Comment by forgotpwtomain | Hacker News Reader

Comment by forgotpwtomain | Hacker News Reader

I think this point of view isn't just wrong it's actively harmful.

It completely ignores what happened - which is that various government agencies skirted around constitutional law, subverted public discussion of the matter [0] and have still not been brought into adequate compliance (Since it's incredibly hard to demonstrate standing and not have the case squashed [1]).

And after all this the author is saying the problem is that the users aren't part of the political process? I would say until powers and programs employed by government agencies are brought into accountability (let's not forget the CIA monitoring the computers of it's own oversight committee [2]) and a transparent debate is allowed to take place, there is precisely nothing better to do and encryption everywhere is precisely the correct populist and democratic way to fight it.

>Just this past week Kazakhstan announced that a "state root certificate" would have to be installed on all computers wanting to use SSL/TLS/HTTPS out of the country.

Yes backwards nations like Kazakhstan might be implementing "state root certificates" but this is just clearly demonstrative that the power is on the side of the users in this debate. Isn't having these kinds of crazy measures on ballot in Western countries precisely forcing the political process the lack of which the author of this paper is criticizing?

[0] http://www.forbes.com/sites/andygreenberg/2013/06/06/watch-t... [1] https://www.eff.org/cases/jewel [2] https://theintercept.com/2014/03/05/congress-intelligence-co...

I think a very important point was raised though.

Before crypto, basically all guarantees where conditional on a judge's say so. With crypto this changes. The issue also comes up, in a clearer way, with crypto currencies. There is no way to deal with fraud or mistaken tranfers in bitcoin.

That loss of intervention hurts, and we gotta think about it. Even though no government has given a satisfactory solution, that doesn't mean there isn't a problem in need of solving.

Before crypto, basically all guarantees where conditional on a judge's say so

Tell that to J. Edgar Hoover.

That just shows they were conditional on things beyond a judge's say so. I.e. security can also be broken by people willing to break the law.

One day I will actually see someone write a response starting with "that just shows" that doesn't boil down to "My being wrong shows that I'm actually right. Somehow."

>There is no way to deal with fraud [...] in bitcoin

Don't we have an extensive legal system to deal with fraud? The fact that a lot of payment providers elect to deal with fraud themselves probably shows that that legal system isn't working as smoothly as it should; but that's really something that should be fixed in the legal system, not in the payment channel.

Part of the way the legal system deals with fraud is by recovering money (where possible). With bitcoin, unless the destination wallet can be obtained, that is not possible.

With cash, you seize the cash. Bank accounts can be frozen, Credit card transactions canceled, wire transfers flagged (http://www.npr.org/sections/money/2016/01/29/464859624/episo...).

With cash, unless the wallet in which it resides can be obtained, that is not possible. I fail to see the difference.

I agree, the legal system doesn't do well when it comes to recovering cash from entities that can hide it. You can get a legal judgment against a company, but if it's fly-by-night, it winds up being a worthless piece of paper, and can become very difficult to recover.

The law does actually have ways of extracting cash from people even when it can't locate the cash in question. It's harder, and less effective, but overall it works more than it doesn't.

Yes, but that works the same for Bitcoin.

It could work the same for Bitcoin. It actually works better with Bitcoin as there actually is a complete transaction log.

However, the law has frequently decided it can't treat Bitcoin like it treats cash, which has created problems...

Looks like they have other ways of handling it with Bitcoin: https://www.usmarshals.gov/assets/2016/bitcoinauction/index....

It's much harder to hide 500,000 dollars in small bills than it is to hide Bitcoin.

Hiding a wallet is much easier. You can even have several copies of it, so that you can hide it so well that you do not even care if you lose it.

Law enforcement has well-developed capabilities to surreptitiously follow the movement of people and objects through the physical world and wield whatever force necessary to sieze them where desired.

Bitcoin moves much faster and much further that cash though.

> Before crypto, basically all guarantees where conditional on a judge's say so.

Sadly, they weren't.

> With crypto this changes.

Actually, with crypto it is much harder to claim that a judge's authorization isn't required to even try to subvert the privacy intent.

> There is no way to deal with fraud or mistaken tranfers in bitcoin.

There's no way because legally the system (for the most part) considers it outside the scope of fraud laws.

> That loss of intervention hurts, and we gotta think about it.

TLS doesn't create a real loss of intervention. It requires that the intervention be far more explicit and/or targeted.

> with crypto it is much harder to claim that a judge's authorization isn't required to even try to subvert the privacy intent.

No doubt, but the issue is that a judge's authorization can be ineffective in the face of crypto.

> TLS doesn't create a real loss of intervention. It requires that the intervention be far more explicit and/or targeted.

That targeted intervention is really easy to block for those who want too. All it takes is not revealing a key/passphrase.

> There's no way because legally the system (for the most part) considers it outside the scope of fraud laws.

Not just that though, it is also impossible because the conditionals introduced by a court system are very hard to encode in a crypto system, even more so if you want to do this and be able to retain trust.

> No doubt, but the issue is that a judge's authorization can be ineffective in the face of crypto.

Only when it should be. Crypto just means you need a secret to access the data. If you can't compel the secret, why should you be able to compel the data?

> That targeted intervention is really easy to block for those who want too. All it takes is not revealing a key/passphrase.

You say that like it is an easy thing. Try keeping your key/passphrase protected against an adversary that can replace/intercept your keyboard, your touch screen, your mouse, your network, etc. It's just a lot more intrusive and a lot more work, which is why it has to be far more explicit and targeted.

> > There's no way because legally the system (for the most part) considers it outside the scope of fraud laws. > Not just that though, it is also impossible because the conditionals introduced by a court system are very hard to encode in a crypto system, even more so if you want to do this and be able to retain trust.

Those conditionals were all choices made by the legal system. Those rules can always be changed if so desired. If they aren't... maybe it is because they shouldn't be. ;-)

This point is a tired falsehood. There have always been phenomena with physical rules that supersede any human decision. The legal system recognizes and works with such boundaries - some even end up codified in our highest laws such as the 5th amendment.

We're only in this present quandary because of professionally irresponsible "web 2.0 services" that siloed up user data for their own surveillance goals, creating enticing targets for governments' desires of control.

When everything is decentralized, it's going to be much harder to enforce any rules about no-crypto. For now things like social networks are cenetralized, leading to all these temptations to mine the data. It's not just government who wants youd data - it's every organization!

http://magarshak.com/blog/?p=169

Until very recently 99.99% of conversations where completely private and society functioned just fine. Even with crypto everywhere the governments have far more access to what people say and do than they had for thousands of years.

People with power pretend if they just had more power everything would be better. But, reality is if everything on a computer where private not much would change.

> Until very recently 99.99% of conversations where completely private and society functioned just fine.

Until very recently, most conversations were plausibly deniable, and off the record. The internet and cryptography changes that.

Edit: [Both plausibly deniable in the formal/cryptographic sense, and in the informal "You're misremembering, there's no record in writing of me saying that" sense]

> The internet and cryptography changes that.

By Internet, I think you mean phonographs, magnetic tapes and pocket recorders.

Cryptography brings some of that back.

To get access to those, you needed warrants, and you didn't get to listen in on every conversation, only targeted conversations. On the Internet, you have full access to pretty much everything I've ever said to somebody.

They aren't equivalent.

> Until very recently 99.99% of conversations where completely private and society functioned just fine.

In fairness, until very recently it was also extremely difficult for two people to privately plot to murder hundreds. Technology has changed the balances on a lot of scales.

To be clear, I'm not advocating for government back doors in encryption here. But I am saying we should not pretend that encryption just puts things back "the way they were". It most definitely does not.

> until very recently it was also extremely difficult for two people to privately plot to murder hundreds. Technology has changed the balances on a lot of scales.

Was what McVeigh did highly technical?

It was dependent on having modern technology, like a truck and cheap fertilizer; it's a lot harder to kill 168 people (and wound almost 700) with an oxcart and a crossbow. Even with a carriage and black powder you need a _lot_ of people to be in on the plot. (Black powder was not cheap, not deniable, and not easily transported or concealed.)

What about poison? or fires? I'm sure there's other means I can't think of right now

Modern tech has made it a bit easier, but it was still fairly easy throughout most of history.

It was not easy to get enough poison to kill hundreds. You get some hemlock, dump it in a well, and it dilutes enough that a dozen people get stomach aches and one elderly person dies. It's feasible to pull this off on a larger scale, but you'd need a lot of poison, and therefore a lot of resources.

Fires? I'm not sure it was that easy to set an out of control fire that would consume enough of a town to kill hundreds. Does everyone just fail to notice as you splash pitch on buildings all day leading up to the final act of starting the fire at night?

Pitch, or indeed much premeditation at all, wouldn't be necessary in a village with five foot alleyways and dried grass roofs. Additionally, tar was a common caulking material. Oh, and running water/chemical foam wasn't usually a thing. And, you know, no trained quick response teams, or well maintained roads and swift transit systems to escape on, or any way to tell the difference between the smoke that always hung around the fireplace and the smoke from the roof being on fire... I could go on.

Face it people -- terrorism is a problem of technology.

Exactly. Primarily the technologies of mass media and democracy.

Of terrorism, the newscasters on TV who sensationalize the act and the politicians who fan the flames for their own benefit are just as guilty as the person who blew themselves up.

(Though obviously, only the latter is guilty of murder)

I take your point, but I think it's important to distinguish two issues that are raised by modern technology here.

One is about information. Modern technology allows new kinds of communications and learning and collaboration. It also means that for the first time it is technologically feasible for governments and other powerful organisations to monitor entire societies pervasively.

Another is about automation and efficiency and scale. Modern technology allows individuals or small groups to do many wonderful things. It also potentially allows individuals or small groups to cause far more damage than they could before, whether that is physical, emotional, economic, or any combination of those.

There is a certain amount of overlap. For example, some of the implications of widely distributed surveillance are concerning, but some of the implications of not just monitoring but actively influencing large groups of people by modifying (overtly or otherwise) the information and communications they receive are downright terrifying.

However, I think these are two quite separate issues in principle, and they each introduce risks of their own along with the benefits. Encryption is the counter to one form of risk in the communications area, and in that sense it does roughly speaking bring back the privacy that everyone used to enjoy automatically. Restoring that privacy necessarily means that bad people can use those communications channels to discuss doing bad things as well, but they could discuss those things in other forums instead if they needed to. We need other ways to control the risks of new technologies enabling larger scale threats, regardless of what we do about communications and encryption.

What are you talking about? There have been plenty on bombings and shootings for hundreds of years.

200 years ago there were "plenty" of bombings and mass shootings orchestrated by groups of 1 or 2 people that resulted in hundreds of lives lost? Can you provide a few examples of this?

https://en.wikipedia.org/wiki/Gunpowder_Plot

Over a dozen involved, zero lives lost (excepting the conspirators).

That's a rather artificial limitations, few modern plots are from 1 or 2 people with hundreds killed. Historical examples are generally things like fires not bombings / mass shootings. As an example of the risks: https://en.wikipedia.org/wiki/Great_Fire_of_Meireki for example was extremely deadly, though estimates of ~100,000 dead where probably exaggerated.

We have also forgotten a lot of this crap. EX: 40 to 81 people killed from a time bomb in 1875. https://en.wikipedia.org/wiki/Alexander_Keith,_Jr. Note, this was for insurance money not terrorists, but I don't think the dead care.

Some historical examples where donations where probably accidental. The Delft Explosion of 1654 killed 100-200 people. In 1769 ~3000 people where killed when a depo @ Bastion of San Nazaro in Brescia was struck by lighting. A lighting also sparked another detonation in 1856 killing ~4,000.

Not to mention smallpox and other infectious diseases where one person could kill thousands easily, even if unintentionally (with some instances intentional).

> That's a rather artificial limitations, few modern plots are from 1 or 2 people with hundreds killed.

I've heard small conspiracies called out to justify mass surveillance. You don't really need mass surveillance of your own citizens unless you are concerned about very small conspiracies. Bigger conspiracies can be uncovered with more traditional detective work.

Freak fires are also not a great example. An arsonist cannot cause hurricane force winds to assist and push a fire through a city. This is effectively an act of God. The time bomb is a better example, though again I would ask for evidence that this was a fairly common event the way mass shootings and similar terrorist attacks have sadly become.

> I've heard small conspiracies called out to justify mass surveillance.

The fact that something is a convenient and effective narrative for propaganda purposes doesn't mean it is a significant and common threat.

200 years ago world population was 1/10th what it is today making most events more common now. Something like 5% of everyone ever born is alive right now.

In terms of most kills by a single person one sniper is credited with ~700 kills and a mounted knight in plate was expedition to equal 100 peasants on average so 200+ in exceptional cases seems reasonable especially considering the Spanish conquest of South America where in the 1500's where ~300,000 vs 2000 and the 2000 won with minimal casualties. https://en.m.wikipedia.org/wiki/Spanish_conquest_of_the_Inca...

Mandela spent 20 years in jail for plotting to blow up a hotel, I believe? And that was nearly a century ago.

Your timeline is skewed. Mandela was born nearly a century ago. His involvement in bomb conspiracies was in the 1960s, about a half century ago, when most conversations were not guaranteed private. Wire taps have been around as long as phones.

Gunpowder plot of 1605 is a rather famous counter example: https://en.wikipedia.org/wiki/Gunpowder_Plot

That's really not a counterexample to my point. This was a conspiracy involving over a dozen people. It's a plot that unraveled because someone leaked the details.

This is why technology has changed the balance. You don't need dozens of people to gather enough explosive material to pull this off today. And with fewer people involved, traditional detective work is harder, and those involved are less likely to leak details.

There are plenty of accidental detonations that killed hundreds which could have easily been caused by just one person as noted in another post. Though, wartime sabotage is not exactly terrorism.

Again, if you want to kill lot's of people disease, and fire where the most common historical examples. It's actually much harder for a lone person to kill thousands today than it was 500 years ago due to effective methods of firefighting.

When you talk about two people plotting to kill hundreds, do you mean any two people or do you mean two people who are otherwise physically apart but are brought close together by technology? Coz if the medium of communication used by your two would be mass murderers is VoTA (0), that technology is as old as human beings.

0. Voice over The Air. :-)

> There is no way to deal with fraud or mistaken tranfers in bitcoin.

This is a social problem of bitcoin. The bitcoin people consider this a feature.

However, it can be dealt with at an exchange. "MyBitcoinExchange(tm) only allows transfers to other members of MyBitcoinExchange(tm) and yourself. All transfers will be escrowed for 5 days to ensure that they are not fraudulent."

The problem is: nobody in bitcoin would use that exchange.

People who want non-traceable will stay away. People who are using it to sequester wealth out of countries will stay away. People who want fast liquidity will stay away. etc.

So, who's left? Um, people who are perfectly adequately handled by existing credit cards, thanks.

> With crypto this changes.

It shouldn't. Listening in on a conversation should be illegal whether or not the conversation is encrypted. But we realized that if we weren't encrypting our conversations that our rights were being illegally infringed upon, so we started taking matters into our own hands.

Well, two things:

1. I've always maintained that privacy is a political and social problem, not just a technological one. Nothing stops a more powerful actor from breaking encryption with a rubber hose if they really want to. Technological solutions are still needed since there's a variety of actors involved, only some of which will obey the political doctrine.

2. I often wonder if encryption will make us think of privacy a bit like we think about rule breaking in sports. Since there's an official, we assume that you can break the rules if you don't get caught, at least on the field. There's no ethical pressure to "do the right thing" because we've offloaded that burden to the officiating crew. Has technology become a bit like the official in this case? Our physical spaces have strong and protected rights, at least in the US, but our digital ones do not, and that may be because we are protecting them in other ways.

How does a rubber hose work when the people are talking via Tox or via .onion sites?

Rarely does encryption exist in a complete vacuum. Just as we cannot necessarily listen to every one on one conversation, those conversations become important when related to something occurring withing the "physical" world. I think we've seen that Tor does not protect you from someone who's determined to find you (see: Silk Road).

Indeed, at least with Tox, all I need is a separate key exchange, somewhere. In person works wonders. So does random email, or instant message, or you name it.

At least for communications that need that layer of security and anonymity, it's certainly there. And as I implied, rubber hoses won't work on anonymous and encrypted digital identities "somewhere else".

At least regarding Silk Road, that's because he made amateur mistakes by posting his real name and then changing it on StackExchange, about making a e-commerce .onion website.

And that was the beginning of his fuckups, with the final one by paying for an assassination attempt.

> And that was the beginning of his fuckups, with the final one by paying for an assassination attempt.

That's my point. At some point he needed to leave the digital world and live in the real world. There is no perfect false identity, not one that does anyone any good. Whether you are looking to overthrow a regime, sell drugs, or just keep your recent pregnancy on the down low, at some point that leaves the digital space and enters meat space. It's at that point that powerful actors can get what they want from you.

Encryption might help you evade detection, but it is not a guarantee of privacy, even for a regular person.

Believe it or not, I expect we agree in the end. The existence of encryption is precisely why dragnetting doesn't work for catching actual bad guys. What brought down Silk Road was old fashioned police work. Get people on the inside, record some chats, get some warrants and so on. They seized him after he had decrypted his laptop, even. It's also precisely why I think political solutions are going to be the most effective in the end. The expectation of privacy we have established outside the digital space (say, no wiretapping, no illegal searches or arrests) still gives tools for protecting the public. Perhaps we need to somewhat tweak those tools for the digital world, but that doesn't mean we need to give the government the keys to our house.

I think OP was talking about torture in order to obtain keys.

And the GP is asking who will be tortured if you can't get the identity of the interlocutors.

You can even hide that a conversation took place.

They might own a significant share of Tor relays and exit points, likewise for Bitcoin.

I've also worked on quite a few methods of using Tor and Hidden Services, where my whole network is in .onion .

I can send data to .onion addresses natively via DNS ( https://trac.torproject.org/projects/tor/wiki/doc/LinuxDNSre... ). I can also receive data from Onions as well.

Right now, I have Node-Red communicating data from my IoT infrastructure, publishing data to my MQTT broker elsewhere (my server onion). I have my 3d printers in another place, and part of the MQTT data is pictures and GCode readouts from them.

It sounds complicated, but Hidden Services allows me to treat a network of disjointed machines (all behind NATs) to be thought of as being on one huge ethernet hub. And I can code against an [hash].onion and know that my data will reach that machine, no matter the IP.

Oh, a purely tool based solution to privacy has failure built in to it. It's a necessary component of a larger system.

Rubber hoses don't scale. With our technological solution, we make it so that the government can't access 99.9% of all conversations. If they want to access a conversation here or there, they can use a rubber hose. But if they want to see every conversation in a person's social network (including conversations they weren't directly involved in), then they aren't going to be able to do that anymore.

Technology helps a lot. It brings us from a panoptikon to a situation where enforcement has to be done on an individual basis. That's a big win for freedom in my book.

This is an important point that the author, even in his comments in this thread, has no answer for beyond scaremongering that "the state" will "break" us all. You can't secretly beat everyone with a rubber hose to let you spy on them.

To "backwards nations" you can add the following, all of which have floated laws to weaken/backdoor/ban crypto:

France

UK

USA

Russia

Put USA at the top. Clipper chip says "Hi".

Reply on news.ycombinator.com