Hacker News Reader

Comment by thinkmassive | Hacker News Reader

> it would benefit both technically and ethically from using non-Tor networks

Obviously the performance of tor is reduced when traffic increases, but performance isn’t the goal, privacy is. Increasing the traffic of "normal" use cases (as opposed to those targeted by state-level opposition) improves privacy for the entire network.

I agree that using tor for all your video conferencing and streaming needs is probably irresponsible, possibly unethical, but the performance is likely to be so poor that it’s not a good use case anyway.

Using tor for everyday text chat (by real humans) is unlikely to have a negative performance impact, and it does provide more noise for attackers to sift through, so it could arguably improve privacy for those who need it most.

View on HN

>performance isn’t the goal, privacy is

I agree, but if performance gets bad enough, people will look elsewhere and to possibly less-private solutions. It's parallel to the whole idea of a tradeoff between convenience and security that I've had to deal with for a big part of my career (and personal anecdotes aside, why adoption for tools like PGP never picked up, at least in part).

Anyway, my original comment's mostly moot as an argument (though I'll keep it around for discussion's sake) -- the founder replied and it sounds like they're working with Tor to make sure the network stays healthy, which is all I really care about.

> adoption for tools like PGP never picked up

nit: Every competent developer I know uses PGP on a daily basis, for signing and validating commits. The responsible ones also use it to validate security-critical software is coming from the expected source. I do agree though, that the original use case of encrypted email is largely dead.

Good point, though in Git's case there's the benefit of: - Users are at least somewhat technically minded - The ecosystem built around Git generally has good automated tooling for verifying signed commits (Gitlab and Github in particular surface that information pretty well)

With security-critical software, specialists take on the burden of knowing how to use it as part of being in that field.

Email sadly has none of those, and the one client I've used with a "normal"-user-friendly GPG interface (Nylas N1) is now a subscription service with limited adoption. Enigmail was better than the CLI, but still not super intuitive for the average person.

if Quiet doesn't catch on to provide that noise, would it be beneficial to flood Tor with LLM-generated conversatinos? We're really good at "cheaply generate near-infinite quantities of human-looking text" these days...

Presumably the content of the message is irrelevant to Tor. If the humanness of the content matters then the game is over anyway because that means it’s being read by a third party.

What matters for Tor is plausible looking traffic and uninteresting people doing uninteresting things with it. Then you can’t use traffic analysis to identify Tor users as points of interest.

If everyone on Tor is evading an oppressive government then oppressive governments can just use Tor to identify the people evading them.

Reply on news.ycombinator.com