Rebuilding DeSci: Trust, Funding & the Future of Open Science

Transcript

Opening & Introduction to DeSci

Well, uh, we're going to talk about DI. I don't know if someone of you have an idea what DI is, but first we'll present ourselves. Uh, well, I'm Rodrigo Nunes. You could call me Roto. And you know, my traditional background, it's quite interesting. I will share this story. It's interesting because my traditional background as a researcher I'm a corologist. I have master's degree in that and since five years I have been quite exploring how can we create proper use cases of technology using blockchain and I just find this concept of open science you know what is the evolution of open science integrating blockchain and token engineering and it was quite a long journey so far you know I just met

these guys five years ago and I have been contributing with their projects so basically this is the second iteration of the open science movement and we're going to share some ideas and exploration that we have been done since five years or so in that particular concept but first of all I want to know more about you know you present ourselves in a few sentences if you want to okay I'll just go first hi guys I'm Eva from the side and I before I dive into the DI host DI ecosystem I've been working with uh layer 2 they VC's and about three years ago uh I exploring the DA things and fully dab

Speaker Backgrounds (Open Science + Blockchain)

into the DA area and I found this uh community called RSAI and now we're officially transforming not just a DAI focused community to a native infrastructure for open science um which transforming the scientific intent to a funable asset. Hey. Um, yeah, my name is Joshua. You may know me from the Bonfires app, which is live at the event right now. Uh, but Bonfires is actually a protocol built by DI World and DI World was built in 2021 to at first attempt to bring more funding opportunities to novel scientific endeavors. Uh, there was a lot of excitement in the 2021 bull run.

Lots of money, lots of communities doing great things. Well, you know, NFT type things. And I thought why not use some of that energy and that money towards like science in particular uh and let's say less institutionally backed scientific endeavors. Um, as it turned out, we went into quite a long bare market and DI also basically never experienced a bull market. And so, DI has been a very interesting space to be in over the last 5 years where some of the most interesting ideas, some of the most important ideas in fact in in my opinion around how we can use blockchain technology in particular the Ethereum ecosystem to further the scientific movement similarly to the open science movement. Um, but with relatively little traction and liquidity, it's been quite a journey and I think we're going to

talk about it a bit today on this panel as to how DI uh the intentions of DI, but also how it's grown and how it's faltered and maybe perhaps where we can go with it in the next, let's say, 5 to 10 years. Uh, yeah, at DI World, we've been focusing a lot on community uh onboarding and getting like more researchers and more web 3 natives into the DI aspects. uh looking forward to talking about it today. >> Oo that's great. So uh it's quite interesting because you know that's my personal opinion but many uh decentralized protocols that we have seen usually incubate in labs you know of applied computer science you know or engineering itself you know so uh like arbitrary we have seen many well-known

Bonfires & Early Funding Experiments

protocols L2s and L1's that usually have a really good solid research group particularly but you know in token engineering and stuff uh that is interested in them for them But uh you know regarding to uh um basically blockchain stuff there there are many institutions worldwide that actually have active labs in apply cryptography that are quite exploring decentralization itself you know um how how can we create a proper incentives but you know one of the key aspects of the foundations of blockchain is actually DI you know the fir we have uh seen that many OGs in the space I consider myself part of the second wave of DAI in delivers or enthusiasts uh if

you want to see that way. But um by the end of the day, you know, we have seen many initiatives like Protocol apps, you know, like the bio ecosystem, like uh other guys that actually have been quite active, you know, the financial aspect of the centralization itself in storage, you know, IPFS for example, a a vertable column of many NFTs, you know, what is the state? I want to ask you what is the state of the eye right now? So proceeding if you want to >> sure um so yeah right now DI is in a relatively dormant state there was a large movement of let's say capital uh well the most amount of capital I suppose that DI has received um and

attention and interest around 2021 2022 uh there was a lot of good ideas a lot of intentions of basically creating a novel scientific infrastructure that institutions researchers and degens alike can all partic participate in and do verifiable science. Uh many people had great intentions. Some people raised money. Very few were able to execute on that because of various tailwind reasons. Market driven the technology wasn't really there. Um but ultimately the DI movement has been focused very much on uh let's say the technological or the technical as the technical build that you can use to improve the scientific uh processes. So for example uh how can you do decentralized storage, how can you do decentralized funding,

The Current State of DeSci

how can you do uh privacy zk preserving uh transactions um and generally like governance through dows and tokconomic aspects. These were let I would suggest um ambitious and the right direction but just ultimately unsuccessful because few researchers used it few institutions used any of these aspects. uh it's quite immature and so we're at a point now where DI has mostly shed all of the let's say tourists that are in the space more so than I think any other vertical in in in crypto right now perhaps maybe NFTs more so um and so there's a a small community with little adoption but with a very clear vision of of how science can be improved using blockchain um so

for example Rodrigo and Ava we met about four to five years ago uh we know many other people from that sort of period of time who have pivoted and may gone to AI, gone to data, gone to sort of verifications and sort of but at the core of it really want to have a a research application of of their of their technologies. Um so that's where we're at in my opinion and I have my own personal ideas of how we should progress but I'd like to hear your your points of view. Um I think from what I see like this definitely like goes through different waves and the latest wave is like end of 2024 and be of 2025 right there's so many tension like talking about like this is a meme. So what we are going through right now is kind of rebuilding

Why Early DeSci Failed to Gain Adoption

the trust for this side like how could we get more re researchers to trust the DI concept and we're not just a meme cultures we're not just about like the token uh like uh value and also like all those things. Um what this is truly are is to like uh providing the I'll say a new method for the researchers to accelerating their research process. So uh and also with our exploration with all those scientists we've been talking with the founding the reputation and citation are the most thing they are like care about. So that's why we want to explore the whole thing like in the design and I know right now there are

less noise in the design area. So it it's just a right time to building things. Yeah, it's a I believe that DA is like a particular niche that needs to you know have this kind of ethos that we're building on the long run. It doesn't matter what will happen you know because science has been here since millennia you know technology since the fire that's why we're quite aligned with bonfires in that and knowledge is basically a market you know science is one of the biggest market in the world so we need to have how can we have a proper approach to decentralize this kind of frameworks that we already have and how can we involve more individuals

for example you know I will share this use case you know we have been exploring in terms of conservation you know of coral reefs one of the most delicate ecosystems in the world. So essentially our exploration so far is that okay we need to build or to allocate capital in various ways you know uh financial capital intellectual capital very important we need to mitigate this kind of digital device situation because uh how can we build this essentially with decentralized networks it doesn't matter where you are if you have internet connection or if you have a proper device that could allow you to participate on the network and contribute in a system you could have

Rebuilding Trust in Science Using Blockchain

proper incentives or a part of the incentives you could mitigate all these aspects regarding to the sharing knowledge many knowledge is quite you need to pay to acquire that knowledge so I believe that this could solve that you know in at least in the mid and short term you know we have seen many universities that are quite adopting the technology and have really active groups here in the US in Australia you know in many parts of the world of of Europe Europe for example and what you know here comes my next question you know and this is more like an inside question for you of all the things that we have seen in the space how can we reimagine DI you know in at least in five years you know

>> yeah I mean can I take this yeah >> so I have a question uh to the audience hands up if you've heard of DI before okay wonderful uh hands up if you've ever like interacted with a DI project bought a DI token haven't done anything with a DI. Okay, one, two. Okay, two. Great. So, um the the interesting thing I think about the term DII, it has kind of been sullied in recent terms uh in recent years because of some of the like let's say extractive activities. But the thing about DI is it's it's specifically focusing on the decentralization of the scientific processes through blockchain technology. Now a lot of as Rodrigo mentioned a lot of institutions, universities, research organizations are already using some of these technologies in some way and will continue to adopt

ZK blockchain uh AI systems whatever um and even potentially like decentralized funding through more centralized chains example but like let's say DI on Ethereum like what differentiates that? Why would a researcher participate in this much more decentralized open network rather than through these other more centralized solutions which provide the same attestation frameworks? They provide the same data security. They provide the same like ability for you to interoperate around the world. But why would you do DI rather than like onchain science? Um so from from my perspective at DO world and from the work that we've done the biggest problem we see with academia and institutions and research program research in general is the

Research Incentives: Funding, Reputation & Citations

social aspect of how people coordinate and it's like we're trying a lot of DI has been trying to build technical solutions uh rather than build the social solutions or to engineer social like change uh within the research communities and I think that this is the biggest unlock which is okay as a researcher or as like let's say a expert group let's say you know um like a professional society perhaps back when research or science really truly was decentralized where people could participate they get the same knowledge they can make the same sort of like the same sort of reputations go through the scientific method in this open way uh that is sort of I think where DI needs to go back to and that inherently leads into openness of data that inherently

leads into permissionless access through for example uh the Ethereum blockchain And I think that um focusing on this this sort of particular aspect of right right now science and academia is incredibly institutionalized for many good reasons. There's many great things that we need to preserve around that relationship. But ultimately the scientists are entirely relying on the institutions for money, for knowledge, for access, for reputation. And that means that the scientific community is the of the institution, right? Like 100%. And then the community, let's say the regular 95% of people that consume the science, they consume the pharmaceuticals, they they consume the research, it informs their daily life, it informs the policies that they're

Decentralizing Knowledge Access

governed by. They are completely separate from the scientific community and there's basically no trust between the two. Like maybe a thousand years ago, you would have a a like a healer in your society or you'd have a super smart guy in your society and he would be getting free food, he'd be getting free accommodation, but he'd be doing everything right. people would trust his word because he's super smart. That sort of breakdown between the general community and the scientific community is one of the worst things that's happened with the institutionalization of science. And I think that if we develop new systems of reputation, decentralized reputation, decentralized funding, decentralized access, then we can break down that barrier between the scientific community and the let's call it non-scientific community and make the scientific community increasingly less

reliant on institutions. to have poor incentives. Like if there's some kind of feedback mechanism here rather than just this, then I think that's how science progresses and that's why I'm excited about DI because I think that's the one thing that can can potentially do that. >> Yeah. >> Go ahead. >> Yeah. So what I see is like the funding side because the funding is the very crucial part just before you start the research. So within our side like we are having this like funding uh release via the milestone based mechanism and which brings more transparency and also involve with like AI agent to accelerating the whole process. So like you know funding is the most important part from what we see. I totally agree

Citizen Science & Real-World Examples

with that you know funding is a crucial part and how can we uh create flow systems you know in funding that it could be you know we could apply you know or gify in a way with all these applied theory that we have in the space but you know we need to develop you know in science usually we have many stakeholders and all the stakeholders like Josh said you know usually want something you know or they want to support you but they want a certain kind of agreement with you you know in terms of okay you will share knowledge with us and the IP situation I believe that this in these days have many verticals you know one is the storage is quite well known u scene in the space for example

many universities actually have apply cryptography you know in their servers right now I just had the opportunity to be in Boston last year and I just have some insights about okay that's really cool you know and many journals usually have that kind of storage you know I did do not share hear it because it's quite intrinsic you know they just do it for security reasons but by the end of the day another aspect that it would be quite good to to explore more and how can we create a proper coordination you know because coordination is everything and that's why I'm quite aligned with Gitcoin and this kind of allocation mechanism that we have seen since a couple of of years but by the end of the day you know in this era of AI for example how can we integrate all this

the infra that we already have in the space but in a particular niche that is not explored yet because we lack about certain credibility I believe you know for the web three communities because all the products that are quite creating in the in the space are for users web three users how can we on board properly researchers and independent researchers that actually need to use the technology you know And I will share another experience. You know, I'm from Mexico.

What Makes DeSci Different from Academia

I'm from a tiny island in the in the Caribbean, Mexican Caribbean, Cosmo. It's well known, you know, uh we lack a lot of uh you know, we lack a lot of situations regarding infra and all that. So, we have been used blockchain. I just decided to use blockchain when I just entered to my masters like four years ago you know to allocate capital I you know and to have another incentive you know because I'm a microphotographer too you know and we need to create this kind of awareness uh using the blockchain technologies to boost citizen science in a way I believe that that will be a proper way to on board researchers you know with art that's why we're quite aligned with the all this the the tracks

that right now are here in Boulder. Rene has has been doing an awesome work in funding creatives because science is a creative process you know we are creators itself that's why we merge you know since the re renet you know so I believe that AI agents play a crucial part I want to know uh what are your insights how can we adapt or adopt the stack that we know about Lisa you know all these kind of blockchains have developed many many explorations with AI agents and you Josh that you have been working a lot since five years or four years you know in knowledge uh and right

Social Coordination vs Technical Solutions

now you're quite pivoting with bonfires in terms of that how can we create better coordination in communities and how can we bridge researchers and inspire other independent researchers to continue their their stuff you know their research lines in in isolated places you know So go ahead, Eva. You want to share or Josh? >> So your questions? >> Um again, I'm going to I'm going to play a little bit on the social aspect a bit. Um we have this phrase internally which is there's no difference between learning, work, and play. It's all research. It's this kind of idea that if for example the Boulder bonfire that we

have right now, that's going to be a persistent knowledge artifact that anyone can use forever, right? if well however long however long it's hosted um there's so many insights that are going to be in that that somebody else around the world who's designing I don't know a localism chapter or is perhaps thinking about AI and society they could query that bonfire and get really high quality knowledge directly from the source of people here talking now that's valuable to the research process and maybe there are researchers here but this bonfire is not a specifically a specific research institutional bonfire it's just a thing that a a bunch of people have made together that has a very powerful research application And that is really what the foundation of citizen science is. Is being able to allow people to participate in the scientific ledger in a way that science

today can trust it and use it like access it and also verify it. So there's a few things that we're missing you know in the DI space in particular around high quality uh reputation mechanisms because that's that's really a big thing. Um but ultimately we need more participation of very smart people in this space who might not consider themselves a researcher uh to basically generate knowledge that can be used in this decentralized way. There are a few maverick renegade researchers that I know personally who are very interested in this technology and who are using it not necessarily us but just the blockchain and and thinking about these open systems. Um and we need to grab those people, incentivize them and really look after them because they're the bridge. But also we need much more examples on our side as to how this

Institutional Dependence Problem

technology is already being used to generate research artifacts. It's already being used to inform the collective intelligence. Right? And so from our perspective with AI, we've built bonfires which allows for those that don't know uh it allows information to be aggregated in real time from various inestion points like Telegram, Discord, whatever, wherever your data is. Turns it into a knowledge graph and that becomes memory or shared memory for AI agents. And we will network these bonfires so you can access an a whole network of knowledge uh generated by people in real time. This the original title the original uh name for bonfires was research collectives because it was born out of DIY world and the intention was give a group of people a way to

capture their knowledge and present it in a way that researchers can use. And so this is the foundation of what I think we need to do to uh to empower the scientist in general is to kind of redefine what is a scientist and what is a researcher. of course without um you know denigrating the quality of science and the the research process which has to be robust but we can recreate robust systems uh we just need to kind of are you a researcher like have you gone and gotten your qualifications then you're not a researcher I think that that needs to change immediately like like I'm not a researcher but I've done a lot of research and I think the stuff that I know is very valuable but I'm certainly not qualified as as other people would hope right

AI + Milestone-Based Funding

>> yeah that's awesome so what is your perspective uh Eva in terms of how can we use this a AI agents uh essentially to verify uh you know and create like an attribution layer you know to to the funding aspect of things you know because uh as I know uh we're quite working into developing something you know regarding the reputation systems in that you know with that perspective so you you could elaborate more if you want to >> uh well I think uh with the agent era like the hypothesis can be verified or um I mean not only verified proposed like 10x or 100x than ever right so with the infrastructure we're creating it's more like a um like credit layer between

the capital is between the capital and scientist and blockchain technology and also with the like mechanism we just I just mentioned like the milestone based funding mechanism um while like AI agents runs very crucial part in it like its auditor like the uh each intent each like hypothesis and also once the like you know the progress be complete the it automatically like releasing the funds so the all those like progress will be accelerating with AI agent and I mean with a new era I think that's the the thing that we are just creating and to helping the whole process to be faster

than ever >> agree I want to add thing first, you know, uh AI, you know, usually people believe that AI could be only used for, for example, for knowledge or for funding, but AI actually could be used, for example, as a as a researcher in Cor ree, you know, in conservation, AI for climate, you know, we could verify systems and, you know, ARWAS are trying the thing, you know, in the blockchain space. So for example there are many devices that do not have right now for example in terms of monitoring environmental battles we could essentially create incentives you know in those devices basically to monetize and continue doing the science you know the battle science in those research groups. So I believe that AI could be

Storage, Coordination & Credibility

used at many levels you know for ecological modeling. We have uh some aspects of machine learning that could be quite implemented too you know to develop metrics and certifications for example in terms of bonding carbon you know uh by diversity credits in the future. So we have a long run you know in terms of all the research that could be done you know properly on a decentralized way you know in terms of energy for example um the DAC aspect of the web tree space you know Vitalik just say that is quite related on the long term foundational aspects of DI itself you know so um yeah go ahead Josh if you have some

>> yeah I just I think that AI levels the playing field for everyone everyone right in every way you need two developers not 20 to build a protocol. So you need one researcher not an institution to figure out something really interesting right um the auditing process of AI is going to be very powerful for that because research needs verification and audit trails and AI can audit pretty easily and very increasingly so cheaply. Um so that's going to allow a lot more participation from citizens who have verified contributions that academia itself can listen to uh which is which is very exciting. Um sorry you said something just at the end then which I wanted to touch on but forget >> real world assets these >> yes this is it. Thank you so much. Thank

AI Agents in Research Verification

you. Um so the the interesting thing is that the research process is going to be very much like it already is the research process is entirely in integrated with the technological proc uh technological growth process right so like AI research and everything like this this is all directly institutionalized it's directly ac like in the academic uh system and really we're looking at new paradigms with AI for everything and so we're looking at new paradigms with AI with participation in the research process and I should be able to go to a person who's doing machine learning analysis because he knows how to do it because he's talking to his claude all the time and pay him the money so that he can basically improve my model and I shouldn't have to go and get it published in nature for it

to be verified. I shouldn't have to go to a university and pay them a grant in order to get access to that knowledge. And I think that ultimately for better or worse AI blows open this whole thing. Um and if we apply it specifically to research then it's going to be very powerful. Although I am just saying now live on stage that I do think the DI branding is something that we need to think about as a space. >> Agree. So I would love to continue with this question. You know what blocks desire adoption tech or trust?

Knowledge Graphs & Bonfires System

I think I think I think people need what I would love is if the community just recognizes that it's all research and that like DI could be a very specific vertical with very specific integrations with it with with academic institutions or whatever. But it's like we're all participating in the research process all the time just by doing our work. Like if we are recording it with AI and it's going to open AI servers or it's going to Google servers, let's just capture that same data. Let's make it available to everybody. Let's incentivize it properly. That's that's the biggest thing probably is how to get people to incentivize them to participate in this process.

>> Yeah, I agree. What do you think Eva? >> Yeah, >> I think the trust like we have so many mature technology nowadays layer one layer twos the like ticket proofs we need to build a trust layer. >> Yeah, properly. So you know uh we just decide to create this track. Personally, I just decided to create this track because you know uh we have seen many allocation mechanisms you know in web three you know and I believe that boulder is a good way to inspire people to explore new use cases you know in blockchain itself particularly to these kind of niches you know so in terms of allocating you know the mechanism that we have in allocation quadratic voting

you know uh quadratic funding you know I personally have explored many of them I usually have been quite at many levels how we explore uh those um those queries and repos but by the end of the day what will be the most suitable if you don't have the answer don't doesn't matter because there are like 40 different you know uh allocation mechanism down there you know that we need to explore and we have only have done like five or something uh what will be the best you know allocation mechanism isms for DI you know >> you should go >> uh well like

Redefining “Who Is a Researcher?”

I just like I don't know like there's so many ways we need to try out and I think one thing is need to point over here the best way is always like um how could we scouting and how could we make the whole like I would say allocation like speed faster and how could um the process more easier for the uh for the scientist to get the funding. I think that's the best way. >> Yeah. How can we avoid this or create new frictionless applications or easy on boarding? You know, we have seen that with account obstruction uh with many mechanism that have been quite implemented, you know, in terms of AI, for example, uh for payments. You know, right now we're quite building the new

infra or the new standards, you know, in terms of AI with blockchain, you you know, with the Treium Foundation, but uh you should go ahead um Josh. Yeah, I mean um I totally agree that making the funding easier and more equitable and more available is a very important step and I'm very glad that people are focusing on that because there's a lot of great work being done on various funding mechanisms. Um from my perspective, I've been shouting from the rooftops for about two and a half years of something we call a knowledgebacked economy which is this concept of finding a way to incentivize knowledge as a as an entity or as an asset. um it's very difficult and abstract to take what is knowledge and turn it into some kind of asset. Um but now more than ever with AI systems becoming the standard for people

AI Auditing & Independent Researchers

to interact with these technologies uh the knowledge that they are they are informed by is entirely let's call it our IP and I think IP is a very uh antiquated system that needs to be thrown out in many ways but right now our knowledge is being extracted wholesale across the world everybody super smart people are using AI and that that data is being sort of wholesale extracted by these large these large companies and there's basically no ownership, there's basically no provenence and there's certainly no incentive. In fact, you have to pay to have your information used and to train the next models. Um it makes absolute sense and I think it's one of the most important things we can think about in the next couple years uh is to how to correctly um prove the or attribute or

attest the contributions of your knowledge to AI systems and to incentivize their usage with those with that provenence tracking because if not uh systems will be used against us. we're going to be completely disenfranchised and only those who already have money or who have access to frontier models uh will be able to access the future and the rest of us will just be farmed for our knowledge entirely. So we are triing this uh with some really interesting novel concepts with bonfires. I encourage you guys to contribute to the bonfire because we're going to trial the knowledge economy and people who participate get to participate in the in the economy of it.

Biggest Barrier: Trust vs Technology

Um but like if this interests you, please come talk to us. We have a great system for tracking knowledge contributions to AI brains uh through the graph and we want more people to get excited about this because I think if we don't do it we are completely screwed. Yeah, for sure. I believe that one of the key aspects of of DI uh these days, how can we get into the or how can we develop applications for scientists? You know, I believe that that could be one approach and the other one is how can we inspire people because the people that are quite interested in DI particularly are the ones that are developers, have a good open-source, you know, background. for example, as me and

the ones that are looking for opportunities, you know, and are quite like essentially, you know, just ended uh her BA their bachelors or they just start a masters and they just figured out that, you know, it has they have a a long way to run, you know, in terms of the competitive, you know, about how is science right now, you know, even if you decide to that's why many, you know, many DI contributors and many uh people that are building in this space just pipe it you know in a way because they do not like the the system that we have and we need to create a new hybrid system and you know it's quite

Funding Mechanisms & Allocation Experiments

interesting to see in top journals for example science uh frontiers uh you know uh nature you know DI decentralized science you know as a as a topic you know or as a branch frontiers in blockchain for example It has like seven different categories, you know, to publish papers, you know, and many of them are frameworks and many of them are write are written by people that are quite independent researchers and they just explore and decide to dig on that rabbit hole, you know, of what is the eye? But by the end of the day, I believe that you know I want I want that you share as simple as possible what are the key aspects that we need to improve

in DSI you know because we're in builder you know this is the land of bers you know of builders in web three you know since years so what do you think that are the key aspects that we need to scale you know in terms of the token engineering token engineering aspect or foundational aspect you know ethos attribution you know in terms of DI >> you want to >> uh key aspect >> um I think we are talking about a lot of things already like we're talking about like funding talking about like trust layer we're talking about like social side right so another thing I think uh need to be add on is how could we as

Reputation Systems & Attribution Layers

aitation the skill else right those researchers there are so many there are so many professional knowledge how could those um like scientific progress be quantified so those are the things I would like to see in the future >> yeah go ahead >> I think probably the thing that is most impactful in the next year is like a reputation system a much more robust reputation system that doesn't just track what your existing institutional reputation is but like what your ongoing academic contributions and research reputation can Um I have some ideas. I think there's people building on that. But ultimately because the trust I think is the biggest problem with the decentralized community in general, DI being a part of that and just generally any radically open or new

Final Thoughts & Workshop Invite

novel academic applications. >> Uh is the trust system. So reputation is very valuable for that. Uh but also we need uh novelty. I think novelty is the big thing. >> I believe that it's quite good. Some people just told me that we already ended this the panel but it's okay. I believe if you want to dig more about this we will have in terrible turtle you know a tiny workshop regarding how can we explore new ways to do some >> uh design at 100 p.m. today. So >> thanks.

>> Thanks me.