https://news.ycombinator.com/item?id=25030085

First, my bona fides: I am a huge proponent of decentralization, empowering people and giving them control over their own data, relationships, and identity. I distrust large states and organizations and hold them to a very high standard of not harming people. I have put my money where my mouth is and reinvested nearly 90% of our company’s profits to build open source alternatives to Big Tech companies, and routiney give away our software on github. We have built probably the most useful and battle-tested open source alternatives for Web2 and Web3, in the world: https://intercoin.org/overview.pdf

Now, having said all that… as someone who designs distributed systems that reach millions of people across 95+ countries, I have had to seriously consider my responsibility in designing the systems. It would not be very difficult to circumvent whatever laws various jurisdictions have. But regardless of the laws, consider what your tech is enabling. (I wish FB and others did this, but the capitalist profit motive keeps them from doing it, they have to extract rents and distract you at dinner with notifications and get you addicted to arguing online, and suck you into virtual reality or they lose money).

OK, so now to the point

If you are relying on end-to-end encryption to protect you against state-level actors or police, you have already lost and have been reduced to sneaking around. The real solution is to work together fix your democracy and make it a more liberal democracy, with more sensible laws that allow greater freedom of actions and make the punishments fit the amount of harm it caused, with punitive multipliers that account for the probability of not getting caught.

End-to-end encryption, if you uncompromisingly apply critical thinking and call a spade a spade (which is what we should be doing as designers of distributed software) is just an abdication of any sort of governance about what to do about any speech. Freedom of speech is certainly a lofty goal, and personally I don’t think the CSAM in and of itself is the problem — rather it is the acts before and after the content. Terrorists plotting an attack for example, or any group organizing to harm people. Even financial collusion.

Forget states and think organizations. Consider that organizations find it desirable to know whether an employee was using their messaging system and giving away company secrets or plotting to harm the company. A dating site may want to know if a predator is luring women into a trap or duping elderly people into giving up their money. Sex trafficking and many other harms can be investigated.

Now what is the proper way to handle encryption? Due to dropping costs and minituarization we will soon have ubiquitous cameras and surveillance everywhere anyway (including college dorms etc. to solve allegations of rape). The recorded info should all be encrypted at the camera, and anything sent over the network must be encrypted. BUT…

There should be a process to decrypt specific minutes from specific cameras, following a process that involves a complete audit of those trying to access the footage. For example: only if a court case is brought and the video is subpoenaed can the keys be produced, by having the judge, lawyers and the tech companies come together, and only for specific times and specific cameras. In other words: the answer is watching the watchers to only access the info for the correct reasons and always ahve audit trail, rather than having no possibility of watchers in the first place and not knowing whether a rape occurred or not.

Based on this example with cameras, we can extrapolate to communication and groups. If there is suspicion of a group, our society should have the means to decrypt its messages, but ONLY via mesns that highlight WHY, and WHICH times. Certainly it should be possible to do after a crime is alleged to have been committed, and false claims of a crime would be punished too. The remaining question is rather about “precrime”, and whether we should “chill” speech of eg determined would-be criminals plotting something, rather than letting them discuss it and catching them before they commit their destructive acts. For that, there is still open discussion.

But for the rest — the EFF is wrong. We can have freedom of speech, and yet ways to have due process to investigate speech that was tighly related to crimes committed before and after it. What we should REALLY be doing is making sure our agencies (which don’t have to be top-down run by the State, they could be fulfilling yearly contracts paid by neighborhoods or cities) are using the Accountability software the software industry should standardize. The agencies serving us should be more transparent. Rather than citizens sneaking around, they should demand their government become more transparent. The tradeoff of secrets vs transparency which should be discussed is that of GOVERNMENT. 99% of the time, government secrecy harms society, why do we allow it?

To summarize about how our society SHOULD ideally function:

1. Have neighborhoods exercise consumer choice in agencies and courts, let those face market competition. The vouchers used by neighborhoods can be a single payer system by states, but neighborhoods choose to renew contracts or not.

2. Agencies should act transparently. In the case of a court case, responsible platforms should allow decryption by agencies following specific procedures and the public should always have access to the entire audit details

3. We still need https, ie encryption in transit. We still need decentralization and resiliency, so content cannot be taken down, and people’s identity and choices aren’t controlled by specific third parties.

4. Content can harm society (and be exacerbated with botnets retweeting stuff). Individuals do not have the right to unfiltered megaphones, responsible publishing platforms should require peer review (like in science, or wikipedia talk pages) before disseminating information.

The profit motive and capitalism prevent #4 and co-opt ideals like “freedom of speech” to allow pushing messages to groups and radicalizing them. It is not an accident people globally are increasingly divided and hateful politically due to the Internet.