It is time to build a new internet
Every morning that we open our phones and laptops, we are staring pitifully at a necrotic digital organism.
It's sad and fucked up, no way around it, and there is no reason to go on what I'll call the Obvious Internet at all anymore IMO unless you are morbidly curious about how far we have fallen.
There are tiny safe zones, like secret closed invite-only forums (i am part of one and it is great), Hacker News is still solid, and bearblog has lovely little corners. Kagi makes it easier to find these remnants of our more honest past. So all is not yet lost. But it is so, so close to being lost. And there is no longer a true decentralized public commons for discourse in the digital realm.
I keep wondering if maybe I should make a private, invite-only forum where I can invite some interesting friends and acquaintances (including zany founders, reclusive poets, eccentric engineers of all kinds, high school teachers, homegrown philosophers, garage tinkerers, and beloved drug-addled futurist artists), but I think this would quickly be ruined too. I would want the network to become tree-like, each initially-invited friend or acquaintance of mine inviting their own interesting friends, but eventually someone with a profit motive would wiggle their way in. And as long as companies and the people who work for them have access to the same internet you are using for your forum, there is a profit motive to ruin that forum. It is hopeless to try to keep them out.
I want somebody to build a new internet, straight up. I'm going to post this to Hacker News to see if a wayward, bored, ambitious computer scientist will take this on. I am not a computer scientist; I understand some basic theory stuff but only insofar as I've read a lot about it.
This will be a shitty/abstract PRD for the new protocol and how we should govern it.
Problem
I do technical writing and founder content for work and I agree with everyone that marketers are a virus (for the most part. I'm biased obviously but I do think certain kinds of founder content and technical content, when done right and in good faith, is fun and interesting.)
I'm not even trying to pass any kind of moral judgment. Viruses aren't evil. They're just doing what they do: replicating within an open host system.
That said, it does suck for the host system, of course. We keep trying to fix the problem of this virus by creating new platforms and open forums and stuff, but that is like trying to stop a virus by generating new cells within an infected host. Obviously it's not going to work.
We had Reddit, which was good for a time, but now every other comment is clearly AI generated or at least heavily AI assisted. And any community that has users that can be funneled toward a product is full of astroturfed product endorsements. Twitter has decent profiles and threads but there's obviously lots of trolling, bots, and mindless AI-generated shit there, too.
Discord, group chats, and certain subreddits are decent. But LinkedIn as a good-faith networking site is trashed, Reddit is trashed, Google itself as a search engine is trashed, and I have no idea what Instagram/TikTok/Facebook look like right now because it has all sucked for so long now. Threads just pissed me off from the jump, so I don't know what's going on there.
The internet experiment is yet another depressing example of humans failing to be conscientious stewards of an open commons. The things that made the internet great (democratized access & contribution permissions, low barrier to entry, and ease of use) has also made it insanely easy to pollute.
I don't know if we can moderate our way out of this one. I think we have to throw it away and start fresh.
Of course, this doesn't need to be a mass exodus. Those who are satisfied with this version of the internet are more than welcome to stay. I am not interested in bankrupting Google or the companies who depend on Google. I just want to read interesting things, learn, and have gratifying discussions and debates with people who want to do the same.
Goal
This is going to sound absurdly over-earnest and idealistic, but that's just the kind of person I am and so it's unavoidable.
I want a living, growing, evolving archive of knowledge, curiosities, and stories akin to the most ambitious, collaborative, and good-faith intellectual explorations ever created. Something similar in architecture and spirit to the Talmud or the Buddhist sutras.
The Talmud contains centuries of impossibly well-considered commentary and debate, preserved with a precise emphasis on faithful representation.
Buddhist sutras are another great example of what an ideal, good faith commons of knowledge transfer and commentary would look like: it contains contradictions and axioms and collaborations that span hundreds of years, and, like Reddit when it was good, commentary and debate are a core facet of understanding and engaging with the texts.
Then, of course, we have the best collaborative scientific traditions, where papers are in conversation with prior work. This is true of all academia, ideally.
None of these are perfect examples, but you get the idea. Obviously, each of these collaborations has its failures, politics, and suppressions (which is innate to human-generated efforts at community and work.)
But my point is, basically, they didn't have ads in them, they were generated by humans, and the texts that survived were written, studied, and discussed largely in good faith.
For a modern example: Wikipedia and large swaths of Hacker News. By and large, those who participate in and moderate those sites and their contributions are doing their best not to corrupt it, because they understand these places as somewhat sacred. There is some awareness that polluting these places beyond recognition runs counter to virtually everyone's goals, and therefore people try to keep things decently cleaned up.
So, the goal is to create a digital arena where good faith engagement is the norm. We know this is possible, at least for a time, because the internet was like this in the early days. There were obsessive early web forums and mailing lists where people did serious thinking about subjects they were passionate about.
Everyone involved took their participation seriously. At this time, moderation was basically a byproduct of the general peer pressure that comes with being part of a self-selected community with a shared goal of connecting over their interests.
Non-functional requirements
What we basically need is a completely new protocol stack that is not interoperable with TCP/IP and won't be for a very long time, because making it interoperable would be 'near impossible', or require a huge technical lift that is clearly not worth the squeeze.
We'd have to make it so our internet is as incomprehensible to the current one as Mandarin is to a dog raised by a family who speaks only English.
Getting a dog who only vaguely understands "walk"/"dinner" and a few other words to understand a completely disconnected language would require fundamentally reconfiguring how dogs work.
It'd have to be THAT hard to move packets of information between the current internet and our new internet for this to be even a short-term solution to what's happening online.
So no TCP/IP, no DNS, and no HTTP/HTML. I don't know what would replace it, but there must be no infrastructure overlap at all.
Getting access to this new internet would also necessarily have to be a pain in the ass. Part of the reason the current internet sucks is because accessing it and contributing to it is absurdly low-friction. So why not litter? It's as easy as dropping your McDonald's bag on the sidewalk.
We'd need to do something new with the backend to make this possible. Then we'd have to make it so that parsing what needs to be done to access this arena is somewhat technically fraught. That way, unless you are EXTREMELY invested in getting into it, you really can't.
How we would keep it from getting enshittified
This would be nearly impossible lol. But just incase it is possible, I want to quickly outline what would be required of participants for this to work.
We would have to not treat the commons as a market, ever. This is difficult for us. We would have to not look at this new internet as yet another mine for carving out, processing, and hawking one type of value for another.
This was somewhat easy for the early internet because its relatively high barrier to entry selected for someone with a technical curiosity that was greater than their commercial ambition. Way early on, the general public didn't realize computers had so much money in them, so they attracted hobbyists mostly.
So one way to keep it from getting enshittified, again is to make the barrier to entry just a touch higher, or add friction.
The other thing we'll need to do is somehow shift the incentive of participation to something altruistic, like honoring the commons. We'd have to somehow encourage people to optimize for integrity vs. engagement or some other exploitable metric.
I am aware that this sounds like I am designing a network for saints, and that saints are not the median user. But I don't think the median user is a good thing for us to design for in this case. The median user will probably just keep using Google for stuff.
This network would be for a specific kind of user who values community and thoughtfulness more than money or popularity. There are probably not a huge amount of people like that. So the network will necessarily be small.
I don't know, this project would be hard for animals obsessed with optimization like ourselves. We universally have a hard time acting against our short-term interests even when we know it is destroying something we value/need. But we might as well try, and implement the best guardrails we can, even if it first the network we create is clunky, overzealous, and non-user friendly. Some things are more important than frictionlessness and a sleek UI.