optical tweezers,
10:31
they can do what they call high-rate error correction,
10:35
where rather than having to have a dedicated cluster of error correction qubits, call it,
10:42
for each logical qubit, you've got a bunch of physical qubits error correcting each other to get one logical qubit.
10:46
Instead, in this neutral atom architecture with the optical tweezers, they can, in theory, have one cluster of error-correcting qubits that corrects a whole bunch of active computing qubits.
11:00
And that's why they got that remarkable result in the paper of basically, if this high-rate non-local error correction works, we can go down from 500,000 to 10,000 physical qubits needed to implement Shor's algorithm.
11:13
And that's basically my summary of a lot of these quantum papers is if this thing that hasn't ever been done works, then we can do this easy thing.
11:23
And that's yeah. So that's where the relationship is and how why you see these radically different relationships between number of physical qubits and logical.
11:30
Because it depends dramatically on what types of error correction you can do on this physical architecture and whether those work.
11:38
And you need to do the error correction because these physical qubits are beholden to entropy, right?
11:42
Yeah, they're notoriously flaky.
11:45
Yeah.
11:46
And so that, like on the physical qubit side, like we're still trying to figure out a way how to make it so they don't deteriorate in real time.
11:53
Right.
11:54
And then also in this neutral atom architecture, again, because they can physically move qubits, they have in the paper, they accept that physical qubits will deteriorate.
12:04
and they design a theoretical neutral atom quantum computer
12:10
that has a reservoir of pre-coherent backup qubits
12:15
that they can grab with the optical pleasers
12:16
and stuff into the circuit to replace one that deteriorates.
12:20
So exactly as you said, these physical qubits,
12:22
they tend to break down.
12:24
And so the big question is how do we stabilize them
12:27
and make it so that we can kind of continue doing computation on them?
12:29
And different researchers in different parts of the quantum field
12:32
have different methods of kind of holding that shit together for longer.
12:37
And how much energy does all that take?
12:40
Yeah, I mean, that's the other big question that I think Bob McElworth and I were talking about on X this morning, right before this.
12:47
You know, even if you linearly scale up the energy needed, it's still huge.
12:54
He was saying it's something like 100 megawatts based on his calculations.
12:57
You need to cram into whatever device.
12:59
if you linearly scale up the energy needed per qubit
13:02
based on current technologies.
13:04
But my point that I think is a valid point,
13:07
if you think about the complexity
13:08
of moving around more and more qubits
13:11
in the neutral atom architecture
13:12
or in cooling more and more qubits,
13:15
it's not going to be a linear scaling.
13:17
And we've seen this even with classical computers
13:19
that the more dense you make devices,
13:22
the more leakage you have.
13:24
And therefore the more energy it takes per bit
13:27
to keep it cool.
13:28
for supercomputer architecture.
13:31
And so even, let's say, as linear,
13:35
it's still a ton of energy to cram into a device.
13:37
And if it's nonlinear,
13:39
it may become completely intractable
13:40
to power a device that can do meaningful quantum computing.
13:47
Yeah.
13:48
So, I mean, again, going back to the confidence
13:52
of the other side, like, what do you think is driving?
13:57
I guess the whole conversation as it pertains to Bitcoin is this perceived urgency that is being thrust on the developer community specifically to upgrade to a quantum-resistant cryptographic primitive that would secure private keys so that they wouldn't succumb to an attack via a quantum computer.
14:27
And that that's been like the confounding, not confounding, but it's been sort of the thing that's really perturbed me is like that you have this very.
14:43
Confident and.
14:48
So what a manic I think manic is a good way to describe it a group of people are saying we need to fix this devs do something right now And it frustrating for two reasons Number one quantum has been thought about within Bitcoin since Satoshi was around
15:06
I mean, he mentioned it.
15:08
If you've been reading the mailing list and looking at some of the research being done,
15:13
there are people trying to figure out, okay, if quantum computers do come,
15:18
what's the best way to transition to a quantum-resistant address structure?
15:23
And there are a ton of not only coordination issues, but standardization issues that exist.
15:30
There's a ton of, for lack of a better term, tech debt that exists.
15:35
Like if you are going to transition to a quantum resistant address structure that is going to disrupt a lot of the infrastructure that's been built today.
15:47
So you talk about the Lightning Network, PSBTs, multisig, all this stuff.
15:51
You need to think through how do we transition to this address structure without disrupting all that if possible.
16:00
And if that is going to be disrupted, what do we do?
16:03
And my biggest worry is that you rush a change.
16:10
And it just not only disrupts all of that standardization, that infrastructure that's been built to date,
16:15
but you haphazardly rush to a change that has been well thought through, well tested and leads to a bigger fallout in terms of disruption to the network than having done nothing because quantum potentially doesn't manifest on the timescale that these people are saying it well.
16:36
Yeah, I mean, I think it's a legitimate risk.
16:38
And frankly, it's why it's one of the reasons I try to stay on top of all of this, because we should be realistic here.
16:44
the most likely outcome is that at some point,
16:49
SecP256K1 or elliptic cryptography in general will fail.
16:54
Like that's the most likely outcome.
16:56
I don't think personally that it's quantum that's going to break it,
16:59
but, you know, crypto systems have failed over the years.
17:01
We've seen it.
17:03
And it seems likely that at some point it will.
17:07
So what do we do about that?
17:10
And so I try to stay up on what the other kind of new cryptographic research is so that when there's something suitable for Bitcoin that can support all the infrastructure we have, wallets and Lightning and everything that you mentioned, we should actually add it to Bitcoin.
17:27
And we should maybe even do that before there's a real threat so that people have options.
17:32
And that's fine.
17:33
That's good.
17:34
There was a whole discussion on the mailing list about the benefits and downsides of having options.
17:38
And I think Peter Wola is a bit concerned about people having options because there could be fighting over what the correct option to use is.
17:46
I'm less concerned about that because we already have that essentially where some custodians use MPC that's kind of cross crypto.
17:55
I think Coinbase does and others.
17:57
But some custodians use on-chain multisig, right?
18:00
That's fine.
18:01
It's OK for different people to choose different tradeoffs and how they secure their coins.
18:05
So yeah, if we have suitable cryptography, we should totally put it in Bitcoin.
18:09
And there's a decent argument to say that we're getting very close to having something
18:12
suitable to add, at least as a backup in the work that Jonas Nick and blockchain research
18:16
and others are doing.
18:18
So I'm not out here saying we shouldn't do anything about the potential break of SECB
18:24
256K1.
18:26
And I think that's where, kind of to your point, these confident, almost manic people
18:30
in their quantum side are like, devs do something.
18:33
well, the devs are doing something.
18:35
Even people like me that don't believe in quantum at all
18:38
are out there doing research and evaluating research
18:42
on what we could use as another crypto system for Bitcoin
18:45
and to chill out, I guess.
18:48
Well, it's not only that.
18:49
It's like they're treating Bitcoin Core
18:53
like any Bitcoin developer is a monolith.
18:55
Like, go do something.
18:56
And it's like, well, there's only so many people
18:59
who are qualified to understand quantum physics
19:02
and the cryptography sec p 256 k1 like that's like a very niche part of like the bitcoin
19:10
development process like obviously you have the p2p layer you have the wallet layer you have the
19:14
gui you have many different facets of of the protocol that make up bitcoin and like finger
19:21
wagging at the whole dev community like do something it's like well not everybody within
19:25
the developer community is going to be able to do anything about this because they're not
19:29
like their core competency isn't the cryptography and taking it further like making the cryptography
19:38
quantum resistant like there's very few people that are equipped to work on this particular
19:43
problem and to your point and what i've been saying for the last six months since this has
19:48
become a huge meme is like the people who are um equipped and able to do this work or seem to be
19:55
working on it maybe it's not the pace that you want but you can't pull out a whip and make them
20:02
work faster nor would you want to like yeah there's a great post uh stew stew txo out there uh but
20:10
there's other like since since this post it was in december uh there have been five new uh post
20:17
quantum cryptographic algorithms published and zero new numbers factored by an actual quantum
20:22
computer. And I just thought that was, that was so great. It's like, so, so really we're,
20:26
we're upset about the rate of progress when month to month to month, we're seeing new algorithms
20:31
that are kind of progressing the state of the art for post-quantum cryptography in Bitcoin
20:34
and no new physical quantum devices being built that run real algorithms that could maybe even
20:41
someday break Bitcoin. Uh, no, I think the progress is at a, at a, at a good pace. We might
20:45
even be over-investing if you, if you ask my honest opinion. Yeah. Well, I think I saw something
20:51
yesterday where somebody was saying like you could do the factoring of these numbers by hand
20:55
faster than the quantum computers could right now. Yeah, I checked with my six-year-old and
21:00
I see seven now. Anyway, he can factor more numbers than a quantum computer for sure.
21:05
I actually just set my six-year-old up with the synthesis math tutoring that app and he's
21:11
been playing with it. So I actually did see him factor some numbers this morning that
21:14
that were faster than a quantum computer.
21:18
But again, so the Google paper specifically,
21:24
I think it's another thing to touch on
21:26
is the fact that they didn't actually release the results.
21:30
They released the zero-knowledge proof
21:32
that they had done something and said,
21:36
basically framed as like,
21:37
we didn't want to release the results of this
21:38
because we don't want black hat quantum developers
21:41
to get access to this and to go build a computer that disrupts the world.
21:48
And then the advocates for this paper were saying, well, now this is the warning.
21:55
They're not going to tell us the advancements that they're making
21:58
because they're worried about releasing them to the public
22:01
because it'll be used against us by nefarious actors.
22:06
Is this a marketing scheme, a way to develop a budget,
22:10
Or do you think there's some legitimacy here that the smartest people in the world working on this problem are truly worried that we're hitting an inflection point, a tipping point that could accelerate the progress being made within quantum computing?
22:24
So it's it's a really interesting problem to think about.
22:28
I had that moment of pause as well when I read the Google paper.
22:31
um the thing about the google paper though is that there is no fundamental physical change
22:40
that they're publishing like i said they published a mathematical change which reduces the number of
22:46
sorry it reduces the complexity of the physical device you need to build in order to run this
22:50
algorithm but no one actually made a more complex physical device so alex pruden who i'll be
22:58
debating at Bitcoin++ in a couple of weeks, I think actually looked pretty cogently about this.
23:04
Of the pro-quantum guys, he's one of the more pleasant to talk to. So I'm glad I'll be talking
23:08
to him at Bitcoin++. And he was pointing out that let's say, right now we're at five-ish
23:15
logical qubits have been demonstrated in a very short microsecond duration coherence.
23:21
Okay. Well, what if through the mathematical improvements, we get to the point where it only
23:26
It takes just 256 logical qubits in coherence for only five milliseconds in order to run
23:34
Shor's algorithm on a Bitcoin private key or public key to get that private key.
23:40
If they, on the mathematical side, compress the time and the number of qubits so much,
23:46
then we could see a situation where it's very much in reach of the physical progress.
23:50
So if physical progress is zero, but the mathematical progress is extreme, we could see it break kind of in that direction where the mathematical progress brings it down to the point where the physical guys have a very short step to take.
24:03
And that's a great point.
24:05
Now, that said, the Google paper puts us about six orders of magnitude away from the needed coherence time and about three orders of magnitude away in the number of qubits that need to be in a physical device.
24:17
So I don't know how you want to look at those two different dimensions of orders of magnitude, but we're somewhere between six and nine orders of magnitude of improvement on the physical side away from implementing Google's paper.
24:28
So we need to be really realistic that, yes, eventually the mathematical progress could bring it into reach of the physical, but we're still very far away from that.
24:38
I guess that begs the question, what are the advancements from an order of magnitude perspective on the physical side been like in recent years?
24:47
is that a big leap?
24:51
It's obviously, it is,
24:53
but have the advancements over the last 10 years,
24:57
was it 100 orders of magnitude away?
25:00
Five years ago, was it 1,000?
25:02
And are we now getting to the point where
25:04
six to nine may seem big, but it's not that big?
25:08
We haven't.
25:09
I mean, even the best kind of longest held stable devices
25:16
have up to a couple thousand physical qubits,
25:20
which is maybe a one or two orders of magnitude improvement
25:25
in the last 10 years.
25:27
But I hesitate to even say that
25:30
because I think it paints it a little bit too rosily
25:33
for the physical quantum guys,
25:35
because the biggest improvements have come
25:38
by implementing whole new architectures.
25:41
And so there's no evidence right now
25:42
that any single architecture
25:44
can increase the number of physical qubits.
25:48
And that's where I think the pro-quantum manic folks
25:55
get way over their skis.
25:57
They're like, all we have to do is scale up.
25:59
But you can't just scale up
26:02
when every scale up that you're doing so far
26:04
for the last many decades has been
26:06
because you developed a whole new architecture, right?
26:09
You're like, you're just discovering
26:10
a whole new sets of problem spaces
26:11
because you're developing a whole new architecture still.
26:13
And so I made it very clear in like, when would I start to think we need to take action in Bitcoin?
26:19
And that's when we see a single quantum architecture increase the number of qubits you can hold in coherence for a longer time over several cycles, at least two.
26:30
we need to start seeing a trend in one architecture scaling up and then maybe it becomes just a
26:36
scaling problem where we just have to do the decades long industrialization to go order of
26:42
magnitude order of magnitude order of magnitude now we need to start taking action but until you
26:46
see that they're like oh okay maybe maybe this one is going to be the one but until there's evidence
26:52
that it actually scales why would you take any action we don't have the evidence to say to justify
26:56
action. Yeah. And that's where it gets very confusing. I think that's been the most
27:04
disconcerting thing observing this over the last year as the narrative around the quantum threat
27:12
to Bitcoin has hit the market. Again, I'm not a quantum physicist. I'm not a cryptographer.
27:18
I know enough to understand, like I can explain, like a logical qubit does like the theoretical algorithm, the physical qubit sort of creates the space where that stuff is computed, for lack of a better term.
27:36
And I understand that there is coordination on the physical side and energy needed.
27:41
And it seems clear to me that it's not where it needs to be and not anywhere close to where it needs to be to effectively run this stuff persistently to wage the necessary attacks.
27:53
But then again, going back to the confidence game and I hate to say like the sky is falling
27:59
perspective that is very reminiscent of like climate change and things that I'm just using
28:07
like PsyOp pattern recognition.
28:09
And it feels like that to me.
28:10
It could be wrong.
28:11
Using heuristics here.
28:13
Again, don't understand the math and not a physicist by any means.
28:18
Study economics.
28:19
but yeah I think that's the narrative side of things it's very very easy to socially attack
28:26
people and it feels I'm not saying that I do believe the the other side is earnest in their
28:32
beliefs and that that's like the interesting part of this whole discussion and observing it
28:37
over the last year is like they're so ardent on their beliefs and people who are just skeptical
28:43
like, hey, I believe that you're genuine, but I'm not seeing it.
28:48
And there's this narrative leverage, this asymmetric leverage they have
28:53
because they can project fear onto the market.
28:57
And if you're not afraid, you're not doing enough and you're actually stupid.
29:02
Yeah, I think it's the classic FUD game.
29:05
And it does make it hard to distill.
29:08
Again, I didn't want to be, I'm not a quantum physicist, by the way.
29:12
I'm just a person with engineering background and with a lot of experience for many decades of reading research papers.
29:17
So I come at it not as an expert in the particular field, but as someone experienced in reading papers and kind of understanding the real implications of what some new publication means.
29:29
So, yeah. So I think one thing I want to say from what you were just saying is that.
29:33
I recommend really that everyone go read the papers,
29:37
because in many cases, the papers that underlie
29:42
the big manic posts about quantum, the sky is falling,
29:46
the papers are much more conservative in what they claim.
29:51
Now this Google one is kind of an exception to that but I think that also a good way to gauge the progress in quantum computing is look at what the actual results in the papers are that justify these very high excitement posts in social media and in kind of science journalism let call it
30:09
And what you'll find is that the real results being published in the academic papers are small, like little nuanced improvements in things.
30:19
And again, this Google paper is kind of an exception to that.
30:22
But the vast majority are, especially on the physical side, actually, I think exclusively
30:27
that's true on the physical side, they're these small progressive improvements.
30:31
And so you can kind of tell where we are in quantum computing by the fact that what gets
30:35
the hype in the social media and science journalism side are these tiny little improvements on
30:40
the physical side.
30:41
Okay, so if that's what gets the hype, then we are a long way from getting all the way
30:46
there because when we're getting close, you're going to see the Majorana particle architecture
30:54
scaled up again and now went from X logical qubits to Y for the third time in three years.
31:01
We're going to see like, oh, these are big improvements. They're building devices that
31:06
can run Shor's algorithm on bigger and bigger, even if they don't really run it, to be fair,
31:10
they don't have to run the actual algorithm, but the physical device is capable of running it
31:13
on a key this big, on a key that big, you know, you'll see, I guess, bigger results actually being
31:19
published with less hype. And right now we're seeing tiny results published with huge hype.
31:23
And that's one of the ways to kind of tell how far we are and how much it is just hype.
31:30
Yeah. It's, yeah. And again, I think really nailing down the risk of trying to rush
31:39
a change to Bitcoin to appease the people who think that this is coming faster than others
31:48
believe it is? What are the risks of upgrading to something in haste because you're worried about
31:56
this? And alternatively, not alternatively, but on top of that, is there a line in the sand we can
32:03
draw? It's like, hey, we hear your concerns. We'll take this seriously. We'll continue the
32:09
research that we've been doing and make sure that we advance that.
32:13
But if we get to 2030 and we're at this state of quantum research,
32:19
we're not going to take you seriously anymore.
32:20
Like what is the line in your sand?
32:22
Oh, that's a good question.
32:23
I haven't really thought about it from that angle of, yeah.
32:25
Like when do we stop paying attention?
32:28
Like it's like Greta was like Greta Thunberg again for the good using the
32:33
climate change analysis and analogy was in 2015.
32:37
She was like, by 2022, Miami is going to be underwater.
32:40
It's like 2022 came and went.
32:41
Miami is not underwater.
32:42
It's like, OK, we can't take you seriously anymore.
32:44
What is that?
32:46
What is that line in this conversation?
32:50
Yeah, I think with quantum, it's really hard because the reality is, again, I think quantum
32:54
is not physically possible, but I can't.
32:56
That's an emotional thing.
32:57
So we can't really do that because there could always be a new architecture developed that
33:05
makes it possible, something we haven't thought of before.
33:11
But I think the good news is that, as I said, something is eventually most likely going to
33:17
break our existing crypto.
33:19
And so we need to keep doing this research and building new crypto systems for Bitcoin,
33:24
regardless of quantum.
33:25
And we should keep doing it at the right pace.
33:28
And so one of the ways to, I think, ease the tension here is to say, look, Bitcoin actually
33:34
is developing towards quantum resistance, regardless of your level of concern or my level
33:39
of concern. I'm not concerned, you are. It doesn't matter. Bitcoin is developing towards quantum
33:44
resistance. You know, BIF360, Payton Merkle route is advancing. I think it's fairly likely to get
33:50
activated on the network, which opens up the door to building quantum resistant new crypto into
33:55
Bitcoin. Right. So that's like step one is actively happening. And why is it happening?
33:59
because it's actually a good change for Bitcoin.
34:02
And pretty much everyone in the Bitcoin developer community,
34:06
there's like a couple of very tiny exceptions,
34:08
but really pretty much everyone's on board with that change.
34:10
And it does move in the direction of quantum resistance.
34:13
Does it matter that it's quantum resistant?
34:15
Not to me, but it does to some people and that's okay.
34:18
And we'll see that continue, right?
34:19
As other cryptographic research is done
34:24
and as we get to the point
34:26
where some alternative cryptographic primitive
34:29
is appropriate for Bitcoin,
34:31
I think it's absolutely a good thing.
34:33
And I think almost everyone agrees
34:34
to have people,
34:38
to give people different ways to secure their coins
34:40
depending on their goals, right?
34:42
So someone who's trying to secure their Bitcoin
34:44
for some kind of dynasty trust, let's say,
34:49
they might want to put it in a way
34:50
where it's secured by both elliptic curves
34:52
and let's say hash-based signatures
34:54
because they don't care about the signing cost when they go to spend that Bitcoin.
35:00
They care that it is almost 100% secure for 100 years.
35:06
And you really can't get 100-year security from one cryptographic assumption.
35:11
You need a couple of them.
35:12
And so I think Bitcoin is going to go that direction.
35:15
I think that's my take is just let's do the right thing for Bitcoin.
35:19
And in the long term, the right thing for Bitcoin is also going to happen to make quantum resistance an option.
35:25
We just don't need to rush.
35:27
And as you said, there's a big risk to rushing.
35:29
You know, Satoshi notoriously chose a elliptic curve specifically that wasn't published by NIST to help mitigate the risk of kind of a backdoor crypto going into Bitcoin.
35:41
And right now people are like, no, let's push this NIST published post-quantum architecture into Bitcoin.
35:46
And everyone who knows what Satoshi did is kind of, no, let's not put the NIST thing in.
35:53
Let's develop something from our own first principles that's appropriate for Bitcoin.
35:57
And when we have something good, we'll put it in.
36:01
Yeah, and that seems to be what Jonas and Mikael are doing with Shrinks Plus.
36:06
Yeah.
36:07
Jonas published Shrimps last week, which is sort of an advancement of Shrinks Plus.
36:14
because the way I understand it, shrimps, if you were just doing pure shrinks plus private key,
36:23
transferring a private key or recovering a private key on another physical device would be quite
36:29
burdensome, but shrimps makes it so you can begin recovering seeds on multiple devices using the
36:36
same seed phrase. Yeah, exactly. I actually wrote it up for Optech. So check out Optech
36:43
tomorrow morning. And I summarized shrimps and also some other work on post-quantum crypto by
36:48
Conduition about isogeny-based crypto for Optech. And that'll come out tomorrow morning. So there's
36:53
two new post-quantum systems detailed in Optech tomorrow by me. I'm the anti-quantum guy, but I'm
37:02
out there actively doing the work to publicize the post-quantum stuff because like I said,
37:07
we need it eventually regardless of quantum.
37:12
And so what are your thoughts on shrink splats and trims?
37:14
Like, is it a solution or is it a step toward the solution?
37:20
They're like, okay.
37:23
You know, it, it would significantly impair the, um, the development of things like lightning
37:31
and pay join and, and tap root and, and, you know, all the cryptographic primitives that
37:36
we rely on, Frost and Musig, and even the MPCs that are kind of non-Bitcoin specific
37:41
for multi-signature and threshold signature, any of these would still significantly impair
37:48
those.
37:50
So I think Jonas's work and blockchain research in McHale is exceptional, and it's definitely
37:56
moving the state of the art forward.
37:58
But as of now, I would be hesitant to put any of these in Bitcoin because of the combination
38:05
of their size, which would impair the number of transactions we could do if we were to use them,
38:09
and the fact that they're not really compatible with our existing wallet infrastructure that
38:14
people are depending on. I wouldn't be upset if other people in the developer community thought
38:20
they were a good thing to add, especially if we're thinking about, as I said, a world where
38:25
we do have multiple crypto systems available in Bitcoin and people can choose whether they secure
38:30
their coins with one, the other, or both, potentially. I'm not going to be mad if they
38:35
go in in that context. I think the shrimps in particular shows a lot of promise in that it,
38:45
I guess the way I would put it is that it's close to compatible. It works decently well with,
38:51
as you said, being able to restore seeds on multiple devices and the ways we use Bitcoin for
38:55
real, while still being not so huge that it completely destroys the usability of the chain.
39:00
or requires massive block size increases or something.
39:05
So I love the direction that Jonas and crew are working there,
39:08
and I hope that it just continues and we kind of take our time
39:12
getting something better than these before I put it in.
39:14
That's my hope.
39:16
Yeah.
39:17
And then there's a whole discussion around hash based and lattice based.
39:21
I know Jonas and Mikhail are working on a lattice based
39:26
research paper right now.
39:28
So what's the, I guess, what's the consensus or lack of consensus around which direction to go in when given those two options?
39:41
I think the consensus right now is if for some reason we decide to or need to do something in the near term, let's say the next five years, it would probably be something hash based because it doesn't require any new cryptographic assumptions.
39:54
It relies on things we already trust and know in Bitcoin.
39:58
with the downside that hash-based has certain downsides in terms of calculating keys that make it harder to use with our existing infrastructure in various ways.
40:13
So yeah, so hash-based would be the thing to do soon.
40:16
Lattice-based requires some new cryptographic assumptions, but has definitely certain benefits in terms of the flexibility of the math.
40:24
I don't understand it as well, if I'm being honest. I haven't read deeply about it.
40:28
And then the other piece, actually, for Optech tomorrow, is about isogeny-based crypto, which is a whole different kind of crypto that also is quantum resistant, but also requires a new cryptographic assumption.
40:40
But it works on elliptic curves still, but using a different kind of key that's not vulnerable to quantum.
40:47
So I think if it's not hash-based, we don't know what it would be.
40:50
It could be lattice.
40:51
It could be isogeny.
40:52
It could be something else.
40:53
and research should continue and is continuing to kind of get to the point where we could do
40:59
something other than hash based. And my, again, my hope is that we have much more time than the
41:05
quantum doomsayers are predicting and we can get to a really good alternative crypto system
41:10
that broadly works within the Bitcoin ecosystem and supports all the stuff we want, you know,
41:17
whether it be signature aggregation or silent payments or HD wallets, like whatever the stuff
41:22
we wanted. We want a crypto system that supports that.
41:26
And I think
41:26
the research is going quite rapidly
41:28
towards having some options there
41:30
in the next, let's say, maybe 10 years.
41:33
Yeah.
41:34
And so, this seems very
41:36
reasonable. It seems like a very reasonable approach
41:38
to upgrading Bitcoin.
41:40
And again, that's been the most frustrating thing.
41:42
Debs do something.
41:44
Bitcoiners aren't focused on this. It's like the research
41:46
is being done.
41:48
What...
41:49
I mean, I can't speak for them, but
41:52
that's what i've been trying to figure out is like what pace would be sufficient for you guys like
41:56
what changes do you want to see and then there's a circular logic where it's like hey we're working
42:00
on this like give us feedback on um the the work that we're doing whether or not you think it's
42:07
sufficient for the security that you deem necessary for these quantum advancements that are being made
42:12
and if you don't agree like do you have a solution and if so will you propose it and they're like
42:16
we can't propose it because bitcoin core is controlled by a cabal of five or six developers
42:21
so we'd never even propose it because we'd get rejected immediately.
42:25
It's like, well, is this productive at all?
42:28
Like, what is going on here?
42:31
Yeah, yeah.
42:32
And I think what they're actually saying
42:34
and what they're reflecting when they say that is
42:36
that they know their crypto proposal would be rejected
42:42
for good technical reasons.
42:46
And so it's not, oh, Core is this cabal.
42:48
It's really Bitcoin core holds the line on the quality of technical contributions accepted so high that nothing right now, no new crypto system for Bitcoin right now meets the bar.
43:02
And that bar, of course, can move if the threats to Bitcoin's existing crypto system get closer.
43:07
And so that's, I think, where the disconnect is, let's say, is that no one in the Bitcoin core maintainership has yet come out and said, the sky has fallen, quantum is in three years, we have to do something now.
43:23
And if we had to do something right now, then some of these existing crypto systems would be accepted by core.
43:27
But the cabal won't accept them because they're not really good enough and they'd only be accepted if it was an imminent threat.
43:35
but what is good enough in the eyes of
43:38
people who do believe it's an imminent threat
43:41
do they have solutions because that's the one I haven't seen
43:45
they've said basically that we should just take
43:49
Sphinx plus into Bitcoin even though the signatures and keys would be
43:53
combined like 10 kilobytes per spend or something like that
43:56
and like that's not a completely
44:01
unreasonable argument if the thing really
44:05
was around the corner. Now Sphinx Plus is a NIST standard under the name SLHDSA, I think it is.
44:12
So the question is, is the sky falling enough that we would accept an unmodified NIST standard
44:19
into Bitcoin with these significant trade-offs in having 100x the key signature size versus our
44:29
And I think as we discussed at length here,
44:31
the sky is not falling nearly enough
44:33
to accept that kind of a trade-off.
44:37
Okay. So that's the one thing I've been wondering.
44:39
I thought, I mean, you mentioned Alex Pruden,
44:41
he's with Project 11.
44:44
They helping blockchain systems transition to post I believe working with Solana and Ethereum I was curious if they had a specific solution
45:00
that they're putting forth for Bitcoin that...
45:03
I don't know if they specifically have one
45:07
that's targeting Bitcoin directly.
45:11
They published a paper on some improvements they've worked out.
45:15
I think it was Lattice-based,
45:17
where it does support a lot of the Bitcoin wallet infrastructure.
45:21
Like I said, Lattice has more key math ability than hash-based stuff.
45:30
So yeah, I think if they were to propose something for Bitcoin,
45:33
it would be probably Lattice-based.
45:35
And the Bitcoin folks would currently say,
45:39
the Lattice stuff is too new,
45:41
and we don't yet fully trust the cryptographic assumptions it makes for Bitcoin.
45:44
we'd want to see more time
45:47
more threat modeling
45:51
more proofs
45:52
more different systems
45:53
based on these same assumptions
45:54
that all are shown to work
45:56
that's the kind of thing
45:57
we want to see for Bitcoin
45:58
we don't want to put a new system in
46:00
that's vulnerable to some classical attack
46:03
in the attempt to defend against a quantum attack
46:06
yeah and this was similar to what would happen
46:10
I mean you mentioned
46:10
ECDSA was chosen for a certain reason
46:13
Schnorr was on the table, but I believe it was patented at that point.
46:17
And I think Satoshi even said, hey, it probably needs more time to be in the wild before we adopt something like Schnorr.
46:23
And then, what was it, 13, 14 years in to the protocol, Schnorr was included into, including into Bitcoin.
46:34
Yeah, exactly.
46:35
There's a track record here that Bitcoin, it's so strange when they're like, Bitcoin should this, that, and everything.
46:42
And demonstrably, Bitcoin has a conservatism that is appropriate and will adopt new cryptographic primitives or assumptions as appropriate to the protocol.
47:00
One thing that relates here is it's interesting to see the difference between centralized things and decentralized, right?
47:06
So centralized systems can just upgrade their crypto.
47:11
They can take a new assumption.
47:12
And if it goes bad, they can turn it off.
47:15
And that's a low cost thing when you're a centralized system.
47:18
And so they have a different math here where if it's easy to change because you're a centralized system, then the risk of taking a bad assumption is much lower because you can just change again.
47:30
But Bitcoin isn't like that, right?
47:31
Bitcoin is a massive global distributed decentralized network.
47:35
And so the costs of taking a bad crypto system into Bitcoin are much higher than for something like Solana or for Google internally or some web server.
47:44
You know, some web server turns on SLH DSA today.
47:48
They can turn it off tomorrow and that's OK for them.
47:51
We can't do that in Bitcoin.
47:54
No, we can't.
47:56
And that's, again, the more frustrating.
47:58
And Bitcoin gets picked on.
47:59
And that's like, and that being honest,
48:02
and I'm happy you said that because that's one thing
48:05
that I think Bitcoiners who don't believe it's a big risk,
48:10
like the whole line of like, and I used to say this too,
48:13
hand up of if quantum comes,
48:19
like Bitcoin's not the only thing at risk.
48:20
Like, yes, that's true.
48:21
But to your point, like all these centralized systems
48:23
can trivially incorporate and rip out these cryptographic systems rather trivially because
48:32
they're centralized.
48:33
Like Bitcoin does have a big, unique problem in the sense that it's a distributed system.
48:40
We need to get consensus.
48:41
Once we put something in, it's hard to take it out.
48:43
And the risk factors to Bitcoin are certainly unique and arguably higher than they are to
48:51
other systems.
48:53
Yeah, so we have to pay attention and we have to move at the appropriate time, for sure.
48:59
Yeah, and to your point about lattice-based, I just wanted to bring this up.
49:02
That's why I'm looking at my other screen over here because we wrote about it yesterday.
49:05
But going back to lattice-based schemes and the fact that they're not as battle-tested as some hash-based solutions.
49:15
So lattice-based schemes offer advantages of verification speed and signature aggregation,
49:18
but to carry a trade-off, they rely on newer mathematical assumptions that haven't been battle-tested as long as hash functions.
49:24
In fact, NIST tested 69 post-quantum Canada algorithms during its standardization process,
49:30
and two of them, Rainbow and Psych, were broken with classical computers during testing.
49:36
And so that's four, what would that be?
49:39
That would be like four and a half percent or less than that, like three and a half percent of these,
49:45
these, or maybe like, yeah, 3.5% of these post-quantum lattice-based systems were proven to be insecure.
49:56
And so that's, if you're going to incorporate a lattice-based system into Bitcoin,
50:01
and you have a 3.5% risk of it being insecure, I think that's pretty high for a trillion-dollar network as well.
50:10
Yeah, for sure.
50:12
And I mean, this gets back to...
50:15
what I was saying earlier, like, what's the line in the sand? Like, how do we have a,
50:19
a more level-headed conversation about all this with, with the, uh, the people who are
50:25
convinced that this is coming faster than, than we are?
50:29
Yeah. I mean, my best way is, is to, to rely on, on evidence-based decision-making. And that's why
50:39
I keep posting kind of every few months, I guess I probably post something about,
50:42
Look, I'll worry about quantum when I see, like, here's a list of things.
50:47
I think it's scaling over two generations, less than exponential scaling in the time needed to solve progressively larger keys on the same quantum system, and beating classical in any cryptographically relevant, even small size problem.
51:08
And to date, none of those three things have happened in any quantum architecture.
51:13
And so there has to be evidence.
51:15
We can't, as I joke about it being unicorn fart based engineering, but the reality is
51:21
that anybody can FUD anything about Bitcoin.
51:24
And if we can be caused to make a change to the protocol based on claims and not evidence,
51:32
then Bitcoin is vulnerable to the most obvious of attacks, right?
51:36
Bitcoin can't be subject to change without evidence that it needs to.
51:42
That simply doesn't make sense.
51:44
And so we can set a pretty clear evidentiary standard for when a quantum architecture shows these three or maybe four.
51:53
Like people can argue about exactly what the criteria are.
51:56
But we can set pretty darn clear standards for the evidence required to start taking immediate action.
52:03
And of course, in the meantime, we're going to take progressive action anyway.
52:06
So it's not like this is a, oh, we're going to do nothing until.
52:08
It's just we're going to take a slow and steady approach until there's this level of evidence that we have to move faster.
52:15
Yeah. And do you think.
52:19
Yeah, we have to think of opportunity cost, too.
52:21
Right. Like what else could be could we be working on in Bitcoin that is necessary to low hanging fruit that undeserved attention to the quantum question could take away from?
52:34
Yeah, that's a really important point.
52:37
And I think even more than not working on the right things, it's essentially fudding the amazing innovations that are still kind of nascent, not widely deployed in Frost and silent payments and Musig and even DLCs.
52:54
All of these things are classical elliptic curve based protocols that are really valuable for Bitcoin.
53:00
You know, Craig Ra is working on getting silent payments into Sparrow Wallet recently.
53:05
Coldcard just shipped music to support.
53:07
And these things strictly depend on the existing elliptic curve cryptography.
53:13
And they're great.
53:15
They're huge improvements in the usability of Bitcoin in a couple of different ways.
53:18
I'm not going to get into them because it's not important right now.
53:20
But when you're saying we need post-quantum tomorrow, people just say, well, then why would I bother developing silent payments or music when we're going to replace the existing crypto?
53:29
in a year. You shouldn't, if that was true. And so I think it is very important that we push back
53:37
on this quantum FUD and say, look, as of now, there's no evidence that we'll be kind of replacing
53:42
the basic elliptic cryptography in the next decade. So we should keep building silent payments
53:47
and music and frost and DLCs and everything based on the existing cryptography. It's going to be
53:51
around for a long time. So, so keep building. Yeah. Oh, it's also, it's hiresome. Do you think
53:59
this is a social attack or man? I don't know. Intentional, intentional social attack. I guess
54:06
if you believe the quantum's not, not coming as quickly as they believe it is a social attack,
54:11
but I guess the question is intent. I, I tend to be optimistic on people's motivations.
54:18
And so I don't think so. I think it's more just that people love to panic. And I mean, we've seen that in the real world in so many ways in recent years. The need to panic, I think it relates to the fact that life is too soft. People don't do hard things.
54:37
And so they need to find things to be worried about to satisfy their nature, their natural evolutionary need to be worried about something.
54:48
And so we just get prone to panic and it's easy to rile people up with this stuff.
54:52
Yeah.
54:54
Any parting notes here?
54:56
Anything we didn't touch on that we should probably mention as it relates to this quantum discussion?
55:03
Oh, I already shilled it once, but I'll shill again.
55:06
read OpTech. I write the changing consensus section of OpTech every month. And I think it'll
55:11
actually put you more at ease about quantum because we cover quantum a lot in there. And you'll see
55:17
the kind of remarkable progress being made, which very likely means that long before,
55:23
because possibly never, but long before even a realistic timeframe for quantum,
55:26
assuming it started scaling today, I think long before it gets to a production relevant quantum
55:33
computer, we'll have a better system in Bitcoin. Like it's happening actively.
55:39
Well, this will be published the day after Optex published. So you're not, you're not
55:43
spoiling anything. Is there any specifics you want to expand on there?
55:49
Oh, sure. So I read this developer Conduition has been posting a lot about cryptography to the
55:56
Bitcoin mailing list recently and to Delving Bitcoin. And he did this big write up earlier
56:02
this month or last month, I guess, about isogeny based crypto that I mentioned earlier.
56:07
And he basically made the argument that Bitcoin developers should be paying attention to it.
56:13
And so I read his whole thing and wrote a summary for Optik about it. And isogeny based crypto is
56:20
very interesting because unlike cash based or lattice based crypto, it's only about twice the
56:26
size on chain of the existing elliptic curve stuff. And part of that is because it's also
56:31
elliptic curve based. But unlike our existing stuff, it doesn't depend on the hardness of
56:37
reversing points to keys, to secret keys, in order to be secure. It has a totally different
56:45
security assumption. It's just based on the same shapes of curves on a graph, right?
56:51
And so I think that is very promising. And I think people should read Conduitions whole post
56:56
if they're kind of even technically interested in this kind of stuff, because he does a great job of
57:01
bringing it down to a place where Bitcoiners who kind of understand elliptic curve
57:05
cryptography, the classical kind, can also understand isogeny-based cryptography.
57:10
And
57:10
because it also works on elliptic curves,
57:16
some of the machinery we already have in Bitcoin could be applicable to it. So we kind of take a new
57:21
cryptographic assumption for the hardness, but we can use some of the same
57:25
optimized elliptic curve math that we have already to work on these systems.
57:30
And so that might be promising for Bitcoin.
57:33
You know, if I'm being totally honest about it, I would guess that if isogeny based crypto were to come to Bitcoin, we would probably still want to do something that's not elliptic curve based as well as a backup.
57:43
So that if there's a fundamental break in elliptic curves themselves, which there hasn't been any evidence of it yet, we'd have a fallback.
57:51
fallback. But it's just promising to see a totally different
57:55
avenue that's not Lattice, not Hash, also being brought to the fore.
58:00
And I'm glad I got to write about it.
58:02
Do you think there's enough top-tier cryptographers
58:07
well-versed on these subjects focused on Bitcoin?
58:10
Do we need more?
58:13
Not an isogeny-based crypto yet. And that's exactly what Conduition writes about, is that more
58:18
Bitcoin folks should be looking at this and seeing if it's suitable, because if it's suitable,
58:24
it would have a lot of really good properties that apply to Bitcoin and let us keep using
58:28
Bitcoin the way we want to. Yeah, great to know. Thank you for all your work on the front lines of
58:36
having the conversation and helping add context, because again, as somebody who's not well versed
58:43
in quantum physics and knows enough to be dangerous when it comes to cryptography, it is
58:47
I don't want to say it's easy to get bamboozled, but it's easy to begin
58:52
questioning, and you should always question, but getting a
58:55
well-rounded perspective on both sides of this quantum debate
59:00
as it pertains to Bitcoin, I think is important. You've been doing an incredible job of
59:03
providing much-needed context. Thank you. Yeah, I
59:08
love funbusting. It's been a hobby for a long time, so glad to be out there doing it on a new topic.
59:12
Awesome. Where can people find out more about what you're working on?
59:16
yeah check me out on x usually at reared encode um i sometimes post on the mailing list as well
59:22
and um if you're building a wallet i offer consulting reviews and stuff for bitcoin
59:27
wallets and similar kind of on-chain bitcoin stuff uh and you can just hit me up on on x about
59:32
that my dms are always open all right awesome brandon i hope you enjoy uh enjoy your day and
59:39
hopefully we can do this again you too man well great talking we'll draw a line in the sand we'll
59:44
say if quantum hasn't progressed in six months.
59:47
Six months.
59:49
Yeah, I'm kidding.
59:50
Peace and love, freaks.
59:52
Okay.