Centralized vs decentralized information systems-NISO Plus
Centralized vs decentralized information systems-NISO Plus
https://asa1cadmoremedia.blob.core.windows.net/asset-e3a30d91-deea-4ded-ab27-79d1428fff6e/Centralized vs decentralized information systems-NISO Plus.mp4
MARIA STANTON: Welcome to our Niso Plus session, centralized versus decentralized information systems. I'm Maria Stanton, director of production with ATLA, and I'll be the moderator for our session. Persistence, transparency, privacy are goals for scholarly communications. Many have created robust centralized systems. But what are the challenges to our centralized systems today? What does the next wave of technology, the decentralized web, have to offer?
MARIA STANTON: In this session, our panel will discuss the evolution of Web3 and explore some projects using this today. I'm pleased to introduce our panel. Wendy Tanimura, she/her/hers, is the director of partnerships at the Internet Archive. Interestingly, since 2015, Wendy and the Internet Archive have led a movement called the decentralized web, or the de web, movement, bringing together the technologists building the new thing called the decentralized web.
MARIA STANTON: And she's going to share with us an overview of that world to ensure we share some common definitions in history. Chris Sholem, he/him, is the executive director at ORCID and a principal at Swiss house information. During Chris' 25 years in scholarly communications, he has helped start several scholarly infrastructure organizations we all rely on today, such as Crossref and ORCID. Relevant to today's topic, ORCID is committed to persistence, transparency, and privacy as a provider of persistent digital identifiers.
MARIA STANTON: ORCID does not currently rely on the technologies of the decentralized web. Chris will share the perspective of an organization effectively meeting these concerns today while considering how these technologies might help us in the future. Ambar Camerino Furnez, they/she/he, is an open source software developer and researcher. They have submitted their PhD at the Universitat Complutense de Madrid.
MARIA STANTON: Ambar is an activist and advocate of open source software and decentralization. They note that they face typical researcher's challenges, and are optimistic about how decentralized systems can help. Ambar is the founder and director of Decentralized Academy Limited, and they led the development of the blockchain based software, decentralized science, and courts open access. Ambar will share some of this research with us.
MARIA STANTON: I will now turn it over to Wendy.
WENDY HANAMURA: Thanks so much, Maria. I'm so glad to be here with you, because the decentralized web is something I care deeply about and have been working on for about six years. As you mentioned, I thought it would be good for us to lay down some common vocabulary, some common definitions. So think of this as decentralized web-- an introduction. Let's start with some basics.
WENDY HANAMURA: Centralized versus decentralized. Centralized systems, centralized networks, are all controlled by that central authority. That's what you're looking at in the diagram on the left. That authority can be at a company. It can be a government. It can be a person. But that entity has to maintain the code, the servers, and the data.
WENDY HANAMURA: Most of the internet we use today, let's face it, are run by centralized applications. And ultimately, we have to trust that those central authorities are doing what's in our best interest. If you look at the far right-hand side, we're looking at a distributed network. Now, in a distributed network, all of these nodes, are connected and they're working together, communicating to get done one central goal.
WENDY HANAMURA: Computation can be spread across all of these different nodes, but one entity ultimately can still control those nodes. Finally, in the center, we have decentralized networks. Just like the distributed networks, the data, the computation, is distributed across many nodes. But here's the core distinction-- no one node can tell all the others what to do. So that's centralized, decentralized, and distributed. Let's take a look at some examples.
WENDY HANAMURA: Facebook, your classic centralized network, controlled by one all-powerful company. Amazon Cloud, a typical distributed network where everything is distributed across nodes. But ultimately, Amazon has control. And a decentralized network, IPFS, or InterPlanetary File System. Now, this is a protocol that spreads things across many nodes, but the creator of this protocol, Protocol Labs, has no control over where you put things or the work that you do across the nodes.
WENDY HANAMURA: So if that's centralized versus decentralized, what do we mean by Web 3.0? Well, I think we have to understand a little bit of history first. Let's start back with Web 1.0. It comes online in about 1991 when Sir Tim Berners-Lee creates the World Wide Web. That's what we call Web 1.0. Back in those days, it was very democratic, very decentralized.
WENDY HANAMURA: Anyone could host their own website. Everyone could be a publisher. But typically, the site was only hyperlink text. Maybe a few applications or files. You could only read those static web pages. That's why people call Web 1.0 the read-only web. Then around 2000, you start to see the big platforms come online. There's Facebook.
WENDY HANAMURA: There's Twitter. We start to be able to interact with content. You can share it. You can comment on it. So it becomes a much more interactive web, and that's why we sometimes call web 2.0 the social web, or the read-write web. However, this is the web we use today, and we have seen that major companies are starting to control much of what we do see, the data that we have out there.
WENDY HANAMURA: Finally, you have web 3.0. That comes online maybe around 2010, around the time of the first Bitcoins. The central innovation of Web 3.0 is that all of these technologies should be verifiable. Some would say this is a trustless system, because instead of having to trust Facebook, I now trust that the data lives on a blockchain, perhaps, and the blockchain is verifying that data.
WENDY HANAMURA: So I have to say that what web 3.0 is, is very controversial. It's very much up for grabs. So consider this. This is by someone I trust a lot and really respect-- Jay Graber, who now runs the Twitter Blue Sky project. Her point, mainly-- and I want you to remember this, if you remember nothing else-- is that Web 3.0 includes blockchains, but there's much more to it than that.
WENDY HANAMURA: If you think of them all as self-certifying technologies, then what's in Web 3.0 includes blockchains, but it also includes non-blockchain based technologies like IPFS and Secure Scuttlebut, Hyper Core. What's common to all of these things, what makes them self certifying, is that you're using some kind of cryptographic ID-- that's how you log on-- and that the content is content address data, as opposed to location address data, which is what we use now.
WENDY HANAMURA: So if this hazy concept is Web 3.0-- also, I have to say some people will throw in the metaverse in there. And some people throw in AI in there. What is what I've been working on? The decentralized web. OK, think about the web is a stack. What if you could decentralize every level of the stack?
WENDY HANAMURA: So, storage. You could store things across many nodes. The transport layer could be decentralized. That's what IPFS does. You could have applications or decentralized apps, sometimes running on the Ethereum blockchain, for instance. You could have self-sovereign identity, so that you held on to your own ID information instead of logging on and they hold on to it.
WENDY HANAMURA: You would hold it in a wallet. This is being done quite a bit already. What if you could decentralize payments, so instead of paying through PayPal or a credit card, you could make automatic micropayments to creators? Some people even think that we could decentralize governance through DAOs, or decentralized autonomous organizations, where you use technology or you use some kind of group mechanism to make decisions for your organization.
WENDY HANAMURA: If you took all of these things together, plus some others, this is what we think of as a decentralized web. Later on, if we have time, I want to show you some examples. But there are magazines, like Compost Mag, that uses decentralized technologies to bring you the kinds of things you're so used to-- articles, magazines, books. So that's my introduction.
WENDY HANAMURA: Thanks for listening. I'll turn it back to Maria.
MARIA STANTON: Thank you, Wendy. That's very helpful. Now let's turn to Chris.
CHRIS SHILLUM: Thanks very much for that really helpful overview, Wendy. And as Maria mentioned, one of, I think, the aims of the decentralized web movement is to make it easier to create services, systems, even organizations, that inherently have privacy, persistence, and transparency. Those values are very near and dear to ORCID. They are the things that we've striven for the past 10 years since we started.
CHRIS SHILLUM: But we don't currently employ any of these technologies. So I thought I'd give you a quick overview of how we have hopefully tried to live up to some of those values. But before I do that, a very quick introduction to ORCID, in case you're not familiar with us. We are an independent non-profit organization open to participation by all. We've been around for about 10 years. We're sustained by fees from our member organizations.
CHRIS SHILLUM: And we have pretty good adoption after 10 years. We have about 7 and 1/2 million active researchers from all around the world, nearly every country, apart from three tiny island nations. And we have members that support us in 54 countries. And our mission is to enable transparent and trustworthy connections between researchers, their contributions, and their affiliations by providing a unique, persistent identifier for individuals to use as they engage in research, scholarship, and innovation activities.
CHRIS SHILLUM: So how do we do that? Well, we offer basically three services that combine and work together. We have the ORCID ID, which is a unique string of numbers, an identifier, that any research can obtain for themselves and for free and use as they choose. We have an ORCID record, which is essentially a collection of metadata, which is linked to the ORCID ID, that enables researchers to describe their contributions to scholarship, whether that's publications, service, employment, affiliations, et cetera.
CHRIS SHILLUM: And then we offer a set of application programming interfaces, APIs, and a set of community support services that allow everybody else in the scholarly community to make use of those identifiers and that data. And because we deal with personal data and because we want to have participation by as many researchers as possible around the world, establishing trust is absolutely being key to our success.
CHRIS SHILLUM: And in fact, one of the reasons for ORCID's founding was that there were several proprietary privately operated research identified systems around at the time, 10 years ago, when ORCID came along. And none of them could get traction and kind of broad adoption, because nobody wanted to give their personal data at some corporation and be worried about what might happen with it. So we operate according to a set of values, of openness, trust, and inclusivity, which are pretty well aligned with the things that Maria talked about-- transparency, privacy, and persistence.
CHRIS SHILLUM: But as I mentioned, we don't use any of the modern decentralized web technologies. We have a very traditional tech stack. We operate a centralized database. We have a user interface that researchers use to come to us, sign up for their IDs, and maintain their data. We have a set of APIs services and a public data file that the community uses to make use of that data. So how we approached building and winning that trust of the community.
CHRIS SHILLUM: Well, by fairly traditional means, we've relied on a set of legal constructs and a set of social constructs that I'll talk about briefly to give you some idea of how we went about that. Legally, it's a traditional structure. We're incorporated as a non-profit entity in the United States. But the original board and the team thought carefully about how we built that structure to make sure that we were able to act in as open, transparent, and trustworthy way is possible.
CHRIS SHILLUM: So for example, the US non-profit law means that we can't ever be sold to a private enterprise. Our bylaws, which our Constitution require that our governance board is operating, is constituted a majority of people from other non-profit organizations. So in setting up this kind of legal stack, if you like, that start with our establishment, our legal establishment, our bylaws, the various agreements that we have with the parties in the community that we work with, we've borne in mind how to tie ourselves legal principles as strongly as possible.
CHRIS SHILLUM: But I think maybe more important is the set of social constructs that we've put in place. And that's really us committing to behave in certain ways, and then building a track record of sticking to those principles and being able to demonstrate that we act as we say we will. So very early on, in our foundation, the first board created a set of founding principles that were discussed with the community published, and ORCID has stuck to those ever since.
CHRIS SHILLUM: Things like, we're fully committed to researchers having control of their own data. We're fully committed to all our software being published open source and our data being published as fair and open. To give some guarantee of persistence, that if anybody doesn't like what we're doing in the community, they can simply clone it and start an organization that behaves better we have a commitment to an equitable, sustainable business model.
CHRIS SHILLUM: And importantly, we hold ourselves accountable by working as openly and transparently as possible, and do a lot of work to engage in continuous community discussions and dialogue, and maintain consensus about choices we have to make. So I hope that most people in the audience who know us would agree that we've done a pretty good job of building trust and participation over the past decade, but it was hard work.
CHRIS SHILLUM: It took a lot of careful thought. It's really taken 10 years to get to the level of participation that we've got so far. And there's more to be done. So I guess the question is-- and hopefully we can get some interesting discussion going in the panel-- is if DWeb technologies have been around when we stared, or if we started to adopt them now, would it enable us to do more quickly?
CHRIS SHILLUM: Or does the kind of hard work that we did in terms of building that trust have to be done anyway? And with that, I'll hand back to Maria.
MARIA STANTON: Thank you, Chris. That was very helpful to understand more about ORCID and ORCID's history. Now we'd like to welcome Ambar to share some of his research with us and his projects that are more in the decentralized web space.
ÁMBAR TENORIO FORNÉS: Thank you, Maria, and thank you Wendy and Chris. So what I would try to do is to introduce some of the projects that are trying to use decentralized technologies, or blockchain, particularly, to support open access publishing. And I will start by introducing two of the projects I am part of-- Decentralized Science and Quartz Open Access. And then I will give an introduction for what other projects are doing to give you a wider picture of the space of blockchain and open access publishing.
ÁMBAR TENORIO FORNÉS: So the first project I want to introduce is Decentralized Science and essentially science is a project that aims to support open access journals to find the best peer reviewers, meaning the reviewers that are relevant, reliable and good, and also to help reviewers to get proper recognition and rewards for their, until now, free work. And we do this by introducing a marketplace of peer reviewers, but not any marketplace of peer reviewers-- a decentralized one.
ÁMBAR TENORIO FORNÉS: There, reviews can share their reviews. These reviews are shared using decentralized technologies, such as IP affairs or blockchain, so the reviewers and journals do not depend on a single organization to access and publish their content, but on a decentralized network that relies on many different infrastructure providers. For instance, different universities could be hosting this content.
ÁMBAR TENORIO FORNÉS: This gives the system resiliency, transparency, censorship resistance, among many other desirable features. Additionally, the journalists can find the best peer reviewers in the system and access the openly accessible data that is stored in the blockchain, such as the timeliness of the peer reviewers or the actual peer review reports. Finally, reviewers can obtain economic rewards for their reviews.
ÁMBAR TENORIO FORNÉS: But to illustrate how blockchain can enable these economic exchanges, let me introduce the next project, Quartz Open Access. Course Open Access is a project that aims to bring economic sustainability to independent open access journals and publishers. With Quartz, users can donate to their favorite journals. Thanks to blockchain technology and web monetization technology, and in the ledger, we can enable readers to send micro donations to the journals they read.
ÁMBAR TENORIO FORNÉS: Small amounts of money can be automatically sent from the user browser to the journal. And being a decentralized technology, Quartz Open Access is just one of the many potential service providers that can be in this network. This way, we aim to bootstrap a new economy for open access policy in a system where journals can, for instance, offer economic rewards to their viewers that later will send micro donations to their favorite journals.
ÁMBAR TENORIO FORNÉS: After introducing two of the projects I'm part of, Decentralized Science and Quartz, I want to briefly share some other uses of blockchain technology that are used for economic policy and open access. One of these uses is to verify authorship and offer proof of provenance of data Thanks to the blockchain, we can create transparent and immutable records where we can verify that something is authored by some person, or that the data and results that are shared comes from a specific origin.
ÁMBAR TENORIO FORNÉS: One of the projects that is proposing to use blockchain this way is Open Science Chain. Other of the potential uses of blockchain is to offer innovative funding and investing tools. As an example, Molecule offers investors the possibility to invest in the research of new docs. This enables the initial research while offering the same of future potential profits to investors.
ÁMBAR TENORIO FORNÉS: Finally, blockchain and other decentralized technologies have also been proposed to share data, to share research data. One of these projects that aim to implement this idea is Eureka. All of the projects that are introduced, and many of the blockchain for science projects, are based at an early stage of development and adoption. So we can have some discussion on the role of decentralized technologies and academic publishing, as well as their potentials and limitations.
ÁMBAR TENORIO FORNÉS: As Chris was saying, is this technology really needed, or is it enough? Or do we need some other legal, social, and technological tools to build this trust that we need? And finally, what are the challenges that blockchain technologies will face before they can provide widely adopted solutions? So thank you very much for your attention. I hope that these introductions and discussions can help us understand these potentials and challenges of blockchain, and that we can have an interesting discussion.
ÁMBAR TENORIO FORNÉS: Thank you very much.
MARIA STANTON: Thank you, Ambar. Actually, I think you've set us up very well for the next part of our session, and I really want to thank Wendy, Chris, and Ambar for centering our conversation and providing us with a shared understanding of what we mean by centralized and decentralized systems, with examples for both. And at this point, the panel has agreed that we would spend a little time discussing these issues, the issues of the decentralized web, and what do we mean, and how we should be thinking about it for scholarly communications.
MARIA STANTON: So I'd like to bring the panel back together and start the conversation.
WENDY HANAMURA: Yeah, I'm so impressed, Chris, by the trust, the transparency, the privacy that you've been able to achieve over the years. And I think ORCID is a model organization. But as you said, it took you-- trust is hard earned, and what we're talking about in Web 3.0 is a trustless web, where I don't have to trust the board of governors of ORCID. I trust the blockchain. That is what it does, right?
WENDY HANAMURA: It creates this trustless system in which the data being on chain is where you get your trust. What do you think, having heard what Ambar is doing with it today in your field? Do you see applications that you'd like to take ORCID in that direction, or other organizations?
CHRIS SHILLUM: I don't know. That's the answer to that question, because I think, as I'm sure we're all aware, there's so much hype at the moment about crypto and investments and all the rest of it. But it's really hard to kind of see through that and see what value the underlying technology can offer to people at ORCID that are trying to advance open research. Fundamentally there's this alignment, right?
CHRIS SHILLUM: There's alignment between the values we have, between the values that a lot of the open research movement has. And what the technology is creating, this set of technologies are trying to achieve. But it's also quite hard for non-technologists to understand. It's really quite, I think even for traditional technologies, it's creating a whole new set of terminology, a whole new way of thinking.
CHRIS SHILLUM: So it's quite unapproachable and quite impenetrable. My suspicion is, just like every wave of new technology over the past 30 years, once all the hype dissipates, there's some interesting and useful stuff there. But I really don't know what it is yet, and I can't really put my finger on it.
WENDY HANAMURA: Well, we can imagine that we take ORCID, and we're bringing it into the decentralized web. Well, the transparency of the data and the trust now lives because it will be on a blockchain, and you can see every single interaction, right? And it has a hash. And in fact, every time something has changed, there's a new hash. So this is what we mean by content address data. Right now, if you have a URL, you can have link rot, or you can have data drift, right?
WENDY HANAMURA: Because it's sending you to a location on a server. Here, it's sending you to an immutable hash. Privacy is achieved because I'm not handing ORCID my information anymore. There's these hashes, right? There's encrypted IDs that we log on using our own systems. Persistence, I have to say, in the decentralized web could or could not be better. But because it's kind of that locks concept, because it's not just sitting on ORCID's servers, it could be on Ambar's servers and Wendy's servers and ORCID's servers and Maria's servers.
WENDY HANAMURA: If one of those go down, the chances of persistence are greater. They're not guaranteed, however.
CHRIS SHILLUM: Yeah. I think it's really important to understand the underpinnings of the trust, though. Right? And again, my knowledge is not expert here, by any means. But it seems to me that the underpinnings of trust in the case of DWeb technologies, at least on the blockchain, are economic incentives. Because at the end of the day, we're talking about people and we're talking about machines.
CHRIS SHILLUM: And machines that are operated by people, right? And the incentives that are created are the economic incentives to continue to operate the blockchain. And I guess we have to understand whether that creates an alternative trust stack, if you like. But in the long run is more sustainable than the trust act that we rely on today, which is underpinned by a couple of centuries of legal thought.
CHRIS SHILLUM: Or I think more likely is that these things will work together, right? I mean, I think I can certainly see the idea of distributed cryptographically validated ledgers being a valuable technology, right? But I think already, there's this idea, if I understand it correctly, of permissioned blockchains, where essentially there's some external construct. There is a set of agreements between a set of parties about operating the blockchain.
CHRIS SHILLUM: I think it's more likely that the technology will become a useful component alongside some of those more traditional structures than completely supplanted. But as I say, I'm by no means an expert. And that's why I'd love to hear Ambar, your and Wendy's opinions on this.
ÁMBAR TENORIO FORNÉS: Yeah. So I I'm somehow between what Chris and what Wendy are saying, because I do believe there is some potential of decentralized technologies. But I'm sure decentralized technologies are not enough at all to build something as powerful as ORCID, for instance. You need to build all this trust, and it's not enough to trust the technology. You can have a fancy technological tool, but then you have to gain the trust of all the community.
ÁMBAR TENORIO FORNÉS: What we can do, however, with blockchain is to have a more resilient infrastructure. Imagine ORCID proposed a standard for how to have IDs, and we shared this standard with other ID providers. And then I can have my ID not only supported by ORCID, but also by my university and the federation of european universities I'm part of. So people can see that my contributions to the field are not only backed by ORCID, but by other organizations that have other ways of verifying that I am authoring what I say, also.
ÁMBAR TENORIO FORNÉS: And this way, we are not transferring the trust from organizations to blockchains, but we're enabling the community to trust a wider variety and establish good protocols so that if ORCID becomes evil or disappears for whatever reason, we still have the tools to keep operating with IDs. So I guess this is somehow where we are heading, towards a new set of standards, but we still trust institutions and organizations.
ÁMBAR TENORIO FORNÉS: But we have technologies that help us have more transparency, more resiliency, more interoperability, things to set standards. And I would like to see where we are heading, and hopefully it's in this direction.
WENDY HANAMURA: I agree with Ambar. We're very much in this hybrid mode. The decentralized web is still pretty nascent. But I want to correct. I think the public view is that this decentralized web is 99% cryptocurrencies and blockchains, and a little bit of something else. But when you look at the projects that Ambar is working on, he's using IPFS. No blockchain in that, right?
WENDY HANAMURA: That's a protocol. Hyper media protocol that does not run on blockchains. He's looking at data sharing things. But one of the most performant data sharing protocols is something called hyper core, used to be called DAD. Also not a blockchain, also run by a nonprofit. So the sense that blockchains are evil because they ruin the environment, which is a problem, but you're throwing out the baby with the bathwater because Ambar's built this beautiful decentralized thing that doesn't use blockchains necessarily at all for much of it.
WENDY HANAMURA: And if that's all you take away from this 30 minutes or hour, that will be very valuable.
CHRIS SHILLUM: Yeah. I guess one thing I'd encourage the technologists to understand to do is help the rest of us understand it. And not on the technological level, but on the social level. For example, I tried to Google a little bit, and I couldn't find the answer. What fundamentally is the trust underpinning of IPFS? What are the incentives? And this is just an example. I'm not saying I want the answer to that.
CHRIS SHILLUM: But it would be great if I could go to a web page and understand in non-technical language, you can trust the IPFS because the people who operate servers are incentivized to continue to run the servers for this reason that I can understand in plain language. And I think that's like all these technologies. It takes a while for these things to become understood, but technologists naturally resort to technological terms to explain things.
CHRIS SHILLUM: I know from having worked in authentication for a while. It's really hard to explain how authentication protocols work to lawyers without talking about the technology. It's really hard. But I think it's worth investing the time, energy, and effort to try and help the rest of us understand that.
WENDY HANAMURA: Well, Chris, maybe I could share this example with you. I'm now in the Brave browser. Anyone can download it. It's a typical browser just like Chrome or Firefox. The one difference about the Brave browser is that it has enabled IPFS natively in the browser. So I'm looking at this magazine that I mentioned, Compost Magazine. It's a new magazine.
WENDY HANAMURA: It's an experiment, but it's built on something called distributed press. A distributed press takes any web page or any website and it distributes it across the decentralized web. So I'm looking at issue one in the World Wide Web. But if I go to this tab, I'm now looking at it through the decentralized web. And you can see that I'm looking at it through an IPFS URL.
WENDY HANAMURA: Now actually, that's a pretty URL that's hiding a long hash. That's really a hash. And some of the differences are that when I'm looking at it through IPFS, all of those different pieces are being hosted by different people. I can also very easily in one button save this and host it myself. So let's say I'm on an airplane and I want to read this. Now I can just read my local copy.
WENDY HANAMURA: Or I'm offline in the Amazon, and I don't have internet. I can still read this. But you're right. What's to say that once I read it, I get rid of it, and then nobody's hosting it? Well, then they created an incentivized cryptocurrency storage system called File Coin. And there you do bring in cryptocurrency to incentivize-- well, they're called miners-- to persistently store this stuff.
WENDY HANAMURA: But in and of itself, IPFS has been running for years without File Coin. And people like to host it. IPFS hosts it. Compost Magazine has a server where they're hosting it. So there's no guarantee in persistence, but it can work. And you see that little mushroom there? Well, that little mushroom, if you have Coil enabled, which I do, as long as I'm on this site, it's going to feed little micro-payments back to Compost Magazine.
WENDY HANAMURA: I have, like, $5 a month in Coil. And depending on how long I sit on this site, it will feed tiny little micropayments. So they're trying at Compost to get around the advertising model. How do you pay for this stuff? Well, this is so experimental, but it's trying to solve a lot of different things. Persistence, payment, mutual value for creators.
WENDY HANAMURA: I think you can imagine the same model being used in scholarly communications.
ÁMBAR TENORIO FORNÉS: Yeah. Indeed, we are using similar technology to Coil to enable micropayments to journals using courts and access. So that's exactly what we're proposing. Now we need to build the trust of the community, though. So if you are part of a publisher or journal that wants to try us, yeah. We have to start building it little by little. And we don't need to build only the technology, like we were discussing with Chris.
ÁMBAR TENORIO FORNÉS:
CHRIS SHILLUM: I guess one of the things I'd be interested to have a quick discussion about is, with Web 3.0, how do we ensure the same thing doesn't happen with Web 2.0, Web 1.0 to Web 0? Web 1.0, as you said, Wendy, was fundamentally decentralized. And the incentive structure was such that as the kind of set of features that are available, in the way it got more sophisticated.
CHRIS SHILLUM: There was a very strong pull towards centralization, and a small number of very large actors controlling things. And I think that's due to a couple of fundamental dynamics, particularly around the network effect and the fact that more engagement and more participation accrues to things that are getting more engagement participation.
CHRIS SHILLUM: And also to the pace of change, where I think it's fundamentally true that things that are decentralized are quite hard to change, because how do you get everybody the agree to move to version two point something at the next time? So again, I don't know much about this, but I may be interested in any thoughts you might have on how we might stop this same kind of pull of centralization for this new set of technologies that is starting out with a very strong intent to be decentralized.
WENDY HANAMURA: I'm sensing from Maria that maybe it's time to bring in our audience. Should we--
MARIA STANTON: Actually, I was thinking this would be a perfect jumping off point for us to invite our audience into the conversation. What we've discussed so far has been wonderful, and a great addition to the presentations. But yes, let's invite the audience in. Thank you very much, Wendy, Ambar, and Chris. [MUSIC PLAYING]