Skip to main content

The Simple Goal

Source: Max Tegmark keynote

Our goal is not unique nor original as it is equivalent to many other amazing initiatives. Like others we want to create a collaborative instead of a competitive unequal society that we have today. We live in a system that promises opportunity, yet in practice amplifies inequality. Starting points are unequal, and because wealth generates more wealth, advantages compound over time. Access to capital, knowledge, and networks determines who gets ahead—and who gets left behind. The issue is not merely wealthy individuals, but a global system of capital and institutions designed to concentrate ownership over time. If you want to have a little 101 economics explanation of how this came about then please check our Status Quo chapter, but for now we recognize that the problem exists. 

So how do we solve such a global problem ? To put it quite bluntly, we believe that each for profit organization should eventually become a non profit. This is obviously easier said than done, because well how are you going to survive if you spend your time volunteering for non profit organizations. On top of that not all sectors of the working life are as enjoyable to volunteer for as others, and not everyone has the time to volunteer even if they wanted to. We recognize these problems and therefore it won't be a switch over night. It will be a gradual process with lots of hick ups and obstacles on it's way but well that is also what makes it worth fighting for. 

As stated, due to the economic system not everyone is as privileged with free time to work on this goal than others. We must recognize that if you are in a privileged position e.g you are from a rich country, have wealthy parents, are a student, can work part time to survive or you are retired then you will have the time to contribute to such a collaborative society. However, we must recognize that the more non profit organizations that exist the more likely others can gain time to contribute as well.

So where to start this journey ? Well, the lowest hanging fruit in this space are sectors where people can create massive impact with the least amount of effort. This space is by far the BIG TECH sector, because it is the sector that is the most wealthy, all other sectors are built on top of it, it replaces/disrupts peoples jobs, but most of all it is the easiest one to fight with a small effort of each individual. The idea is this, if you can change the most influential sector in this world then other sectors will follow suit. It is not a surprise that Deternet is focused on this. 

Now we are actually on the cross roads of another activity that accelerates the whole problem. Yes the elephant in the room indeed AI. The ultimate goal is to create systems that are millions times more intelligent in all aspects than human beings. Yes, we are not there yet but all capital is this moment invested in it. It is something that globally people work towards and will inevitably work towards due to the economic system. To get out of it now would mean catastrophic economic collapse. 

So we need to tango an ever increasing inequality together with AI that will accelerate that equation even more. Now, why then open source ? Basically for two reasons now that AI is also in the room which was not there before

    For a collaborative world in which To make sure that if we create AGI somehow that we are giving it an initial value system that works to be loving, benevolent and has compassion. Because if we continue with the current economic system than it will be focused on killing, gambling, selling and spyin

    LLM output

     The fear is about what could happen if very capable systems are given tools, access, and authority—or are used by people who have them.

    People worry about a few different things:

      Misuse: A person or organization could use AI to do harmful things more easily or at greater scale, such as cyberattacks, mass persuasion, or assisting with dangerous research. Bad automation: An AI put in charge of important systems could make a consequential mistake quickly, and people may struggle to intervene. Loss of control: If a system can pursue goals over many steps, use tools, and act with little oversight, it might find ways to keep pursuing a poorly specified goal even when that conflicts with what its operators intended. This is a debated future risk, not a settled prediction. Power and speed: Companies or governments competing to deploy AI may cut corners, while AI can make existing harms faster or cheaper.

      So the concern isn’t simply “a smart machine will want to escape.” It’s the combination of capability, access, autonomy, imperfect goals, and human decisions. And not everyone thinks superintelligence is imminent or that catastrophic loss of control is likely; people disagree about the odds.

      AI on top of the economic system

      Now, on top of this we invented AI throughout the years that add another layer of power on the people that already hold the wealth. It all started somehow with beating humans at chess, than at Go and then eventually LLM's came along that changed the whole ball game. Now, to some extend these basic LLM's from 2023 gave people access to writing essays faster, make presentations faster, code faster, make artworks faster etcetera. Basically, it empowered people and therefore companies as well to do things more efficiently and faster. 

      Throughout time we started to interact 

      Now, we also have self driving cars, we have AI tutors, AI beating the best human math olympiad, run faster than usian bolt and we can continue on the trajectory of things that AI can do better and more efficient than humans. 

      Now, yes to some extend this matters because it will replace lots of jobs and hence make the people that can invest in these technologies even richer than they already are. 

      Capitalism as a catalyst for AI disaster

      Everyone knows that each tool can be used in multiple ways. I can take a knife to cut vegetables or kill someone. In the end it is up to the human using the tool what they want to use it for. Now, in the best case scenario

      We can better use AI for good i.e to solve resource problems. Imagine a world where AI agents are interacting with each other to provide information on how to divide resources globally, in which locations we can best build factories locally, with the least amount of impact on the environment. Yes, just as with the argument for using AI for profit also here still exists the problem that agents could execute actions that in the end are optimized to reach the goal but do something catastrophic for us humans that we did not anticipate to reach it. 

      The point in the end is this! Do we want to let these AI's explore in a world where there is fierce competition, lying, betraying, resource hoarding, wars etcetera to basically fight against each other or to optimize our collaboration. Of course it is not certain but the chance that AI goes rogue on us in such a society is way more likely than in a society where AI agents from different countries, sectors are optimized to see the bigger picture and not only act out of the interest of the corporation owning them. 

      Where do we want to go from here ? 

      1. Are we pro-human and do we want to create a collaborative society, cure diseases, solve problems
      2. Or pro acceleration i.e the silicon valley path, tech bros, investors, VC's, billionaires that replace all of us 

      Race to Replace jobs, relationships, decision making, power, species or

      AI tools that make us prosper, no AI girlfriends, humans are in charge, we collectively own the intelligence 

      Verifiable Code i.e vericoding

      They say that you can verify if a piece of software is actually doing what it supposed to do by doing a formal proof. How, can we trust such a formal proof if it is checked by some black box tool 

      Get Oligopolies

      If we don't do anything we will get oligopolies which you see happening all industries. Large corporations are buying out there competition. Just see paramount buying Warner Bross for 111 billion dollars. Google, Facebook etcetera all buy up the small entrepeneurs. 

      Human Verifiability

      If you ask any person on the street that you should be careful of what your LLM tells you is the truth than 

      LLM

      An AI can produce something that looks confident and true while in reality it is wrong in the end. To prevent this is of course to check the sources. But lots of questions given to AI are actually not so black and white and can be used for inspiration. At least all the non scientific questions can provide interesting insights that you would not have thought about before. Then once you obtain the new subjects you can go off and study these subjects which you would have otherwise not thought off. The important thing is that we always need to keep scientists and researchers checking what AI produces is actual true or not by doing tests that we humans can comprehend. That is really important, since doing a test done by another AI that we also don't comprehend then we cannot formally proof if something is true or not.

      Why would I work for free

      An understandable argument is the fact that why would I work for free if the owners of the organization I am working for receive all the financial returns. Obviously, there is no non profit organization where volunteers would be motivated if they know that because of this the owners/investors of the organization are getting even more rich than they already are. Hence, what we mean with non profits is that the whole non profit is voluntarily based. Pretty similar to sports clubs and other student associations. People are in the board because they want to be in it for the cause of it and not to earn money on the backs of other volunteers. Once, you create an organization, learning group, sports association for the greater good with ownership by the users than you will enter in such a great environment of happy people that all want the same. The likelihood of jealousy will decrease, lying to other coworkers, leaving information out, kissing bosses asses all of this will go out of the window. This would definitely create a healthy work environment because everyone is there because they are proud they can contribute and not just to get their paycheck in the end of the month. 

      Human values

      Sources: 

      Human values completely change over time. Not so long ago you needed to follow a religion depending one where you grew up. If you did not do that than you were seen as wrong. Let's not even talk about the LGBTQI community that was also surpressed, or Jewish people were considered wrong. Women were not allowed to vote before, slavery was accepted by the owners of slaves. Or when a white woman came to prom with black guy was not accepted. Today we still accept the current economic system that we should not only fight against other countries but even within our country. Therefore, to say that AI will keep our human values is straight up nonsense. 

      Just like when you try to raise a kid it will develop its own moral values eventually regardless with which starting values they have come to be due to their parents and environment. 

      Killing, Spying, Selling and Gambling

      What about raising AGI and super intelligence trained on compassion, love, understanding, helping, collaboration than we will maximize the odds that AGI will be compassionate for us. This will provide a way a higher chance for raising a benign AI instead of a killing machin or large Advertisment agency. 

      Even if you don't have full guarantee, then at least you can increase the odds for a positive future of humanity. 

      Convince people

      Now, it all may sound nice and great in a person understanding these issues, but how do you convince an average joe that they should change phones, change operating systems, or at least change to open source applications if they are already used to the applications they grew up with. They then need to learn again a new application, which by the way is not that hard it just takes a little of explanation. Therefore, we also have our documentation page that step by step explains how all of these apps work. On top of that you also have the official websites of these applications as well. 

      It is just hard to explain to the layman the concept of the impact AI will have on our daily lives, how people will be and are already laid off and convince them that  way out of a future where all resources are in the hands of a few corporations and governments is to go open source. That then simply starts by changing the apps you are using on your phone. By doing that you are fueling energy into a future of collaboration, no data extraction, no profit maximization etcetera. 

      I like the parenting analogy: nurture the "young AI" for compassion, love and understanding. But parenting has had mixed results: many children have evidently been shaped into hostile, hateful, predatory adults. So I'd like to believe that we'll build super-intelligence for the good, but for sure some will do so for the bad

      Bioweapons

      Cybersecurity threats

      Recursive capability

      Making the new version of itself. 

      Palentir

      For reasons of competition AI is constantly developed and will not be stopped for the sake of competition. 

      End Goal

      The end goal should be to become fully human! That is to constantly challenge our physical and mental bodies. Yes we can use AI for this to improve ourselves, but we need to stay great story tellers. 

       

      Muti-Agent AI systems

      AI agents can collaborate effectively even with the same skillset by dividing tasks among themselves and working in parallel, which allows them to tackle complex problems more efficiently. This parallel processing enables them to cover more ground and find solutions faster than a single agent could manage alone.

      What does it mean to call something super intelligent ?

      Ben Goertzel defines superintelligence as a machine that can achieve any computable reward function in any computable environment. This concept emphasizes cognitive abilities that surpass those of humans, rather than simply being a collection of AI agents or robots operating within a datacenter.

      Intelligence alone doesn’t give an AI physical reach. To affect the world, it needs channels to act: network access, software permissions, robots, sensors, factories, energy, and people or automated systems that carry out its plans.

      So a superintelligence could be extremely capable while isolated—solving problems in its computing environment but unable to change anything outside it. If connected to the internet or operational systems, its reach would depend on what those connections let it do; if embodied in robots and linked to manufacturing and logistics, it could potentially act physically too. Those are separate questions from whether it is “superintelligent”: capability is one thing; access and agency are another.

      That fits with Goertzel’s interest in both AI and robotics, and with his proposals for shared or decentralized AI infrastructure—but it doesn’t mean his definition of superintelligence specifically requires a datacenter full of agents and robots.

       

      The initiative
      The Deternet initiative seeks to build a global digital commons—one that is decentralized, open source, and collectively owned. This ecosystem would span hardware, software, telecommunications, hosting, and power infrastructure on a worldwide scale. Its approach centers on empowering a network of interconnected local hubs by providing them with the resources, knowledge, and support needed to grow and sustain this shared infrastructure. 
      Each hub would take responsibility for developing and maintaining systems within its region, while also helping individuals transition toward open-source tools and services in their everyday digital lives. The broader vision is to create ecosystems that function like open internet protocols—accessible to anyone, shaped by contributors everywhere, and free from centralized control. In this sense, it mirrors technologies like HTTPS, IPFS, Bittorent and DNS: universal standards that remain open rather than being owned by large tech corporations. 

      While much of the necessary infrastructure already exists, it is currently concentrated in the hands of a few dominant players. Part of the initiative’s mission, therefore, is to encourage established companies to participate and support this shift. This may be challenging, but with enough momentum, a decentralized movement could reach a tipping point—where organizations are faced with a choice: adapt and join, or risk becoming irrelevant.
      The rest of this paper will discuss briefly these 5 different sectors that make the tech and therefore our modern society run. However, if you want go in depth and actually contribute to one or more of these sectors than enjoy the dedicated websites, community and documentation pages we have made for each of them. That is were you will find specifics, project roadmaps and contribution guidelines. 
      Software
      We said that Big Tech is the easiest to change to a non profit compared to all the other sectors. Now, that statement is largely built on the software aspect of the digital infrastructure. That is because software can basically be replicated for free without others loosing access to it. That makes it a non rivalry good i.e no scarcity. However, why do we then have all these for profit applications that either charge a fee, extract all your personal data or they do both ;) ?
      Ecosystems vs You Craft your Device

      From device full of apps at the start to a black screen with no apps and start from grass roots


      Today, BIG TECH focuses on creating ecosystems and preferably ecosystems that only interact with devices and accounts of the same corporation. Think of Apple for instance, there ecosystems is completely locked down and is not made to interact with other brands. You cannot share an Apple calendar with a person that uses android or another Calendar app. The same for photos, notes, email, iMessage, Airdrop, Find my all of these apps are integrated to only work with Apple devices. The same can be said for Google, Microsoft and yes even Proton is doing the same thing right now as well. The objective is clear ecosystems that don't interact and keep users locked in. 
      If you buy an iPhone then your phone already comes preinstalled with all the Apple Apps, iCloud and Apple Siri that works 
      The idea is not to have one corporation making all the apps but have an open source app made by a different group of volunteers for each type of app. For instance, a great photo app is Ente or Immich, while a great Calendar app is Silentsuite and a great note taking app is Joplin. Hence, it is not called Notes, Photos, Calendar etcetera as it is with current operating systems. No it is named all differently because it comes from different groups of volunteers. However, the goal is though for them to interact seamlessly with each other as wel. For instance a great journaling app like Journiv greatly integrates with Immich photos. In this way you don't need to store photos twice but have them stored once instead but used for different apps at the same time. 

      Community Cloud vs Home Lab
      Apps and data needs to be stored somewhere in this physical world even though it might all seem non physical. Now, data can either be stored one someone else's computer or on your computer. Eventually, it is up to you how much you trust someone else's computer to store your data on. Finally, also there is a big difference between public and private data. One is completely happy that Wikipedi is public and that it is stored on other people's computers. In the end you wan that your contribution is seen by the world and not only to your self privately. 
      Nevertheless, today we are used that all is stored in the cloud. Sure, we have some laptops, phones local storage etcetera but most of it we safely store as a "backup" on someone else's computer i.e Apple, Facebook, Googlee's or Microsofts computers. 
      Now, we would like to create community run data centers that run open source apps on redundant infrastructure that everyone from that region can use for free. Today, there are no community driven data centers yet but there are community run open source application instances that people run for free completely based on donations. 
      A great example of this is vaultwarden.net and vaultwarden.uk. They are group of volunteers that went into a datacenter and actually installed three servers and run a kubernetes cluster instance of Vaultwarden for the world to use. It is completely based on donations to keep everything up and running i.e the costs for renting the datacenter space I suppose. Now, it must be said that this is an example of an application that barely uses any data. Hence, it is extremely cheap to host this for the whole world. However, it is still not free though. The reasons this also works is because the app is E2EE which is definitely a must if you let someone else host your precious data. 

      Now, not all apps are E2EE and hence that is where self hosting or a home lab comes in. With a home lab you are the only one that sees the data and no one else. Neither metadata that a maintainer of another server could see as well. For private data that is not E2EE it makes sense to do this. However, once it is E2EE then someone else could do the maintenance and you only donate to keep the project up and running just like Wikipedia for instance. 


      Different Colors
      To discover new apps there are several platforms you can use. However, the platform Atlernativeto.net definitely stays one of the favorites. They have a dedicated section of new apps with the newest on the top and then getting older over time. Event though they host 100 thousands of applications on there website in the end there is only a handful maybe a 100 that people actively use and maintain. Especially today with the use of AI new apps spin up almost every day but actually do the exact same thing another open source app already did as well. The only difference between apps is often the "colour" i.e the user interface. The great thing is that everyone can choose their preferred colour but in the end a project that is heavily maintained with a great team of humans behind with scheduled release dates is a way more safer bet than using any new app that comes along. At homelab.deternet.org you will find for each category of apps an x amount of examples with for one category more than another but most often there is only one or two that are actually leading. Also, we should not be surpriced if most open source apps eventually converge to one app per category. In that case all the attention can be on improving the functionality of the application and people can build different user interfaces on top of it.

      AI Agents War

      Now, what will AI agents be mainly used for ? Well companies will use them to sell the most amount of products and users will use them to get the cheapest deals. The economic system will still be exactly the same 
      It will be one AI agent of a corporation that interacts with your AI agent. You AI agent will try to fight the companies agent to get the best deal and the AI agent of the company tries also to get the most amount of money out of its user
      You will have AI made to manipulate people to buy the latest gadget .
      But not only AI marketing for humans but as well we will have marketing that should manipulate other AI agents that act on behalf of a human. 
       
      If we use our AI agents to think through a problem for us then it will always take into account the economic system. So if you want to solve a resource problem then it will think about how to convince investors investing in your project, a business model, how to manipulate people in buying your product or to convince other bots to buy your product or vote for you. 

      Super Intelligence

      If super intelligence actually becomes a reality then we rather have an environment that is collaborative and not competitive. Not optimised to make weapons, make people addicted to screen time, gambling on the stock exchange and now cryptocurrencies as well. 

      Human Alignment Problem
       

      What about governments ?

      Now, we could also say why not let governments solve this problem ? Well governments still work in the current economic systems. Even if public infrastructure is created then there is again a for profit company hired, paid by the public to build and maintain the infrastructure. Then we don't change the system, since it is not publicly owned and maintained but only publicly funded. The goal of Deternet is to create true non profits where people voluntary contribute their time and resources to make this happen. We need to be able to show that a non profit can exist and run without the backing of governmental support.