you are viewing a single comment's thread.

view the rest of the comments →

[–]magnora7 3 insightful - 1 fun3 insightful - 0 fun4 insightful - 1 fun -  (3 children)

Voat is an example. They basically created a "founder-effect" setup, where the early users got more power than the newer users. What ended up happening is neo-nazis hijacked it through dedicated long-term efforts, and the site for a long while literally was just a neo-nazi site. And it couldn't really be corrected because the new users didn't have voting power and had to get 100 upvotes before they could downvote others. So once the echo-chamber was hijacked, it was all over.

Even reddit is another example. There are so many administrators and moderators, and some of them are very bad, and are literally selling their services (the ability to promote or remove certain content). This has created an atmosphere of censorship, and also has filled the site with advertising disguised as content.

With voat a bad community got control. With reddit, the admins and mods sold out. (Which just requires a percentage of them to sell out, not all)

Both are vulnerabilities that are avoided by the "Good King" approach.

What is the real issue? It's what you said- in the end it comes down to a human. New thinking here has to start by wrestling with this fact, because it's not going away.

The real issue would be either automating the administration, so the whole site could be distributed with no one at the helm. But that may be impossible. If humans have to do it, then the only solution is to have trustworthy humans. And I'm not sure of any vetting system that would filter out people who are just seeking power, or people trying to subvert the system, or people trying to sell out.

The only real filter I see is the people who are dedicated enough to spend the time and effort to build their own site, like d3rr and I, or like goldf1sh who built www.notabug.io.

A great solution would be if there were so many active reddit alternatives that if one got taken over or compromised, it wouldn't matter so much because people could go to a new site very easily. But unfortunately people tend to congregate all on one site, like reddit, because they want to go where there is the most activity. I am not sure how to address this particular problem, but there might be a solution.

[–]MostlySunnySkies 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (2 children)

If humans have to do it, then the only solution is to have trustworthy humans. And I'm not sure of any vetting system that would filter out people who are just seeking power, or people trying to subvert the system, or people trying to sell out.

In computer science we have this concept of an NP Complete problem and another concept of a reduction of a problem. An NP complete problem is a problem that is as hard as any hard problem in its category; it's category is a miserable one, for which no known "efficient" solution exists. What we know is that, if a solution for any of the NP complete problems exist, a solution for all of them does also exist because they are all "equivalent" to each other; we say they can be "reduced" to another of their (miserable) category.

This problem you're citing here, the trustworthiness of humans, is an NP complete problem in human affairs. No known solution to it exists, but if it did, then organizations like the FBI would not have to worry about the Robert Hanssen's of the world. But they do.

The thing with NP complete problems is, most developers don't waste their time trying to prove they're not what they seem to be- asymptotically intractable. They (probably) are; so use your life's energy proving something else.

The same thing applies here. If we want to depend on a Good King then our solution will come to incorporate something akin to an FBI vetting process- absent the cooperation their very special powers and reputation elicit from the world. Either that or it will succumb. I just can't see that working in the long term, although in the short tterm you're doing a great job in my opinion.

Small blockchains suffer from an easier to exploit 51% problem. I have some Reddit sub numbers here, and I think they're big enough to make Saidit a hard target for a blockchain takeover, supposing everyone defects and the site continues to grow.

s/announcements 56,238,906 (everybody I presume)

s/AskReddit 27,580,454

s/ funny 30,094,540

and so on with lots of 20 and teens millions.

magnora, early on I was very heavily involved with the web. We had a very early startup; we preceeded everyone in a ton of later to be sucessful technology. Details matter and I am not one of the tech billionaires :( but it's not from a personal lack of vision or work. Online, I got massively burned on a discussion group / news website and I had to decide to walk away from it . I truly feel Saidit's ex-Redditor's pain.

I thought- I'll never do that again ! and psychologically, I haven't; none of my posts are anything I have some big attachment to, as in olden days.

But it got me thinking, like it did a lot of people, about what's wrong with the web.

I concluded we had taken a wrong turn early on and instead of sharing data with each other's connected computers, we aggregated ourselves into what amounts to complex honeytraps run by psychopathic authoritarians.

I never joined FB or Twitter or any of that. I don't even use a smartphone. I take much-more-dramatic-than-average precautions against online threats and treat all my posts and emails and web activity as though they were fully public to my neighbors and I might one day have to stand trial and defend myself for something I said or did. That's my mindset. I don't expect that my activity is public; I reserve my legal assumption of privacy and anonymity, but I act as though a worse case scenario could happen.

Since those days, I've come out of a kind of hibernation of hope. I feel like something's in the air like it was back then. Things are changing and a Big Change over a few decades is going to happen and it can be a Good Big Change if people of good will can be coaxed into applying themselves.

I feel like this problem is interesting to me again. If would be very happy to kick around ideas - nothing is too wild- with anyone who wants to. Feel free to dm me and maybe tell me what /s covers these kinds of things.

[–]magnora7 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (1 child)

I am thinking a good idea might be to have an essay when registering a username. The user must write a 5 sentence (or longer) essay about a given question. Then I can judge the quality of the essay, make sure it is unique and not copypasta or computer generated, and use that as admission.

It doesn't solve the administration problem but it might go a long way to increase user quality, which might make it easier to let the community make decisions because it wouldn't be so easily hijacked.

[–]comments 2 insightful - 1 fun2 insightful - 0 fun3 insightful - 1 fun -  (0 children)

interesting idea for a gatekeeping thing that isn't (easily circumventable anyway) email