I always wonder… can a truly open and free new internet be built? What would be the options in doing something like this? Maybe running on existing hardware (fibre, towers etc.) to a certain extent…
Sure, we could do something clever with mesh network access points, or use tunneling (VPN) to build a pocket network on top of the existing Internet (TOR does something generally like this to create a more anonymous Internet). So if this were simply a matter of infrastructure, the tech is already there.
However, there are two problems. The biggest problem is adoption. What service can our little pocket network provide that would convince the lay person to tap into such a network? How are we going to advertise this to others? Even if we had our own copy of the current internet’s infrastructure, we would have a cool webpage and spread by word of mouth and they would still have advertising dollars. Either we need a killer feature (that they can’t simply replicate) or else they’ll just win over the average person by the pillow talk of advertising bucks.
However there’s also a philosophical problem. To create a open internet, it has to be available to everyone and our problem is that includes the asshole corporations we don’t like. The fundamental nature of an internet is to be an interconnected network. By building our own separate network, we’re fundamentally creating a walled garden network, not an open network - it’s essentially defined by who we’re keeping out.
But I’m not going to leave you without a solution. Here’s the framework of what I think we need to do to fix the internet†:
We need to stop treating internet access like a consumer good. It needs to at least be treated as a utility, i.e. as something that has an inherent monopoly and doesn’t self-regulate through the process of supply and demand - there is only one internet, no substitute exists. Heck, I’d argue that internet access should be a human right, a tool that fulfills a basic need for connection and communication.
We need to restore network neutrality, ISPs need to be content neutral, because if they can pick winners and losers, they’ll make private deals and pick the winners that work best for them (often another branch of themselves). Since we lost network neutrality formally in the USA less than a decade ago, the internet still looks kinda mostly open, but it’s eroding slowly.
We need to separate ownership of the physical network equipment from the ownership of the information services. Let’s call these ‘equipment ISPs’ and ‘general access ISPs’. The physical equipment should be owned and maintained by small companies, ideally with about 5-10 field technicians (the physical footprint that covers will vary based on the setting, dense urban settings will need more companies than sparse rural ones). These small equipment ISPs will not be allowed to negotiate directly with the consumer. The Access ISPs will be the ones that will lease an IP address to the general public as well as basic services such as DNS, and will compete on general service quality (up/down/latency speeds) that they’ll have to negotiate with equipment ISPs to ensure quality of service, access ISPs can also sweeten the pot with things like offering an email address or bundling with media services(e.g. Netflix), etc. Equipment ISPs should be expected to have deals with multiple service ISPs, and be prevented from having exclusivity deals. Ultimately, the goal is to allow the general public to have options about which ISP they choose that’s not fundamentally limited by where they are at, and the service ISPs are then on the hook to work with the equipment ISPs to fulfill those promises. Equipment ISPs are being given a small monopoly, but if they perform shoddy there’ll be neighbors on all sides to shame them, also they’ll have to work with at least one or two access ISPs to have any income at all.
Start choosing people over brands. The biggest crime corporations perform against humanity is to take credit for the work that is ultimately done by unique, talented people, then internally treat people as fungible assets that can be let go once they’re not useful. lemmy.world is administrated by @ruud and a small team of admins (check your instance’s sidebar for more details). If @ruud and lemmy.world split and he created a new, different Lemmy instance, I’d follow @ruud to the new insurance because he’s proved his talent at weathering the problems of keeping a service up and running in the modern internet, whereas lemmy.world … is just a domain name. Google wasn’t nearly as evil when it was still run day-to-day by Larry Page & Sergey Brin. Valve rakes in money, but Gabe Newell keeps the company priorities on actually being a good game platform. By contrast Steve Hoffman is hated partially because it often feels his job is to be the face of an otherwise obscure board of directors and he serves them in a way that he doesn’t serve his employees, the moderators, or the users in general.
Overall, that’s four things we can do. None of them are easy. One is on the global level, one on the national level, one on the state or local level, and one on the personal level.
†I live in the USA, so my perspective is through that lens, but I’m trying to offer ideas that should generalize.
Yeah. This project doesn’t go down as far as hardware, but it did take pains to make it as difficult as possible to extend it in bad ways: https://gemini.circumlunar.space/
I always wonder… can a truly open and free new internet be built? What would be the options in doing something like this? Maybe running on existing hardware (fibre, towers etc.) to a certain extent…
I’m gonna argue ‘no’.
Sure, we could do something clever with mesh network access points, or use tunneling (VPN) to build a pocket network on top of the existing Internet (TOR does something generally like this to create a more anonymous Internet). So if this were simply a matter of infrastructure, the tech is already there.
However, there are two problems. The biggest problem is adoption. What service can our little pocket network provide that would convince the lay person to tap into such a network? How are we going to advertise this to others? Even if we had our own copy of the current internet’s infrastructure, we would have a cool webpage and spread by word of mouth and they would still have advertising dollars. Either we need a killer feature (that they can’t simply replicate) or else they’ll just win over the average person by the pillow talk of advertising bucks.
However there’s also a philosophical problem. To create a open internet, it has to be available to everyone and our problem is that includes the asshole corporations we don’t like. The fundamental nature of an internet is to be an interconnected network. By building our own separate network, we’re fundamentally creating a walled garden network, not an open network - it’s essentially defined by who we’re keeping out.
But I’m not going to leave you without a solution. Here’s the framework of what I think we need to do to fix the internet†:
We need to stop treating internet access like a consumer good. It needs to at least be treated as a utility, i.e. as something that has an inherent monopoly and doesn’t self-regulate through the process of supply and demand - there is only one internet, no substitute exists. Heck, I’d argue that internet access should be a human right, a tool that fulfills a basic need for connection and communication.
We need to restore network neutrality, ISPs need to be content neutral, because if they can pick winners and losers, they’ll make private deals and pick the winners that work best for them (often another branch of themselves). Since we lost network neutrality formally in the USA less than a decade ago, the internet still looks kinda mostly open, but it’s eroding slowly.
We need to separate ownership of the physical network equipment from the ownership of the information services. Let’s call these ‘equipment ISPs’ and ‘general access ISPs’. The physical equipment should be owned and maintained by small companies, ideally with about 5-10 field technicians (the physical footprint that covers will vary based on the setting, dense urban settings will need more companies than sparse rural ones). These small equipment ISPs will not be allowed to negotiate directly with the consumer. The Access ISPs will be the ones that will lease an IP address to the general public as well as basic services such as DNS, and will compete on general service quality (up/down/latency speeds) that they’ll have to negotiate with equipment ISPs to ensure quality of service, access ISPs can also sweeten the pot with things like offering an email address or bundling with media services(e.g. Netflix), etc. Equipment ISPs should be expected to have deals with multiple service ISPs, and be prevented from having exclusivity deals. Ultimately, the goal is to allow the general public to have options about which ISP they choose that’s not fundamentally limited by where they are at, and the service ISPs are then on the hook to work with the equipment ISPs to fulfill those promises. Equipment ISPs are being given a small monopoly, but if they perform shoddy there’ll be neighbors on all sides to shame them, also they’ll have to work with at least one or two access ISPs to have any income at all.
Start choosing people over brands. The biggest crime corporations perform against humanity is to take credit for the work that is ultimately done by unique, talented people, then internally treat people as fungible assets that can be let go once they’re not useful. lemmy.world is administrated by @ruud and a small team of admins (check your instance’s sidebar for more details). If @ruud and lemmy.world split and he created a new, different Lemmy instance, I’d follow @ruud to the new insurance because he’s proved his talent at weathering the problems of keeping a service up and running in the modern internet, whereas lemmy.world … is just a domain name. Google wasn’t nearly as evil when it was still run day-to-day by Larry Page & Sergey Brin. Valve rakes in money, but Gabe Newell keeps the company priorities on actually being a good game platform. By contrast Steve Hoffman is hated partially because it often feels his job is to be the face of an otherwise obscure board of directors and he serves them in a way that he doesn’t serve his employees, the moderators, or the users in general.
Overall, that’s four things we can do. None of them are easy. One is on the global level, one on the national level, one on the state or local level, and one on the personal level.
†I live in the USA, so my perspective is through that lens, but I’m trying to offer ideas that should generalize.
Yeah. This project doesn’t go down as far as hardware, but it did take pains to make it as difficult as possible to extend it in bad ways: https://gemini.circumlunar.space/