4 ms·
I was an author of one of the first papers examining bitcoin's lack of anonymity. People understand things a lot more now, but from this thread, in my opinion
by feral 9y ago
I was an author of one of the first papers examining bitcoin's lack of anonymity.
People understand things a lot more now, but from this thread, in my opinion people still don't 'get it'.
The traceability issue is not magically removed by mixing - at least not in steady state, once the system matures.
To see this, consider a future where a single large exchange, lets call it Xchange, has substantial market share - (maybe like Coinbase). Let's
assume they are a respectable company regulated by the laws of the USA, and that law enforcement tells them they have to comply with existing laws on stolen property.
Some company Bobcorp has $10M of bitcoins stolen, and pays someone to follow the coins. If those coins ever show up in the Xchange's addresses, Bobcorp is going to call the police and file suit to recover them.
Obviously there aren't literal coins in bitcoin, but Bobcorp is going to go to court and explain their block chain analysis to the judge, and the judge is maybe going to ask Xchange where the coins come from. And that'll be an interesting court case. But it seems unlikely the judge will be persuaded by ideological arguments about how this not-really-fungible currency should be treated as if we all pretend its fungible, and will instead be asking Xchange for its KYC records about the customer that lodged the coins.
Now, here's the important bit: the above doesn't have to actually play out. All that has to happen is that Xchange starts worrying about the above playing out, and starts getting a tiny bit wary about accepting funds that come from reported thefts - starts seeing them as a tiny bit riskier, a tiny bit more likely to end up in court, than other bitcoin.
Once any non-trivial amount of entities start thinking like this, (and it seems obvious they will - I would if I was running any business; but imagine the most conservative bank who eventually reluctantly participates in the maturing ecosystem) then the 'stolen' coins instantly become worth less than obviously clean ones.
And as soon as you live in that world, what happens mixers?
Suddenly you run the risk of putting in your good clean coins, into the mixer, because you care about privacy, and getting back tainted coins, implicitly worth less.
Suddenly mixers become a lot less useful.
It seems inevitable to me that if bitcoin continues to grow that this is how things will play out. It's still a super interesting system but, like I've been saying for years, it's got a ton of downsides around anonymity and privacy, and in steady state may be the perfect Orwellian currency rather than the perfect cypherpunk one, and I don't think mixers make this go away.
- skycaptain1 9y agoCan you post a link to the paper? If not, just the title would be helpful. Thanks!
- feral 9y agohttp://arxiv.org/abs/1107.4524 http://arxiv.org/abs/1107.4524 (This is also in my profile.)
- MichaelGG 9y agoSo it's in everyone's interest to fight this every way possible. Coinjoin or better on the tech side, lobbying on the legal side. Otherwise you end up with a traceable currency that can retroactively be stolen. What if someone cashes out before Bobcorp notices? At which point does Bobcorp get to file a theft notice? It can't just be an open registry since then anyone can temporarily block anyone's coins. After a court order? That could take forever. What about when they wanna seize money from Silk Road 5 or MegaUpload3? Plus you'll end up with coins dirty in one country but not another. China and Venezuela probably want to claim any coins that went through their country in some way - they're violating currency export controls no? At least Monero provides far more fungibility.
- feral 9y ago>What if someone cashes out before Bobcorp notices? People (exchanges etc) will be wary of accepting large transactions which have moved a lot recently from untrusted sources, whereas coins that have sat in the same wallet for a few weeks won't be discounted against as much etc. To give some intuition advice here, think of it like how a risk management department of a financial institution would, trying to price grey risks; not like a distributed systems engineer does, trying to reason about edge cases as either Boolean correct/incorrect. ("Hmm, guess we'll have to discount fast moving coins from people we don't know" vs "If the hacker sells it fast enough the whole taint thing is unworkable and everything will be fungible again") Analogous to https://xkcd.com/538/ https://xkcd.com/538/ - regulators and financial institutions will take the $5 wrench approach!
- lawn 9y ago