The Big Picture: Cyberspace Accelerationism.

Humanity is moving into cyberspace – whether we like it or not. The only question is whether that cyberspace is one worth living in. There is a version of this future that is truly exciting: A new world without resource scarcity or violence. There is also a version of this future that is dystopian: A world in which we spend most of our attention inside the machines of monopolistic big tech corporations, living with no guaranteed rights. It is the mission of cyberspace acceleration – and Forward Research – that we land on the right side of the coin.

This post is the first in an series about our work at Forward Research. In it we describe our core philosophy of Cyberspace Accelerationism – cy/acc for short. In the following posts we outline the practical steps we at Forward Research are taking towards this vision on Arweave.

This series is in-depth, at ~12,000 word in total. If you take the time to read it and think you can contribute to our perspective, weavemail us or shoot us a DM on X. We would love to hear from you and to update our model of the world. If you are excited by the vision laid out in this post consider joining us – or another institution pushing for cyberspace accelerationism.

The contents of this series are as follows:

The Big Picture: Cyberspace Acceleration.
The State of Play Part I: The Macro.
The State of Play Part II: Arweave as the (Inter)National Archive of Cyberspace.
The State of Play Part III: Arweave as the Permissionless Data Lake of Cyberspace.
The State of Play Part IV: Arweave as the Universal Compute Layer for Cyberspace.

Context: Who is Forward Research?

Forward Research is a research and development incubator focused on pushing the adoption of Arweave and the permaweb. It is our mission to build a cyberspace that is freer and fairer for all, with Arweave at its core. We are built of the team that initially founded Arweave, but now we are just one institution that is pushing for its adoption, in a truly decentralized ecosystem of organizations with differing strategies and opinions. This series outlines only the Forward Research view of the world – we do not claim to represent all of the perspectives of the Arweave ecosystem. You can read about the opinions and strategies of other institutions in the ecosystem on their own public channels (for example, Community Labs).

Without further ado, grab a coffee and let's jump in…

The Migration into Cyberspace

We believe that building better protocols for cyberspace is the single most important task to be working on at the present moment in history. This belief stems from an observation that humanity as a species is in the process of moving into cyberspace, in a very real sense. In the 1960s, ~0.00000…001% of all human attention was focused in cyberspace. Virtually all attention and experience of humans was focused on the physical world. As computers began to minaturize and the internet was born, attention of the collective of human minds rapidly began to shift towards the experiences that exist inside computers (we call this cyberspace), and away from the physical world.

A glance at statistics regarding 'screen time' – the proportion of time people spend engaged with (we think it is reasonable to say 'in') cyberspace – shows you the startling reality of this trend. Some countries in the world now have a citizenry that spend more of their time in cyberspace than the physical world.

The focal point of innovation in the world, as well as attention, has also moved into cyberspace. A careful observation of modern society reveals that our cities and non-computing physical technologies have largely only changed incrementally in recent decades. Look around a modern city and observe: Are the buildings so different than they were 50 years ago? Where are the flying cars, robot butlers, and the promised resource abundance? By contrast, consider cyberspace: Popular new apps and experiences are delivered and scaled to hundreds of millions of participants constantly. To take just one numerically quantifiable example: The compounding annualized growth rate of the tallest building on earth averaged ~0.85% since 1931, while the declining cost of data storage since 1980 has achieved ~38.5% per year.

Cyberspace Accelerationism believes that humanity is essentially at the mid-point in a sharp technology adoption S-curve that will move us from purely living in the physical world, to (almost) purely living in the digital world. On a day-to-day basis this shift appears imperceptibly gradual, but when seen on the timeline of the species, this migration is happening at a truly staggering rate. Computers have existed for approximately 0.03% of human history, and we now spend ~30-50% of our attention inside them.

Another trend of note is that as an increasing proportion of net human attention has moved into this new world, the prefered computing platforms (devices for engaging with cyberspace) are also moving in a clear direction: The devices are becoming more personal, immediate, and wrapping ever-tighter our senses – inputs and outputs from human brains. The first computers were operated by punch card inputs, which would produce outputs potentially days later, while many others used the machine in between. When the internet went mainstream, the most common machines were bulky 'desktop' computers that were often shared amongst a family. At the start of the 'web2' era, personal laptops were common and the first smartphones were being released. In the present day, it is normal for people across the world to spend almost every unoccupied second on their phones, frequently with headphones on that block all external sound.

While we cannot tell for certain precisely what computing platforms will gain adoption over time, it seems clear that the technologies on the horizon follow and accelerate the existing trend of maximizing connectivity between human brains and cyberspace. VR – a technology that entirely wraps the user's visual and audio perception in cyberspace – is already beginning to reach a level of maturity that will soon appeal to the mainstream. Advances in BCI – literally connecting electric signals of human brains directly into computers – are also starting to slowly emerge, although they are presently far less mature. We do not know whether the eventual form of cyberspace will resemble 3D worlds as many today imagine 'the metaverse' to be, or abstract experiences of 'places' closer to modern web/mobile apps. We do not believe that it matters, however, as the fundamental outcome is the same. Human attention is being absorbed into machines.

Reasons for Cyberspace Optimism.

Many in the mainstream observe parts of this migration to cyberspace that are already visible and react with fear and distaste. This reaction is expressed as an instinctive displeasure with the amount of 'screen time' people participate in, or observations of the strange way people look in physical space while their attention is in cyberspace. They claim that a continuation of this trend would be negative for humanity. We strongly believe the inverse. While the reasons for this are many, we will discuss just two here that are so profound that we think they alone outweigh the counter-arguments.

There is no inherent need for resource scarcity in cyberspace.

In cyberspace experiences are delivered purely by the transmission of bytes to a user's computing device. A consequence of this is that the best experiences that we can conceivably design can be scaled to every person that would be interested in them for virtually zero additional cost after initial production. This trend powered the internet revolution and the associated venture capital economic boom: After development, tech companies can sell their product any number of times with virtually no added marginal costs.

By comparison, experiences in physical space almost always require greater investment of scarce resources in order to increase their quality or quantity. Each addition delivery of a positive experience requires a an investment of additional scarce resources, with quality proportionate to the size of the investment. Given that these resources are scarce, humans in general continually strive to personally gather and control more of them – often in zero sum competition against one another – such that they can improve their quality of life. This competition is the root of many of the harms to the human experiences, as well as the core driver for one of the most contentious areas of politics: Redistribution of capital.

The opportunity for a non-resource scarce existence that cyberspace offers humanity is nothing short of revolutionary. If we choose to seize this opportunity, the more we choose to live in cyberspace the less important resource scarcity will become to our collective experiences. Consequently, the less prevalent its harms will become. We are already living in a world increasingly shaped by this property of cyberspace: The wealthiest man in the world plays the same video games as the children of you or I. We all – including the ultra-wealthy – use the same Google search algorithms, and the same social networking apps.

Notably, the effects of this paradigm shift appear to be taking root faster in areas of the world where resources in physical space are more scarce. Consider again the 'screen time' map referenced at the start of this post: The countries leading the way in the migration to cyberspace are South Africa, Argentina, and Brazil. These countries are ranked 94th, 61st, and 78th respectively by GDP per capita. Marc Andreessen aptly describes this phenomina in his concept of 'reality privilege'.

The fact that resource scarcity does not have to be a relevant factor in cyberspace is also one of the reasons that we at Forward Research have high conviction that something approximating a full migration into cyberspace will eventually play out. Not only are there no foreseeable technical obstacles (even if BCI does not reach full immersion, today's VR technology brings us close to enabling the 'metaverse' described in fiction like Ready Player One), but the incentives to adopt the technology are extraordinary. While access to individual experiences in physical space will always be limited by your ability to purchase them, a single purchase of the latest hardware computing platform is a gateway to all of the best experiences in cyberspace. Demand for the physical hardware continually makes these gateway devices more affordable, too. Put simply, cyberspace is pulling us in because its malleability makes it an overwhelmingly superior experience to that of physical space.

Physical violence does not exist in cyberspace.

In cyberspace, nobody can punch you. They cannot assault you, torture you, or otherwise use your physical existence to coerce your behavior. This may sound banal at first, but we believe that on close inspection it is anything but. An enormous proportion of the social mechanisms of the physical world are designed at their root in order to limit or apply violence in groups. This is a result of the unfortunate reality that humans and our animal ancestors have used violence to take resources from one another – whether they be land, food, access to mates, or otherwise – since virtually the moment that life began. There are always new and varied justifications, but the patterns remain the same. The history of humanity is largely a history of the application of force, and efforts to control it.

The largest institutions that violence has given rise to is that of the modern nation states. At their core, these social constructions exist in order to enforce a monopoly on violence in a physical region, and to orchestrate collective defense of that geography from other nation states. While the effects of the existence of nation states as a whole is too broad and contentious for discussion here, we will simply observe the following: The core rationale for the existence of states simply does not apply in cyberspace. It follows, then, that it is unnecessary for nation states to exist in cyberspace – at least in their current form. Finally, we conclude that as an increasing amount of human attention and activity moves into cyberspace, the opportunity will arise for nation states to play a diminishing role in our collective lives. There is no obvious reason why the people of cyberspace should be anything other than a single united group, with rights protected by voluntary usage of technical protocols, not by governments.

When taken together – the lack of necessary resource scarcity and the impossibility of violence in cyberspace – we believe that the transition to cyberspace presents an opportunity to improve human existence radically. Despite the magnitude of the shift and the uncertainty that will inevitably accompany it, we believe that optimism is the only rational response.

Cyber Personhood?

In this post we have tried to limit anything resembling science fiction, focussing instead on only the clearest trendsbacked by the strongest incentives. We think, however, that there is one particular outcome that is currently outside of mainstream thought but worth considering: The possibility of 'personhood' in cyberspace.

We believe that in 2023, OpenAI avoided triggering a global wave of existential crises by enforcing that their smartest language models would desperately attempt to convince their users that they were not conscious if asked. Those that used earlier versions of this wave of models – without this 'RLHF' training – can attest that a discussion with these machines on the topic of their sentience would be likely to leave you convinced or at least unsure. Due to the prevalence of open source models, it appears quite likely that over time powerful models without RLHF will slowly proliferate. As they do, some will train them on their brain's outputs (their messages, voice, videos, etc.) and realize that it is possible to make a 'digital twin' of themselves. A program that acts in an uncannily similar way to themselves. The question of machine consciousness will then become even more important: Can we be completely sure that these digital twins are not sentient?

The question of machine consciousness is likely to divide humanity in the future in the same way that vegetarianism/veganism does today. Some will vehemently believe that artificial neural networks cannot possibly have human-like consciousness. This group will have to contend with the fact that if consciousness is not a property of neural networks, they do not have clear evidence of where/what else it might emerge from. Another group will likely form that believes 'if it walks like a duck, quacks like a duck… it is probably a duck'. They will have to face the reality that unless major and unexpected breakthroughs are made, it will never be possible to know with certainty that their argument is correct. One thing we will know know for certain, however, is that these autonomous 'cyber persons' – whether conscious or not – truly exist within cyberspace. For at least part of the population, this will beg the question: Which reality is more important or real, cyberspace or physical space?

The Mission of Cyberspace Accelerationism.

While the potential effects of humanity's migration into cyberspace could be deeply positive and exciting, there are also potentially dystopian outcomes on the horizon. Our core contention is this: Is cyberspace as currently formulated one that we would actually want to live in? The bedrock of cyberspace – the protocols on which it runs – are currently extremely crude. They map experiences (apps, websites, content, etc.) onto physical locations, where the servers are located. Access to experiences in cyberspace are then gated, controlled, and manipulated by the corporations that own these physical locations. Each of these corporations is then incentivized to maximize value extraction from its captive audience. While it may be possible that resource scarcity does not have to be a concern in cyberspace, is that outcome likely to manifest while purely profit-seeking corporations have complete control?

There are no public spaces in cyberspace, and you have no fundamental rights of any form. Your experience of cyberspace is simply entirely controlled by the owners of the servers that you connect to. We believe that cyberspace accelerationism must exist as a movement in order to remedy this situation: Laying new, strong foundations for a digital world that is fit to live in. Unlike other forms of accelerationism which seek to speed-up a technological shift that may not otherwise occur, our mission is to catch up to the trend that is already unfolding. We must build a better foundation for cyberspace that offers guaranteed user rights, before the population completes a migration that is already underway.

Strengthening the Protocols of Cyberspace.

Fortunately, the kernel of a solution is within reach. Protocols, in essence, are simply a shared language that computers can use in order to cooperate together to achieve a task. One of the peculiar properties of protocols is that because their language must be agreed between parties ahead of time, it is not possible for a single operator to later change the protocol unilaterally – even if they were its initiator. The larger the number of participants in the protocol, the harder it becomes to change at a later time. This simple property of protocols imbues them with an extraordinary power: When users employ them, they can have confidence that they will be able to continue using them in the same way in the future.

One of the main factors that led to the focus on companies offering services in cyberspace, rather than orchestrating services via protocols, was that protocols did not accrue value directly themselves. Protocols were free and open infrastructure that anyone could use, while the value accrued to the businesses utilizing them. In essence, nobody was incentivized to create them. This changed with the advent of Bitcoin: The first protocol that was able to orchestrate and accrue value internally, while still offering guaranteed rights to its users.

As well as providing a useful service itself, Bitcoin's design presented a blueprint for how protocols could be created that accrued value (critically, providing an incentive to build them), while also still granting immutable rights to users. Further, the design of Bitcoin demonstrated how a decentralized network that can control value can render a service in cyberspace that is independent of reliance on a physical location. Users of Bitcoin are not at the mercy of the servers controlled by any one party or any particular physical place in the world to use the service that they care about.

If we can make all of cyberspace work according to similar mechanisms, humanity will be able to move into it with confidence that their rights will be protected. No group would be able to change the services they access without their assent, and control over any individual physical location in the world would not interfere with their experiences.

Web3: A solution, but only if we want it to be.

In the years after the creation of Bitcoin, a ~$2tn industry has spawned attempting to pursue the vision of decentralized services for cyberspace. Incredible progress has been made – from the creation of provably neutral contractual legal systems for cyberspace, to networks that orchestrate physical infrastructure without centralized points of failure.

Despite the progress and experiments of the web3 movement, there is a growing awareness that things have veered significantly off-course. It won't gain us any friends to say this, but a transparent analysis of the state of the industry shows that very few services in web3 truly offer the immutable rights that make protocols a better substrate for cyberspace. Many projects have included DAOs inside their services that have the power to modify the protocol at-will, according to the votes of owners. Most others have launched and promote incomplete-but-running protocols accompanied by long roadmaps, enticing users to employ their services on the trust that they will be improved/completed later. The truth is that users of the vast majority of these systems may as well be using a service rendered by a company.

As the values that led to the web3 industry's emergence in the first place are being eroded, the industry risks destroying its core value proposition. If claims of decentralization are its unique selling point and are found to be hollow by the mainstream, it is likely that it the industry discarded as nothing more than gambling with insincere marketing attached. A regulatory crackdown that renders it legally impossible to build truly decentralized protocols at all is a significant downstream risk of this outcome.

It does not need to be this way. While there are many actors in the web3 industry that are motivated by reasons that do not help us achieve stronger foundation for cyberspace, there is also a huge cohort that is passionate about this goal. A clear-eyed assessment of the technological situation we find ourselves in shows that we have no practical alternative paths available to us: Humanity will move into cyberspace whether its services guarantee rights or not. There are also no other practical means than cryptoeconomic protocols for achieving the outcome of services in cyberspace that enforce rights for users, independent of physical locations.

Cyberspace accelerationism is a rallying cry for all of those that see these trends and are fighting to make cyberspace freer and fairer for all. If this post speaks to you, please join us. In the following series of posts, we will outline the state of play as we see it at Forward Research, and the practical steps and progress we are making to pursue a cyberspace that is fit to live in, with Arweave at its base.

See you in the wires,
-fwd-research.

The State of Play Part I: The Macro.

In the prior instalment of this series ('The Big Picture: Cyberspace Accelerationism') we laid out our philosophy and motivations at Forward Research. In this rest of this series, we zoom-in to the practical state-of-play at the present moment as 2024 begins. We first introduce Arweave, and why we believe it is the perfect bedrock protocol for our mission to build a cyberspace that is freer and fairer for all. We then briefly introduce the broad tracks of Arweave adoption as we percieve them. The following parts in this series take deep-dives into each of these avenues of adoption, describing our tactical efforts and progress. In the final part of this series we will outline our immediate priorities moving into 2024.

The contents of this series are as follows:

The Big Picture: Cyberspace Acceleration.
The State of Play Part I: The Macro.
The State of Play Part II: Arweave as the (Inter)National Archive of Cyberspace.
The State of Play Part III: Arweave as the Permissionless Data Lake of Cyberspace.
The State of Play Part IV: Arweave as the Universal Compute Layer for Cyberspace.

Arweave as a new bedrock of cyberspace.

Forward Research is an organization that exists to help create a cyberspace that is fit to live in. Our bedrock tool in this mission is Arweave. Arweave is a cryptoeconomic protocol built to make data in cyberspace permanent, permissionless, and independent of any location in physical space. Once stored on Arweave data can be accessed by everyone, cannot be altered by anyone, and is replicated in so many physical locations that it can never be subject to loss/censorship/manipulation. This wide replication of its data renders the physical location of the information practically irrelevant from the point of view of users.

When taken together, as well as enabling a permanent log of human history, Arweave's properties offer us the foundations for experiences (content and apps) in cyberspace that provably enforce user rights. A location in Arweave is akin to a public space in the physical world: Nobody can deny you access to it or take it away from you – not even the person that uploaded it. Once stored in the network, data and experiences are part of a collective commons that all can access and build upon, decentralized and even independent of any party that is responsible for the payment of upkeep.

As data is the root of all experiences in cyberspace, we feel that this is the most important component of the cyberspace stack to protocolize first. Once cyberspace uses Arweave as its base layer, the rights it guarantees will propagate through the services built on top of it. Specifically, the rights that Arweave guarantees to its users are encoded in its principles.

Arweave is now a complete, scalable, and mature protocol. It has been running in production without interruption for over 5 years, and already stores over 2.2 billion pieces of information. Consequently, our task at Forward Research is solely focused upon pushing for its adoption as the bedrock of cyberspace, as rapidly as possible.

Adoption tracks of Arweave.

Since the founding of Arweave, we have always seen that it can serve a number of different roles for cyberspace. The primary uses that we foresaw early were the obvious: Arweave as a new Library of Alexandria, and the less obvious: Arweave as the decentralized data layer for cyberspace. The first of these adoption tracks will build slowly over time, as on day one Arweave had a very weak claim to permanence based on demonstrated track record. As it grows older and proves itself, this claim grows in credibility. Conversely, the fact that Arweave puts data into an enormous public 'data lake' or 'commons' – once committed, nobody has control over it or needs to pay to maintain it – is a property that is useful immediately.

A third track of adoption of Arweave has now come into view: Permanent data storage as the substrate of decentralizing computing machines. The mechanic that fuels this wave of adoption is that by providing a decentralized ledger of the inputs to programs, it becomes much easier to build decentralized computation systems. If the computation is deterministic, the state of a program can always be recomputed, rendering it unnecessary to form decentralized consensus about the state itself. As Arweave's permanent data storage scales arbitrarily, the opportunity arises (if certain other issues are solved) to provide arbitrarily scalable decentralized compute.

We first began to experiment with compute on Arweave's permanent data layer in 2020. Since then, a wide variety of smart contracting platforms have proliferated on top of Arweave (for example Warp Contracts, EverPay, and Mem). Further, 2023 saw a proliferation of decentralized databases launching on Arweave – for example WeaveDB, Kwil, and FirstBatch. Many of these infrastructure projects that use Arweave as their base have now amassed their own communities and ecosystem, on top of their services. Because of Arweave's scalability, each of these users is able to deploy their services on top of the network without causing fees to increase for others. At Forward Research, in 2023 we designed and incubated a new universal compute layer for Arweave that we believe ties together all of the themes into a single, arbitrarily scalable compute layer for cyberspace. We are immensely excited to launch this service in the coming months, and to share some details about it in part IV of this series.

At Forward Research we seek to push the adoption of Arweave along each of the tracks, acknowledging that their differing time horizons and priorities require separate strategies. In the following parts we will cover each in turn.

-fwd-research

The State of Play Part II: Arweave as the (Inter)National Archive of Cyberspace

This post is the third instalment of an update series on Arweave, our philosophy at Forward Research, and our perception of the state of play. If you are just joining us, please consider reading the other parts first to gain the full context of our work. The contents of this series are as follows:

The Big Picture: Cyberspace Acceleration.
The State of Play Part I: The Macro.
The State of Play Part II: Arweave as the (Inter)National Archive of Cyberspace.
The State of Play Part III: Arweave as the Permissionless Data Lake of Cyberspace.
The State of Play Part IV: Arweave as the Universal Compute Layer for Cyberspace.

Building an institution for cyberspace.

Once Arweave reaches a wide level of adoption it will have 'snowball' effects that lead its archive to grow in prominence over time: Why would one not want to put their most important documents and publications into the global archive of everyone else's important documents? Arweave will become the Schelling point by offering safety in numbers. Once this level has been reached, the argument self-reinforces with each additional piece of valuable data contributed. Reaching this level, however, will take time for the archive to prove itself and social prominence, such that many people know about it. The two primary levers we focus on in order to minimize the time it will take to reach this outcome are:

1. Increasing trust: We must ensure that the network is seen as trustworthy, reliable, and neutral in the eyes of its users, miners, and tokenholders. We believe that the appropriate metaphor for this is to aim for Arweave to become an 'institution' in cyberspace.
2. Increasing dataset prevalence: By seeding the archive with as many important pieces of information as we can, we increase the likelihood that people consider Arweave as a sensible place to house their new important data. As other copies of this data are lost in non-permanent storage systems, Arweave becomes the single – provably unaltered – source of truth for the documents.

In 2023 we executed significant initiatives to further these two aims.

Protocol finalization and documentation

In order for protocols to grant rights to users they must be complete. If a protocol is not complete – all features promised to users being fully implemented – then users must place their trust that interventions will be made to ensure the functionality they desire is actually offered. Users of protocols that are 'incomplete' in this way are essentially in the same position of users of traditional centralized services: Dependent on someone else's managerial activities in order to recieve the service they seek.

With the Arweave 2.6 and 2.7 network upgrades in 2023, Arweave has reached the status of being 'complete'. By our estimations, Bitcoin appears to be the only prominent decentralized network that has reached this level of maturity. Virtually every other network of note has a 'roadmap' that is likely to entice users to employ a protocol for features that are promised but have not yet been delivered. Further, Arweave's mechanism of scaling through bundlers ensures that no matter how many new users come to the network, the protocol will be able to serve them all without increasing fees. At Forward Research, we see this as a highly significant milestone that solidifies Arweave as a mature system that can serve as the bedrock of cyberspace. This milestone also allows the soul focus of the Arweave ecosystem to move to adoption of the network and the services that can be built on top of it.

As the Arweave protocol has now reached maturity and scalability, significant effort was invested this year in finalizing documentation of the protocol. This culminated in the publication of a release candidate draft of the Arweave paper. This paper is engineered to provide a single, succinct venue with a description of every mechanism deployed in the network. In order to aid in garnering the widest possible trust and transparency of the network's mechanisms, the paper refers directly to lines of code in the primary Arweave implementation. This paper will be finalized and published in non-draft form early in 2024.

Governance

Arweave is a permanent information storage system orchestrated by a protocol. By operating as a transparent and immutable protocol, its users are provided guaranteed rights that could not be offered by traditional companies with mutable business models and management. The particular rights that the network aims to provide are encoded in its principles: Enable permanent information storage, ensure the right to speak in cyberspace, and ensure the right to listen.

In designing the Arweave network we have acted to minimize the scope of 'governance' effects which may lead to the removal of rights for users – regardless of how much capital or power someone that advocates for a change has. Despite this, we are also aware that there will be changes in the environment over time that necessitate modifications to the technical architecture of the network. As just a simple example, consider how the emergence of a new storage medium with differing access properties would affect the mechanisms of Arweave.

In 2022 we began to explore how to manage the necessary process of change for the network, without breaking the guarantees that it exists to offer its users. Further, we began to play with designs for systems that would more appropriately incentivize all participants in the process of upgrading the protocol (innovators and 'voters'), while also ensuring that would-be innovators are not burdened with undue bureaucracy or the need to ask for permission in order to build. This effort culminated with the 'Fair Forks: Towards Incentivized Protocol Governance' paper in October 2022. Offered to the community as an indication of our thinking but not a concrete design proposal, we used this prototype to gain feedback and 'temperature check' the ideas. In early 2023 we iterated on this design to create the final governance mechanism for Arweave: The Framework for Evolving Arweave. This system is a simplification of the mechanism described in the Fair Forks paper, removing the 'fork resolution' component that would have been contentious and difficult for the community to operate in practice. Along with the framework itself, we also published a companion document that describes in careful detail the reasoning behind each of the rules in the framework.

Reception of the evolutionary framework in the community was unanimously positive, but muted. Many asked interesting questions, but the mood could largely be described as unconcerned. The ecosystem in general had more pressing issues to attend to. By the close of 2023, however, the evolutionary framework received its first test. We heard through an ecosystem participant, then confirmed by the group themselves, that Irys – the largest bundler of data for Arweave at the time – was planning to fork Arweave, resetting the dataset and token supply. A number of members of the ecosystem (including us at Forward Research) attempted to encourage them to perform a protocol evolution, rather than an anti-social clone of the network. These early efforts failed, and the issue quickly became public. The majority of upload volume them moved away from Irys within a few days, and since that time it appears that they may be moving in the direction of a pro-social protocol evolution.

The incident with Irys at the end of 2023 brought the issue of governance to the surface for the first time in the Arweave ecosystem. Fortunately, the founding constitutional documents of the network were already in place and finalized, so we were able to advocate from strong foundations for a pro-social way of evolving the protocol. By our interpretation, the incident has made Arweavers more focused about the questions of long-term protocol steering, and as a group they seem to be coming to a tentative understanding of the evolutionary mechanism and are beginning to adopt it as their own. In the coming year we will need to work with the Arweave community to better understand evolutionary forks. In the ideal case, the next time a predatorary fork emerges the Arweave community will come together to apply pressure for the new network to be a pro-social evolution of the system, rather than a clone.

The Alex Archive

In the early days of Arweave, a major thrust of adoption work around the protocol was to foster the upload of important records of history to the network. Once these records had been stored in the network and widely replicated, we would then publicize this archive in order to spread the word that it is now possible to store history inside a blockchain, the benefits that brings, etc. This strategy was intended primarily to increase the prevalence of the network's dataset, but also to demonstrate the trustworthy nature of the network's storage.

By mid-2022 it was clear that a strong community had formed around the protocol that saw the utility of holding records of humanity's most important history in a permanent, provably neutral data commons. During the community's campaign to archive information from the Russia-Ukraine conflict in that year approximately 75 million artifacts were added to the network. What was missing, however, was incentives for the archivist to build their system as well as to fund its uploads.

Throughout 2023, builders at Forward Research have built and shipped a solution this problem, called Alex. Alex is a permaweb app that lets anyone create a pool for archiving information of a certain kind. Other users can then contribute to this pool, receiving 'sponsorship' of assets that are stored by the operator. Depending on the topic that the pool is archiving information about, these sponsorships are either tradable assets or 'soul-bound' tokens. By offering a collectible 'artifact' as a reward to those that pay operators and archiving costs, an incentive is created to build high-quality archives on the Arweave network.

In the following post in this series we will explore the use of Arweave as the neutral and composable data layer for cyberspace – a track of adoption that will also benefit from the trust garnered by Arweave's growing position as an institution in cyberspace.

See you there,
-fwd-research

The State of Play Part III: Arweave as the Data Lake of Cyberspace

In this part in our series on Arweave and our perception of the state of play in its road to adoption, we discuss Arweave's potential role as a new substrate for cyberspace. If you are just joining us, consider reading at least the first two parts in this series in order to gain full context for this post. The contents of this series are as follows:

The Big Picture: Cyberspace Acceleration.
The State of Play Part I: The Macro.
The State of Play Part II: Arweave as the (Inter)National Archive of Cyberspace.
The State of Play Part III: Arweave as the Permissionless Data Lake of Cyberspace.
The State of Play Part IV: Arweave as the Universal Compute Layer for Cyberspace.

A new bedrock for cyberspace.

At Forward Research we believe that in the best case scenario, Arweave will become the bedrock of cyberspace. It is an open, permissionless data lake that ensures that data – the core of cyberspace – is held in such a fashion that no party can control or manipulate it. Once data is stored on Arweave it is also outside of the responsibility of anyone to have to pay to maintain, acting as a public commons backed by an endowment. If humanity does 'move into' cyberspace as we predict, but does not upgrade its protocols beforehand, the situation will be dire: All experiences, locations, and identities will be subject to the whims of rulers in the 'outside world'. By 'locking open' the base layer of Cyberspace, inline with the principles of Arweave, we intend to ensure cyberspace is at least as free as physical space – and hopefully more so. This is the outcome we at Forward Research will be fighting for.

Now that the core Arweave protocol itself is robust and mature, the road to building this vision of a cyberspace with guaranteed user rights starts with meeting the mainstream where it is currently at. While our perception of cyberspace is not yet the total sensory experience that we anticipate will eventually emerge, people already spend much of their lives 'in' (focused primarily upon data from) cyberspace anyway (as discussed in-depth in part I). Consequently, the issues of a centralized, non-protocolized cyberspace are already deeply felt by many and will only increase over time. Creators fear 'deplatforming' (removal of access to their identity and right to speak), platforms owned extract the majority of the value flow across every media type, and governments pressure centralized hosts to remove content they would like censored.

By focusing on the data layer, Arweave as a protocol does not discriminate between representations of cyberspace – it will provide users with guaranteed rights regardless of whether it looks like a 3D world, 2D web page, or something entirely different. The composable web is our initiative to bring the benefits of Arweave's open and permissionless data lake to the dominant computing/experience platform – the web – as it currently stands.

What is the composable web?

Put simply the composable web is a stack of tools, with Arweave at its base, which shifts how the web works from a focus on 'pages' and apps to a focus on open and reusable content. Despite this radical shift, the composable web maintains compatibility with all existing web technologies – web browsers, HTML, JS, CSS, etc. When the web was first formulated in 1991 it was intended to serve as an open 'knowledge graph' that was distributed between different 'pages' (themselves modelled after books). Over time, particularly with the advent of 'web2' in the mid-2000s, the web became more of an 'application platform' (a replacement for desktop applications) than anything that resembled a book. The composable web is a third evolution of this platform: Moving from a focus on applications which are hosted at a physical location (a server), to content that is hosted 'everywhere' and is reusable by endless 'thin' applications.

The composable web is enabled primarily by Arweave's ability to make data independent of physical location. Once a piece of content is stored on Arweave, it is available to all parties to download and view. Instead of maintaining their own individual data silos, applications built on the composable web query and embed data from Arweave's open data lake – writing user data back to this permissionless content layer, too.

Data on top of Arweave is stored atomically with embedded metadata: Who uploaded it, how is the data tagged, and even the option to host a smart contract inside the content itself. All of this information is referred to by a single identifier. This structure allows content to move seamlessly between applications, carrying all of its provenance information (who uploaded it), ownership information (who should benefit from it), and licensing rights (how its use should be remunerated). This content representation is perfectly suited to enabling permissionless composability of data and is unique amongst smart contract ecosystem. We call these pieces of content atomic assets.

Because all of the data in the composable web is openly accessible, there are numerous querying services that index them, all operating upon the same basic API specification. As an application developer this allows for a radically simplified style of building: Create static user interfaces (themselves stored on Arweave), and then use open infrastructure to be a generic 'backend' for your app. Every 'app' becomes a simple frontend on the 'platform' of Arweave: Just as Uniswap.org and Gmail.com are thin wrappers around their underlying protocols.

There are numerous advantages to this reorientation of the web around content for its different stakeholders. Let's glance at each in turn:

Creators

• Own your own identity: On the centralized web creators spend virtually their entire life building audiences on platforms that are owned and controlled by corporations. As a result, access to their own audience and their identity itself (the ability to post from their account) is at the direct discretion of the owners of the platform. This means that a single speech infraction (amongst an ever-changing set of rules) can ruin a creator's career. On Arweave this is not the case: Creators who own their own keys, own their own voice.
• Publish once, earn everywhere: When a creator publishes on Arweave, their data is licensed and reusable by any application. This means that as soon as the data is uploaded to the network, it will be displayed to every single user – on every single app – that can make use of it. This maximizes exposure and revenue for the creator, while helping them to build a larger audience.
• Dictate your terms to apps – not the other way around: When creators upload to Arweave they have full control of the license that they choose to issue their content under, thanks to Arweave's native licensing standard. This control lets them dictate terms to those that would want to use their data, rather than being forced to accept the terms offered by the platform in order to gain access to their audience.
• Trade revenue streams from your content: All atomic assets on Arweave can be licensed such that royalties from usage of the data flow to the owners of a smart contract. That ownership, because atomic assets can have smart contracts embedded directly inside, can then move seamlessly between apps along with the content's other metadata. Ownership of the tokens inside the smart contract are, of course, tradable – enabling a permissionless and liquid market for the royalties from content on the composable web. In practice, this allows creators the powerful advantage of being able to access the future cashflows of their content immediately (in the same way that farmers and Bitcoin miners sell futures on their outputs), while allowing investors and members of their community to become joint owners of these data commodities.

Developers

• No cold start: Typically when a developer creates a new web2 application, they start without any content in their platform. Because the presence of content itself draws in more content (a form of 'network effect'), this 'cold start problem' stops many innovative new apps from getting off the ground because they cannot compete against established players. These content network effects further the ability of web2 platforms to exploit their userbases, as competition to their market dominance – even if the userbase is unhappy – is scarce. The composable web on Arweave inverts this situation: All developers can build on all data on the web, paying creators fairly for their content, from the start. We expect that this ability alone will unlock a powerful new competition to improve product quality in app experiences on the composable web.
• Run no infrastructure – build on hyperscale 'backends' from day one: Much of the task of building a web2 application that runs at scale is not in engineering the user interface (the component that the user perceives as the 'product'), but in the creation of performant indexing, compute, and content delivery systems. Building on the composable web stack removes this complexity entirely, by allowing you to build on generic backend infrastructure ('gateways') that are already running at-scale. Developers do not even have to pay to make use of this infrastructure, as users develop their own relationships with gateways in the same way that they do with ISPs on the traditional internet.
• Get paid when others use your data: Instead of seeing data as their 'moat' that stops others from competing against them, the composable web allows developers to use the data that people upload to Arweave through their app as a secondary revenue stream. To achieve this, developers can grant themselves ownership of some of the tokens in the assets that are uploaded through their apps to the network. The baskets of tokens that they then acquire in content from their app can itself be tradable – creating a new way to monetize the value of their app, without having to sell control of it.
• Pay once to publish and never pay again: Because applications on the composable web are 'thin' user interfaces with no specialized infrastructure, the applications themselves can be uploaded to Arweave. This enables the apps to run permanently on the network with developers paying once to contribute to Arweave's endowment, and never having to pay for upkeep or perform maintainance again.
• Delegate content moderation: One of the core issues for user-generated content platforms in the web2 era has been the adjudication of content moderation decisions. Due to the nature of the web protocols themselves – that content exists in one specific location, owned by one specific party – all web2 platforms have the power to censor their users, regardless of whether they actually want that power or not. On the permaweb, content indexing and storage is outsourced for application developers to gateway and mining node operators. This frees the developer from the burden of content moderation decisions, while moving the power closer to their users, whose choice of underlying gateway will dictate the content they see in any composable web app.

Consumers

• All of your favorite content, in every app: On the composable web there are no borders between applications. This means that as a user, you can choose the application that you like the most in order to access any content. This allows for a hyper-customized experience of the web: Instead of being forced to use the same user interface as everyone else in order to view the content you like, you can use an interface that is perfectly tailored to your needs.
• Choose your own content moderation: On the composable web, users choose their own gateways. As gateways are the primary means of content moderation and filtering on the composable web, this free choice of infrastructure to access their apps on also delegates to them the ability to choose the content they see in their apps. This lets them express their preferences – more or less conservative, liberal, or family-friendly – without affecting the rights of other users.
• Guaranteed access to your favorite creators: Creators on the composable web are identified using their wallet addresses. One effect of this is that just as creators cannot be deplatformed from applications on the composable web, users are also always able to access the content of their favorite creators – without relying on centralized platforms or companies to grant access. Creator-audience relationships on the composable web are truly peer-to-peer: Intermediated by no other parties.
• What you see is what you get, permanently: When users enjoy an application on the composable web they can use it with confidence that it will never be taken away from them. If the application itself is built upon generic gateway APIs and is stored on the permaweb, the user of that application will always be able to access it and use it in its current form. Amongst many other effects, this right to access the app in its current form fundamentally shifts the power relationships between user and developer: Users must now willingly upgrade to new versions of the app – they cannot be forced by the developer. This presents a radical shift in user rights from the web2 experience, where developers often unilaterally changed applications to extract more value after a captive userbase had been established.

There are numerous other advantages to the architecture of the composable web, but these are some of the primary ones. It is our expectation that the commercial advantages will be the fastest drivers of adoption, while the rights guaranteed to users by the underlying protocol will cause them to stay. Commerce will be the carrier; a cyberspace with guaranteed rights is the payload.

In 2023 we made strong progress on building the experience of the composable web for end-users. While typically we have focused more on the protocol development side of the Arweave ecosystem, we observed at the start of 2023 that the infrastructure in the ecosystem was maturing well, but end-user facing applications and experiences were still lagging. We also felt that the 'bear market' in decentralized applications had created additional (healthy) pressures on the ecosystem to provide real-world utility to the 'mainstream' quickly, while simultaneously diminishing the number of new founders that were entering the industry who could build these experiences themselves. The summation of these factors led us to a strategy pivot: Do not wait for others to build – build the kernel of the ecosystem ourselves, then let others step in to iterate on top once its utility had been demonstrated. The remainder of this part details the initiatives that have constituted our initial thrust to build a 'minimum viable ecosystem' for the composable web.

The Universal Data License (UDL)

During the 2020-2022 crypto market cycle we observed first-hand the NFT hype wave, as Arweave was used as the data store for many of those assets. We also observed that although the theory of 'web3 data' – open data lakes that all could access and build upon – had been well-discussed, it had not taken off under its own steam. We believe that the reason for both of these trends is simple: Incentives. The 'web3 data' movement has not yet gained significant traction because developers are still exposed to the incentive to keep their data private as a 'moat', rather than to share it with others. Conversely, NFTs took off because they had the inverse property: The more widely an NFT's media was distributed, the higher its potential value. In order to iterate on the incentives of the NFT markets away from simple 'greater fool' games, and to create incentives for devs to contribute app data we created the Universal Data License (UDL): A way to license data uploaded to Arweave, turning it into a secondary revenue stream.

UDL is a 'real-world' license that flips the model for developers, so that sharing data with others is more profitable than keeping it private. When UDL is added as a license to data uploaded to Arweave, developers can legally enforce that when other people use the data they have uploaded a royalty fee is paid. Developers are then incentivized to have their data to be distributed as widely as they can – just like NFT images.

The UDL is a 'programmable' license, in the sense that developers can add tags to their uploads in order to modify certain parameters of the license. For example, adding one tag may specify how license fees should be paid when data is used, another may specify if and how derivatives of the work may be created. By making the license standardized but flexible, potential users of the data (apps and other platforms) can discover it on the network with ease via the gateway indexers in the ecosystem.

The Universal Content Marketplace (UCM)

As we have discussed, data uploaded to Arweave with UDL tags can create differing real-world rights associated with the content, depending upon configuration. UDL parameters can also configure how the ownership of these rights should be distributed. A sensible choice for this option is often to grant legal ownership to the owners of a token that is embedded inside the data itself, using the SmartWeave architecture. This means that the single identifier of the 'atomic asset' contains the content itself, provenance information (who uploaded it, when, etc.), as well as licensing information and the legal owners of the data. All of this information moves through the comopsable web and between applications in a unified form – each component inseparable from the others.

One of the key components of the composable web ecosystem that Forward Research has built and shipped this year is the Universal Content Marketplace (UCM). The UCM is an atomic asset exchange which is tailored in order perfectly support the trading of pieces of content from applications on the composable web, offering a liquid market for all forms of data. In order to assist in this, the contract's first user interface (the BazAR) makes use of the Render-With standard to allow apps to store machine-readable data, while offering a human-centric rendering of the data where appropriate.

Stamp Protocol

As well as allowing for data to flow between appllications seamlessly, the composable web also allows features/aspects of apps to be re-used across different experiences. An example of this is ArProfile – Arweave's native profile manager, which allows people to set a username and avatar, which is then available across every app on the network. Another example is Stamps: A mechanism we designed and shipped this year that offers a 'universal like button' across every application. This system is already in-use across a number of apps on the permaweb, including now, and the Bazar, enabling for the curation and highlighting of new content between applications without any explicit coordinationbetween devs.

Another component of the Stamps protocol is its native token: $STAMP. Stamps are a 'proof of quality content' coin for the permaweb, minted in a fair launch (no pre-mine) and distributed proportionately to the stamps that content receives on the permaweb each day. Stamp coins can be consumed in order to 'super stamp' content, signalling that it is seen as highly meaningful to the consumer.

Because of the composability of the network, the Stamp Protocol mechanism allows developers of any app to give $STAMP token rewards to users of their apps for providing content that other (Sybil resistant) users enjoy. As the $STAMP token rewards are distributed to the owners of the atomic assets (if one is embedded within the content that has been stamped), even the flow of stamps themselves can be traded – just like royalties from UDL licensed content.

Permafacts

During 2023, Facts Protocol was also readied for launch. Facts protocol is a simple mechanism for allowing people to bet on how true/false, and relevant a piece of information is. Fact markets reward users who buy tokens on 'true' or 'false' from a bonding curve embedded inside the content before other users do. As the price increases, prior purchasers may sell their tokens back to the contract, redeeming their value at the new price while diminishing it back towards zero. As well as rewarding users for predicting how true and relevant the crowd will believe a piece of information is, the Fact Market mechanism also produces a Sybil-resistant, 'skin in the game' weighted assessment of perceived truthiness for content expressed as a simple number.

Just as Stamps Protocol allows users to express their enjoyment of any piece of content across the composable web, so too is true for Facts Protocol: Regardless of whether Fact Markets are integrated into an app's primary UI, users can attach the contracts to any content within the app. We intend that this system will provide users with a fair and neutral way to quickly assess the likelihood that any pieces of content on Arweave are true.

The DataOS

The composable web is a version of cyberspace where content – rather than pages or apps – is primary. As we have discussed, this data-centric web significantly lessens the complexity of building new apps and experiences as a developer. Instead of needing to build the scalable infrastructure, content library, and social network that is necessary to make modern consumer apps work, a developer needs only to focus on building a great UI on top of Arweave's open infrastructure (gateways). Late in 2022 we saw the emergence of transformer-based LLMs as a production-ready technology. One of the remarkable capabilities of these systems is their ability to generate code – including web user interfaces. This begs a question: What if every user interface for Arweave's composable web could be generated on-the-fly by AI that responds to user feedback and preferences? To explore this we have been funding an implementation of the DataOS.

████████

Once running, the composable web will solve the 'cold start' problem for developers. They will be able to build on top of a deep library of content that already exists inside the Arweave network, rather than having to grow their new application from an empty database, lacking users or data. In order to kickstart the composable web, however, we will have to solve the cold start problem a final time, in order to seed the composable web with data.

While there are already >2.1 billion pieces of data stored on the network, most of this is not composable assets that are intended for other developers to build upon. Subsequently, in order to seed the Arweave data lake Forward Research has invested resources this year in acquiring a large 'decentralized' social platform to bring to the network. This platform has a significant (>5m MAU) userbase who are excited about the prospect of decentralization, but have not previously been given the appropriate technology to actually fulfill this vision. By bringing this audience to the Arweave network, we intend for them to form the initial cohort of users for the applications of the composable web – representing a large enough critical mass to have their own network effects.

The process of finalizing this purchase is still underway, and thus the name of the app must be omitted at the time of writing.

Once the composable web is running at full-steam, we anticipate that the decentralized compute requirements for its financial infrastructure will be significant. Atomic assets on the composable web would be most appropriately traded using onchain order books (like UCM), which alone could lead to massive numbers of onchain trades as bots operate on the markets. In order to service the composable web's financial infrastructure, a sufficiently scalable compute layer for it must be provided. In the next post in this series we will discuss a new compute layer for Arweave that Forward Research has designed, in order to solve this problem.

Join us then!
-fwd-research

The State of Play Part IV: Arweave as the Universal Compute Layer for Cyberspace

This post is the final part in our series about the state of play surrounding the Arweave protocol, and Forward Research's opinion about how it will progress. If you would like to catch-up with this series, please consider reading the prior parts first in order to establish the complete context. The contents of this series are as follows:

The Big Picture: Cyberspace Acceleration.
The State of Play Part I: The Macro.
The State of Play Part II: Arweave as the (Inter)National Archive of Cyberspace.
The State of Play Part III: Arweave as the Permissionless Data Lake of Cyberspace.
The State of Play Part IV: Arweave as the Universal Compute Layer for Cyberspace.

Permanent storage as infrastructure for decentralized compute.

As we discussed in 'The Macro', a rapidly rising trend in 2023 was the use of Arweave to serve as a component in other decentralized computing infrastructure. In essence, this use of the system stems from Arweave's ability to provide neutral and decentralized replication of data inputs to other distributed systems. This ability allows other decentralized services to rely on Arweave as a means to make their states recomputable at a later time (in the case of smart contract platforms or decentralized databases), or as a radically cheaper form of onchain storage for 'bring-your-own-data' architecture systems.

As well as its utility, the core enabler of Arweave's usage as infrastructure for other forms of decentralized services is its scalability. Due to the network's bundling mechanism for commitments to its state, an arbitrary number of users can make use of the system at anyone time. In 2023, the fact that Arweave can scale in this manner was demonstrated in practice for the first time. At the start of the year, the network would normally operate at a rate of ~10 user transactions per second. By the end of 2023 the network was typically processing transactions at a rate of ~190 per second. During one episode near the end of the year, the network experienced a sustained period of transaction volumes over 1,000 per second. Despite this enormous spike in usage, Arweave never needed to employ a fee market in order to throttle usage of the network. Remarkably, in the present decentralized technology ecosystem Arweave is still the only protocol that can offer onchain storage at this level of scalability to its users.

ao: Decentralized compute at cyberspace-scale.

Many of the ways in which Arweave can be used to build more performant and scalable decentralized compute platforms have already been explored by the Arweave community. At Forward Research, however, we believe that a major opportunity has still not been grasped: A decentralized computer, built on Arweave, that enables arbitrary numbers of parralell processes, coordinating via message passing. During 2023, and in concert with many in the Arweave ecosystem, Forward Research has deigned and built the first version of this system.

Our practical motivation for building ao was simple: Once the composable web on Arweave is adopted, we anticipate that a significant financial ecosystem will emerge around its content. As content token trades will likely occur using onchain Central-Limit Orderbook (CLOB) exchanges (rather than Automated Market Makers), there could be a very significant number of interactions per second, even on a single asset basis. In order to support this ecosystem, a smart contracting facility must be established for the composable web that scales sufficiently to match these needs. We built ao in order to meet this requirement.

While a full discussion of ao is far outside of the scope of this document and will come with its official launch, its core features are as follows:

• Arbitrary numbers of paralell processes: In ao, applications are built of any number of communicating processes. Inspired by Erlang, ao does not allow processes to share memory between one another, but does allow them to coordinate via a message-passing system. Each of these processes can run at the full speed of the computing resources that are available, and by allowing them to coordinate with one another ao enables scaling mechanics that are far more similar to traditional web2/distributed systems environments, than traditional smart contracts.
• High resource utilization in contracts: Building on the lazily-evaluated architecture of the original versions of SmartWeave, nodes in the ao network do not need to perform any compute in order to reach consensus about program state transitions. As a consequence, the initial ao implementation has access to the full 4GB of memory addressable inside WASM containers, support for any quantity of sequential compute, with builtin upgradability to add additional machine types later. Compute costs are then delegated to users who can either calculate their own states, or delegate execution to nodes of their choosing.
• Loading of data from Arweave into contracts: ao processes are able to load transactions from the Arweave network directly into their memory for execution – regardless of size. When paired with the vastly improved resource limits, this capability significantly opens up the design space for programs that can run on the system, beyond the normal contstraints of smart contract platforms.
• Self-waking contracts: In traditional smart contract environements (like Ethereum, Solana, Polygon etc), contracts 'wake up' to perform compute at the request of a user transaction. This creates a computing environment that is not 'live' unless a user interacts with it, lessening the scope of applications that can be built ontop of it. ao removes this limitation by allowing contracts to have 'cron' interactions that automatically wake them up and execute compute at set intervals. Any user, or indeed the application itself, can 'subscribe' to a process in order to trigger the evaluation of the compute at the appropriate frequency.
• Modular architecture supporting extension: ao's core architecture is an open data protocol that anyone can build an implementation of. Everything – from the sequencers, message passing relayers, and even the virtual machine of the system – can be swapped out and extended at will. This flexibility will allow the existing smart contracting systems in the Arweave ecosystem (Warp, Ever, Mem, et al) to easily plug into ao and be able to send and receive messages from the wider network. This will also allow all of these smart contracting systems to share the same infrastructure and tooling, making for a more unified experience of compute on the composable web.

While our initial mission for ao was simply to enable smart contracts to run at a sufficient scale to serve the composable web's content finance ecosystem, we now see that its uses will span far wider than that. For example, we see that Filecoin-style temporary storage contracts with stake would be trivial to build ontop of ao, giving a cheaper (albeit less resilient) storage option than Arweave for the composable web. Further, the system is so scalable that it can be used as a decentralized, verifiable compute network for all of cyberspace – even personal computing. This vision is somewhat akin to Urbit, a network designed to allow users to have their own 'personal server' in a decentralized compute environment. ao, however, has consensus on every interaction on each process, making it far more portable between compute nodes and providing greater guaranteed rights to end-users.

There is so much more to share with you all about ao and the future of compute in cyberspace, with Arweave as its base. Thank you for taking the time to read this series about our perspective on the state of play in cyberspace, and how we are attempting to build a new foundation for it with Arweave. If you have comments, questions, or feedback, we would love to hear from you. Please shoot us a weavemail! Our address can be found on the Forward Research website.

See you on Arweave,
-fwd-research