The Bitcoin/Crypto Thread

[Doctor Phibes]

I wish these people would speak in plain English rather than use all this technobabble.

[Axtremus]

1. Plain English has been failing at keeping up with technology advancements. Heck, practically all natural languages have the same problem. There will always be subject matter that less than .01% of the population understand. "Plain language" is developed to accommodate the middle one to three sigmas of the general population. There will always be cases when plain language simply cannot do a subject matter justice.

2. But if you really want that compromise to get some aspects of a message across to the middle one to three sigmas of the general population, sure, try a "plain language" approximation. AI can probably do that quite well most of the time these days.

[taiwan_girl]

[jon-nyc]

Well it’s not technically just in its current form.

Shor’s algorithm breaks the math underlying most public key cryptography. But today it requires millions of years of computation whereas a sufficiently powerful quantum computer could do it in hours or less.

[Axtremus]

@taiwan_girl said:

Why does this mean blockchain is doomed?

Only blockchains of certain implementations are doomed. Google's work is fairly specific about cracking the ECDSA 256-bit encryption method in a reasonable amount of time using a quantum computer that Google believes will come into existence sometime in 2029. Blockchains that use the ECDSA 256 (or fewer) bits encryption methods will be "doomed" if/when their underlying encryption can be broken in a reasonable amount of time -- using a quantum computer that may come into existence in 2029.

But blockchain as a general concept will continue to have a place in technology and industrial applications. The U.S. government, especially through NIST, is spearheading the world-leading effort to standardize a bunch of new encryption methods that are expected to be "quantum resistant" -- i.e., encryption methods that even quantum computers cannot break for a very long time. New blockchains can be implemented using these quantum-resistant encryption methods and continue to be secure in the face of quantum computers.

[Klaus]

@jon-nyc said:

Well it’s not technically just in its current form.

Shor’s algorithm breaks the math underlying most public key cryptography. But today it requires millions of years of computation whereas a sufficiently powerful quantum computer could do it in hours or less.

With a 500K qubit computer, which is completely unknown whether that's ever going to work. Nobody knows whether quantum computing will ever be practically useful. I highly doubt we'll have a 500k qubit computer in the 2030s.

Also, cryptographic algorithms can be changed. It just requires a "hard fork" of the blockchain.

[Doctor Phibes]

@Klaus said:

With a 500K qubit computer, which is completely unknown whether that's ever going to work. Nobody knows whether quantum computing will ever be practically useful. I highly doubt we'll have a 500k qubit computer in the 2030s.

More importantly, will it be able to run Crysis at 120fps?

[Klaus]

Quantum computing is way overrated in my opinion. Even if we'd have working quantum computers with millions of qbits, nothing really extraordinary would change. We'd have to upgrade some crypto technology to be quantum resistant. We'd set a few new records in finding large primes. And a few narrow applications, such as certain simulations, would become faster. But it's far from being a general purpose "make my program run faster" machine.

[jon-nyc]

I think you’re too dismissive. It’s true the applications are narrow but that doesn’t mean they’re not impactful. Quantum computing could totally transform drug discovery, catalyst design, or battery chemistry for example.

[Klaus]

OK, yes, there are some quite useful application domains, but in the public media it is often presented as an "everything will be a million times faster" thing, which is completely false.