Tarun Chitra vs. DeSci Debate with Benji Leibowitz

Transcript

are you guys funding other teams are you secretly cooking a better design you're pure reviewing the D ecosystem and you're not being very nice to us on Twitter but it doesn't seem to disincentivize you that's the part that feels very dupli this that's kind of why I don't like the the current system the aim was to lower the barrier of entry for peer reviewed signs why is that so problematic from your view welcome back to the rollup your home for technical talks and open conversations that shape the modular future nothing on today's show is to be construed as Financial advice you can see our full disclosures list at the roll up.co Ventures sit back relax and enjoy today's episode after a quick word from some of our awesome Partners imagine if the world's GDP

moved on chain guys say GM to safet the newest protocol by safe isn't a blockchain and it's not a new layer to it's a transaction processing Network built to send money anywhere and swap anything on thousands of layer twos from a single Universal balance in sub 500 milliseconds it's fast it's smooth and it's simple and it's built with safe multi Sig level security in mind safet unlocks smart accounts across multiple chains allowing users to supercharge security scalability and speed so get started today at net. safe. Global and remember to follow at safe on X welcome back to the rollup for today's episode

we have a dside debate with tun cheet CEO of Gauntlet and partner at robot Ventures a big economics and math guy with a bit of science background and H Benji uh who is running pump. science and big into d d studying over the last couple of years guys great to have you on thanks for being here excited to for having us yeah same here yeah welcome to the show guys Benji welcome on tun welcome back and uh pretty excited about today because it's kind of going to be a debate we have these once in a while they're always spicy and fun and uh I feel like when people disagree that's when uh everyone

ends up learning the most So today we're going to be talking about dii decentralized science um if you're familiar with all things on crypto Twitter and unless you've been living under a rock then you know tun's got a bit of a qualm with Dai and so tun maybe we could uh pass it to you first for kind of opening remarks if you will give us your interpretation of D and tell us your perspective on it yeah for sure thanks um so I I guess like it's probably best to kind of start with a definition and in my mind the goal of dii as I see it right now generally is as a funding mechanism for

ensuring that um research that is not so easily funded uh in the current Institution environment research that might be um not funded because it's not in Vogue research that might be uh not funded because um the institutions that fund it are like only giving kind of Grants and uh money to establish researchers and it's a a new field um as well as sort of uh research that might just be in things that are not in the Mandate of sort of classical funding agencies nonprofits or even corporate uh research funding um now there's a lot of things that can go beyond just the funding

right there's things that are measuring accuracy kind of having ways of uh measuring you know using blockchains to kind of store verifiable information about experiments so that people kind of know what the outcomes of experiments are there's a lot of stuff that goes down the the pure more crypto route but I'd say the majority of what exists today is really f fused on fundraising and so I think there's a natural tendency to say well that's a really great Noble goal right like GoFundMe was uh you know helps a lot of people out um however I think the main problem I sort of see with it is there

is a lot of stuff that's being promised um generally in these funding me uh kind of setups that don't exactly seem realis istic or likely to return to the buyers of these tokens what they think they're getting um and I think people have tried to to kind of broadly argue that things in dii are things like there's some IP that is generated from research that is funded uh and that IP is sort of owned by the Dow uh that uh holds this coin so this is sort of the bipr protocol molecule uh side of things um there's

other stuff uh which are just pure meme coins um which I guess Benji will talk about um I uh I I think like that's fine if you don't promise anything I think the reason meem coins are in my mind fine and great whatever is They Don't Really promise some utility to the end user right they just promise you a funny ticker and a bunch of memes in Discord um however these research projects promise you a lot and there's no real way of measuring whether they are accountable for doing the things they say or what they promise um and science is hard science oftentimes is very nonlinear you might

start a research project and find quickly that your hypothesis was very wrong and you pivot to something else and then you kind of try to keep moving in that direction maybe that fa fa fails so then you pivot again and I generally think that type of system works well when you are able to be accountable right you know the next time you go for a grant you have to like say why something worked or didn't work and in general I see the current setup is having basically zero accountability whatsoever it's pure Dow social accountability uh and the second thing is that there's no sort of enforcement

of penalties for doing things that aren't what you're promising to the end user so with that I think I will uh rest my opening remarks yeah Benji would love to hear your interpretation of Dai um as well as any rebuttal and just broad thoughts 100% so I think um tun kind of nailed it honestly on like what D is for the most part what the potential applications of it are it's just basically adding a you know crypto technology and blockchain Technology on top of Science and figuring out how can we make it more efficient like the whole scientific uh apparatus more efficient um and yeah I think like what what what

what tun started off with was basically saying that most of the dii projects are around funding and I think that that in and of itself is untrue I think what's interesting is that it's the it's the D projects that are focused on funding that get the most attention there are a ton of project objects that are focused on cryptographic storage of data from science verifying that uh different instruments are um basically streaming data in an untampered way um there are basically like compute to data like lots of very interesting applications of uh cryptography and blockchain to science but there doesn't seem to be quite the market for it yet and it's the funding

layer that just gets the most attention so I I kind of disagree with him there already that that's most of what D's focused on I would say actually it's a minority but in terms of attention it it looks like it's the majority of dii so I think that's Point number one point number two around like the IP enforceability and accountability I think that's a super important Point as well it's like you know if you're promising something with funding you better freaking deliver on it and what is the accountability mechanism to delivery and what's the penalty for not delivering that so I think you know each each project can speak for themselves right but on pump science we're live

streaming the experimental data as it's being generated and everything is timestamped in real time so for us that's the best form of account it's actually not the best form of accountability the best form of accountability is streaming it directly to the chain right but we're doing the the next best thing that was the easiest to develop which is just simple live streaming right but us that's the most transparent way to show that you know like where's the science like here's the science right here's what you funded here's what it's going toward um and you can you know there's there's not quite like there's accountability and then there's like cryptographic proof we're we're providing at least the accountability layer where the data is getting streamed I don't know of a single other d project that's doing that

um I also don't know of any anyone in science who's like live streaming experimental data it's like pretty rare to find and TOS I mean like I agree with Trin there completely it's the most important thing um you're a lar if you're not you know showing what you promised uh as a funding any sort of funding mechanism so I think that that um having that kind of scrutiny on anyone who's accepting funding is is absolutely valid um and then on on the IP layer um I think frankly the IP system and how that interacts with dii is set to evolve I think we have our idea of how the IP system can be disrupted um it's our goal to kind of change how IP gets created um and

represented on chain and transacted but um I think tun's points are pretty well taken which is like it's just not clear yet um how this will manifest so you know I don't want to like there's there's nothing to respond to there other than there's design space and want to innovate and and experiment and um yeah that's what we're trying to do um but we absolutely do not have it figured out yet by any means got it okay um appreciate that and so uh the the two core points here is effectively the the verifiability of understanding what happens with with the funds and this whole funding process and and how that money is being used but then also tun's getting at like the

outcome of what happens here for those who are being promoted idea of decentralized Science and don't quite understand it because they don't understand what they really investing into or what the outcome is is going to be T maybe you can elaborate a bit on on that specific point why is that so problematic from your view yeah so maybe I'll I'll give a little bit of background of why I feel so uh compelled by this kind of issue and it comes more from like kind of practical experience with um research

that was like completely funded outside the current institutions and how hard it was to kind of like get anyone to take you seriously um so I spent six years uh from 2011 to 2017 working in a place called D Shaw research uh the story of deaw research is basically uh David Shaw the founder in the 19 1980s uh he was working on what are sort of called non-von noyman computers so non Von Norman computers are sort of computers where uh the architecture of how the CPU processes uh an instruction cache and a data cache are different than kind of what you see in an Intel trip and in the

1980s you know he he was assistant professor at Columbia and he didn't get tenure because at that time intel was was like you know having's law start and so basically everyone in Academia is like we're not going to fund these like esoteric architectures um you know so he basically had to to to leave doing research and he went to uh into finance and then you know 20 years later he's a billionaire but he basically was like still kind of pissed off at the academic uh establishment uh for kind of like not

investing in this type of uh computer now you might say hey like was would it have been a waste for society to have done that um arguably right now all of the AI chip kind of marketplaces and custom accelerators and stuff are basically nonv machines so I would say he was just probably too early but what he did realize was that there were a bunch of problem s in science um namely in molecular Dynamics and protein folding uh that if you did build custom architecture chips you could actually you know improve the speed of computational biology and like also make drug Discovery potentially do sort of computational drug discovery which was

sort of I think right now is very hot but has you know like AI goes through like Winters and Summers um over time so long story short once he was a billionaire so he kind of got story is kind of didn't get tenure kind of got screwed by the institutions became rich and then funded his own research and then we built this as6 this custom Hardware it's actually ironically indirectly how I got into crypto was because I was working in ASX in 2011 2012 and the only people doing that then Were Us and Bitcoin miners whereas now it's like everyone doing Hardware stuff but um the uh idea is we built this thing we were able to like fold these proteins really quickly and publish a bunch of

Nature and Science papers and so even then even with this guy who is this billionaire or whatever even then it was still kind of like hard to get you know kind of acknowledgement by The Establishment and in the long run I think he spent he ended up spending over a billion dollars I mean they're still around so he's they're still doing a lot of research um but uh and and also a lot of their alumni are sort of like you know one of their alumni won the Nobel Prize this year stuff like that so it's like people who have been in places outside of the institutions and have been successful in science and there are a couple points about this that I want to

make out which is one science is a a field in which you need to be able to cohesively uh convince people that if they redo an experiment that a certain result will happen a certain thing is an invariant of nature is always sort of true um and the second thing is that there's a sense in which it's not just the funding that is important it's like the the quality of the research it's kind of making sure that there's some measure of This research being used by someone else or This research influencing other other kind of parts of the world uh in a in a kind of positive way and I think one of the things and

and the current system there's obviously a million things that are broken right like peerreview kind of unincentivized lots of problems with it um funding right like pis who barely do any of the work and like make and have an army of grad students who do everything you know obviously but they they get all the grants from the NSF NH that's obviously not great either right like there's there's a lot of stuff that's wrong but I think the problem is just by saying hey look anyone can fund it and or uh kind of participate without acknowledging that it takes a bunch of expertise you also need to be able to have some measurement of whether your outcome is real or not and you also have

to have some some sort of accountability for for fraud or things being misrepresented and if you don't have those systems then you're just basically not really doing any any kind of scientific you're not creating any scientific progress that will persist beyond your lifetime or even Beyond next week and I think that's the Crux of the reason I feel very motivated by this because I've worked in privately funded science right because like fundamentally dcii is basically privately funded science it's F science is not funded by you know a government or a nation state it really is sort of a group of people agreeing with fun science and a lot of

science is funded in in mechanisms that look like dii via certain types of nonprofits um obviously they don't have as easy of a job at doing capital accumulation or being able to to kind of raise money as efficiently um and if you look at them over time they have all developed these kind of systems and the problem I see right now is like everyone who is Shilling dii stuff does not at all seem to care about that whatsoever uh and and it at that point it feels a little bit like duplicitous and a scam like it if it was just a meme coin that's fine but like the fact you're promising the stuff for things that are hard to convince the

world that these you know results are real and should persist forever it it that's the part that feels very duplic to me and then that's kind of why I don't like the the current system within D I think that makes a lot of sense I think all your points are landing um turn I guess like the the analogy I'm kind of drawing here is like we're happy to shill the therapeutic but we don't have a diagnostic to know if the therapeutic is actually working if you analogize deside to like a therapeutic for the scientific apparatus um and I guess like my my question to you would be like if you were us if you could redesign the system I think you talked about like broken

incentives in terms of funding um in terms of accountability within the existing you know academic system how would you redesign it and if you were us like what metrics would you use to um measure quality impact all the things that you would want yeah I mean I I guess I think I wouldn't start with like trying to design a drug so drug Discovery is like one of the hardest kind of fields for Success so there's actually a really funny um sort of uh uh casually named law which is called e- Rooms law e r o Ms do you know who coined the terms law that's a good question actually I

don't his name is Jack skel and he is on molecules board so let's ignoring kind of the the the that the the kind of history of this let maybe we should say what it is first which is Mor's law say something akin to the number of transistors doubles every 18 months or so on a chip E Law as the the amount of dollars Andor time depending on which scaling size you want to use uh to make a drug doubles X years I forget what the time interval was but you know it's like mors law is like exponentially increasing this is like exponentially decreasing in terms

of like the efficacy of each dollar of funding into drug Discovery um I basically think that going straight for drug Discovery a thing where like the capital and resource costs are blooming exponentially with something without any accountability kind of doesn't make that much sense I think the places that make the most sense and that are the easiest to start with that could actually be really useful are a uh verification of existing results so like you know a lot of the the famous academic fraud cases of the last uh decade were found by independent researchers and individuals who are just

rereading papers trying to replicate the experiments or trying to to understand kind of where their data came from what kind of setup they had and then they were unable to do it and I actually think that is the type of Citizen science that you see a lot of impact with where it's like people around the world can try to replicate experiments try to to see if these results still hold and if they don't then they can kind of collectively act as an Oracle and say oh no actually this results false and I think in general the fact that that has been done by like individuals uh Outside The Establishment and they've done done it without any incentives if you created a system for incentivizing people checking

results doing all the stuff that's kind of like maintenance right like you don't get a scientific career right now if you publish negative results or you verify old results and show that something is wrong why not why not I mean this is true for a lot of fields right it's like everyone wants the shiny new object they don't want to they don't want to see the old thing overturned but this is the type of thing that decentralized systems decentralized means of coordinating people and incentivizing them to actually do these things would actually be very useful and would make a lot of sense I agree how like who would the um customer there be to your point it's like if you want to

create an incentive you have to provide one who who who would buy verification of science that they just want to get published so actually going to the uh drug Discovery uh world there's there have been a number of um studies that are sort of uh done in pharmaceutical companies or biotechs or like not not academic institutions where they've tried to replicate a bunch of drug Discovery claims and basically they get paid by these pharmaceutical companies to rep at Old claims to decide whether certain targets are worth going after right so so of course there's been there was a kind of the many hundreds of billions of dollars have wasted in the

amid plaque Alzheimer's type of stuff and a lot of the way that people found that that thesis that Target was not correct or like there wasn't statistical accuracy in you know that thesis for why that causes Alzheimer's uh came from people who are basically paid at their day job to validate old results effectively and then they were unable to valid them so I I actually think there is a sense in which people who are doing drug Discovery actually would pay uh for kind of replication of things that are the reason they're working on certain targets right but couldn't they do that already like what's stopping them from

doing that right now anyway I I can keep going cratic method there's another thing I guess yeah to to your point these things are are connect ConEd like another reason is that you're not getting you know the currency of Academia is some combination of citations and where you get your papers in right because in a lot of ways academics generally get paid less but they get paid in status right there's this sort of notion of like you know you you you're the Godfather of this field yeah James Simons always would call it Gentile poverty but uh yeah like I which is kind of true it's like you get to like look like you're the you the high

class person but you're like getting paid nothing um and it's such an Arbitrage by the way on on intelligence that academic institutions are pulling off it's incredible I mean they're getting cheap labor right like I mean the real losers are like the posts and like I feel like they're at the bottom of the totem pole but yeah either way like it's not that I disagree that all these things are wrong that that's not my problem I think it's just more I think it feels like okay we're throwing money after the sexy thing like the drug Discovery but we're not throwing money at the things that actually decentralized things are good at doing like solving these Oracle problems like validating whether some set of conditions holds right like that there's

verifiability those things don't seem to like ever get you know a lot of funding like every you know if I look at the you know fing why not ra rather like I hear you and I agree with you but why not the answer is because there's not really a market for it yet any Pharma company could go and try and replicate their results with with a cro that is you know incentivized just to run the experiment and not for the outcome to be any certain way ex except they will only hire the cro that make the data look like how they want it to because that allows the asset to progress you have progression bi us all through these Pharma companies and it's because all of the people if they're part of a program that gets dropped they get fired or

reassigned or they don't have as good of a story to tell so the incentives are all aligned around progression and maybe this is the root cause of E's law which is what you're talking about right and so to say that you know drug Discovery is hard and expensive and therefore we shouldn't touch it to me is the exact like it's it's the it's the opposite right it's because it's so inefficient there's a huge Market opportunity there I sort of think like in a lot of things that require long term funding funding over decades you generally need to have bite-sized portions of things where you Pro you you raise money you promise something you deliver a result you get

an outcome and based on that you are able to raise more money or or not right the problem is right now it's being sold as we're going to raise a bunch of money and you're going to get a drug at the end great which is like not true right it's like not accurately presenting the the probability of success a and then then B it's not presenting the set of steps along the way that would get you there right it's just like oh we're just going to like make this drug right like and I I think and like get it through clinical trials and whatever right like it's very unreasonable when you read some of these proposals right that that's the part that is wrong as opposed to like you know I I think like if there were a me a means again you could have

conditional funding mechanisms over time that are used to you know on each milestone fund either give more money or less or whatever right like that uh a you have some sort of like Milestone based funding and to do that you kind of need to have a very long- termist view the problem is everyone in crypto is buying these tokens thinking that they're getting the the the the return tomorrow and the the the the duration mismatch between the long-term is View and the bite-size Milestones that need to go into a big project like making a drug versus the like I'm buying the token and like I'm expecting like this thing in

two months to to to make this drug you know get approved somewhere I think that disconnect is the part where I think people will get burned they will view this as a scammy funding mechanism and then it will have trouble raising money in the future for real things which do need this funding right like like the reforming peer review like the um kind of uh uh things about kind of replicating studies and so I I think like the problem is that if this funding mechanism just gets viewed as sort of purely a scam kind of the way that like junk bonds and biotech uh in the 90s and early 2000s were kind of viewed as a a kind of scammy way of funding biotech

and then you know you saw everything kind of moved to the private Equity VC model from the fully Public Market model you kind of will destroy its credibility and I think starting with things that are a approachable B have a high probability of success and C can actually you know deliver on those in a reasonable time is a way to prove that the funding mechanism Works versus kind of like promising the world uh at the beginning that that I guess that's my sort of I couldn't agree with you more honestly on every single one of those points um I think that coming out and saying you know give us money and we'll give you a drug in 2 months is obviously an absurd expectation that we'll never ever be

able to be fulfilled like and you're just going to burn a lot of people in the process um and so I don't want to speak for any other projects but like can speak for the one that we're working on which is you know we are taking these exact bite-sized chunk approach with Milestones that you're talking about where it's like you know put money into this token out you get a a celegans experiment on the small molecule you know more money goes in than you get a fly experiment so there's no promise of a drug it's a promise of that an experiment will be executed it right and you know hopefully that provides a data package that could then maybe not be used for a drug but potentially a supplement right and these supplements are like much much easier to get to Market than a new chemical entity right

so I think that you know other projects can speak for themselves this is what we're focused on which we think is much more near term and therefore aligns with like crypto you know time Horizons and we're also not promising supplements we're promising experiments to get run and um I actually have a question for for our hosts yeah which is uh which is you know from your perceptions just reading things on Twitter what do you think PE what do you think dii is and what do you think people's expectations are as sort of kind of people who have been around and have seen some of it but kind of are yeah maybe more external yeah I mean I think um I I from

my understanding just from outside is looking in is the funding of yeah science that is going to lead to you know not not to the Grand scale of like actually rivaling what the the you know big Pharma or or like Pharmaceuticals today but the decentralized funding of science at scale that's going to lead to breakthroughs of um you know things that weren't previously possible through the slow inundated processes of traditional Pharmaceuticals my understanding of dii was not so much funding but more so around the information flow and so here

like I I was of the understanding decentralized s science was to uh displace or disrupt peer review for the most part because I kind of heard about it around the same time that I was learning that peer review was gatee kept um you know there there's expert peer reviewers that kind of fact check all these articles that ultimately get published um but there's also pay walls and like sometimes like pretty pricey pay walls on these peer-reviewed articles to either publish or at even just read some of these journals and so my understanding was that decentralized science the aim was to lower the barrier of entry for peer reviewed sence and

maybe digging into that a little bit a little bit more is like look at look at the incentives of peer reviewers they call them like three and on randos right you have no idea who these people are they're not accountable to the outcome of their peer reviews and yet that is like the standard for getting a a piece of research published right by people that are not accountable to the outcomes of their like advice and their review to me it's like the best peer review is putting money behind something because you're putting it stake you're putting your reputation at stake your resources like anyone should be able to do that and then organically you should have it emerge who the best pure reviewers are where the resources will ACR to them and the reputation will

ACR to them organically so I I agree with you DCI has the opportunity to um to disrupt this what I worry is that the peerreview system that D creates doesn't actually properly align incentives but I'm super curious to hear tun's thoughts on this too Redstone is def's modular Oracle supporting over 100 clients on over 50 chains offering 1,000 assets Redstone is the leader for Bitcoin staking and yield bearing assets they offer the most robust customizable price feeds available in any model join the Redstone exhibition [Music]

today welcome aboard on our journey into the singularity this is the endgame road map for fra and fraxel the new L2 out of the frax ecosystem that is totally based on fractal scaling fractal a modular L2 blockchain built by fra with modular and fractal scaling in mind fractal scaling is based on the mandal BR set of fractal images and it explains how l2s become l3s and the entire app chain life cycle goes through its Evolution the frax ecosystem is made of some of the top builders in the space and has stemmed

from its OG days of the frax stable coin it is now in the LST space the LRT space and this frax L2 joins the existing Suite of fra products including fra Swap and frax lend you can see frax fairy the bridge to go from any chain arbitrum optimism or main net into the fra main net Bridge your tokens and bring your FRA assets into the fractal ecosystem join today to get the best out of fraxel you are still early join today take part in our Expedition into the singularity yeah I mean great great question so I I

will say on a personal level like one of the reasons I kind of got into crypto research in like 2017 2016 was like I was just like reading a bunch of people's uh papers and I was like wow there's like actually good sort of like I wouldn't have called it theoretical CS but like on the border of like theoretical CS work that people were doing in crypto and it was kind of fascinating that people were doing all this research on their own right like I would say crypto research especially in defi or even in uh layer one cons layer one consent is probably less so it's like way a little more like

establishment academics but like in defi um in um some of the L2 research a lot of rollup research is done by non- phds right like I'm I for instance I I don't have PhD I like dropped out and then never went back um and I think that the ethos and crypto the idea that like most of the research is and AI is a little bit like this too I think in the early days uh a lot of the research a lot of the highest cited research was like someone who just posted something online and it kind of took off right and the peer review there was the marketplace of people being incentivized by their investments in different crypto things to read the

research and like criticize it and whatever right now of course that also leads to FUD and you know the usual like stuff like you have to kind of you know the marketplace of ideas does sometimes need constraints but I think crypto research always fascinated me because it started on its own as this already open thing there was no like real gatekeeper per se like and I remember in the earliest days that I was writing papers where they would sort of be in the the middle of sort of Quant finance and and sort of algorithmic Game Theory I would like submit to these classical conferences and like no one the reviewers I get wouldn't even read the paper like they're like like ah this cryptocurrency thing all must must

all be a scam and then like you know two years later I would find out the like the reviewer wrote some some paper on Defi and like suddenly was like ah it's real now you know so like I I I totally get this problem I just sort of think the the harder part is somehow making sure that you can keep the quality of the review High while also being able to incentivize people like the problem is if you start paying people you turn it into a volume game and less of a even more so than it is now right like so you have to you have to find like a good middle ground and I'm not saying that uh you know I I and

iens I don't know I I I'm I think like there needs to be some form of slashing this is my other problem with a lot of the DS St so if you look at bio protocol it literally is just a Yi Farm like it it has the token and then it has the subow tokens and wait wait wait we were talking about piew we were talking about pie let's let's finish let's not we're getting we're getting to why we're getting to why to to no no no no but you were you were so close you were so close to like nailing the point and I just want to make sure we get there which is like once you have tokens and you have markets around the science now you have an incentive to to actually give good review because it's basically like a short report or a long report by a hedge

fund right you put out the research there because you are incentive advis to I I do sort of agree but I also think there's like one other problem with markets that is sort of a problem that you have a little less of in the current peer review system even though there is definitely some of this is a n the natural notion of collusion amongst Market participants um where they're incentivized to to push up the price of the token even for for research that's false because it happens to be the most convenient one to get a return from and like that is where you have to really be getting the collusion resistance piece is where I think the the hardest part is because you know I don't know if you

follow this in AI so in AI actually because there's just such a insane number of papers most of the review process is done in open review now which is like basically the the paper the reviewers uh reviews are even though they're still blinded they're posted online so you can like read any of their reviews so the idea was like it would be like socially embarrassing for you if like someone could see that you were just like a huge in your review and like on you know cuz like there's a lot of ego in the review cycle stuff where like people will just be like oh well so so so so the natural reason for this is you know I send a paper on topic X to a journal or conference the Journal of the conference will find someone who has interests near

paper x to read to be the reviewer well they might think I'm front running some research of theirs or like that I'm I'm kind of like saying something that they already thought of or and so like there's this natural like beefing that goes on in the review process and I think the AI world is a good example of like actually just doing the dumb thing of making reviews public people started getting a lot nicer uh you could kind of see that people sto being like this is absolute garbage like like they would say oh I don't think this is correct you know you could see the and people have people who cares people are mean or nice because you're you're you're you're pure reviewing the D ecosystem and you're not being very nice to us on Twitter but it doesn't seem to disincentivize you and

that's okay and that's okay be mean we can we can use the criticism be mean what I'm saying is that's not the best accountability metric is likeor I didn't I I didn't get to one other thing which is like uh then people once these reviews were public started taking started trying to classify people's reviews by their like language style and this is actually one of the reasons that the thing is important and people did a bunch of stud where they're like oh are these reviewers all like citing themselves and then also like blocking other people and they were able to classify like groups of like authors who like you know you can see their public citations and correlate them to their reviews and the open review thing actually the the the the tone change was

actually a strategy taken by authors who now realize that they could be doxed as reviewers by like being too much too aggressive and so I think there's there's actually like some interesting Dynamics in Ai and crypto research that are much closer to testing these kind of game theoretic things than most other areas of research mainly because the volume is high and they're already completely open right like anyone like they're comparatively you know like in BIO it's like I have to submit to a journal and it takes a year to get a review in some journals like that it's like kind of crazy right so my point though is that I I I agree fundamentally with the idea that markets are more efficient at at kind of this

type of curation I think the harder part is just making sure they're collusion resistant and they don't have these like kind of indirect second order incentives where it's like oh in order to like get my if I site this other paper that had a successful peer review token a bunch then like the buyers of that token will all buy my token and so then like I've made the reward for reviewing mine higher even though my paper might not be as good you know like there's a lot of little stuff like that you have to get right and I think that's where I think I haven't seen any like comprehensive thought but that's where I think you again that's like a great usage of a decentralized mechanism right like you and the reason I brought up the the bio

protocol thing is everyone is focused only on rewards like paying out tokens to x no one is focused on slashing and penalties but you need penalties to make any of these types of mechanisms work especially when you think about these collusion type of properties right like one of the reasons a lot of things in crypto work is because there's some notion of a smart contract being able to penalize you right and so far I've seen very little in the way of implementing penalties and penalties for doing something wrong in D and that that's what I mean by this sort of accountability thing I think the peer review stuff also kind of is missing that right now if that makes

sense it makes perfect sense um I I agree with you like nobody I mean I think frankly when you're in a system that you're trying to grow you know coming out and saying yeah we've got all these like slashing opportunities isn't exactly going to track very many people into the space I think once you have I Igan got $15 billion and literally the main main usage was is is for slashing now of course it started with the points reward stuff but I'm just trying to say that like a lot start with the rewards but these crypto systems at least promise the design of uh you know these slashing kind of rules to enforce these kinds of kind of mechanistic properties that you

want that's another reason when I like look at the code bases of a lot of dite things I feel like they they haven't learned any lessons from crypto protocols in the past and like may maybe this is 2017 but it doesn't feel like they've learned that lesson super briefly here I think the important Point here is that there is design space here and do we have it figured out probably not but like I think we should be allowed to experiment um and you should be allowed to criticize and like in that process we'll Converge on a better system than exists now and what we have today right and so like to come out and say like d is all a scam you know even though you're acknowledging here that

there are interesting potential mechanism designs that aren't figed out I have I have acknowledged that on Twitter too but once the tokens started coming out and everyone was suddenly everyone stopped caring about those uh I don't know about everyone or it's just your you're it's the noisiest people this is the problem it's like and and it's to your point it's like oh GE it's only about funding like it's not it's just these are the ones that get attention whenever you put like resources and money behind it a lot more people care right and then they care to P riew they care about prediction markets they care about all these other they care about cryptography and verifiability the point is is like you have to create the Tok tokenize the research first get

people to care once they care now there's a market for all of these other components that you would consider more noble well I don't consider them Noble I think that they take advantage of decentralized systems these things that don't need decentralized systems there's no like what's the point of the token other than to just be a mcoin and I think that's the real criticism I have is like if you're claiming to do this higher more noble thing than a meme coin and promise all this utility you have to have a way of actually guaranteeing some delivery on it and this is what I mean by like penalties are important you can't just like only have rewards yeah understood so tan when you were looking through the um code bases of some of these projects what did you

find I found a lot of like multisig Plus+ where it was like there was a multisig effectively there was like kind of like some white list for some participants who could send money in One Direction or another there was sort of um for instance the bio protocol stuff just reminds me of like Aerodrome or convex Finance where like basically they have these rewards so they have the bio token and they have the subow token so for the subow token to join you have to pay uh 6.9% uh of your uh uh token supply for 100k of Bio so it implicitly values all of the the BIOS at 1.4 million to start um and so you first off have this weird

thing where the subd Dows are kind of Tethered to this main token it reminds me a lot of ironically bit clout where you had the bit clout token and then you had the user tokens and they had to do this kind of same bonding curve thing it looks it actually looks more like big clout than anything else um and then there was there's rewards these bioac rewards which are sort of like um the reason I'm bringing up C convex Finance or Aerodrome is that they're basically this type of thing where you lock up uh your either bio or subow token and you get higher rewards uh and then you can use that to direct funding like new token you know like the biod Dow gets some fees from like selling tokens and

from these new tokens and it you you use that to direct the uh funding somewhere so like maybe I'm in sidow and and I uh sort of use I kind of vote to to redirect the the funding that bu has into side out instead of into invited out or something and this is exactly like the bribes in Aerodrome or convex Finance right where like basically you would vote on like where the reward which amm the reward should go to um the difference is with amm they're generating real fees their fees aren't like we're selling our token uh they're act fees from usage uh and so I kind of feel like this mechanism is is just

going to look like a pure Ponzi scheme because it doesn't have any real fees unless these IP nfts magically pay out but then as I've mentioned earlier there's no accountability they rely on this wh list thing they don't really have a guaranteed payment there's no like legal there's sort of some legal structuring around it but like compared to like RW it's like 100 times less um and it's like sort of this weird thing where it's like oh like you'll own a right to the IP scroll down oh actually not you're not guaranteed any payment unless the IP holder decides to send it to you it's like there's like a lot of stuff that I don't know it it it's it's kind of misleading and also I I like I said I think the aerodrum and convex

financings are interesting mechanisms in defi for things that have real yield right they have real fees there's like a reason to actually use the mark token Market to like say oh okay this amm Market should get more rewards because it's not getting enough liquidity but it's generating a bunch of fees right I don't I don't understand how that helps in in a scientific thing where you're you should be like long-term minded thinking about uh kind of making sure you hit certain Milestones as opposed to like locally pumping the token price like like a lot of the stuff just looks like pure yield farm that I would see on binance Smart chain in 2021 like that that fundamentally that's what it looks like benjes what do you uh I I I know that you're not representing bio

protocol directly but clearly you have deep ties with and obviously knowledge so what's your what's your kind of rebuttal there yeah no I mean I think I agree with Chun that like um yeah you have to be super careful about any incentives that you create and rewards and making sure that you tie those with outcomes and Milestones versus like inputs um I I believe and like please don't um quote me on this but I would I would verify this with the bio-protocol team but I believe that their Launchpad will be fee generating uh the Launchpad for you know launching new new bodow so there is some sort of like yield component there um from true Launchpad fees I believe but um you'll have to fact check me there I

haven't I haven't connected with their Tech Team in a second so um but as for what tun is talking about in terms of you know having these bribes but rather for inputs for outputs I think that makes a ton of sense I would agree with him completely I think you have to be super careful about how you dilute your token and making sure that um yeah you're just incentivizing the right things so I agree with him there completely I think it's it's good to scrutinize you know what's being incentivized and how it's being incentivized and making sure it actually aligns with what we want which is like higher quality science coming out the back end not just like funding and token price pumping yeah so tun you've got voices like Adam Draper being and I quote

annoyingly pro dessi this year and I think you quoted I'm going to be annoyingly anti- d d this year in its current form um and you kind of put this list in I think this is probably a good way to to come to to the wp as we've got about 10 or so minutes left one of your points was that D has a job and if it fails you are 10,000% responsible as a proponent I think that's kind of more or less what what we covered today um with regards to the verifiability and accountability of actually uh getting results um and and and the other one we just kind of touched on about kind of the real value creation so this isn't this is a sector that I think is

particularly polarizing because it is actually and this is even more like pessimistic but it's actually one of the few crypto sectors that Bridges the gap between offchain and onchain in a real way uh like RW have uh and and are doing a lot of what we're doing today in crypto is very much onchain oriented we're swapping we're speculating we're buying and selling we're operating within communities um so trun I would pass it to you and say if you were to to uh you know look ahead to the next 18 months of dii what would what would change your mind on the progression of some of the protocols mentioned today as well as some of of this broader idea is

it just this Milestone based work is it better smart contract verifiability is it a different way of building this design are you guys funding other teams are you secretly cooking a better design like like what is going to change your mind and like that's what I want to know how is this going you know I just don't yeah I like look I actually in my mind the types of things that I think are worth funding are like experiment replicability like actually really doing that the peerreview reform I think like this is why I like things like research Hub I feel like they're trying to get they're trying to build simple small tools that like slowly move you towards that as opposed to like promising the

world I think that to me is like kind of what I want to see is like dii right now just feels like mcoin pump and dump uh and if it could actually be incremental progress to things that are achievable in short periods of time that provide a very tangible Improvement like that that obviously is better but you know instead I I I opened Twitter and I I just like you you read what people are writing and it's like okay clearly the people buying this stuff buying these tokens are diluted either by not reading the docs carefully or just you know whatever uh and then the people promoting it are kind of

like promising this kind of you know the world and there's this huge gap so I think things that are are trying to like take small gradient steps that are well defined are things worth kind of um investing in I just don't think I've seen that much I think like some things have been kind of interesting like like yes no error like having AI fraud detection um you know like I I think like that that's that's like the type of I I don't see more of that type of stuff right like but I haven't se it feels like and maybe this will because the there's a a nice nice little downturn like maybe there'll be like they'll wash out some of the frothy stuff but I I I think like there is a world in which

these smaller things actually a generate value and B are able to generate like sustainable economics like and if if that's true then I think like you know there definitely are businesses worth funding Co T may ask a follow-up question if you don't mind so um I think you've mentioned this a few times on podcast previously where you criticize dii um and when people try and say oh is this defi 2017 and one of the things that you say is that would really distinguish is Dai from defi um both in their early days is the lack of metrics in dii and that with defi you had tvl right like that was just a clear it was it wasn't just tvl yeah it's more

actually that you have like I can verify the state of what the protocol is doing right like in general uh if you if you think about unisoft I mean one of the reasons I even got into writing crypto research papers was that I was just going on chain looking at the contracts looking at the statistics of what people were doing and then like initially I was like naively this shouldn't work uh you know in like traditional traditional economics system like there's perfect adverse selection and all these other things and then over time once you saw people actually using these things and it's working empirically there was like okay there must be some like Theory some math reason why this thing is working in

general that is the type of stuff you can get out of defi like you can actually you can almost think of it as like science in Reverse like people are doing the experiments economically with these protocols where the designers of the protocol don't even understand how or why it might work uh they they just are like kind of this this seems cool and people put their money in and and and if it works you can verify whether you know if if like they don't lose money or like they facilitate trades or like some sort of more expensive metric than tvl like volume traded or open interest or something that's less manipulable you can actually really verify it and it's on chain it's cryptographic ly verified now for science again it's like harder to do that right like what you're describing

you're doing with the streaming and stuff it's like a lower standard of manipul anti- manipulability it's like it's much cheaper to manipulate that than it is to manipulate um you know certain metrics in defense now obviously and this is why I brought up the binance smart chain yield Farms it's like yes eventually people figure out ways to fake things and make it look like it's doing something real and and I I that's sort of why I I think I've been a little so so negative on a lot of the dside contracts they remind me of these like yield Farms that didn't do any productive defi work that showed up after defi summer where they were just you know giving you rewards for doing nothing and then everything really relied on just the reward token going up

and if it didn't then it' blow up but the thing is when those things blow up you know you you know it happens whereas I feel like in dii it's like yeah rug pull it's like yeah and I think the verifiability component is the most important part I think obviously for a lot of experiments that's going to be impossible like making attested devices for like every experiment might be impossible but yeah like there's stop gaps potentially like what you're describing um of making a many different ways of generating proof that you know a an experiment was done or you know Etc so it's not it's just that Finance is more implicit

explicitly you know you can you can you can verify a lot more than in science where it just takes a long time and and it's off too it's also offchain right like you can like on Unis swap like you just make smart contracts and then everyone interacts with that and then now on metrics we have to create a tested devices right like yeah we don't even know if people care about the experiments yet right and so yeah exactly exactly yeah so I I it's it's like I think there's a world in which the good stuff can come out of it but it just it unlike 2017 like a while 90% of stuff failed some things that were really good

got quite a bit of funding right and it still persists to this day right I just have yet to see anything in DSA that looks like that and may maybe I look I want to be proven wrong as you can tell I like care about this this issue right it's just more that I just don't like the things that are like seem like just clear scams look I get it and I would say like you know just to flip it back um we're super appreciative for all your criticism I think it's a massive net positive to the whole industry um I don't think any of your takes are bad I don't think any of them all of them are directionally correct I think the important piece is like you know we as a d space feel like

we're being CRI ized for experimenting um but I agree with you that we need to be held accountable you have the right to give your opinion um and I appreciate you sharing your feedback so publicly because it really forces us to level up and so this is what I mean it's like I think it's a huge net positive and you know at least you care right and at least you care so much to give your you know deep criticisms and yeah agree with vast majority of them well thank you I was very very very uh kind of you I I I just you know I it's good to hear that there are people who are like trying to to be uh be more cognizant and like I know they exist right like a lot of my

my old colleagues who are in uh either in academic jobs or work in um research of different forms uh just like they they message me about this and they're like oh is this the this is great like I can get this this I can get this random side project funded I'm like that's not that's not that's like you're you're buying into the mcoin thing too much and I think like there's kind of this thing where like it does it has that Ico boom feel where like people like cryptocurrency I don't know that is that ever gonna happen but yeah the token goes up you know like it does have a little frenzy aspect which I just think if that dies out the real best

things to invest in that are like long-term and will last or probably when when the market crashes so take the bad the bad takes away the good as well yeah I hear you gentlemen thank you so much for joining us today I think we're I think we've pretty much covered the most important points in a in a very respectable and productive as well as educational manner uh frankly this thanks for having us yeah this entire sector has just kind of came out of really I mean I I I know that you know for example tun you've been bu this space for a while but I feel like for the broader crypto space it's came out quickly um so this was definitely a primer for Robbie and I and our audience and uh would like to do

more and learn more and um Benji I hope that that the that team watching listening and also just internally has uh obviously thought about these these kind of criticisms and working towards a more uh you know a more refined version as as the years go on and thinking long term I think is very important here so again thank you guys I appreciate it so much thanks for having us yeah feed feedback while taken and yeah I would love to do this again if tuns down thanks so much yeah for sure amazing thanks