By Sticky Keys

There are two possible futures for the web.

The first is a web that includes many of the technologies being talked about, used, and hacked on in the web3/dweb/blockchain/crypto ecosystem today. It’s a web in which users have more control, apps are more transparent, and economic opportunities are distributed to more of the web’s users. This is a web of greener grass. It’s not a radical departure from the web we currently have, but it’s different, it’s newer, it uses crypto, and it rewards a new generation of tech entrepreneurs.

The second possible future for the web is something that is almost impossible to imagine or describe, because we’ve never seen anything like it before. It’s a web that includes many technologies, most of which are yet to be built and many have not yet been conceived of. I think that’s exhilarating.

Both of these versions of the web seem ideologically quite similar, while their practical implementation is necessarily divergent. This second version of the web is far more radical than grass that’s a deeper shade of green. It’s the one that will increase the quality of human life, having a positive impact on billions of people around the world. And this second version of the web is crucial for humanity. That’s because it is the collective vision of the most underestimated communities in our society today. This is the web built by queer and trans communities of color and our allies.

I’ll be referring a lot to QTBIPOC communities. That stands for communities made of queer, trans, Black folks, Indigenous folks, and other people of color. But for the sake of this discussion I am also including in this definition allied communities who may share some of those identities, as well as allied peoples who are marginalized, facing crisis, mistreatment or oppressive rule.

While I acknowledge that web3 and dweb are not the same. For the purposes of this discussion they are similar enough to be considered roughly synonymous. So I’ll use those two terms interchangeably and refer to them as the Type I web. The Type I web ultimately is any iteration of the web that is not created by the aforementioned identity groups.

The only web truly different from existing social paradigms, and truly radical, is a web built and driven by Black trans people. This phenomenon is enough of a departure that it will knock the current trajectory of the web out of its orbit into infinite possibility.

Even without knowing in much detail how the Type II web will work, we can already have a sense of its impacts. The impacts are clear based on the values and strategies of QTBIPOC+ communities to solve the problems we are facing today. One of the core values is to make it possible for the most underestimated and marginalized people in society to thrive. Which means everyone in the world thrives. And that’s the gay agenda!

Today, web technologies are already a tool and resource that supports the strategies that QTBIPOC+ folks use to make thriving possible. It’s important to recognize that these strategies have been developed and learned out of necessity. The community thriving is essential to the community surviving.

The dweb movement has a set of guiding principles broken down into 5 categories: technology for human agency, distributed benefits, mutual respect, humanity, and ecological awareness. As noted on their website as of March 15, 2023, these are related to others developed by marginalized groups such as the Indigenous data governance principles (https://www.gida-global.org/care), the principles of the Detroit Digital Justice Coalition (https://www.detroitdjc.org/principles/), and the Feminist Principles of the Internet (https://feministinternet.org/en/page/about), to name a few. The overlap and resemblance between these and the dweb principles does not come as a surprise to me because one of the two primary stewards of the dweb movement is not a cisgender white man. She is a person of color and a woman, Mai Sutton. She has spoken and written about seeing the value of these principles in part because of her personal identity and life experiences.

An attendee of the most recent of the annual dweb camps, 2022, recalled their experience at the camp describing it as well-intentioned but strange because many of the conversations there were about groups of people who overwhelmingly were not represented among the attendees (for example people of color or formally incarcerated folks). This attendee is a journalist and a Black woman. Her experience and observation points directly to one of the major problems of the Type I web: it almost entirely excludes from the development process the people it plans to onboard.

The Web3 Foundation is striving towards similar aims as the dweb but from a more technology centered approach than a human centered one. As of March 15, 2023, the About page of their website states: “Our passion is delivering Web 3.0, a decentralized and fair internet where users control their own data, identity and destiny.”

Five years ago, Gavin Wood, the founder of the Web3 Foundation gave his first so-called “radical” talk about web3 in which he said, “we are by our nature centralized beings. We see the world in a centralized way.” This is an extremely Eurocentric view of humanity. It is not a fact. It is a reference to a culture shared only by roughly 20% of the world. This is another great example of the problem at hand: while the ideologies of the Type I web draw heavily from QTBIPOC+ movements throughout history, the people trying to build based on these ideologies have comparatively no experience living them.

QTBIPOC+ communities are the only group of people suited to create the Type II web.

For this reason, QTBIPOC+ communities are the only group of people suited to create the Type II web. Our communities have laid the ground work for what the Type I web so often claims to be striving for. The work of the civil rights, gay rights, and trans rights movements have firmly placed our community in a position to make the Type II web a reality.

Now let’s examine the practical overlaps, areas of divergence and things that we might expect to see in the Type II web that lack technical precedents today.

Communities are sources of resilience that enable us to survive and sources of joy which allow us to thrive. When we are in a community of people who can empathize with us and who we feel understood by, we are able to have the support we need in the face of adversity. As we seek and find media and experiences that validate or reflect our own experience, we gain a sense of belonging and connection.

Strong communities are made of people who develop mastery over the interface by which they communicate. This means learning how to most effectively communicate with each other, make decisions, settle disagreements, and adapt to the unexpected together with consideration for individual differences in communication from language, to tone, to speed, medium, etc. This only works if you consider each person in the community to be worthy of this type of attention. Strong communities care for each person inside of them equally and they also acknowledge that treating everyone the same is not an effective solution to meeting the needs of the collective.

In the context of a decentralized web, mastery over the interface means developing a protocol by which peering nodes communicate that accommodates both the expected and unexpected. This is already present in a lot of peer to peer systems that are designed for fault-tolerance. By using redundant communication, we ensure peers have ways to coordinate in the face of a node going down or a node sending duplicate messages whether accidentally or on purpose.

What is missing is the sense of equity and equality between peers. In web3 today we tend to treat peers the same but do not see them as equally valuable, resulting in explicit or implicit hierarchy. Our expectations for how each peer operates is static and any peers that do not meet the expectations become ignored or replaced.

However we can re-frame peering relationships and we have a set of tools to do so in a practical sense. This is the first place where we see a gap between the Type I web today, and what queer and trans BIPOC communities do to stay strong. The Type II web can use mechanisms to distinguish between peers in order to increase good behavior in a decentralized system and minimize disruptive behavior. This includes localized reputation systems to know who to trust, fraud proofs to detect bad actors, random sampling as a form of accountability, and designing systems where there is little to no reward for not acting honestly. With these existing techniques we can maintain a structure where each peer is of equal value to the system as long as they coordinate under the agreed upon protocol and with some honesty and integrity.

I think many American children grew up hearing the phrase “sharing is caring” but somehow it is so far removed from modern day American society that it is almost impossible to see in action. Instead of sharing, we expect people to earn so we can exchange. Instead of sharing, we expect people to beg so we can donate.

Instead of sharing, we expect people to earn so we can exchange. Instead of sharing, we expect people to beg so we can donate.

So what is sharing? I think of sharing as a practice to give people useful things. It comes in many forms: it can be an act of aid during a time of crisis, an act of gift giving, or an act of resiliency to prevent a single point of failure due to the centralization of a resource.

QTBIPOC+ communities share because of mutual understanding of their experiences, empathy, and compassion. In 1969, the Black Panther Party started the Free Breakfast Program which resulted in feeding 20,000 children that year. This was the same year as the Stonewall Riots in Greenwich Village where trans women of color fought against police brutality. Then Black trans women started sharing houses in New York City in the 1970s as the Ballroom scene developed, doing the work to help keep each other out of homelessness. Today these communities among many others who are systemically mistreated and marginalized continue to share home cooked meals, clothing, skills and helpful information.

The concept of sharing is vocalized a lot in web3 and the dweb, but it is still considered under the framework of exchange when we talk about sharing compute or storage, and it is still considered under the framework of charity when we talk about grants and public goods funding as ways to get communities what they need.

In the Type II web we can expect peers to share storage and compute freely, because decentralizing these resources allows for greater redundancy and resiliency. We can also expect to see peers sharing responsibilities as this is known to be an effective practice for marginalized communities.

In the Type I web nodes are hyper specialized, limiting potential peering relationships and the total set of actors that contribute to decentralization. In contrast, when a network has a fluid topology in which roles rotate between peers, skills can be more distributed and accountability can be shared. This prevents concentration of knowledge and or power, keeping the group structure flat and it also keeps peers engaged in meaningful rather than trivial participation. All of which results in a network that is more resilient to external forces or internal conflict.

Once a community is formed, governance practices are used to determine how the community should work together and collaborate. QTBIPOC+ communities determine their form of governance intentionally through agreed upon norms, expectations, and decision making frameworks that support their goals. Trust is then established as community members learn how to utilize their governance structure and make progress.

Cryptography happens to be super useful for developing a form of digital trust and of course digital identity. Both are essential for effective collaboration. So we can have digital applications in which community members can identify each other with a high degree of confidence, as we already do in both web2 and the Type I web, and from there members can collaborate online in a way that translates to actionable outcomes offline. What I’m describing here is basically what the chat applications Signal or Element are today and also how DIDs are being used in the Type I web today.

So given we have effective collaboration tools, what does governance actually look like? One effective form of governance is consensus. While many of us in the web3 space are familiar with consensus in terms of what it technically means and how the nodes we run eventually reach it, few of us actually practice consensus based decision making in our daily lives. So, what is consensus? Consensus is where everyone agrees.

Dean Spade is a queer and trans liberation fighter with two decades of experience in racial and economic justice work. In their book entitled Mutual Aid, they describe consensus as “a radical practice for building a new world not based on domination and coercion”. While the standards groups and DAOs that we participate in today have governance mechanisms to create proposals, iterate on them, and turn them into actions, these groups overwhelmingly operate based on majority rule.

Mutual aid groups that support people in the most vulnerable times of crisis understand consensus intimately because they practice it in order to keep their organizations alive. Consensus takes into account all views to sufficiently meet the needs of the collective. The practice of consensus is difficult and can be time consuming but it works when everyone in a group is committed to the same cause and value system. Without consensus, groups stop treating everyone equally, which results in churn and power imbalances that lead to dissolution.

The Type I web is at a point of acknowledging the power of global and local forms of consensus in decentralized systems. However in the case of governance I think we will find that majority rule is not enough to build apps that are truly in service of their users or DAO members.

In the Type II web the use of consensus in governance will lead to constituents having a higher level of education about the systems they are using, so that they can make informed decisions. When people know that their opinions actually matter and that they have the power to directly impact outcomes, there is a stronger incentive to engage deeply in the decision making process.

Self-determination is the ability to have control and agency over your own life. Though there are times when a person can not change an outcome, being able to identify and act on the power they have while maintaining self-authenticity, promotes their resiliency.

Having self-determination online means having the autonomy to opt in or out of a system or feature within a system at will. This is only possible if there are meaningful alternatives available for use. It means that there is no vendor lock-in forcing individuals to engage with a set of tools or technologies that are not valuable to them.

High switching costs is one of the biggest barriers towards digital autonomy and it leads to centralization that can be fatal for a community or network. While open source software makes autonomy possible, it isn’t enough for it to be possible for users to switch to a new service or application, it must also be easy and convenient to do so.

It isn’t enough for it to be possible for users to switch to a new service or application, it must also be easy and convenient to do so.

High switching costs can be prevented when we have interoperable models for decentralized data. In this case we can use things like exit and migration protocols. Yet today we are still lacking technical tools that make switching practical because it is very difficult to do something useful with large amounts of data within a reasonable amount of time and without a large amount of effort.

Teia, the Tezos NFT marketplace which was a fork of HEN is a good example of when this has worked. The community got together after the HEN project was abandoned without notice, and were able to set up a replacement service that pulled in the existing data and kept the functionality of HEN. ENS is a good example of when this failed to happen. A leader in the organization turned out to be toxic for much of the community, but a fork was not able to be created probably due to the difficulty of forking and adopting a suitable replacement system.

One technical solution we might expect to see emerge in the Type II web is self-replicating systems that provision the services needed for the system to make progress but act upon isolated data sets, resulting in independent sub networks, or even the same exact data set so you simply have a back up service.

The final important practice that I want to point out is activism. Activism is a form of self-defense from oppression. It is designed to stop and prevent harm. And it is among the most powerful things a community can do to protect itself. Independent of the outcome, simply engaging in activism serves to promote resilience for individuals and communities by contributing to feelings of self-determination.

A digital equivalent to the practice of activism would be protecting a system from bad actors. As already discussed there are many ways to do this in a decentralized peer to peer system from data redundancy, to cryptographically verifiable audit trails, to making it extremely difficult and unreasonable to perform a bad action in the system.

For me, one thing that stands out as one of the most effective ways to make a system resilient to an attack is the ability for the system to reform, stronger than before, after breaking down. This could mean reforming a group after removing a member, splitting a group because people disagree about what is acceptable conduct, or even changing a group’s purpose entirely to better satisfy the needs of the group.

As previously mentioned, this is where we see the power of forking. The open source movement puts us in a position where reforming a digital system or community is possible. If an attack happens in a system we can minimize or nullify its negative impacts by forking and adapting the system so that the attack is no longer possible. ETH Classic is an example of this. Forking is also already a political stance, and potentially an activist action within the context of the technology where it occurs.

Beyond forking we have the ability to make decentralized systems self-healing for example by restarting after a failure, we can put in place things like dynamic scaling based on traffic demands as a form of self-preservation in a system, and we can migrate actors in the system and their data to new protocols or digital services.

I’ve laid out the practices that marginalized communities use today in order to reach a state of thriving, explored how they relate to the technologies we use in the Type I web today, and I also presented a few open problems where technical precedents do not yet exist. The remaining question of this discussion is how do we fill the gaps and make the principles of our queer and trans BIPOC communities a reality in the next version of the web? While it can be difficult to imagine what this Type II web really will look like, it’s not actually impossible to imagine or to make real. It’s already happening. The reason that I got into this space in 2017 was because I saw Black people doing this work in the space at that time. I saw trans people doing this work in the space at that time.

We don’t need all the answers right now, but we do need to continue asking the right questions. I personally have more questions about where we’ve been and how that relates to where we are headed. At this point, I’ve learned that I need to change the way that I think about leadership in this space, and the lens by which I examine technical solutions to determine if they really are living the principles that we talk about or not. And if not, what needs to change and who is actually the best person to make that change possible?

[This work was first presented as a talk that I gave at ETHDenver 2023.]

By Sticky Keys

There are two possible futures for the web.

The first is a web that includes many of the technologies being talked about, used, and hacked on in the web3/dweb/blockchain/crypto ecosystem today. It’s a web in which users have more control, apps are more transparent, and economic opportunities are distributed to more of the web’s users. This is a web of greener grass. It’s not a radical departure from the web we currently have, but it’s different, it’s newer, it uses crypto, and it rewards a new generation of tech entrepreneurs.

The second possible future for the web is something that is almost impossible to imagine or describe, because we’ve never seen anything like it before. It’s a web that includes many technologies, most of which are yet to be built and many have not yet been conceived of. I think that’s exhilarating.

Both of these versions of the web seem ideologically quite similar, while their practical implementation is necessarily divergent. This second version of the web is far more radical than grass that’s a deeper shade of green. It’s the one that will increase the quality of human life, having a positive impact on billions of people around the world. And this second version of the web is crucial for humanity. That’s because it is the collective vision of the most underestimated communities in our society today. This is the web built by queer and trans communities of color and our allies.

I’ll be referring a lot to QTBIPOC communities. That stands for communities made of queer, trans, Black folks, Indigenous folks, and other people of color. But for the sake of this discussion I am also including in this definition allied communities who may share some of those identities, as well as allied peoples who are marginalized, facing crisis, mistreatment or oppressive rule.

While I acknowledge that web3 and dweb are not the same. For the purposes of this discussion they are similar enough to be considered roughly synonymous. So I’ll use those two terms interchangeably and refer to them as the Type I web. The Type I web ultimately is any iteration of the web that is not created by the aforementioned identity groups.

The only web truly different from existing social paradigms, and truly radical, is a web built and driven by Black trans people. This phenomenon is enough of a departure that it will knock the current trajectory of the web out of its orbit into infinite possibility.

Even without knowing in much detail how the Type II web will work, we can already have a sense of its impacts. The impacts are clear based on the values and strategies of QTBIPOC+ communities to solve the problems we are facing today. One of the core values is to make it possible for the most underestimated and marginalized people in society to thrive. Which means everyone in the world thrives. And that’s the gay agenda!

Today, web technologies are already a tool and resource that supports the strategies that QTBIPOC+ folks use to make thriving possible. It’s important to recognize that these strategies have been developed and learned out of necessity. The community thriving is essential to the community surviving.

The dweb movement has a set of guiding principles broken down into 5 categories: technology for human agency, distributed benefits, mutual respect, humanity, and ecological awareness. As noted on their website as of March 15, 2023, these are related to others developed by marginalized groups such as the Indigenous data governance principles (https://www.gida-global.org/care), the principles of the Detroit Digital Justice Coalition (https://www.detroitdjc.org/principles/), and the Feminist Principles of the Internet (https://feministinternet.org/en/page/about), to name a few. The overlap and resemblance between these and the dweb principles does not come as a surprise to me because one of the two primary stewards of the dweb movement is not a cisgender white man. She is a person of color and a woman, Mai Sutton. She has spoken and written about seeing the value of these principles in part because of her personal identity and life experiences.

An attendee of the most recent of the annual dweb camps, 2022, recalled their experience at the camp describing it as well-intentioned but strange because many of the conversations there were about groups of people who overwhelmingly were not represented among the attendees (for example people of color or formally incarcerated folks). This attendee is a journalist and a Black woman. Her experience and observation points directly to one of the major problems of the Type I web: it almost entirely excludes from the development process the people it plans to onboard.

The Web3 Foundation is striving towards similar aims as the dweb but from a more technology centered approach than a human centered one. As of March 15, 2023, the About page of their website states: “Our passion is delivering Web 3.0, a decentralized and fair internet where users control their own data, identity and destiny.”

Five years ago, Gavin Wood, the founder of the Web3 Foundation gave his first so-called “radical” talk about web3 in which he said, “we are by our nature centralized beings. We see the world in a centralized way.” This is an extremely Eurocentric view of humanity. It is not a fact. It is a reference to a culture shared only by roughly 20% of the world. This is another great example of the problem at hand: while the ideologies of the Type I web draw heavily from QTBIPOC+ movements throughout history, the people trying to build based on these ideologies have comparatively no experience living them.

QTBIPOC+ communities are the only group of people suited to create the Type II web.

For this reason, QTBIPOC+ communities are the only group of people suited to create the Type II web. Our communities have laid the ground work for what the Type I web so often claims to be striving for. The work of the civil rights, gay rights, and trans rights movements have firmly placed our community in a position to make the Type II web a reality.

Now let’s examine the practical overlaps, areas of divergence and things that we might expect to see in the Type II web that lack technical precedents today.

Communities are sources of resilience that enable us to survive and sources of joy which allow us to thrive. When we are in a community of people who can empathize with us and who we feel understood by, we are able to have the support we need in the face of adversity. As we seek and find media and experiences that validate or reflect our own experience, we gain a sense of belonging and connection.

Strong communities are made of people who develop mastery over the interface by which they communicate. This means learning how to most effectively communicate with each other, make decisions, settle disagreements, and adapt to the unexpected together with consideration for individual differences in communication from language, to tone, to speed, medium, etc. This only works if you consider each person in the community to be worthy of this type of attention. Strong communities care for each person inside of them equally and they also acknowledge that treating everyone the same is not an effective solution to meeting the needs of the collective.

In the context of a decentralized web, mastery over the interface means developing a protocol by which peering nodes communicate that accommodates both the expected and unexpected. This is already present in a lot of peer to peer systems that are designed for fault-tolerance. By using redundant communication, we ensure peers have ways to coordinate in the face of a node going down or a node sending duplicate messages whether accidentally or on purpose.

What is missing is the sense of equity and equality between peers. In web3 today we tend to treat peers the same but do not see them as equally valuable, resulting in explicit or implicit hierarchy. Our expectations for how each peer operates is static and any peers that do not meet the expectations become ignored or replaced.

However we can re-frame peering relationships and we have a set of tools to do so in a practical sense. This is the first place where we see a gap between the Type I web today, and what queer and trans BIPOC communities do to stay strong. The Type II web can use mechanisms to distinguish between peers in order to increase good behavior in a decentralized system and minimize disruptive behavior. This includes localized reputation systems to know who to trust, fraud proofs to detect bad actors, random sampling as a form of accountability, and designing systems where there is little to no reward for not acting honestly. With these existing techniques we can maintain a structure where each peer is of equal value to the system as long as they coordinate under the agreed upon protocol and with some honesty and integrity.

I think many American children grew up hearing the phrase “sharing is caring” but somehow it is so far removed from modern day American society that it is almost impossible to see in action. Instead of sharing, we expect people to earn so we can exchange. Instead of sharing, we expect people to beg so we can donate.

Instead of sharing, we expect people to earn so we can exchange. Instead of sharing, we expect people to beg so we can donate.

So what is sharing? I think of sharing as a practice to give people useful things. It comes in many forms: it can be an act of aid during a time of crisis, an act of gift giving, or an act of resiliency to prevent a single point of failure due to the centralization of a resource.

QTBIPOC+ communities share because of mutual understanding of their experiences, empathy, and compassion. In 1969, the Black Panther Party started the Free Breakfast Program which resulted in feeding 20,000 children that year. This was the same year as the Stonewall Riots in Greenwich Village where trans women of color fought against police brutality. Then Black trans women started sharing houses in New York City in the 1970s as the Ballroom scene developed, doing the work to help keep each other out of homelessness. Today these communities among many others who are systemically mistreated and marginalized continue to share home cooked meals, clothing, skills and helpful information.

The concept of sharing is vocalized a lot in web3 and the dweb, but it is still considered under the framework of exchange when we talk about sharing compute or storage, and it is still considered under the framework of charity when we talk about grants and public goods funding as ways to get communities what they need.

In the Type II web we can expect peers to share storage and compute freely, because decentralizing these resources allows for greater redundancy and resiliency. We can also expect to see peers sharing responsibilities as this is known to be an effective practice for marginalized communities.

In the Type I web nodes are hyper specialized, limiting potential peering relationships and the total set of actors that contribute to decentralization. In contrast, when a network has a fluid topology in which roles rotate between peers, skills can be more distributed and accountability can be shared. This prevents concentration of knowledge and or power, keeping the group structure flat and it also keeps peers engaged in meaningful rather than trivial participation. All of which results in a network that is more resilient to external forces or internal conflict.

Once a community is formed, governance practices are used to determine how the community should work together and collaborate. QTBIPOC+ communities determine their form of governance intentionally through agreed upon norms, expectations, and decision making frameworks that support their goals. Trust is then established as community members learn how to utilize their governance structure and make progress.

Cryptography happens to be super useful for developing a form of digital trust and of course digital identity. Both are essential for effective collaboration. So we can have digital applications in which community members can identify each other with a high degree of confidence, as we already do in both web2 and the Type I web, and from there members can collaborate online in a way that translates to actionable outcomes offline. What I’m describing here is basically what the chat applications Signal or Element are today and also how DIDs are being used in the Type I web today.

So given we have effective collaboration tools, what does governance actually look like? One effective form of governance is consensus. While many of us in the web3 space are familiar with consensus in terms of what it technically means and how the nodes we run eventually reach it, few of us actually practice consensus based decision making in our daily lives. So, what is consensus? Consensus is where everyone agrees.

Dean Spade is a queer and trans liberation fighter with two decades of experience in racial and economic justice work. In their book entitled Mutual Aid, they describe consensus as “a radical practice for building a new world not based on domination and coercion”. While the standards groups and DAOs that we participate in today have governance mechanisms to create proposals, iterate on them, and turn them into actions, these groups overwhelmingly operate based on majority rule.

Mutual aid groups that support people in the most vulnerable times of crisis understand consensus intimately because they practice it in order to keep their organizations alive. Consensus takes into account all views to sufficiently meet the needs of the collective. The practice of consensus is difficult and can be time consuming but it works when everyone in a group is committed to the same cause and value system. Without consensus, groups stop treating everyone equally, which results in churn and power imbalances that lead to dissolution.

The Type I web is at a point of acknowledging the power of global and local forms of consensus in decentralized systems. However in the case of governance I think we will find that majority rule is not enough to build apps that are truly in service of their users or DAO members.

In the Type II web the use of consensus in governance will lead to constituents having a higher level of education about the systems they are using, so that they can make informed decisions. When people know that their opinions actually matter and that they have the power to directly impact outcomes, there is a stronger incentive to engage deeply in the decision making process.

Self-determination is the ability to have control and agency over your own life. Though there are times when a person can not change an outcome, being able to identify and act on the power they have while maintaining self-authenticity, promotes their resiliency.

Having self-determination online means having the autonomy to opt in or out of a system or feature within a system at will. This is only possible if there are meaningful alternatives available for use. It means that there is no vendor lock-in forcing individuals to engage with a set of tools or technologies that are not valuable to them.

High switching costs is one of the biggest barriers towards digital autonomy and it leads to centralization that can be fatal for a community or network. While open source software makes autonomy possible, it isn’t enough for it to be possible for users to switch to a new service or application, it must also be easy and convenient to do so.

It isn’t enough for it to be possible for users to switch to a new service or application, it must also be easy and convenient to do so.

High switching costs can be prevented when we have interoperable models for decentralized data. In this case we can use things like exit and migration protocols. Yet today we are still lacking technical tools that make switching practical because it is very difficult to do something useful with large amounts of data within a reasonable amount of time and without a large amount of effort.

Teia, the Tezos NFT marketplace which was a fork of HEN is a good example of when this has worked. The community got together after the HEN project was abandoned without notice, and were able to set up a replacement service that pulled in the existing data and kept the functionality of HEN. ENS is a good example of when this failed to happen. A leader in the organization turned out to be toxic for much of the community, but a fork was not able to be created probably due to the difficulty of forking and adopting a suitable replacement system.

One technical solution we might expect to see emerge in the Type II web is self-replicating systems that provision the services needed for the system to make progress but act upon isolated data sets, resulting in independent sub networks, or even the same exact data set so you simply have a back up service.

The final important practice that I want to point out is activism. Activism is a form of self-defense from oppression. It is designed to stop and prevent harm. And it is among the most powerful things a community can do to protect itself. Independent of the outcome, simply engaging in activism serves to promote resilience for individuals and communities by contributing to feelings of self-determination.

A digital equivalent to the practice of activism would be protecting a system from bad actors. As already discussed there are many ways to do this in a decentralized peer to peer system from data redundancy, to cryptographically verifiable audit trails, to making it extremely difficult and unreasonable to perform a bad action in the system.

For me, one thing that stands out as one of the most effective ways to make a system resilient to an attack is the ability for the system to reform, stronger than before, after breaking down. This could mean reforming a group after removing a member, splitting a group because people disagree about what is acceptable conduct, or even changing a group’s purpose entirely to better satisfy the needs of the group.

As previously mentioned, this is where we see the power of forking. The open source movement puts us in a position where reforming a digital system or community is possible. If an attack happens in a system we can minimize or nullify its negative impacts by forking and adapting the system so that the attack is no longer possible. ETH Classic is an example of this. Forking is also already a political stance, and potentially an activist action within the context of the technology where it occurs.

Beyond forking we have the ability to make decentralized systems self-healing for example by restarting after a failure, we can put in place things like dynamic scaling based on traffic demands as a form of self-preservation in a system, and we can migrate actors in the system and their data to new protocols or digital services.

I’ve laid out the practices that marginalized communities use today in order to reach a state of thriving, explored how they relate to the technologies we use in the Type I web today, and I also presented a few open problems where technical precedents do not yet exist. The remaining question of this discussion is how do we fill the gaps and make the principles of our queer and trans BIPOC communities a reality in the next version of the web? While it can be difficult to imagine what this Type II web really will look like, it’s not actually impossible to imagine or to make real. It’s already happening. The reason that I got into this space in 2017 was because I saw Black people doing this work in the space at that time. I saw trans people doing this work in the space at that time.

We don’t need all the answers right now, but we do need to continue asking the right questions. I personally have more questions about where we’ve been and how that relates to where we are headed. At this point, I’ve learned that I need to change the way that I think about leadership in this space, and the lens by which I examine technical solutions to determine if they really are living the principles that we talk about or not. And if not, what needs to change and who is actually the best person to make that change possible?

[This work was first presented as a talk that I gave at ETHDenver 2023.]