Hacker News Reader

I’ve regressed to my cynical default in the past few days, but “a large institution that I would trust with this authority” sounds like pure fantasy.

I don't know, maybe it is. Maybe the politics of our world simply are that cynical and won't allow for such a thing to exist. But I guess I'm hoping it isn't. I'm trying to imagine how a non-authoritarian institution would work; it wouldn't have the "authority" to compel anyone to turn over their data, but just about everyone would have enough trust in the community and its decision making process to be okay with it.

What I guess I'm asking is ... what if the organizations we built looked more like the WHO, and less like the CIA?

Personally, I think the only way we will retake our privacy is when hobby level, readily available hardware, software, and 3d printable parts enable hobbyists to create “good enough” phones with encryption and privacy options that won’t exist in mass market phones.

Then an enterprising startup can deliver marked-up, preassembled devices, then people might have privacy again.

The goalposts on "good enough" keep moving. When everyone else has a holographic direct-to-eye refraction display, people are going to be as disappointed in a 2020-level smartphone as people were with OpenMoko devices, which still had a pressure-sensitive touch screen with a plastic stylus when the iPhone has just come out with its relatively massive capacitive multitouch display.

I was disappointed with the OpenMoko for many reasons, but the screen wasn't one.

Things like it wouldn't read my sim (bug 666), the stylus was almost required, but there was no place to put it, audio issues, sleep/wake issues, poor battery indication and tricky to charge from a flat battery.

Sure, the screen was kind of small, but the resolution was high, so you could claim hidpi before Apple invented it?

Depends...'good enough' goalposts are for a particular audience. Technophobes seeking privacy are certainly less motivated to upgrade to a holographic direct-to-eye display, than a technophile.

You can already buy a Blackphone. Almost no-one does.

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

You would still have position data based on triangulation from signal level / latency for different BTSes.

Personally, I like the idea of a Fairphone with Lineage/MicroG.

I say like the idea of as opposed to use as currently Fairphone is out of my budget and I can't find a phone in my budget range that's supported by Lineage/MicroG so I just have somewhat reliable <£100 phone with as much de-googling as I can while still having normal android.. (Not much)

Edit: the bonus of phones in this price range is like Fairphones they almost always make it easy to quickly remove the battery so you're certain nothing's running.

What if information could be revoked? Imagine the government suspects me in some investigation. I give them a key that gives them all information about me and my recent activities. Then once some period elapses, I revoke that key and they no longer have any of that information. Nor, of course, were they able to copy or independently distribute it in any way when they had it.

This is probably pure fantasy, but if there was any way this scenario would be possible, I think it would go a long way to assuaging people's concerns. As it is, we know data can be copied and distributed without permission. That's kind of been the underlying problem of spies throughout the history of the world.

I think this concept is where Stephenson was going with Snow Crash, but his scenario for this seemed like fantasy as well to me.

I maintain that copying is fine in some or a lot of cases because stale data has a diminishing ROi and eventually becomes useless. Doesn’t mean the govt couldn’t amass data from multiple sources to provide a complete picture of you. But you just proposed DRM for your personal data and we all know about the analog loophole...

Every time I think about this issue, the closest I can get is a distributed world of APIs, where everyone hosts their own data [however], and for any service whether it be government, commercial, work, friendly, etc, the entity's agent makes an API request and gets a result. But that result isn't the data - it's a privacy-focused result that can be used in place of the data.

The simple example would be "I'd like to send indigochill a gift", which would send a request to your API (with some notification for you to approve or deny receiving said package) and return an id that could be written or printed on an envelope. That id is unique and the mail service would scan the id and the package would be routed. And the mail carrier would scan the id and get an actual address (or a blip on their map of the block or something). Nobody along the way has learned your address or name or anything - nor do they know my address, which is just another unique id.

There are some very interesting implications in that data can become stale so quickly. If you approved of receiving the package but were out of town for the week, the package could be routed directly to you without any interaction whatsoever. Nobody along the way would know the difference between your home or your hotel room.

The same could easily be done for lots of personal data. Phone numbers, email addresses, etc. All your shopping history could be stored on your own private data store and the local grocery stores would send out a mailer "Address this mailer to all people in the neighborhood who have purchased discounted burgers in the past month". The store doesn't know who is getting the flyer, but the people who would be likely to use a coupon for buying burgers would get a notification and likely go buy some burgers on sale - or reject the notification and not be bothered - to the savings of the store and the recipient.

Social Media could still exist with this sort of thing but the services would be querying engines rather than the creepy data warehouses that they've become.

This of course gets significantly harder with the queries this article is about, but the idea is the same. Instead of "give me everyone's location for the past month", the request would be "give me the anonymized ids of people who have been to x locations for x amount of time within x amount of space between one another". The response would provide data that could be used for the purpose in question, but the underlying data remains private.

I've been pondering this idea on random occasion for about 15 years, but I don't care enough about it to make it my life's work. But I sure would like to live in a world with that level of data ownership and privacy.

Sounds like an interesting possible world that might one day get built.

In the meantime, sounds like most likely outcome is the government hits CTRL C, CTRL V on Facebook/google’s user database and starts a China/Tencent style SQL dB on every American.

Or even more likely, this already exists and now the government can start using it more openly.

This could be possible if the data is stored encrypted in a public ledger (blockchain), which has time-locked script functions that permit another key to decrypt certain pieces of data for a limited duration. I think something like this is a primary goal of most decentralized identity projects, some of which have been underway for around a decade.

Of course none of this stops the person with temporary access from making a copy to save forever. I doubt that problem has a technical solution but would love to be proved wrong.

I think the solution would be sending as little data as possible to answer the query, and choosing which/how many queries to respond to. Any large scale requests for data would have to be designed in a way that minimizes intrusiveness or risks being largely ignored. I don't think there's a great way to counter spoofing or sending garbage results without sacrificing user control, but the onus to de-noise should be on whatever institution is attempting to collate data. Additionally, if it's an opt-in system focused on public good there would generally not be much incentive to mess with it as long as users are somehow verified as being in the community (cert generated/granted once after initial identity check, etc). Sadly this is all very pie in the sky at the moment, the amount of cultural change + added infrastructure required for any system like this at the scale of a country would be insane

That's a good point. A common example is how a bar doesn't need to know your birthday, let alone your home address or driver license number, but right now it's common to hand over your entire id. All they need is a yes/no answer to "is this person 21+ today?"

> What if information could be revoked?

What if information can't be revoked on systems outside your control? That's the reality, however many secret handshakes you do, blockchain, etc, eventually some person or process knows something and can do something with it (e.g. copy it).

You could pull a gmail and say, there is no human in the loop, it's just a process of personal data ingestion and ads come out the other end. That might be more comforting if it was open source, but even open systems have things can be hidden in obscure technical details.

The truth is, we are all human, including our leaders. Your average human is weak and susceptible to trading a vast privilege like having control of everyone's whereabouts for favors (e.g. money, position, control, sex, etc...).

I'm a sucker for an open source process (e.g. robot) that would look at the data and tell us who needs to be quarantined, but again, you still have human technical people, with all their failings, controlling the system (e.g. source code). How much do you trust the open source process is a matter open for debate.

It exists. Its not that people "don't allow" non-authoritarian institutions/solutions. Its that they have their own limitations in dealing with complexity. Rationality is Bounded. Know what the bounds are. Read some Herbert Simon - https://en.wikipedia.org/wiki/Satisficing

Look at the EU. There is no dearth of cockups there, even though their instinct at everything is non-authoritarian.

From a sovereignty perspective, that's a nonstarter.

I think you just described Facebook.

I agree but it's also the most compelling fantasy in human history.

I disagree that "large" authorities could be fantastic. Authority, like most things, is better when smaller. For one thing, a small authority is more likely to notice those people subject to it.

Just to be clear, I don't subscribe to it. But I have noticed that pretty much any time anyone has a good idea, their next idea is to force everyone to do it.

I would trust the CCC or the EFF with this I think, and the other paranoid fighters for privacy. But only as long as it takes for the to rotate their members enough for takeover by others.

We've seen that wit smaller political parties as well. When membership suddenly become a viable path to power things change in the structure and people.

>But only as long as it takes for the to rotate their members enough for takeover by others.

This is the key. The power makes joining their ranks appealing to those that would abuse the power.

I agree. Once you cross the point in organizational size where people stop feeling personally responsible for the organization and start feeling responsible to the organization it starts behaving like a sociopath in pursuit of whatever it's official goals are (shareholder value, catching criminals, etc).

You can mitigate this to some extent by having one or a couple people who are personally responsible for the whole organization (royal families in monarchies, the owners of privately owned companies, etc) but then your organixation is not resilient to bad leadership.

I agree. Ironically the measures that should have increased national security are endangering social contracts established around fundamental rights in a way the opposite is achieved. But I guess you are preaching to the choir here...

But that aside, I don't think the measure would be worthwhile even now. Instead focus should be to fortify local health services and provide support where needed.

A damaged believe that fundamental rights are compromised has of course negative repercussions on the willingness to cooperate with authorities of any kind.

Trust cannot be enforced and authority is always limited. I wouldn't call it paranoia, because in that case the respective fears aren't confirmed that often by people now in solitary confinement because some hot shots thought they were in control and did was was allegedly necessary.

>> paranoid about privacy issues

it is not paranoia but an acknowledgement of the times we live in.

>> our secret agencies are actively planning to exploit this crisis to be able to spy on us in the future.

this is based on their past history. Like you said, it is entirely due to utter lack of trust.

Yeah this is kinda how we got the Patriot Act.

Which also has a sunset clause that will never ever be allowed to expire. All this talk of temporary authority is a joke.

I had this conversation with my dad a month ago, are we not already past the point of no return on this? Can’t the government legally buy purchasing data from credit bureaus and other vendors? Can’t they legally buy location data commercially from at&t and other entities? Don’t they already do it anyways to monitor food shortages etc,?

It’s not a question of can they, but have they yet. Don’t police officers already have access to this data therefore the feds definitely do?

The boom in location based advertising a decade ago was the final nail in this coffin because it made the availability of location tracking data required and highly accessible.

> a responsible institution that would use it only for the purpose of saving lives and then delete it when the crisis was over

The UK managed to get rid of ID cards in 1948 on pretty much this basis. The problem is, that war ended, but the one started on 9/11/2001 hasn't. For temporary authority we need both clearly defined end conditions, a belief that they will be reached, and faith in the authorities to work in good faith towards the ending of that crisis.

It just shows that different people have different triggers that makes acceptable for them to restrict basic rigths and liberties. For some it is terrorism or child pornography, for others it is pandemic.

We do not really need to find agreement, it would suffice if mobile operators implement opt-in that could be easily enabled in web portals. Then society can just advertise this opt-in.

But the data already exists. Google has it.

The smallest step up from that is running an open-source algorithm on Google's servers, for corona detection.

From that situation, maybe we could figure out a way that people would trust, enough ?

Privacy should be a right. In the USA it should be up there with the Bill of Rights. The only way it can be gone around is with a warrant or something similar. Cops don't like that and neither do police state regimes.

When they turn in the extra tracking, will non-American data get collected too?

> This is the sort of problem that could be solved in a better world; collecting location data to find interactions between people would allow you to get them tested faster

NO. Global data about who interacts with who and where should NOT be available to anyone. Especially NOT to HEALTHCARE orgs, I couldn't imagine less trustworthy and skewed-incentives institutions! At least you can trust that a country's secret service like your NSA would keep most of that data safe (as in "safe even the eyes of from judges & prosecutors unless you have a national security issue" etc.). Like in I'd rather have +10k deaths from preventable contagious diseases per year and take the chance of being amongst them then having this on a wide scale.

> civic trust in our community organizations and institutions was built up over time

You have no reason to trust anyone and anything. Stop doing so.

> a responsible institution

There's no such thing. Any institution should be considered thoroughly corrupted in the absence of evidence to the contrary.

> our secret agencies are actively planning to exploit this

It's their job to do so! If you want to provide solid security in globally connected world the "protectors" need to be a bit above the law - it's OK as long as what they collect stays in their archives and you prevent them from holding political or financial power.

The point is to build societies that can function TRUSTLESSLY and where everything is SECRET BY DEFAULT unless explicitly shared.

That's the only way to have an open society! Sure, in small isolated pockets you can have your "trust" in institutions and stuff.

Otherwise it's a choice between:

(1) accept framework of trustlessness + secrecy-by-default + broad exceptions for security agencies mostly as long as they DON'T share data with anyone (no, not even to "save lives", unless you're taking more than thousands...)

(2) drop the global interconnectedness, drop most free travel and work (I don't mean relocation - you'd also need to drop the kind of temporary travel that can't be easily monitored!), move to a more localist society - here you can have trust in the institutions of your own "pocket" that you voted for yourself.

Unfortunately it seems that the pendulum will now swing towards (2) as you seem to say you prefer - ...too bad, I'm sure we could've made the global-village thing work too :( probably we need to try again when tech advances a few extra steps, eg. neural-interfaces + pervasive-AI + usable distributed financial tech. and use the tech properly to make the openness and interlinking irreversible!

You describe anarchy. Capitalism exists because such trust is not achievable.