The case against encryption: it is the wrong solution to society’s problems.

> I no longer think that end-to-end encryption is the right solution to human rights problems. If citizens are reduced to sneaking around and denying their activities to survive, their governmental system is way past due for fixing.

I think you and I have fundamentally different, and likely incompatible, views of government. I believe government will always spy on its citizens, on other governments, and on their citizens, to the extent its technologically able. I believe this because this has been the case since governments, monarchies, etc existed. (In the 21st century we can add that "corporate entities" are doing the same level of spying on their members, their competition, and everyone else they can.) I should hope Snowden would have put a fat emphasis on that point. I believe no amount of progress, voting, reform, etc will change that governments do this. Yes those things can, and do, make a difference, but governments will always find a way around or through it, while, for now, governments seem to be able to keep corporate entities in check when they want to.

Pre computers, "technologically able" required lots of footwork. Governments were necessarily limited in their ability to spy on everyone. Today that limit has been removed. This is why I am a firm believer in encryption for everyone, everywhere, always.

I would say not that "citizens are reduced to sneaking around and denying their activities", but that citizens need encryption to check their governments' (and corporate entities') ability to spy on them.

You and I could definitely agree that our "governmental systems are way past due for fixing", but I believe there is no truly fixing them because they are made up of corruptible people. At best, journalism can inform the people of the corruption, and the people can be well-armed to check military might (but good luck nowadays fighting a tank with an AR).

I have idea on how to solve for this. I am currently on a military deployment and would like to turn this into a business when I return. The idea basically involves inverting the web to create a decentralized operating system. If both end points are IPv6 there is no server. Crypto will primarily be used for certificates and signatures as a trust scheme in a new security model. I don’t care what people share as the prototype I am working allows direct sharing of the local file system, like copy/paste in a GUI but different file explorer windows may or may not be different devices/users.

I conscientiously disagree. At the end of the day you can't seperate the obfuscatory encryption mechanisms from those for repudiation, and even if you did, nothing would stop a government from tending to create carve-outs for itself for obfuscatory encryption, and by denying it to everyone else you've paved the way for abuse of those without encryption by those who do.

You're right in the sense Encryption does not fix the problem of securing human rights. That is absolutely true. By definition though, the only safeguard one has from their self-sovereignty (in the John Locke-ian sense, not the Sov Citizen sense) being violated is the tools, technology, and relationships at their disposal. Human Rights security is not guaranteed for anyone from anyone. There is only a collective commitment amongst particular groups of citizens to only breach those rights in certain ways, and the agreements brokered between that particular group, another groups on how things will be done.

Also, to be frank, ubiquitous end to end encryption merely resets us back to the pre-telephony or dragnet metadata collection days of investigation. Before SMS, and internet traffic was all in the clear with enough third parties collecting enough metadata to make a data trail, no one had the ability or means to know your thoughts or social network, or all of your interactions trivially. Before digital payment processors, the couldn't even know what your financial activity was with a quick call to your bank/Credit card company.

This is what everyone out there seems to hope everyone forgets. Investigation was hard. It was necer intended to be easy, and arguably, shouldn't be. Networking, and by extension Third Party Doctrine changed the face of it. Encryption may restore some of it.

People may be uncomfortable with the idea, but at the end of the day, tge scale of atrocity enabled by preventing people from looking after themselves is always greater in scale than that resulting from the small portion of the population that abuses it. You can work around that small group. History has proved that is a rare thing indeed once something has been taken away, for it to either be restored, or for it's restriction to do anything but disempower the law abiding citizen, who was never the issue anyway.

> To be honest ... I no longer think that end-to-end encryption is the right solution to human rights problems. If citizens are reduced to sneaking around and denying their activities to survive, their governmental system is way past due for fixing.

You're arguing with a point that nobody is making. End-to-end encryption is not a guaranteed solution to human rights problems. The absence of it increases the probability of human right violations. There's a big difference.

In the rest of your post you're just plugging your own work and mixing up terms like crypto and encryption and all of that is so irrelevant that it seems that you aren't interested in talking about end-to-end encryption after all.

I solved this problem a few years ago, and applied to YC and was rejected. I struggled with how to make money from it, but have decided that it's not relevant anymore, and will work on cleaning it up and releasing it. Try as the govt will, they have already lost the fight on encryption. It's possible to build apps that use encryption where the govt can never force you to backdoor the app. If they really want to force backdoor then they must be going after the device manufactures.

Why not both? Encryption may not get at the root of the problem, but at least it helps. Never let the perfect be the enemy of the good.

To the extent encryption is a band-aid to help mitigate the real problems, it MIGHT be good. But unlike a bandaid, it might exacerbate the problem, since it allows criminals who would otherwise be foiled early to organize a horrendous act and then invariably someone (mob or government) blames the “type” of person (Muslim, Black, Mexican, whatever) and starts profiling innocent people with false positives, exacerbating a totally avoidable sectarian problem.

The biggest problem is when encryption is focused on to this extent and defended religiously. Groups of people start to believe in the ideological position that encryption is essential for human rights, and this sucks the political capital out of the room for developers and statespeople to work together to develop real solutions.

We see this in other areas — such as when we focus on helping manage health problems of already obese people in the USA when we should be focusing on the massive industries that put sugar, HFCS, aspartame and other things in food, genetically engineer corn to have sugar, factory farms that pump their animals with antibiotics and create superbugs etc.

Or how we focus on end users recycling plastic and banning plastic bags while putting zero pressure on the corporations producing the goods to switch to biodegradeable plastic. They are left free to externalize the cost to the environment because we don’t hold them accountable — precisely because we focus too much on fixing the problem further downstream. The root solutions would include a UBI funded by pigovian taxes.

Same with other “downstream fixes” such as getting more minorities represented in a particular field when they’re applying at 30 years old instead of fixing the systems that they are shaped by growing up...

What is common to all these disparate areas is that the root problem isnt fixed, because we keep diverting attention from it with misguided “downstream” mitigations. We create a groupthink and virtue signaling environment where lots of influential and conscientious people who could have been focused on the root problem instead spend all their money and energy on the premise that if only we did MORE downstream mitigation, it would somehow make things better. It delays the real solutions, perhaps for decades.

What we really need is to decentralize the personal data in many places, and use zero-knowledge proofs for attestation, but that is different than encrypting and hiding information.

Should we make bulk collection of data infeasible in huge amounts? Sure. But let’s see if there are any uses of encryption — as opposed to signing and securing data - that are truly indispensable to society and are the right and best solution to solve the problem.

Within organizations, auditing and accountability are very much desired. Banning users and so on. Even though I am a libertarian, there is no simple solution to “just always have a flat system” - hierarchies always form due to efficiencies of scale (eg to gang up on people and defend against that, etc.)

What we should instead focus on, imho, is programming incentives from the beginning to promote checks and balances among the most powerful entities in the system. Look at the US Constitution and how the system has endured. Look at Bitcoin miners, Ethereum miners and so on. That is what we need - to have benign rules for everyone and make sure it’s super hard to maintain a collusion for long to overturn these rules.

Otherwise you can geek out on encryption all you want while the AI-enhanced government will track all of your physical movements and speech with cameras, gait recognition, speech recognition on vibrating potato chip bags through windows, will figure out what you’re all planning through network analysis and precrime, and nail you with parallel construction and social credit score penalties, and if anyone tries to help you they’ll get nailed too. Think it’s far fetched? China and Palantir are already doing it. Social credit systems are actually more benign, jailing and physical coersion and brainwashing are worse. Encryption won’t save the Uyghurs from re-education camps, nor profiling of Black youth in USA, nor will it help foil the next terrorist attack or mass shooting, so it doesn’t help either side of the equation that exacerbates the other in endless escalation and reprisals. And if you really don’t want those things to happen, you need to focus on architecting the governmen’t technology to be more benign in its rules, rather than facilitating sneaking around and delaying the need for the real solutions!!

Don’t listen to me only. Listen to the venerable Randall Munroe: https://xkcd.com/538/