LFGrow | Social Data & Unstoppable Thoughts

[Music] [Applause] [Music] all right with that we are ready for our next uh amazing discussion so the next topic we want to talk about is on uh social data and unstoppable thoughts uh it's it's a pretty uh broad and abstract sort of topic we'll kind of dig deeper into kind of how uh we think about it and uh for this conversation like to uh welcome jonathan uh from starling lab to uh to join me um welcome jonathan i'm glad that we're able to kind of talk about this and uh hi thanks karthik awesome to be here really excited awesome so i want to start off by just kind of getting uh more context for everybody else and kind of people understanding what you do so we'd love to kind of get an intro on who you are what starling is and uh and just everything that you uh spend your time on sure so the starling lab is based at stanford and usc and we are focusing on how to use web3 technologies to advance human rights in in three domains history law and journalism and we've been at this for about three years so um we've understood that as with everything in web3 you've got to be patient because stuff breaks and you've got to figure out not only what are the possibilities of the tech but um our our big focus has been how to actually integrate different solutions so rather than thinking about one protocol or one solution that can try to do everything instead we're realizing that there are different tools that are out there that can be integrated together to create really robust solutions um and so that has been exciting because you can imagine in the types of things we do across the different domains you need to be flexible right today journalists don't know how to use these tools right straight out of the gate so you need to listen closely as to how they get their work done or for that matter we're starting to work on uh work in the ukraine it's a fast-moving situation it's very dynamic and our we understand what the lawyers need but we also know that web3 tools are struggling to even work in a basic way so we need to simplify and create robust solutions and um and that's really what the lab is all about to just cut through the hype make real stuff happen amazing is that more kind of a comment on like how do you guys decide what to spend your time on is it a project-based thing or is it like a situation-based thing obviously you talked about ukraine but how do you sort of measure the outcomes as well as what to focus next well um we got into this because we realized that there was just natural and intuitive things for us to do so we started by focusing on taking vulnerable data and sensitive data and seeing if we can get that on chain and that started with historical records that was a natural thing but what we realized very quickly was that we could start our chain of custody actually upstream right rather than just taking records as they are when we create new records could we start let's say with video or images with a chain of custody that gets on chain in the camera itself right so part of what guided us was realizing that there were these interlocking things so we have storage we have capture and then finally verification that was another area that very naturally led us to think about how as experts are reviewing information and trying to build confidence and context and clarity about it we needed to provide tools there so that essentially was our our approach to creating a framework and all our metrics are around how we can deploy into case studies that are in the domains i spoke about by using this type of framework again and again amazing and kind of obviously there's a lot of room here to kind of go deeper into all these things but uh i guess specifically focus on just what you what you just kind of are spending your time on now um what are some recurring themes that you sort of get into as you talk to the end users of these technologies like what stands out when they think about what three leverage kind of uh solutions or any recurring themes around like this is what you get complaints about all the time with web3 technologies yes yes um wow it's a it's a big question well let me jump just by talking about the problems in general and we can kind of create the subset for web3 um you know today i understand you guys are focusing on hacking around social media and you know with lens and all sorts of really cool things that are being experimented with um in creating new types of social profiles and methods for sharing information etc and what i can say is that the number one thing that we're addressing is that the social media platforms of today have very little context around the information which is distributed within them so as an example if i send a photograph um on to you on twitter one of the very first things that twitter does is it actually strips out all of the metadata that might help me establish the veracity or the authenticity of that photograph so it becomes a massive problem because people are essentially with that original choice which has some wisdom to it but but um in my view it's outdated um we're now really just struggling to catch up from behind so a lot of the challenges that we're addressing and a lot of problems are that there's just bad hygiene people haven't done proper schema they haven't thought through the interlocking protocols and figure out like actually what's the optimal workflow essentially it's just bad planning so i i would say that the number one challenge that we're seeing in weapon three is that it really it can't actually address those problems um it takes humans to actually come up with better plans and better strategies and then use new web 3d technologies to help build confidence around good choices that you make at the beginning because if you're authenticating something it's it really it's garbage and garbage out right right so i think that's that's like the number one thing that we see is that it's just people have rushed stuff is kludgy they haven't thought through all the best practices so that's in the human rights context while we want to act and we want to be responsible at the same time like the stakes are so high and so we can't afford just to put stuff out there that's rushed that's kind of one one one big piece that we look at just clarity and structure the second which is really important is about performance and thankfully that's getting better with web3 so i know there's you know the story of the last 18 months is surely a story about efficiency in terms of cost and performance and so that keeps getting better and better and i think that we're now at a point where inner chain work is also going to get better and better and so that i'm really hoping becomes a major accelerant and a story of 2022 is that we can start to use different protocols for different things and that maybe i kick it off with an nft but then i immediately go into using three or four other protocols to achieve the type of consensus that i want to um and the type of distribution that i want to and and i think that likely will produce more complaints right because more systems more problems um but the end this is really important because if we're going to keep web 3 decentralized and we're not just going to go and recreate the bms of web 2 right and end up with the same problems that we that kind of got us here in the first place um that requires us to be dedicated to this idea of interoperability and so in spite of all the challenges um it's worth it and we're hoping people keep at it no that's that's an amazing answer um kind of coming back to sort of what you're spending your time on now um i mean this event to sacramento is all about thinking about new ways to create social media platforms that are kind of crypto native or web3 native and there's a lot of ways you can slice that there's a lot of ways you can think about what you want to pick on there's there's i mean there's a lot of text there's a realm of kind of possibilities around just other types of media consumption and all these things are sort of being tried on for the next two weeks but um as you can look at the situation right now that's out there in the world like what do you believe is the role of social media in gun documenting uh anyway war crimes or any conflict yeah well let's give you some context because this is very applied for us we about three weeks ago as the situation ukraine worsened and the war began we realized that we had to spring into action and our lab has been working for years on doing perfecting various forms of documentation for war crimes documentation human rights violation in syria now that conflict which got started 11 years ago at the time it was um if you remember karthik it was the arab spring was dubbed like the facebook revolution right kind of unofficially and you had all of these giant well at the time they were kind of growing but web 2 platforms that were offering a new opportunity for people to start to document things just with their mobile phone right it was this incredible revolution and they thought maybe naively well certainly naively at the time that would be enough if you could just document it online then you're good to go because the truth would be out there and dictators would come down and google has estimated that with that spirit that there are now more hours of documentation on youtube of the syrian conflict then there are actual hours of the war in real life wow over the last seven years that's kind of the torrent of information okay so now 11 years later here we are we have a tremendous challenge in trying to bring people to justice and people are documenting like they've never done before however the torrent of information that's coming in there and also the mechanisms for authenticating the information it's we we still have a really really difficult task ahead just showing the photo is not enough as we saw um zielinski was just featured in a deep fake video that was claiming that he had surrendered so he had a lot of challenges you know just around trying to build trust right at the same time just because you document something doesn't necessarily mean that it's going to get into uh be admissible for evidence so there's a whole set of protocols that you can take to ensure that something has the best chance of being admitted in the court case and if that's your goal for documenting war crimes as an example we need to take preemptive steps and then the last thing and this is really critical the part of the problem with um the rush in web 2.0 was that everyone just put stuff online and they had just no idea as to what to do with dealing with the safety of the people that were featured in the video and i think my biggest concern right now is that we are rushing and doing all forms of documentation which is critical but that record of all the things that is that people are putting up can be very easily taken and spun around from a tool that is just simply a tool of documentation for good to being a tool that now is used to surveil and to condemn people to find people in their homes and take them away and imprison them etc this is the reality that many people in ukraine may face as and how russia proceeds in capturing territory and potentially taking over the country and so while the initial intent is very important to preserve that there's a spark of good here the reality is that this can all go wrong very quickly okay and then the final thing which is that as you're trying to arbitrate whether or not to keep something up content moderation becomes really challenging and what we have found is that youtube and twitter and facebook were put into this really difficult situation in which they had to decide what information do you actually keep up right so if you want to protect people or maybe there's terms of service violations around like the brutality of what people are showing does that deserve to stay up um in spite of it being against common sense in terms of service right and but if you take it down then maybe that would be then destroyed and potentially unavailable for war crimes investigators so trying to find that balance of how you deal with securing the privacy of the individuals involved but at the same time getting information to the authorities that can act upon it making sure it's admissible all these things you could think of as problems we think of good as opportunities and we are really excited because we have been working well let me take that back there's nothing exciting about any of this we are passionate and we are dedicated to the idea that web3 can potentially help and so we're very carefully looking at how the tools that we've been working on could start to make a difference oh thank you so much for uh for that uh just kind of thought process around i mean there are a lot of challenges here and there's there's no perfect or right solution there are multiple ways you can think about solving each of these um and as you look at all the time that you spend understanding like what should be done here as a next course of action he's kind of if i'm understanding this right please correct me if i'm wrong it's kind of the clarification here that this is more of a nuance between moderation and the content existing and the fact that like the data persisting is still separate from what should be done with it and right now the completion is that in in the web 2 world if the data doesn't persist that it means there's no other way somebody would find out about it because suppression is real directly tied to the existence is that a fair statement or is that kind of where we are starting off it's all about authority right card it's like somebody has to make choices right and in the web 2 world the platform have said we're the ones going to make all the choices so we're going to set the metadata standards we're going to set the modes of content moderation we're going to figure out how to deal with identity and security right and that worked for a period of time it kind of bootstrapped us to the point in which you know everyone got on these platforms but now obviously we've realized that well i'm not so sure that i want um mark zuckerberg determining you know the course of war crimes documentation doesn't seem like they may have all of the interests of everyone in mind here and indeed i mean to be generous to some extent like the platforms have been active this time around and doing a lot more um and they've been consulting with the international human rights community but the the same challenge persists which is it's about agency right and so what i'm trying to emphasize here is that if web 3 you know people talk about like cryptography and crypto and all that right well what exactly are the features that we really want to use right so the first part of it is that we want to use the ability to encrypt and when you encrypt and if you have the agency to encrypt that also means you have the choice of who to allow access to this information and you can then choose who can decrypt right so that that first thing if it's putting the power of the users we already think that that's a better solution or if it's put into community where a group of users can come together and make that choice that's great separate but related is the other part of the cryptography coin which is authentication so that i know that as i distribute the information and i want to lock in all the facts around my initial documentation of something i can authenticate it using a hashing algorithm and signing algorithms and then what that allows you to do is ensure that no matter where this goes no matter where it stays stored where it's consumed et cetera that we have some providence and we know where it came from and we know that if it's been manipulated or not which is critical for chain of custody and web 3 is built on these primitives right that you have authenticity that you have some method of protecting uh users with anonymity or pseudo-anonymity with encryption um so as you guys are building and hacking away over the next couple of days like really think about it's like how how can you allow users to try to flex these really powerful tools on their own with their own agency but then also in groups that expand and expand and essentially keep that agency with the users rather than the platforms no this is a this is really good kind of way of thinking about it because what i was going to ask next was just go into a bit more detail around what are other lessons learned from how existing web 2 networks do this and how we actually kind of think about making a list of everything we need to address and as we're thinking about a brand new way to uh experience in a way a social network in in web3 it's kind of like a choose your own adventure uh sort of thing which is the data separate from the users which is separate from the platform which is happening from the user experience which is separate from like the front end and the platform you're going to look at it from and there's multiple things that you can tweak on every layer of that stack and the question becomes like what is important uh to focus on and what is ephemeral versus what is actually variable right um yeah that sounds daunting right because you're you've essentially disintermediated yeah exactly and that's a lot of pressure right if you're if you're a developer and you now have to essentially make all those choices um yeah i'm extremely empathetic in a sense that it's like this is um i think what the engineers are facing the next couple of days are you know they're obviously like a multitude of choices but also a multitude of responsibilities um so so let me give you the the one thing that where we when we consulted with the leaders in the human rights space they said to us it's actually really simple what you need to do is create platforms that allow for you to air grievances and to provide feedback so you have to basically assume that you're going to get things wrong at some point it's assured right every major type of implementation is going to stumble in some way because that's what happens when you work at scale right that's what happens when you make something public and so you know if you look back at like web 2 and you think about well did they really listen were they actually humble in their approach and trying to say you know what we're screwing up and let's make some some choices that might actually make a difference right or did they give agency to users to start to develop solutions around like better information right about what was wrong um i think at the end of the day um we can say web 2 failed in that respect the early experimentations around like voting around proposals and trying to get feedback i mean you may remember this right there was facebook advertised this right as as like a key thing they were going to do right i forget what the threshold was it was something like a hundred thousand votes or something like that could could yield some sort of change you know it should sound familiar right like dow governance and stuff like that right and of course they didn't stay committed to it um in part because you could argue maybe some part of it wasn't impractical but also they were arrogant i think they in retrospect they would admit that so that's what we need to flip on its head right is we need to think about not just governance for the sake of decision making but you need to make people informed through this process and so you need to be super committed to finding ways of getting feedback from users and then creating grievance processes that are structured ahead of time not blatantly once like really bad stuff happens um so i'd i'd leave that message with everyone it's like don't this is not an afterthought it's um this is this is actually core to the success of your platform is being able to hear these types of things absolutely um i mean i i kind of listed some features that i i think are just uh from a framing standpoint things that we need to solve for in this world and uh you kind of have direct experience thinking about and addressing some of these uh in production uh for lack of a better term um so so kind of from perspective what are some of these key requests that you get and or key features that all these new platforms uh should have and actually you kind of think about breaking down or solving or checking for like what is the right check that hey we have a sufficiently good enough solution if that check meets yeah well the first thing is thinking about identity right it's a super core and i'm imagining that everyone's going to spend a bunch of time focused on that and you know identity is complicated because it's obviously multifaceted and i think most people are realizing that personas are what is going to be one way out of the problem with identity which is that you can have essential identity that defines that you can essentially speak for and have agency but that you might want to present your persona in different ways under different circumstances you might want to have a lot of choice around that and that persona in some cases could be anonymous pseudo-anonymous you know you can you can make varying levels of security features around that so i think that's that's one of the the core requirements of a social network i think we're to re-architect this is you know it is to to find ways of establishing a multi-faceted identity uh that then leads to the second piece which is around content moderation which really flows from this concept which is that you want to find a way to generate community here so now we have a set of personas that can come together and they can help manage some complex tasks and um that a core feature really important and when we think about for instance content moderation you know it's one thing for instance to have content moderation happen for a broad distribution of content it's another thing where it would be content moderation for a bunch of war crimes prosecutors right and so you know clearly they have different needs and agendas and um in in future requests so i think that having the ability to shape community with a whole set of different requirements and around what the community requires is um organically that's super important okay and then the last thing which i'll mention is um is really about authentication and provenance um so much of the problem that we have right now is that information is flowing with very little information amid the data accompanying it so you have photos that have no time and date you have challenges around establishing location and even things like the veracity of information like it's like it's underlying hash right hasn't been manipulated or how can i authenticate it as close to the source as possible now this one holds tremendous promise right like we're already seeing for instance images from other wars being recycled in the ukraine conflict because it's just they're evocative right they get people to to think and and to care right but of course those are that's falsified information right that's a big problem did you also find the nuance in that third point about how much of that kind of is already coming in as a future for just being on the blockchain versus the additional information that you would have to amend or append cool yeah i'm glad you paused me there all right well let's think about it so the first thing is about ordering right so blockchains are really good about establishing like the order of events of something so if you can register something we we have at minimum we know what happens before and after right we have some relative sense of time and then we can align that with potentially like real time say this date this time so that that is kind of the first thing the second is that if you in that record can pack in things like location as an example or information about the device that took uh the image et cetera you can help start to build confidence that when someone says hey this actually happened that you're starting to say well i have now additional information there and that's giving me some more confidence that yeah indeed this this happened the way this person claimed that happened or has it's not a deep fake or it's not a a shallow fake or you know some sort of recycled image so um that's really important but i'm going to be a little bit of stinker on this um the immutability right of that type of record is challenging because today i might want to disclose that information and tomorrow um i may not right today i might be free and able to do this type of capture safely tomorrow there may be in a completely different security situation and i now need my anonymity right so choice is very important to ensure so as you're developing these features it's not about um opting out it's about opting in and that's like really important because of the the nature of the immutable nature you know of these types of um devices or systems and i'll tell you by the way karthik this is fascinating this is not abstract fear at all um as we were developing cameras and applications that did exactly this it would take a photo and you would know it's in this exact time and place right it hasn't been changed um our security researchers um let us know that actually 10 years ago north korea actually pioneered a lot of this work and has made that a standard feature of every phone that is shipped in north korea that's approved by the government so that form of authentication powerful amazingly helpful if i choose to do it but when it's forced upon you then you're it's a completely different ballgame and now we're in the in the case of surveillance so that's those are the table stakes right that i think the engineering example of a double-edged sword um wow so i mean i think a lot of it is i feel like otherwise i'll tap out the listing i don't think it's a solved thing i think there's still a lot of improvements to be done it's uh it's more about i think we get a lot of things built in in this world especially on the persistence and the available largely the availability as well as kind of the immutability of information and it feels like a lot of what we're thinking about or talking about here and especially discussing is that a way to think about moderation and different ways to improve uh moderation how do you think about breaking that down like it is just a world where everybody says bring your own algorithms and if you're doing this for war crime then you have a different script you're gonna run on the same data if you're doing this thing for like understanding what could go viral you have a different sort of algorithm are we going to see all this fragmentation and bifurcation here like how does that world kind of change and evolve from your perspective it's complicated because i don't think we've done content moderation successfully even with billions of dollars being spent on it and you know platforms that have maybe not a um a moral interest but a but a profit interest that they put forward to say we got to solve this because this is hurting our brand this is hurting our you know our reputation so um let's let's try to unpack it i think the biggest challenge right now is that there is no way to actually moderate content within community and so i think the the ability to create more robust groups that can make choices around the types of things that are acceptable and the values that they uh want to stand for is is an important step forward now mind you i can go you know that that requires you to make sure that people like actually are doing thinking about their values right and that ideally those values you know are stated they're clear and that governance is built around those so that i think is the very first set of types of features i think that we would want to see here about content moderation is to instead of thinking about it as a centralized function that a platform needs to handle instead we're starting to localize it to affected parties to stakeholders right um the second thing is that their content moderation you know is tricky because obviously we want to promote freedom of speech as much as possible that's a core tenet of democracy but the reality is that not all forms of speech are protected right and so um again like spending time preemptively to figure out the nuances of how you want your community to deal with this and then maybe forming webs of webs right where good practices from another community might be something that you can now have like a dao bao type of relationship here where then you can have those values disseminate and probably the key to that is really transparency so that you can understand exactly what's going on here and remember even that was something that we didn't really know what the standards were in web too right there were these kind of investigative reporters that would uncover like crumpled powerpoint presentations that explain like how content moderators take down content and and they were like insane to see because some of them were really wrong-headed right so um so i think making just a lot of that process more transparent and more governable is going to go a long way won't solve everything but it certainly will be much better than the current state of affairs absolutely so to kind of give you a little bit more context into uh this event um a lot of people so we have 500 developers who are working on creating different ways of reimagining social networks whether it's on the twitter or facebook side or media content or live streaming just different ways we can have more consumer social interactions and uh it's all leveraged on lens protocol and kind of the one key interesting piece here is that kind of what you're fundamentally sort of facilitating is a standard around how people have a shared identity that gives them the ability to just get a graph of all the users who have also interacted in any same realm and i think that's to me a super interesting sort of uh primitive because we kind of have almost everything open uh and transparent on watching you can inspect this you can inspect the metadata you can inspect kind of a to b transactions and destinations but there hasn't been a strong incentive or even um a lot of attempts right now to really figure out a way to think about how do i in a way roll up the activity or or the identity of a user on a different platform and sort of leverage that to make sense of it on another whether that's like a reputation sort of uh score or or aggregation if this is what you've done that comes in and this is kind of one of the bigger problems with social networks too like i can't quit twitter because everybody i followed follows me must do the same thing or or i have sufficiently enough of an audience that i can say i'm going to quit this thing i'm going here please follow me and i hope that majority of them do um do the same principles exist as you're going to think about in this kind of ukraine case or the war crime sort of side like does a network actually matter or is it largely about the outcome of uh kind of the situation and that sort of thing or do you think this actually enables more interesting cases or not wow that's like they're saying there's a lot here so i'll try to break it up well just i'm just excited to think about think about how i can answer this some in a clear way because i think there's a lot of possibility right um yeah the first thing i just want to talk about is is you described here um the possibility of creating a social network that has portability as like a key feature right so i have portability of my data my identity et cetera and um i just want to put out there that portability does not equal interoperability right so i can take my data and maybe i have it but like if i can't use it on the other platform then forget it there's like nothing there and so i think that's really important for people to distinguish and stay ahead of because you saw this like when the hand ringing kind of came out um with web 2 they said well you can just take out your data okay but what can i do with it right and and by the way what type of data is actually being taken out right in many cases the analysis of my data which was arguably the most important thing that the platform was doing that was still kept proprietary so i think portability and interoperability which are now by the way requirements under things like gdpr um they they are are really important to build into the design and um and and i think it we still need a lot of innovation there so that's that's a that's an important piece um the other thing which is complicated about this but but i think an opportunity is that you really need to figure out what what are you betting on and i think what we're all betting on is not that you have portability and interoperability and identity and um uh you know in the ability like robust forms of security but you're really like betting on emergent behavior emergent features and that means that like your design surface is not just around the primitives of how i'm going to do something at a tactical level but that you really want to think about like what are the emergent things that are going to happen when this actually gets deployed in that scale and so i'd urge people to think about like getting the lean prototype experimenting with some things but then really being it's that next stage of the innovation which is actually where you're going to find out what you've got right it's because what you're looking for is changes to the fundamental content and the behavior that is basically that emerges from the features like i'll give you a funny example but it's a good one which is that until netflix decided to go in and basically like stream things in the way that they did and make their catalog available and the way they did um the phenomenon of binging wasn't possible right and so what happened is that you now had a new pattern of viewing right that's the first level innovation but then the content changed right like the actual television product like yet variable length you had like new ways of thinking about episodes and series et cetera i think that's a really intuitive example to say yeah like it used to be that you had linear content which was distributed on broadcast television that occurred weekly in half an hour chunks and you know it's like wait what you know and 20 weeks and then you get to wait another six months for the next times and then here you see 14 episodes dropped in the same night and people going crazy exactly so the the innovation is not just about the fact that okay here's the content it's available to you in one day it's that like oh yeah this story is like way different than a sitcom right or some whatever kind of classical television format and so i think that's a good analogy for what's going to happen here like you really you want to say like okay we're making things interoperable we're making them portable we're making them secure and authentic but like okay now what and so um that's i think the broader conversation that you guys are probably having but with respect to war crimes documentation um i think that there is a really interesting phenomenon that it we're on the cusp of which is that right now the challenge is that this type of work is centralized through the investigative authorities that are doing this work for the most part and i think that there's still some wisdom in in having experts deal with content of this kind that makes sense to me but engagement is still really important and that what i think is going to happen is that communities are going to rise up and they're going to say we want to be a part of accountability here and so as open source intelligence gets out there and people are going to start processing that information i think the emergent behavior is going to be that we have now open forms of analysis as well as just open forms of documentation and that people are going to say two things one i want to use cryptographic protocols to actually retain the integrity of this information and to preserve it and so i think the archive of this war is not going to be centralized in one repository or like one aws instance right i think it's going to be distributed i think that we're going to find that there's i think it [Music] and that's what will make it resilient because if russia even decides to attack or destroy or turn things off okay great well now you know this information is out there and it's the wider and the more diffuse pattern of that storage the better off we are and the emergent behavior is that people are committed they're not forgetting they're engaged right as because preservation is a very high form of engagement and then the second thing is that they're going to look at the content in some cases where it's responsible to allow people to look at things and they're going to find evidence that maybe investigators can't find because they're too stretched um and so those um i think those different types of parallelization which is kind of what we haven't seen in moderation like if you like they're largely i mean there's a fair amount of good but largely i think a lot of not great cases where you see kind of reddit mobs hopping up and they end up trying to solve certain things or find clues from just little information and it does end up often leading to results but that's also one sort of side of moderation at scale and we just haven't been able to do that um largely yeah social exists and central networks yeah and the chaos is basically that it's like this with the analogy it's a mob right so it just forms on its own and it's and then it's instructed to essentially go and kind of through the crowd the thinking of the crowd right go and attack something um i think in this case like that's the power of these types of solutions that are on chain is we can start to delegate we can create more structure we can think about ways of of encouraging participation but not having to be so unruly um so i hope that that's a that's an emergent thing as well um in i can tell you having worked on the syria conflict which is now in its 11th year and that still most of the major war criminals have not been brought to justice in any meaningful way by the way none nothing on the international level and sparing work done in domestic courts in europe um the reality is this is going to take a while and that what we really need to be thinking about here is a solution that's not about these intervening weeks and months um it's about this is a decade plus right type of work on on on account of a growing challenge with no way to call it this is done this will mutate in different forms um i i do want to end on a note that sort of uh helps the attendees for the hackathon sort of uh get some more uh direction towards what they should be thinking about i think one thing you did point out which is a really good thing i should i would like to repeat is just the emergence behavior i think a lot of us when we think about oh let's have a decentralized social media network we're saying okay i need to like have some insurance from getting banned from twitter or youtube so i just need to clone that and have the same one-to-one mapping of what this would look like now just that the data is more distributed that doesn't that doesn't count as emergent i think the emergence comes from the fact that we are seeing use cases that either just are not possible or we can't think of yet or we haven't figured out how to put that right combination in place and obviously nfts did that they did that with music just now and that's kind of an emerging theme we're seeing that with just how people interact and sort of form graphs and that's a really interesting thing but the hard piece about this is that we're hoping that there's something new will come up but there's no way for us to kind of prescribe or or give a category because by definition it'll be impossible and uh and while i i will kind of put you on the spot but i'm not going to have you answer that question i'm curious like if you were to kind of give some guidance to the attendees here uh to think about what they should build or what different ways you would like to see social media networks evolve granted they only have 10 days to try something out and do a minimum viable product what are some things that you think are important um one thing and you know we can direct people towards looking at some really cool things on authentication so proponents is really a cool feature that we are looking for in social media networks so the ability essentially to track content when it's first uploaded and who gets to basically how it gets shared and then finally how it's consumed in the end um provenance is a really interesting area to think about um for innovation so uh and specifically with images and video which are are really important um the other thing i'll just come back to um i i will beat the drum again on um issues around grievances and feedback so really cool mechanisms for um content moderation that are driven through like awesome methods of feedback and um and grievance addressal is like a is seriously the the biggest challenge that social media has at the moment and i think small experiments here might be really really cool to check out like how if you have like a massive queue of complaints for example how do you how do you sort through that right how what signals could you bring in that could allow something to jump up to the top and that that people wouldn't game the system necessarily right so i think some of those kind of dynamics are really cool to think about um and finally um i think the idea of the the personas is really powerful and and going to be a tough one uh for people to look at and really meaty um so i'm hoping people if i'm a user on the platform how can i exist in different spaces and and shape reputation in my activity um in a meaningful way what does the web 3d version of that look like so you're saying that my address or my identity underlying is the same but it morphs based on the platform i'm on because there's a second sort of abstraction of who i am and how do you kind of manifest that on chin is it just like me saying this is a list of topics i'm interested in and that's sort of what i fill in for a different platform or or how would it kind of dedicate my persona i mean declaration is certainly really important because you obviously want to make sure that people can retain their privacy and disclose what they want to disclose so i think that's important um you know the other thing is just um is keeping certain things sequestered in a way right that's another piece of the puzzle right so if you really want to um potentially you may not necessarily want to associate one persona that i have to another persona right so thinking about some innovation about how i could potentially keep different parts of my identity separate that seems really important um i know you know privacy norms were really challenged in web 2 and zuckerberg kind of famously said that you know privacy is like a generational thing um i i think that time has not looked favorably on that quote at all i think there's right a lot that we've learned about why privacy really matters i think a new generation has maybe new ideas about expression that's really awesome but i also think that they recognize that um choice and agency that comes with um intended encryption and you know some form of um uh of security um means that you can actually have more robust expression if you have things like security and encryption because you can in those types of settings maybe be your more authentic self right yeah exists um you're right yeah so so that if we can play with those ideas and now say okay how does that relate back to an identity that practically i can use that that's really the kind of area of innovation there's a lot of things that are actionable from everything you've said today so i want to thank you for taking out the time and making this amazing chat happen and kind of working on something that's super important relevant right now so i really appreciate this and uh thank you so much so great to be here big fan of the community and looking forward to doing even more can't wait to have you back thank you so much jonathan cheers [Music] [Applause] [Music] you