I just found this place on an idle lurk through /r/slatestarcodex. As far as I can tell, it's a surviving remnant of the more iconoclastic elements of the rationalist/postrat/lesswrong sphere, similar to the community surrounding Scott Alexander in the pre-AC10 days.
I've visited a lot of similar communities that splintered off of Reddit and Twitter during purges or censorship escalations, and seen them live out their lifespans. Voat, the original, was the most interesting, since it started out in the era when free speech was still acknowledged as universally a good thing, and had majority support, with Pao being seen as a single bad actor acting independently. Unfortunately, it coincided with the crackdown on the entire infrastructure stack, and it eventually became too expensive to keep online.
Since then, a bunch of other communities showed up, each with technically interesting stories.
- .win, anchored by The_Donald's still fairly sizable userbase, is reasonably healthy but has one big community, a few minor ones that are semi-active, and little else.
- TRP and its surrounding subreddits all made independent sites, some of which came together under the banner of TRP.red. It doesn't really seem active at all, though the big subreddits in it are still not banned, even after everything else was.
- A number of Voat clones, all pretty unremarkable, cropped up here and there. Poal.co, which gradually petered out, Voat.xyz, likewise, and Ruqqus, whose admins decided to ban edgy speech after journalists doxed one of them and closed the site due to lack of interest shortly thereafter.
- On the infrastructure side, Kiwi Farms deserves a mention, surviving attacks (both DDOS and infrastructure loss - related) that would have killed earlier attempts handily, rolling their own DDOS mitigation system when Cloudflare backed down.
- During the peak of Twitter censorship, something called the Fediverse cropped up, aiming to solve the fundamental coordination problem of countless tech-savvy dissidents all being willing to throw their hats into the ring, but no one individual having the pull or funding to do it singlehandedly. A number of servers, all set up and branded independently, could connect with each other and aggregate their posts and users. In practice, a few servers exist, but Gab would have been the flagship and it shut off federation early on. Poa.st is the current flagship, largely hosting a core group of imageboard-adjacent users, but the vaunted robustness of the system didn't count for much when it briefly went down a little while ago and, as far as I could tell, most users just went quiet in the meantime rather than jumping to other instances. Most instances don't have open signups, even. Kiwi had its own instance at some point, but no longer does.
- Nostr took the Fediverse idea a step farther, using a bunch of cryptography to establish a protocol that lets users tie their accounts to a file on their computers, and use it across a wide variety of services. It had all manner of very impressive guarantees against censorship, but the onboarding friction was high enough that the only users I could find were a small handful of diehards talking to themselves and each other.
- Amusingly, a bunch of lefties whose views weren't remotely targeted by Reddit's administration made a bunch of federated Reddit clones during the API fiasco. As far as I know, they still exist and are 90 percent low-effort political slop, with the remaining 10 percent amounting to low-effort memes that aren't very good. I'd expect that activity is largely botted, simply because it doesn't pass the Turing Test.
By virtue of sheer stubbornness and dedication to the (worthy) cause of free discourse, The Motte seems to have a core group of devoted users who followed it here from Reddit, but most of what I see is automatically-created weekly discussion threads. I've certainly seen many individuals out there who share its demographic and general sentiment, but all of them are basically individual bloggers or Twitter users who established themselves pre-Closure of the Internet.
All of that said, it seems to me like the survival of our culture depends on finding a way to not just preserve an archipelago of different communities' remnants, but to find a way to re-create the frictionless, organic internet that could easily onboard new people. It seems like that would be a common topic of discussion here, and that people would have opinions on how to achieve this.
As it stands, my best proposal would be something like the ill-fated Dissenter extension, which added a separate comments section to every website as a tacked-on sidebar, but done more elegantly. Dissenter's failure case was that it was banned from Google's app store, but Brave is used widely enough, especially among our core demo of tech-savvy 20-something men, that sideloading a browser extension is no longer a huge hurdle. As for doing it more elegantly, I was thinking of something like:
- Every major social media website gets its own custom integration, bundled with a bunch of quality of life features that users ask for. E.g. region-blocking and a chronological timeline on Twitter, better post filtering on imageboards, and so on.
- When users post to these social media sites, their content is mirrored to a separate, federated service of their choosing (with a default for users who don't want to set up their own), and users of the extension can see when another Reddit/Twitter user is a fellow user of the extension, as their posts will be highlighted in the native view. This provides a backup of their posts and OC, limiting the chilling effect of a high-effort post potentially being removed or deboosted. Additionally, posts can be made exclusively to the extension's Fedi instance, visible to other extension users but not ordinary users of the social media website being interacted with, so that conversations can be had on e.g. Reddit without the local moderators stepping to say that "Y'all can't behave".
- Users can go directly to a given instance's URL and see (and message) other users across all platforms, with extension-specific notifications visible in the notification feeds of all augmented sites.
- As far as onboarding, I think the various anti-censorship social media people have missed the mark by treating friction as a weed-out. People are exhausted with the state of the world, and aren't going to take an hour of their day to figure out what Nostr is on their own. Instead, a simple, memetically-transmissible onboarding image outlining the process in simple steps, flattering the user as a cyberpunk hackerman rather than a refugee or tech enthusiast, could do a lot better. Dating myself, I remember the fun of the first days of 8chan, where users reveled in being technically hardcore and somewhat exclusive while still being happy to get new lads up to speed. I think that kind of culture is vital to anything that wants to bring back what was lost.
Essentially, the minimum viable product is the equivalent of being able to draw a fish in the sand back in the early days of Christianity - a way to indicate to fellow users on Closed websites that they aren't alone, and that there's a place they can get in touch with you if you get b&.
Do people have other ideas, in this general direction? I feel like we're in a critical time, where enough people remember the free internet that it can be brought back without the need to remake everything from first principles.
Jump in the discussion.
No email address required.
Notes -
I suspect the open Internet is permanently dead. The main impediment in the future will not be censorship it will be an avalanche of AI bots.
We already have some small amount of bots that get on here, they are mostly obvious because they are unsophisticated. It can be a human or a bot, but if all of its posts are about this new real estate fad I can easily know it is spam. What if some bot joined, made an intentional effort to sound human, became a valuable contributor over a month or two, and then made a sob story about needing help on their real estate project.
I can honestly say I don't think I'd spot the fake. Even after it giving evidence that the plot to join was just a ruse to get donations/investment. The only way I'd be sure is if multiple users attempted the same thing at the same time.
That's way too high-effort for a practical spamming operation. It also takes way too long to pay off.
More options
Context Copy link
I feel this, as someone who has been around it for like 3 decades now. I've watched the digital frontier get explored, claims staked, fences erected, gardens walled off, and IDs checked at the door. The last one is happening now, between various proposed legislation worldwide and informal methods used in the name of "fighting bots". Within the last few months, Reddit is no longer usable without an account, and I can't even consistently make Google searches in incognito mode.
Some things have certainly gotten better. Video was once a [BUFFERING] exercise in [BUFFERING] patience. Now it streams in HD while I download whole video games. A large local AI model can replace much of the information you'd be looking for anyway. But I do think something of value has been lost.
I'd love to see some good "closing of the digital frontier" stories in fiction.
More options
Context Copy link
I disagree on both counts. AI detection may be imperfect, and it's true that it's easy to circumvent, but most bot operators don't really try at all. If Reddit cared, they could jury-rig something that got rid of the overwhelming majority of botted comments, but as far as I can tell they're either apathetic, inept, or in collusion with the groups operating various bot campaigns.
Fundamentally, I expect the 'defense' side of this arms race to dominate, especially if you decentralize it a bit. Optimizing an LLM to generate spam that gets past most users' custom filters is a huge ordeal, and, as with ads, the people churning them out may well consider the sort of person who operates a blocker to be less valuable as a target anyways because of their implied skepticism. Probably something like the pre-LLM arms race between spammers on 4chan and the users writing filter lists. I think, insofar as imageboards have a future, it looks like a few different users operating custom Qwen instances to screen threads for spam and shilling, then surface only the stuff they might be interested in, and sharing out their models' filter results through an extension. I have an architecture for that as well, which I'm working on building a demo for.
Of course, all of this is contingent on a few tech-adept people having the agency to set up this kind of infrastructure. We take things like Ublock Origin and Return Youtube Dislike for granted, but they're largely a product of the old, high-trust internet in which people were willing to work for free in pursuit of a common good.
Reddit likely doesn’t care about bots because its business model depends on impressions it can sell to advertisers. The bots are impressions, active users, even if they don’t react to ads. If they purge the bots, I suspect Reddit is probably much smaller than they appear.
More options
Context Copy link
There is a cascading failure route. Big website gets overwhelmed with bot spam. People get sick of it and move to medium size website. Medium size website gets overwhelmed with bot spam. People get sick of it and move to small website. Small website gets overwhelmed with bot spam. People get sick of it and move to tiny website. Tiny website is overwhelmed with bot spam.
At each iteration the website owners might care more about getting overwhelmed with spam. But to crack each layer the bots get more sophisticated. The fact that big websites are already overwhelmed is not a good sign, its a sign of things to come.
In my experience of moderating slatestarcodex and themotte for the last decade is that the moderators have an effort advantage. It is easier to ban a user than it is to spin up a new one. We have a default system that excludes new bots. I had to read your post before I allowed it through the filter. But humans do not have any inherent advantage here, unless you start implementing draconian personal identification rules. Which actually benefits the large websites. I've had to do youtube's creepy 'im a real human' process of taking a selfie video.
I just don't like a defense that is basically "my adversaries have been stupid". They can easily get smarter. Especially when costs are in their favor.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link