Discord Alternatives build and hosted on the Autonomi Network

Press
1

Given this has happened, below…

The danger of using centrally controlled services like Discord which can be shutdown by nation state governments may affect our community’s ability to freely communicate at any time, if we don’t fit the current political narrative/policy of the nation state, in this case the USA being the likely controller given Discord is US VC financed , and HQ’d there.

The topic is set up to encourage discussion and to explore alternatives to Discord, which can be set up on the Autonomi Network to give the community the same capabilities with community continuity (no disruption by nation states where Autonomi does not fit their platform and narrative.)

3 Likes
2

It would be a great project to create a decent place for the community to be able to communicate and collaborate effectively without being tied to old systems.

Given what’s already been shown to work with Friends and Atlas, I expect it’ll be possible with the right skills & focus.

Hopefully Project Decorum will kick into gear before too long now there are some interesting proof-of-concepts functioning on the network already.

With the decorum team + Riddim (Friends) + Josh (Atlas) + others, and surely we can have some awesome communication options on Autonomi before too long :slight_smile:

Having some profile on discord / x / other centralised platforms will still be necessary for outreach and building awareness, but I hope the aim will be to get people on Autonomi hosted systems as soon as possible.

2 Likes
3

I think there is nothing strange about the American authorities asking the heads of centralised platforms what measures they intend to take to prevent radicalisation and narratives inciting violence and political agitation. No one intends to restrict or “shut down” these social media platforms, but since they are platforms that can block users, their comments or delete them, etc., they must also be responsible for ensuring that there is no hate speech, agitation or other negative activities. When Google removes a video that it does not like because of its political tone, that is okay, but when it is supposed to be responsible for verbal aggression by users, it says that it has no influence over it…

In Autonomi, the matter will be clear: the network functions as a public space and the team has no influence on users’ statements or actions, so MS does not have to explain itself, and the authorities will have to prosecute directly the user who broke the rules or the law.

2 Likes
4

There are a ton of open source platforms out there that we can modify. Discord is closed, so we can’t start from there.

The first question is what do we want? Do we want a classic forum approach like Discourse (aka what is powering this site)? Or more of a global social media feel like X, Facebook, etc.? Or maybe we go the 4chan or Usenet route and allow anonymous/unregistered accounts. What about live updates and notifications?

Do we want full featured enforced moderation (like here) or leave that up to the user (blocking, ignoring, etc.)? Maybe both options should be available? Maybe a federated system of moderation that users can optionally subscribe to?

So that’s first, what is the UX. Next is the matter of how to handle the backend. Every platform I’ve encountered has some kind of database running on the server managing puts and gets. We don’t have that option, there is no server. Clients handle this on their own (like what I did for Colony metadata), or we need some kind of zkVM or other trusted compute environment in place to verify the correctness of data being written and handle the transaction for the user. We can’t allow a rogue entity to corrupt the database. A core piece of development will be to create some kind of database leveraging the Autonomi datatypes and network. If we can make the interface be like your standard SQL type database, it will become much easier to bolt this onto the wide array of existing platforms. Maybe this paradigm doesn’t fit Autonomi and we’ll need to create something from scratch, we would need to investigate.

Now all that said, I’ve been thinking about making a system with a UX almost identical to your classic text based Usenet clients. Something like pan, though we could make it prettier. This is something we could build in pretty short order and it would be fully client side, no servers, no trusted compute, just posts shared between users, 100% Autonomi. If this is something people would use, I can get started on it. I actually have the database side drawn up on my whiteboard. But only if people want to use it.

3 Likes
5

In Autonomi, we have the ability to create multiple identities at our discretion – official, anonymous, hidden, etc. – and they are not registered. If you mean network updates, they are automatic and we do not need to pay attention to them, unless you mean application updates? I’m not sure, but I think applications will also be able to update automatically.

Due to the way the network operates, only self-moderation by users will be possible, but it will not be possible to delete public content, only to obscure it. However, depending on the design of the application, some forms of user blocking by the community will be possible. Another form, if we agree to it, will be the possibility of moderation by an AI bot according to implemented guidelines, but I am against this.

One of my thoughts concerns the idea that the forum is an important element of the project, and I believe that the approach you described is the most sensible one. I think that a forum should reflect public space in the real world, where every user would express themselves through their official identity in the same way as they do when talking, exchanging views or discussing in the real world, where people address each other with respect, communicate according to certain rules and maintain a certain level of decorum (fortunately, this is the case on the Autonomi forum!).

In my opinion, statements that are considered false, offensive or violating someone else’s dignity should be obscured by the community, but I would also like to introduce something I call “reputation”. That is, when someone makes a statement using their official profile, and the statement contributes something to the conversation or provides valuable knowledge, the community gives a plus, and otherwise it can give a minus - such a system would mean that everyone would care about the quality of their statements, their substance, and would also try to gain reputation, which would naturally improve the quality of the forum.

Next, I think about an algorithm that would cause, for example, a plus received from a user with an official profile who has a high reputation to have much greater “power” for the person receiving that plus, while a minus received from a person who uses a hidden profile and has a low reputation would have much less “power” for the effects of the user being evaluated. Such a system would naturally create a high level of substance, conversation culture, discussion, etc. The system could even be more advanced.

3 Likes
6

Lets see if we can eat our (related) dogfood.
Is there something in GitHub - dirvine/communitas: Communitas: A decentralized collaboration platform built on the Saorsa P2P ecosystem. Features four-word addresses, end-to-end encryption, and distributed storage that can be built upon?

Lets not reinvent too many wheels unless we absolutely must.

4 Likes
7

Are you able to run Communitas from source? I get this when I try to run:

371zeus ~/communitas                                                            % npm run tauri dev

> communitas@0.1.17 tauri
> tauri dev

     Running BeforeDevCommand (`npm run dev`)

> communitas@0.1.17 dev
> vite

node:internal/modules/esm/resolve:275
    throw new ERR_MODULE_NOT_FOUND(
          ^

Error [ERR_MODULE_NOT_FOUND]: Cannot find module '/home/system/communitas/node_modules/vite/dist/node/chunks/dep-D_zLpgQd.js' imported from /home/system/communitas/node_modules/vite/dist/node/cli.js
    at finalizeResolution (node:internal/modules/esm/resolve:275:11)
    at moduleResolve (node:internal/modules/esm/resolve:860:10)
    at defaultResolve (node:internal/modules/esm/resolve:984:11)
    at ModuleLoader.defaultResolve (node:internal/modules/esm/loader:780:12)
    at #cachedDefaultResolve (node:internal/modules/esm/loader:704:25)
    at ModuleLoader.resolve (node:internal/modules/esm/loader:687:38)
    at ModuleLoader.getModuleJobForImport (node:internal/modules/esm/loader:305:38)
    at ModuleJob._link (node:internal/modules/esm/module_job:137:49) {
  code: 'ERR_MODULE_NOT_FOUND',
  url: 'file:///home/system/communitas/node_modules/vite/dist/node/chunks/dep-D_zLpgQd.js'
}

Node.js v23.11.1
       Error The "beforeDevCommand" terminated with a non-zero status code.
1 Like
8

Yeah, its not building for me either on various machines.
I was going to poke @dirvine about it tonight, but it will likely be tomorrow.

I bet it works on a high-end Mac - said he viciously :slight_smile:
In fairness the last line of the README is “get testing” or words to that effect.
Feeding back to @dirvine will hopefully move it all on.
Its a big project, showing what can be done with the intelligent use of AI by someone using it as a tool rather than a crutch.

However I can hear @happybeing snorting into his PIMMs and laughing like a drain at that failure :slight_smile:
Hope it goes up your nose, Mark :heart_eyes: :tropical_drink:
Don’t fall in now…

5 Likes
9

And why did we need IF if David wants to win outside of the competition? :laughing:

2 Likes
10

He’s been pumping out some code, that’s for sure. It sounds really cool, but I need to play with it. Interestingly the autonomi lib is not listed in either of these projects, just some core networking utilities, but not the storage side of the network functionality :thinking:

2 Likes
11

Because deep deep down David is a Maoist. :slight_smile:

“Let a thousand flowers bloom”

or in perl-speak TIMTOWTDI

There Is More Than One Way To Do It.

And knowing David he would not put his own work into the competition - too many snidey questions would be asked, no matter how fair we tried to make the judging.

At least thats how it looks to me, I have no particular inside knowledge

Lets see lots of approaches, but having seen them, lets converge on what is most useful of these appraoches (or flowers) and can be used as the building blocks for future development

4 Likes
12

I suspect this is part of a cunning plan to get a wheen of folks using Autonomi without actually knowing it :slight_smile:

Or providing a couple of ways for Communitas to work, one of which is waaay more effective than the other.

2 Likes
13

I ‘m all for eating our own related dog food, that said, let’s make the NIST part really modular so it can be replaced? Why?

All current PQC protection is binary math, anyone thinking Q-Day is ten years out needs more research, AI is accelerating design of everything, especially Entangled Quantum Computer Clusters = Decryption of most any Binary Math encryption much faster than the main stream press is saying. AES 256bit is likely ‘toast’ by 2029 as the QC analysts estimate Qubit power will be 256x bigger than now for one QC by then, imagine simple clustering, 1024x faster, means AES gets cracked in < 1 yr. They already have entanglement working in RT in Germany over 90km distances over fiber (this past June).

High Value targets first? yup. All them dormant BTC wallets. Given a QC costs today only US$ 1.5M , expect that price to hold as they boost qubit ratings 2X every 6 months (lower cost of production and energy consumption as this stuff gets sized reduced), the current rate of qubit power increase. Lots of competition building these jazzed up analog gas machines. :wink:

14

I think David is above that, he doesn’t want to waste time and focuses on the best solutions, searching for new ideas and concepts, improving and developing the network…

At least it seems that way to me, I have no special inside knowledge. :grinning_face_with_smiling_eyes:

1 Like
15

more on Q-day movin’ on up

2 Likes
16

I invested in a company called CLAVI about a year ago, they claim to be quantum resistant. They are rolling out their first units now, although they are very expensive. Their target market is hedge funds, family offices, CEXs, and misc HNW individuals.

https://x.com/Clavi_AG

3 Likes
17

Imo NIST encryption is OK until the first ‘Entangled’ Quantum Computer Cluster Shows up, then with PQC Binary Math it becomes a race to the bottom, crack n change, crack an change ;ad infinitum; with huge losses piling up along the way. Some but not all Quantum computer clusters will get re-organized to get very good at guessing using approximation math…making several guesses in parallel in real time, because the payoffs are HUGE and cost of these ‘gas pipes’ is only US $1.5M to US $2.5 M today.

The question is how quickly ‘they’ show up, as the AI driven design continues to accelerate the creation of everything, especially Quantum Computer Capabilities, Decryption being the obvious big reward motivation, aside from total nation state control of their populace.

The expected Q-Day arrival date Range is 4 to 10 years for Q-Day to show up, estimated by the ‘smart’ researchers and analysts in the binary math space.

That said only a Few of these ‘smart’ math heads have have ever heard of :

Discrete Deterministic Chaos Theory which can be incorporated to inject multiple passes of entropy to eliminate patternicity making it virtually impossible to crack by creating True Random Numbers for keys..

In fact if one tries to explain the above to a math head, one is more likely to get in a shouting match with them if one brings up the topic as the topic is not in their ‘wheel house’ … :wink:

also there is a thing called a Mathematical Barren Plateau that when applied to the generation of pure Random Numbers (used to create keys) together with the above, it creates an unsolvable problem in math which is the only thing Quantum computers are good at, doing it REALLY FAST.

My bet is on the latter two techniques, and definitely not NIST.

The PQC Vendors doing the ‘Binary Math encryption sell’ to the customers that they are protected for 100 years is a ‘big stretch’,

however Quantum Qubit power is doubling every 6 months accelerated by AI,

in four years that is a 256X increase by 2029,

then the clever opportunists gang them into clusters and aim they at high value Targets. Like BTC dormant Wallets, or Banks, or if they are into take-overs, Nation States.

Back of the envelop math says AES 256 will be broken in under < 1 year by a quadratic Cluster of Quantum Computers using the highest speed we today have as the starting point for Qubit power before doubling every 6 months. That means companies that are high value targets, BIG ones

have til 2030 to get ready. One thing I do know is while the awareness might be there of Q-Day in the FT 100, it will take those companies four years to get anything done.

Whoops. It’s Y2K all over again and this time Binary Math is not going to help, in fact it will be slowing us all down with giant keys in the ‘guessing game’, latency will really start to suck.

Any way, the homework I have done points me away, far far away from PQC Binary Math based solutions, especially NIST. So that is the path I am taking and also recommending others to take.

As always do your own research though. :slight_smile:

Quantum Computer bonus joke. Just heard it today…

“A Quantum Computer shows up at two bars”

1 Like