Decentralized Science

Transcript

good morning uh thank you first of all um so much for being here this early i think this is the earliest panel or talk that i've ever been on um but was quite excited to be given the opportunity um and so today we want to look at the decentralization of science uh which i think is a really big and important topic there's many things i think in the scientific industry in terms of how scientific research are done where web 3 and decentralization could potentially really accelerate the progress of science and i'm joined here today by my three wonderful panelists maybe it's a few words about myself my name is paul

cos i'm have the honor to moderate this panel this morning um my background actually isn't in science which is why maybe it's a good idea that i'm moderating the panel my background is in economics originally but i've been trying to figure out actually what the economics of science could be so um i work at a company called molecule uh and we also help launch visa recently which is one of the first bigger biotech laws that are really aimed at funding medical research um and with that i quickly want to introduce my panelist so this morning we have with us um bianca biancasama tomo okay okay who joins us here as a neuroscientist and bianca is heading up a company that

is focused on decentralized peer review um and then i'm joined by tyler tan i have worked together for for a few years now tyler's background is in in biomedicine and uh he's he's always been a web 3 skeptic slowly coming around more and more um and and yeah he's also been focused on this use case of how do we actually decentralize funding and research around biotech and then last but not least joined here by a doctor georg yuri stricker who comes with a background in bioinformatics and actually works at a fund called signature ventures which is a leading web 3 fund based in germany and

we thought it would be really interesting to also have someone with a scientific background but who can bring a little bit more of a funder perspective to how scientific research is funded and maybe with that yuri with mike in hand i'm gonna hand it over to you and maybe from your perspective um if you could describe a little bit what are some of the current issues that you see in in how science has worked over the past let's say 100 years and how it works today sure thanks so if you look at science today and let me just give you a couple of numbers just to give a framework to think about one of the leading companies in the publishing of science papers is elsevier and elsevier made last year

a net margin of almost 40 percent now if you think about it that's just a publishing company right most of the journals are paperback and some of them obviously are online but they're making 2.6 billion i believe in revenue and almost 1 billion on it is just the profit now google makes something like 25 facebook makes something like 30 right and they're making at almost 40 maybe 37 right usual publishing houses are somewhere below 10 that's the way it works so how did they do it well all the costs are basically zero because they're using scientists

to do all the work and they're just netting all the profits right so just that as a nutshell so scientists are reviewing the papers they're writing them and they're paying for the access and paying for the publishing itself as well just to give you a framework how it works today and the reason it was possible is because scientists don't care much about money so when science became really like a profession after the second world war uh people just went on like oh that's just convenient let's give it to the publishers they do all the cool work we need we can do the research and all of a sudden for iterations and to raise iterations it became centralized in the hands of a few who control the whole

publishing and because they control the whole publishing they control the way people publish who gets selected who gets the positions and you can publish again so it's a social circle that is happening in science and belief that's also to a certain degree maybe a lesson and how things can get out of hand just because your favorite convenience over decentralization hey everyone um one of the things that i think is really interesting when you look at decentralized science or terms like open science is that almost any scientist on the planet would agree that science should be a collaborative process it should be a process that involves working with people across borders we should be open about sharing data but the incentive structure that exists

currently in science whether it's in publishing in funding in looking how people move unilaterally through their career um really de-incentivizes or disincentivizes these sort of structures you have um the process of science itself isn't necessarily centralized but all the structures that basically dictate how science is done are fundamentally centralized so you have for example funding bodies like government funding bodies like nih or private funding bodies that control the majority of the way that science is actually funded people people write grants they wait a certain amount of time and in the for example united states something like 8 to 15 percent of

people apply for nih grants are funded this dictates the things that people can work on the pace of their research publishing is very similar as a someone working in let's say biopharma or the life sciences your ability to publish in a top journal like nature or cell or science really dictates your ability to move up and sort of climb the ladder to get tenure to maybe move from being a post-doc to to a professor and so all of these these sort of trends incentivize the way that that scientists fundamentally work so if i'm a if i'm a postdoc who's looking to get into a tenure track position

i'm really at the mercy of a couple of large publishing houses and a couple of large institutional funding sources to be able to really have an impact and achieve my my career goals the promise of decentralized science and one of the things that i think is quite interesting is is really facilitating shifts where researchers have more autonomy to fundamentally be able to do the work that they're interested in so for example what would the world look like if researchers were able to self-fund uh one of the things that we think about it and molecule and veda dao is sort of looking at the creator economy that was fostered by this this sort of nft craze over the past couple of years would it be possible to create the same sort of thing for for researchers and scientists for example

and could we move to a structure where you have people all over the world collaborating in an open way and sharing information freely the other problem that these sort of centralized structures create like for example nature or or cell or any of these large journals is the focus on impact factor that they have and the prestige that the journals create also ends up really biasing the the scientific process people tend to be quite careful about publishing negative data it almost never happens because there's not incentives to do so and it these sort of things actually stimulate pretty large institutional crises like the sort of reproducibility crisis that we see in the sciences

so yeah i'm personally quite excited about the ways that decentralized science can begin to cause a shift in how we work together um and and collaborate hi yeah i wanted to stress out that in fact we have to we must distinguish two things one is science and the other one is the research economy so science per se is already decentralized by definition because if we think of science let's imagine let's think just of the mod of the royal society of london a new leosine verb so basically it means don't trust anyone taking their words for granted it's very similar to the blockchain motto don't trust verify so the idea that is that for science there

is a scientific consensus that is peer-to-peer so this verification process should happen without relying on trusted third parties so no central authorities so but this doesn't happen because de facto research economy really relies on aggregators that we have mentioned before these big academic publishers who basically extract all the value from the research that is created and basically they capitalize on these misaligned incentives so scientists are kind of incentivized uh to publish more and more and just to increase their reputation

among the scientific community and they do it completely for free right so we mentioned before the impact factor um the impa factor is centralized by journals in the sense that the more important a journal is the higher the reputation of the scientists will publish there will be and and actually we we we see especially for the preview process that is kind of like this consensus algorithm like in the blockchain but for science specifically that we don't really need journals we don't really need the mediators of editors and academic publishers we just need basically the scientific outputs which could be paper software or data and

actually the scientific community itself already does the job of reviewing and approving the science so we could have the centralized platform like the one that i've been co-developing and tribute in which science scientists can gather together and put their preprints and then other scientists are incentivized to to review them in a sort of token economics system because actually all the economic value that usually goes directly to the scientific to the academic publishers could be redistributed back to the value creators so i just wanted to put this idea that we could talk about web3 science already and right now we are stuck in a web 2

science system i don't know if you agree um yeah thank you so much so i think at a high level there's there's three kind of bigger topics that we can that we can hear out or i think identify in the context of decentralized science uh one of them is funding how is scientific research funded today and then who defines what is being funded in the first place as you said tara and then i i think the second one is publishing publishing of data and scientific results and then lastly reputation um and so now i think we've we've kind of briefly gone into some of the issues and problems um what kind of how can decentralization actually address these issues bianca you just mentioned um so token economic systems developing new incentive systems

um would you guys like to share a little bit maybe at a high level what solutions you see but also feel free to give examples to the audience actually jump into how some of these things can be solved and maybe some of the first examples that you guys are seeing emerge yeah so i think at a very high level the primary thing that needs to be cracked in order for decentralized science to really function well and as you mentioned it's already beginning to our are the incentive structures so if we look at what motivates a career scientist they're not necessarily extrinsically motivated by by money but they are extrinsically motivated by the ability to do the research which in a lot of ways is dictated by their ability to get funding

and i would say the other thing that really motivates people in the sciences are factors like reputation and prestige so i mean impact factor i think has also been something that people have really put a lot of weight into because it does you know it really dictates how you're viewed in the eyes of your peers so i think it's very important as somebody who has a career in academia and becomes an expert in a particular field to really be respected by their peers and also to be able to to maintain a steady stream of funding to carry out the work so i think the question for anyone building a platform in decentralized science is to really think about what are the incentives that we can create to motivate scientists to

work in a fundamentally different way and maybe to speak from from an example you know what we're building with vita dao is basically a community that's focused on decentralized methods of funding research that is in the longevity space so really focused on human health and lifespan and the idea with veda dao is that basically we're distributing funding that comes from a community we held a gnosis auction a community basically formed around this dow to contribute funding to it people can also join by contributing work for example and then that funding is distributed to research labs all over the world that are working on anything related to age-related disease and in exchange for that that funding um the dow actually

governs the intellectual property that's generated from from these research projects but what what um it enables for the researchers one having a funding source that they can get in a relatively short period of time and two basically not having the same sort of strings attached that would be intrinsic to let's say a grant funding source or money that was coming from a vc or a biotech community instead the thing that they're actually engaging is a community in our case of about 2500 people that are fundamentally interested in the work that they're doing and for me this is a really interesting incentive in the sciences so you have now in this context a decentralized community that is engaging with the researcher getting

updates every week sharing ideas they're putting their data and their and they're sort of publishing out in the open into this community and constantly engaging in this discourse um around their work and and and also receiving feedback so i think you know there's also a governance incentive and a token incentive there but i think we really need to figure out how do we create let's say new reputation systems for example in the context of web web 3 that really work well for scientists and in this sort of peer-to-peer context how do we create funding sources where people in essence have the ability to to self-fund and generate um a sustainable stream of of funding for their laboratory in a way that allows them to explore whatever

work they're interested in without having to worry about publishing in the best journal or or yeah being able to move up unilaterally in their in their career can i you have some yeah i think that's that's a very crucial crucial point of the funding for because there is so much reliance on one of the on the publication side for example right that even if scientists know what the problem is and they are very well aware of what the problem is they're complaining about it all the time they cannot do anything they will just continue living in this system because this is the way to get the career and this is the way to get the funding and so one of the way to do that is just to break one of the points that we have because

if you break one of those the circle breaks entirely and so it's very important to attack those vectors and you can see already a glimpse of it in this web 2 world a little bit because before that science was not digitized there was no digital world and so in order to be published you had to publish because i went you cannot be seen and that's a problem for science when you go for big discoveries now with digitization you can easily publish which means you can easily see in the problems around you around there's maybe the noise right also it's about uh content curation at the end but you can see already that for example pre-print servers were done like the physicist i believe it's not a coincidence that they did the first of the preprint servers because in

physics it takes a long time to publish really it goes full circle for peer reviews and so on and so on so you can actually publish your results which might be like years away already when you publish which is insane and so pre-print servers are one of those things and then communication platforms the best everyone discovered during the corona crisis now that people go on twitter and just talk about scientific research especially when it comes to corona and so a communication platform that enables and fosters the discussion is also one of the important parts and i believe bianca is working on something like that yeah i think it's really important to stress out that there is already a decentralized community as tyler was saying and and in fact we saw the

importance of already these during the pandemics because the publications times are really really slow and people really needed fast results to be published so scientists started relying more and more on preprint servers which would provide scientific outputs that is not peer-reviewed yet and and so the the topic came to the attention of of the public like do we need science to be previewed at all i personally think yes but the problem right now is that peer review is centralized by by these journals and and if we could take all these decentralized output engagements that come from twitter from comments from

the community that already exists on the webtooth platform and and bring them to a web3 system in which actually this value is not dispersed but also correspond to an economic value then we could already solve part of the problem right because we could already we already have the science that is produced and we just need people to have not only intrinsic motivation to verify validate the science but also extrinsic motivation because we all know now it's a hot topic in science um open science right everybody talks about open science and now a lot the main academic publishers publishing houses started you know writing this

way of of open science but basically capturing again the value of these and what happens is that substantially researchers need to pay a lot more to have papers published and open access to everyone just to give you an example in major journals like nature uh from the nature group if you want to publish a paper that is open access so that everyone can access it without paying a fee it may cost like 10 000 or even more for one paper right so basically they are capitalizing again on this on this open science and on the fact that people really want science to be accessible to everybody right and

yeah so basically if we could bring this economic value back to the people that are already producing it i think this is the way to go um thinking about blockchain blockchain works well because there are trust machines and incentive machines and we kind of already have these in science already right because the trust comes with the verification process and incentives comes from the fact that you're gonna increase your reputation on the on the system in the platform and so we could we just need to adjust around this um maybe to kind of to round it off for for our audience as well

uh so we're kind of talking about the intersection here of like web 3 and and signs and kind of moving from like science 2 by now to science 3.0 um what do you guys think are the the first and easiest like um fields that the web 3 space and like maybe some developers in this room like if they're interested in like working on this issue what do you feel are some of the simplest wins that web 3 can can help accelerate a new movement in science and then maybe a following question with that do you think it needs the incumbents the old publishing houses to actually jump onto this wave or should it be an entirely new system if so what would that look like

yeah i mean from a from a super high level perspective i think there are some easy wins for decentralized science and it's sort of the things that have an intrinsic crossover with with software for example so things like bioinformatics for example could be very collaborative very decentralized things where you're putting a lot of data into a system where you have verification of software where you can use things like github for example to be able to actually um yeah or some sort of uh you know uh system to be able to coordinate a community that is not necessarily reliant on laboratories for example and not even necessarily relying on funding because they have the ability to work with data that is decentralized and open to your to your second question um wait can you repeat the second part

one more time is it whether you feel we need incumbents yeah so i mean i think i dream of a world where there are you know the i've been thinking a lot lately about biotech dials and and dowels that are really focused on funding research and facilitating research and initially it's a system that is fully reliant on on the incumbents academics work in the laboratory and we need to be able to engage with those laboratories but in the same way that the taoist and and other communities here are setting up sort of decentralized hubs in in lisbon for people who work in the crypto space to come together i think we can imagine a world where decentralized laboratories are being set up people are part of communities that are sort of self-sustaining and self-funding and ultimately researchers are able to work in the confines of this system

without engaging the sort of incumbents yeah i think self-sustaining is the is the key point like the problem of sustainability because we need open science but we also we nee we also need an economy around it right otherwise we will always rely on central authorities in order to have funding in order to have ways of redistributing the science and maybe we'll leave it to yeah i think i think software is a really interesting point because there is a second basic revolution happening to certain degree that scientists maybe we're aware i mean i'm biased here because i'm a bioinformatician but in the natural sciences usually you just go to your lab and you do your thing and then you come out to a bunch of

results and then you go give it to it and the idea like software guys and then analyze it now this is changing slowly because software becoming more and more important and that's why you have bioinformatics as you have uh all the chemistry chemical informatics and so on and so on so in natural science the software becomes bigger and bigger and then software naturally as you see on the internet as you can see on github on stack overflow collaboration is way more natural somehow people just come together and just build things and so there's also a mindset change that needs to happen in science as well generally that they don't treat software as like the iit people anymore but these are the guys who are actually providing new methods and new ways to discover things and that's you can see clearly mostly in

the biotech sector as well thank you so much um depends if we still have time for question i'm going to ask the organizer but otherwise yeah there's a question here in the back uh just about to say corruption in the research industry yeah absolutely i mean so i would say

that as you mentioned is already a problem currently right so uh there's always going to be some number of bad actors in the system there's always going to be some level of intrinsic corruption but i think the way that we can avoid that i mean in the real world right now there's things like retraction watch there's watchdog groups that talk about you know you have like declaring conflicts of interest on on a paper in web 3 i think we need really advanced reputation systems that allow peers to constantly be inputting data around the quality of a scientist and hopefully have some checks and balances that that create a system that is fair and equitable and also able to be judged by by peers i think it it's a problem with um you know we can't necessarily have an anonymity in these systems i think identities need to be

known which is maybe different from a lot of the web 3 space but that could be one way to approach it yeah just around that um so i think the main problem right now is that reputation is attached to this h index on the google scholar right and and this is some this is a number that identifies basically your value among the scientific scientific community but this is a value that doesn't go down so you could be yeah yeah it's it's really so basically it's just based on citation numbers so even if your science becomes crap you will still be a star a vip in the system right so we need to change this and make this reputation dynamic such that you can lose reputation on the

platform if you produce bad science if you if you misbehave so this is something that maybe it can be incorporated with some tokenomics system and it's really like a lot game theory right and we don't have we really need this if we want to implement web3 uh in science and we still don't have much of this vision in the scientific scenario right now um and yeah that was the main thing and another thing is to incorporate for example cryptography this specifically for peer review because there is this open debate like should we have everything like open access but then everybody knows your identity and there's a lot of bias in the scientific community as well if you

come from a prestigious institution or not or if you're a minority etc so if you have a system where you can keep accountability like in a version control system um like using ipfs for example but at the same time your identity is protected so what matters is just the content that you are producing then maybe we could you know adjust for having proper incentives and at the same time like uh protect from misbehavior but also having you know like this control that is decentralized and everybody knows what you're doing because everything is accessible to everyone do we have time for one more question or

yeah all right i can answer the questions for myself one minute left one of the like underlying fundamentals necessary here is what the community and where prayer is already doing let me open source right you have to really be strict about open source publish the code which means in science publish the experiment it might be the process how you get there it might be the code right but publish it because if it's published everybody can verify and you can replicate it and you can address it if it doesn't replicate and for that of course needs the platform where you can address that too right so we're coming back to the way how you foster discussions but those two things necessary and you have to be strict about that so if somebody doesn't

do that it's just not valid enough right thank you so much