Strilanc

15 ms·

181.
by Strilanc 4y ago
Could you cite the "much touted paper" you're talking about? In my view, making the noisy physical magic states is the easy part of the distillation process. You reset a qubit, then rotate it 90° around the Y axis, then 45°
182.
by Strilanc 4y ago
Why are you assuming that measurement is the link between quantum mechanics and the real world? Quantum mechanics doesn't need to be linked to the real world, it is the real world. That quote is simply wrong when it says that MWI &qu
183.
by Strilanc 4y ago
Your code is not properly enforcing the communication constraint between the two players. Try to enter code that passes the bell tester widget included in this blog post and you'll find the problem: https://algassert.com&#x2
184.
by Strilanc 4y ago
I won't defend popsci presentations, but formally MWI is pretty precisely defined. It's defined by assuming the postulates of quantum mechanics without including the Born rule. States are unit vectors, operations are unitary matri
185.
by Strilanc 4y ago
Have people actually constructed a concrete example of a theory that's super deterministic in a way that violates the bell inequalities but is complex enough to contain Turing machines driving the tests? The requirement that you can en
186.
by Strilanc 4y ago
In a decent sized quantum Fourier transform, almost all of the phase gates are rotating by extremely tiny angles. For example, in a thousand qubit Fourier transform, 99% of the phase gates are rotating by an angle that is less than 0.000000
187.
by Strilanc 4y ago
What I'm trying to convey is that the statement "entanglement is instant" is implementation dependent. You can meet the specifications of quantum mechanics using a simulator where entanglement is instantaneous, but you can al
188.
by Strilanc 4y ago
Saying it happens "instantly" implies a preferred frame, but the process is invariant under Lorentz boosts. It also assumes a specific way of representing the quantum state (eg a state vector over all parts) whereas other represen
189.
by Strilanc 4y ago
Fun fact: if you use more and more polarizing filters adjusted by finer and finer algorithms, and assume they are ideal (no loss of correctly oriented photons), that block of filters acts like a polarizing filter that filters to one angle b
190.
by Strilanc 4y ago
You're suggesting that the EM field doesn't need to be quantized if everything it interacts with is quantized, so that the non-quantized parts can never be brought to bear and observed?
191.
by Strilanc 4y ago
I really like quantum flytrap. Happy to see you added a bunch of tools to do basic classical processing of the events, so I don't have to hack around that anymore by like making the bombs conditionally transparent. One feature request:
192.
by Strilanc 4y ago
There are often cases where some small thing needs to be dealt with in one map, like flip one switch, and it's hugely disruptive to have to drag everyone to get it done. I also disagree that it can't be fun to work together separa
193.
by Strilanc 4y ago
That's always something I found unsatisfying about factorio, actually. I always use mods like construction nanobots so that I can start copy pasting from the start. The manual precise placement of early buildings is so tedious otherw
194.
by Strilanc 4y ago
I really like Mindustry. One place where Mindustry is clearly better than factorio is the feel of trying to build things quickly. It's very satisfying to lay something down fast . A stumble there is that the keyboard combinations are
195.
by Strilanc 4y ago
Counting the number of amplitudes isn't really the right way to approach it. Individual amplitudes are basically irrelevant to a large quantum computation. It's all about large scale patterns in the amplitudes. Little exceptions t
196.
by Strilanc 4y ago
Isn't the multiverse where one version of you solves the problem and the other versions don't die strictly better than the version where they commit suicide? It's not like their experiences and memories get merged into the
197.
by Strilanc 4y ago
The QFT is the Cooley-Tukey FFT algorithm [1] expressed as a tensor network [2]. Cooley-Tukey has two main steps that are repeated recursively: bulk replacing a,b with a+b,a-b along bit boundaries, and applying twiddle factors. The bit-boun
198.
by Strilanc 4y ago
Quantum computers don't analyze physical systems by reproducing them 1:1 in a different substrate and evolving them. They can be used that way, but it's inefficient. For example, you might do chemistry computations in first quan
199.
by Strilanc 4y ago
Funny, I was thinking quantum computers are actually replaying the transition from analog to digital. The qubits you get from pure analog control are too noisy (1e-3 gate failure rate). You need error correction to get to good error rates (
200.
by Strilanc 4y ago
I realize it's a joke but I'll explain why it doesn't follow. The amount of measuring required to prevent the water from boiling would look more like some impossibly intimate process that was actively and constantly interroga
201.
by Strilanc 4y ago
The "you need to travel the distance between states" speed limit on quantum computation is called the Margolus-Levitin Theorem [1]. 1: https://en.wikipedia.org/wiki/Margolus%E2%80%93Levitin_theor...
202.
by Strilanc 4y ago
This effect is derived from the mathematical axioms that define quantum mechanics. It has nothing to do with instrument error or strobing. In fact it works best if there's no instrument error. Instrument error creates chances for the s
203.
by Strilanc 4y ago
The Zeno effect can still occur if there's no measurements. Let there be a qubit B and a register of N qubits C[0], C[1], ..., C[N-1]. Now repeat the following N times: rotate B by an N'th of a full turn and then perform C[k] ^= B
204.
by Strilanc 4y ago
I don't think that's really a viable strategy in practice in an ecosystem as complex as python's. There's too many libraries and too many little corner cases and interactions around what the behavior should be. For examp
205.
by Strilanc 4y ago
Sometimes I wonder if the main use of quantum computers will just be to verifiably have no side channels, because any such side channel would act as a measurement (which produces phase errors that you can check for). It wouldn't be ef
206.
by Strilanc 4y ago
Another danger using fork is it duplicates the internal state of pseudo random number generators. It's a great way to accidentally take the same random samples in every process, utterly trashing any statistics you were intending to do.
207.
by Strilanc 4y ago
I would say the issue is which float you want. If the int is larger than 2^24 it can't be represented exactly as a float32 so you may have intended to widen to float64 which covers int32 and float32. Or maybe not; maybe you wanted pe
208.
by Strilanc 4y ago
They explicitly say they trust the results. They're complaining that top labs use lots of compute, so the results aren't relevant to someone who can't. They give an example where a paper used 18K TPU core hours. It's eas
209.
by Strilanc 4y ago
Suppose there's 1 woman who had 10 kids, and everyone else had 3. Everyone then has max kids according to the "1 fewer than mother" rule. After 5 generations that 1 woman has 10 * 9 * 8 * 7 * 6 * 5 = 151200 descendants that w
210.
by Strilanc 4y ago
> Everything has correctness issues somewhere. This is fallacy of gray. The blog post isn't complaining that there are non-zero bugs, it's complaining that when you use the language you hit a lot of correctness bugs. More bug

More ›