YHK - yet another hacker news clone

Vitalik Buterin backs crypto ‘bunker mode’ amid rapid AI math advances

49 points firstcomm 2 hours ago 37 comments

spottedmarley 2 hours ago | parent

It may be a plausible concern at some point in the future but, once it's possible, it would manifest as an extremely expensive and time consuming attack against a targeted address, it would require an enormous amount of compute. So, I suppose the first line of defense would be to never keep more funds locked under one key than it would cost an attacker to decrypt it.

warkdarrior 1 hour ago | parent

Cost of running an attack is super cheap if one buys botted computers (it is/used to be $25/1000 machines).

_ink_ 1 hour ago | parent

Which leaves Bitcoin between a rock and a hard place. The Satoshi funds are probably worth an attack. At the same time they cannot move to post quantum cryptography, because that would also require to move funds.

spottedmarley 1 hour ago | parent

Yeah the Satoshi treasure would be a prime target. Maybe we see those funds get mysteriously split up at some point? But we wouldn't know if they were hacked or protected..

tzone 1 hour ago | parent

There are exchanges, custodians, etc that hold insane amounts of BTC or ETH in a single address.

We are talking billion dollars+ worth in a single address. So if either chainskey security is compromised in a way that it is conceivable to brute force it, there will definitely be plenty of targets.

spottedmarley 1 hour ago | parent

Yes, it would require those exchanges to move their funds around. I'm pretty sure they can manage it.

dist-epoch 57 minutes ago | parent

Stealing a billion dollar address is kind of worthless. Try cashing it out. Even if you are NK.

search_facility 1 hour ago | parent

May be Ethereum should fund 10000 agents to pursue new crypto math with provable strongness/hardness

Lerc 1 hour ago | parent

I think much of the the doomer propaganda is paid for by by the half billion or so that Vitalik Butering gave to the Future of Life institute. It sounds like he wanted it spent on research more than advocacy and is now distancing himself from them.

what 1 hour ago | parent

Why should we believe that LLMs are going to break ECC?

plesiv 1 hour ago | parent

Your probability distribution should be flatter at least.

tptacek 1 hour ago | parent

LLMs probably won't break curve cryptography. Cryptography researchers armed with LLMs are a different story.

contingencies 1 hour ago | parent

Don't forget LLMs can now design efficient high speed hardware systems, a traditional slow-down for many attackers which represents a significant barrier to scalable next-gen attacks falling. Unfortunately memory prices are through the roof, but that isn't always significant.

what 50 minutes ago | parent

Okay and why should we believe that?

tptacek 43 minutes ago | parent

It depends a lot on your priors. I think making existing cryptography researchers hyperefficient and much more mathematically capable is likely to generate some disruptive results, but you may disagree.

nextaccountic 1 hour ago | parent

It's a risk

cidd 1 hour ago | parent

You don't

EA-3167 1 hour ago | parent

It’s always tragic when the new hype train runs over the old one. It pains my soul to imagine those gentle folks with millions in crypto being subjected to the incredibly foreseeable consequences of their limitless greed.

notnullorvoid 1 hour ago | parent

My gut is telling me pushing security model to hashes seems like a far far worse bet if you are worried about advances in maths.

palmotea 1 hour ago | parent

> My gut is telling me pushing security model to hashes seems like a far far worse bet if you are worried about advances in maths.

Aren't hashes far less reliant on math tricks?

My understanding is the problem with public-key crypto is it extremely reliant on a single math trick (e.g. the factoring problem), so if it turns out that trick was weaker than was assumed, the whole thing crumbles.

tptacek 1 hour ago | parent

Yes. Symmetric cryptography doesn't generally rely on the existence of "trap doors", so there's no direct relationship in any sense between the inputs and the outputs. Huge portions of the cryptographic attack surface are foreclosed in these constructions.

MattPalmer1086 1 hour ago | parent

The point is they have less algebraic structure than things like elliptic curves. If you want "math resistant" properties, this is a good thing.

jMyles 1 hour ago | parent

That's the opposite of what my gut is telling me.

Hashes are so simple. There's no place for these crazy math attacks, which rely on rethinking long-standing and otherwise reasonable assumptions, to hide.

Surely we can all agree that there's no way to reverse a modulus.

dist-epoch 1 hour ago | parent

Isn't it more secure mostly because of the mixing than the modulus.

tptacek 41 minutes ago | parent

I think I agree with you in spirit but I don't think "there's no way to reverse a modulus" is the right way to say this. Like that's in a sense what Coppersmith does? (Several attacks on RSA can be thought of as in some way "reversing a modulus").

jMyles 27 minutes ago | parent

> I think I agree with you in spirit

holy heck, the planets have aligned, HN

;-)

tptacek 5 minutes ago | parent

We both work in security, right? That shouldn't surprise you so much.

simonw 1 hour ago | parent

Crypto's greatest weakness has always been security. Even if the algorithms themselves hold, humans need to keep their keys safe. We have decades of experience now showing that humans can't do that.

I wouldn't be worrying about the algorithms so much as I'd worry about all of the end-users who are going to get their keys stolen and their wallets drained thanks to the deluge of new vulnerabilities in operating systems, browsers, wallets, personal agents and other software.

contingencies 1 hour ago | parent

https://xkcd.com/538/

masfuerte 43 minutes ago | parent

It happened in the UK last week:

https://www.bbc.com/news/articles/c6eq84eygz0qo

j2kun 1 hour ago | parent

FWIW, there has already been a lot of effort thrown at AI attacking lattice problems underlying PQC without anything to show for it. I'm not sure why Vitalik is suddenly worried about lattice problems in particular, unless he has some insider knowledge.

dist-epoch 1 hour ago | parent

The AI unit distance conjecture proof used high-dimensional lattices and algebraic number fields, adjacent to the kind of stuff used in lattice PQC. It shows skill in that area.

int32_64 1 hour ago | parent

The guy that made the 'bunker mode' tweet is an Ethereum guy, the chain that they reorganized when the wrong people lost money in the DAO hack.

What good are strong cryptography primitives for cryptocurrency if your "CEO" can successfully make calls to reorganize the chain?

whattheheckheck 27 minutes ago | parent

When it makes sense. Eth classic was a psy op from cardano guy.

pants2 7 minutes ago | parent

That was over 10 years ago now, when the chain was less than a year old. Things have changed. Can we drop it already?

gumby 22 minutes ago | parent

I'm a bit concerned about starting a flame war, but I barely follow any cryptocurrency news at all. So I am curious: would there actually be any negative macro consequence if this were to happen?

I know, at the micro level some people would lose their money, but does bitcoin (or any cryptocurrency) have any practical, legitimate use with visible impact on the global economy (or even the national eonomy of any country with major footprint)? And I'd be sad for some of those people. but would it have any impact that I would notice? My money is in broad, boring index funds.

gucci-on-fleek 1 minute ago | parent

> but does bitcoin (or any cryptocurrency) have any practical, legitimate use with visible impact on the global economy (or even the national eonomy of any country with major footprint)?

I've read anecdotes on HN that it's used quite a bit for remittances to some countries with unstable currencies, but my understanding is that nearly all of these users are immediately converting it to the local currency, so the only money at risk would be that which is in active transit. It would definitely be unfortunate to lose any money being transferred, but remittance recipients don't tend to have any savings at all, so I don't think that it would have any broader economic effects (talking only about this specific example).