The Cost of Entry

Transcript

trying to get um Mark on stage, but let's reserve a seat for him and kick off the discussion. There's plenty to discuss here, so >> yeah. >> All right, before I kick it off, um >> yeah, you can maybe do you want to introduce yourself quickly because you haven't been around when in the morning. >> I guess we all should introduce again because I'm not sure who who here was here the whole day. >> All right. So I was here in the morning. I'm Robert from 1KX early stage uh venture fund. So uh I'm taking the investor perspective a little bit on this topic. Um but yeah maybe you guys can also quickly

introduce again for those who haven't been around. >> I'm Angela. I founded token engineering academy. Um and we ran the first token engineering track at ECC three years ago. It's the third time or two years ago then and our goal was always to establish cu uh cryptoeconomics token engineering rigor in designing tokenbased systems and this has been our mission since then and I'm happy to have great allies here um from both sides practitioners and research uh to organize tours and get this to the next level.

Hi, I'm Chris from Token Engineering Labs. I'm also with the crypto econ institute at the Vienna University of Economics and I'm working in the token engineering space for almost eight years now. And the approach that I'm taking personally is um this synthesis between research um applied prototyping and um and um basically trying and modeling um to to uh to arrive at at a solution that is really stable and um and functioning. So um the conversations that we're having here very much feed in into the methodology of work that we're doing and I very much appreciate that we can have

this this discourse. >> Yeah, that's also how how I end up because I met this in um in academia and then we work together like also in the startup side. So like I'm I'm part-time in academia in the blockchain center of the University of Zurich where we research uh from from the law aspect up to the regulation to technology to the economic impact of these blockchain based systems. So we have quite a diverse group of people doing research there which I'm representing here and I find this discussion very valuable and hence I'm very happy to be here.

>> All right. Yeah. Um I mean we we put this panel a little bit under a provocative question like what happens when research gets crowded out by hype and the context is a little bit right I mean we all know crypto moves fast you know move fast break things um but the rigor often lacks the you know behind this narrative and then what happens I mean we all see this right not only have you boom bust on hype but then even kind of like mechanics where you know this must break right the the most prominent example is you know terror stable coin like everybody who knows mechanism design saw it coming because there was no way this this not going to end bad, right? Um but as seen today in the talks um turser the you know token engineering

um research symposium uh exists to create uh a research first non-promotional space and um yeah to to explore our question I want to start with Angela um you founded the academy as you as you just said you built out the token engineering community for years right you actually kicked off uh this symposium uh so the question to you um yeah what gap did you see that that made tur necessary for you now? So I would like to start earlier when uh we what gap did we see in token engineering because it's so simple I mean in the blockchain space uh very different dynamics come together to form uh an

economic systems it's uh like the fascination about blockchain systems and what we can do with this permissionless systems um uh ownership um sensor assistance then we have the startup up spirit in um okay let's break things and move fast and also this uncertainty so it's very hard to design things from scratch you can't plan for success you have to use the momentum and then we have this okay but we are in fact launching and trying to steer and make successful very complex systems and and this was for us in in the beginning and

still is I'd say a huge motivation because it's hard and hard is fun. Uh so I would say the initial starting point was we have great ingredients and disciplines and knowledge we can draw from but we have to combine it in a new way to tackle these huge range of challenges we see and also the dynamics in the blockchain space. And I think that's still visible here. Um if you're looking at the agenda, uh people from very different disciplines are tackling problems in the blockchain space from different angles. And I found it even

interesting uh for some speakers like I I think Stefan Lush um showed this journey in exploring how can we tackle this problem and and you found new ways and um this is exciting and I think it's this we have a huge chunk of intrinsic motivation in our community and I think um a growing body of knowledge um how economics makes game theory um modeling and simulations can be applied to the blockchain space and it has been a very exciting journey.

>> Okay. Um yeah I mean maybe to uh go go to to Mark uh here because um you just presented your research um and you are publishing regularly right? So, so for sure the researcher here in the group um maybe coming back to the you know hype versus like fundamental research questions do do you actually when you know are going out for for your research feel impacted by by by the hype and narratives um right when you know when it comes to scoping your research um but also yeah how how do you how do you view the gap basically between kind of like academic research and you know what gets discussed and uh implemented in

protocols basically right so that's practitioner academic gap basically >> yeah um I mean like a sensitive question um in in the end I mean for the research I presented I would I would argue it's a bit uh more more fundamental so in the sense of um it's driven less by by hype like um but the problem is always um then the rewards are not that immediate so in some sense so like you need to find these sweets spot where you connect to maybe interesting things which other people are already doing. Um if you do it too much alone like uh academic terms you're not um getting um that much more um momentum in in some sense and um I

just recently got more interest in this work. So like I think there's like uh value in this um and there's also value in hype topics I I believe. So like and and I would say the interesting thing is that um there's also hype in research and there's hype in industry and they don't necessarily need to overlap. So like it can be quite disconnected. I think what we're trying to do is maybe get some interesting things uh maybe overlapping. Um and I think like uh what I like to do is like kind of um these different hypes to to play play with it like because like um hype in research can help you to publish something to have like talking points for industry to maybe change something fundamentally more in industry. On the other hand um

industry is sometimes quicker or discover something new. So you can kind of take successful hype topics to push forward your own topic in academia. And in this sense um makes sense to sustain these hype topics. uh and um publish in in these domains. So I think they can can help each other and um the second part of the question how much it's applied. Yeah. Or like uh >> Yeah. Yeah. kind of like also as a followup because you're also kind of like supporting protocols. So you have both viewpoints, right?

>> Yes. And I mean Yeah. Like so I I think that's I mean it's leading to why we are why we are here. I I do think that like in both sides there's so much value. I mean so many shortcomings but let's focus on on the values and um and um yeah I mean I see it for example from this DAO perspective. Yeah. So like in in Dows um they're like you have like the research community but there's also like very split like you have like the DAO researchers and then you have like um um yeah political scientists. We also saw like from Cleos we like they like to use social choice theory and voting um and so these communities they have

different topics um they are working on and when I see like in the context of Dows we all in the industry but also in the DAO research we focus kind of on on voting but not enough on deliberation or maybe even in um um as later like in conflict resolution. So like these aspects are heavily underdeveloped and in this sense academia has solutions for this like there like researchers really focus on deliberation and everything like since years already and there's like kind of crypto totally ignorant like we believe like um I don't know like uh improvement proposals on GitHub are good enough or something like this.

So, so there's more. On the other hand, I think like the innovation with different voting mechanisms per se in in crypto like a lot of people in academia are just overlooking it. Optimistic governance we heard today and like like these things um could be studied much more taken up by researchers. So I I think there's like a lack on both ends but there's also I mean there individuals pushing on both ends and I think also this group here. So I'm quite optimistic. >> Yeah. I don't. >> Okay. Yeah. I mean Chris you want to ask something because you also basically cover all ranges of like protocol support research academia.

>> Yeah. So totally I think so it is a reframing problem actually or a reframing conversations because um all of this already exists in like those isolated silos or isolated worlds. So you have the academic space where you have been deliberating on problems for hundreds of years. At the same time, crypto is existing, is building, and people are talking about the same problem, but they're using uh different jargon. They're using different um vocabulary. So, um obviously they do not understand each other or are not do not even understand they're talking about the same thing. So, uh you can approach the the thing top down or bottom up. So you can say let's try to establish

crypto in academia or you can say let's try to get people in crypto look more at what has been happening in academia. Um and both solutions are valid and I think that um the space of token engineering has been maturing for the last couple of years and we are now at the intersection where we actually can synthesize those approaches. So we really are mature enough to now have a conversation here uh on topics with um experts from both fields and they would actually acknowledge that yes this is a problem that we are both talking about and yes there are solutions or there are are approaches that we can leverage in this particular blockchain use case. So

it's um not anymore a problem of translation. It's more a problem of synthesis. this is where we stand right now and um and this is what we try to to approach. So um I want to to to end this with like another note which is um there is a distinction between multi disciplinary and interdicciplinary research and I think that this is very important. So multi-disiplinary is if you are capable of going into multiple directions and you can talk about this things using different methodology or apply different methods to solve the problem. Interdicciplinary um research is when you carry over solutions or methods from one discipline into another and I think that we are

entering this stage right now. So we are entering the stage where uh we as a community will start working interdisciplinary to synthesize those things. Yeah. Yeah, that's good. And yeah, now now we also have Mark on stage. Another Mark. Um, and yeah, maybe also you can quickly introduce yourself. Um, but also the question for you, you elaborated even in in a blog post your your motivation why you support the the TUR program. Um, and yeah, maybe you can share from from there what your observations are, you know, as a builder, as a frequent uh conference participant. um you know when it comes to exchange of ideas and the pres presentation or like discussion of research right

>> sure uh yeah so I'm sorry I was late for the panel um my name is Mark Richardson I'm the project lead at Bangor since October of 2022 um before the pandemic I was an organic synthetic chemist and so my you know I was on the sort of tenure track career path um which got thoroughly interrupted by uh the pandemic which is when I made the the switch to blockchain. I only bring that up because I I have been well trained right in the you know traditional um let's say academic caliber of what it requires to make a claim and support it and you know bring that into the scientific record and it's prohibitively high for a good reason and

I think that with respect to to hype and you know one of the things that Mark brought up was that even in research there's hype and that's certainly true like when I was early in my my PhD career things like olphan metaththesis and palladium catalyzed chemistry was like all the rage and there was like this hype effect where if you were applying for grants, you kind of want to apply for things that are trending with funding agencies as well, right? And they governments will usually prescribe things for you that are um that they consider important even if they're not important, you know. Um and I think in in crypto and in uh blockchain mechanism design and protocol design and consensus mechanisms, you actually have some

version of the same uh deliberately introduced bias right into these things when there is um you know like at the moment um privacy obviously privacy is super important but it is also hypy at the moment right like these things get some time in the spotlight and if you happen to be offering privacy technology or focusing on privacy technology. Now is a really great time to come to blockchains with grant proposals and things like that. Um but you know, think about all the things that you remember from the the cycles through crypto that have had that same um time in the limelight, right? Before privacy, there might have been things like futures contracts all of a sudden were all the

rage and to some extent still are, but diminishingly so. And then before that was betting markets, right? and maybe these things came together and you can keep going back, right? I remember sort of the the Olympus Dow and all the Olympus D forks and prior to that like anything to do with DeFi in general, but specifically on Bonding Coast was all the rage and before that there was NFTTS and you know what I mean? Like these things keep coming up. Um and they spend their time, you know, as movie of the week and then most of that stuff disappears even if um even if something happened there that was worth holding on to. And I think that that is the word like that that is worth recognizing that

within that hype bubble there actually are people that accidentally got swept into it, right? That were doing legitimately cool stuff and that should be recognized and and remembered. Um, and I think that the only way to tease out those gems, right, of of actual new knowledge that occurred is to have something like this, right? to have a filter where uh people who are qualified to speak on certain topics get found through some sort of process because otherwise you know it's very especially in today um you know in in in the environment that we have today it's very very easy to

fake um audition on a topic um especially if you're being asked to prepare pages and things chat GPT can do this for you now with like amazing conviction. Um, and it gets very very difficult to I think tease apart um, you know, who who is actually knowledgeable on certain topics and who isn't. And the only thing that can do that I think is is rigor and friction, right? It's not just what someone sounds like when they prepare a blog or or something like that, but how they respond to questions when they're, you know, you can see if they're consulting GPT, you know, to work out what the answer is. Um but then I think also um in terms of lessons learned, right? So there are like great technological concepts that emerge from

these paradigms. Um but then there are also like really dangerous ideas that we need to hold on to and remember why they're dangerous. And I don't know if it's like every generation needs to learn the same lessons because apparently that they do. But things like you know liquidity mining rewards and how unsustainable that is. And you know even things like US from Terra Luna and the the anchor incident um when I was reading about how uh people were thinking about how to diagnose that problem. It turns out that it's based on an old scheme from the stock markets.

like it wasn't even the first time this had happened in history. Maybe the first time in crypto, but it wasn't even the first time in crypto, right? And so there is like even uh I think a need for a a a well curated historical record, right, of what's real and what's not and what's a good idea and what isn't. And the only way you can do that is is by an academic process, right? And that that informs both the research and the development aspects of what we do. Okay. Yeah. I mean already a good uh step towards the the next point about like what TUR you know aims to solve and and and also how um yeah I like that the both aspect right the positive and the

negative of hype maybe also the the question for the other ones right because it always sounds quite negative of like oh yeah hype is crowding out out the research but you were indicating also some positive examples of like yeah it can also attract talent for example. Um yeah may maybe do you guys have other positive examples where you know hype can actually be useful for example for coordination um in that sense like obviously funding I guess because oh yeah the researchers so um obviously if there is interest if there's venture interest um funding comes along and then this helps to make progress I mean um although we are in the midst of a bare market um I believe that there is kind

of like um floor of builders or researchers sticking to the space that keep working. Um so you can basically measure um how how much intrinsic value there still exists um in the space based on kind of like the the people who keep working uh even if there is if there like if there are in different times right now >> and I must say I mean I'm looking at it uh mostly from the field of governance now head of governance at near and also um hand yeah managing trying to tackle um most up-to-date challenges uh in governance and in protocol design and

of course this is at at a very high speed you sometimes I wish I had the time to discuss a certain problem with researchers but many times I have to have you know make a decision on something um before I even have the chance to raise the question to researchers So this is the speed we are talking about. It's not about weeks, it's sometimes about days and sometimes about hours. And this can make a whole research topic go away or suddenly get into uh the focus of attention. And this is certainly something um any practitioner and anyone interested uh in

let's say rigor in in the crypto sector is struggling with. probably not on in crypto but this is definitely a huge challenge. So I wouldn't even say no one is interested in in rigor. It's rather things are moving so fast. Um but then there is also this a positive side of hype. I'm I mean we started in 2018 and there have been numerous concepts. We had um curation markets today um a long list of primitives that have been developed due to hype basically and these primitives wouldn't exist if not uh between I don't know March and June

everybody was talking about the same thing and then it disappeared but the primitive stayed and I think this is something also the the whole governance sector if it comes to governance in permissionless systems in certain numo systems um how can we secure such systems how can we make good decisions now adding AI on top I I must say in this respect I don't even care about hype because at some point uh questions get surfaced and addressed by people who would probably never met each other or never ask these questions without the hype and if we are lucky

primitives stay insight stay and we can pick them up in a ever growing body of knowledge and this is the positive side and sometimes I'm or many much more frustrated about the moments I cannot dive deeper but then also I think we have we can look at a valuable body of knowledge created in crypto that might be relevant again in a different context text. It's there and it's available and that's great.

>> I think as well it's important to recognize that hype is is something that just happens, right? I I think that people um assume and sometimes it's at least partially true, but I think it's got more to do with luck that hype is something that's like engineered, right? I I promise you there isn't a, you know, marketing, advertising campaign or agency in the world that can bring hype to a subject, right? It's something that is just emergent in a population and I think it's an early indication of how interesting a topic material is to a a big community right to like the masses.

Um and there are you know consulting firms that um have characterized this. It's I think it's called the Gartner hype cycle right where like as soon as there's a new technology um you get this kind of immediate increase in attention that absolutely exceeds you know the the level of maturity that that um that that technology has. Um, when I first learned about this, it was actually being critical of nanotechnology research where I remember in high school people like there's going to be like microscale computers and that kind of thing, meaning like whole devices that are the, you know, that are compiled from just, you know, a few dozen atoms, which is of course insane, but that was the kind of thing that you would see in like Wired

magazine or New Scientists or something like that. And of course, none of that was true. And then nanotechnology fell off a cliff. Um I think though that the reason why some of the miniaturaturization that we have today is largely because of the people that were inspired like during that wave of you know new attention that it does bring in um you know it casts a very wide net and from that wide net you actually do stick some people to it um that end up making a career and actually advancing the field in a significant way and I think crypto is entirely that right like if it it wasn't for the hype of the Ethereum launch like we wouldn't be discussing this right now on this stage right there wouldn't be an

international conference in Khan let alone one that's been running for you know nine consecutive years um in Bitcoin to some extent right that first you know where you can measure the height peak um of of Bitcoin with its price you know um and that does bring people in and um you know at least in cryptography you know that there are people today that know about things like hashing algorithms and elliptic curves and that kind of thing that only know about these things because they were interested in whether or not the Bitcoin protocol actually does what it says it it does. I think that that's beneficial, right? It it does raise awareness and it does bring in new talent and that sort of thing. Um, but it takes time and it's

not something that you should aim for. Like what I'm saying is the hype just happens and that's kind of the the the mass social psychology mechanism just doing whatever it does the hive mind of humanity if you like. Um and during that process right is when people become activated and then join a certain industry and so on. So I I I do think that that's positive >> and maybe just to to add to that um so I mean the beauty of research is that it is cumulative in the sense that it's it's independent of the hype. So for example, if there's if there was a hype um 30 years ago about the voting systems like some progress has been made

on this topic, then it has been stored in archives and then today if there is um attention on the same topic but in a different use case, people can just pick it up and built on top of that. So um I mean what what we're trying to do is basically just making sure that the common denominator as an output you know exists and can be then stored for future value. Yeah, I mean I I would argue even within crypto right you me mentioned betting markets as one earlier hype now we have you know the prediction markets picking back back up from from there and yeah if I aggregate this then yeah basically the hype cycles are good in the sense we get a wider net of talent

just you know programs like ters need to make sure that we condense the knowledge so that we generate this cumulative um basis um to to go on from there and yeah maybe with that um outlook look um yeah maybe at you Angela um this was the first symposium today um and we have seen great talks lots of knowledge sharing very different directions very different research fields um yeah what do you see next from tur like um yeah and also the question to the other guys if you can make a wish for the symposium or token engineering space in general uh for the next year or years um yeah what would that be

>> we just started this conversation over lunch and I must say there's there's a lot to process. So let's let's let life brainstorm what should be next for tur. I I'd say I'm I think this building let's say a a channel or a format to have an ongoing conversation between practitioners and research is still something I feel it's not not there yet and e even such a great day with amazing talks doesn't um is not exactly what I'm looking for um because as I mentioned um challenges pop

up at at a certain speed and also um volatility of attention that it's a hard to have the right knowledge or the right conversation at any point in time. So it would be helpful to just have access to I don't know um be able to share practitioners question easily because I think this can be very inspirational for researchers while at the same time is sometimes frustrating you're working for a year on a certain research topic and see the hype gone when once you are ready to publish. Um I think validating against where's where's

the space moving or what are practitioners thinking about this certain topic is always relevant and helpful to not uh stay in in a researcher's bubble while vice versa um for practitioners and people around me that I'm having conversations with what should be next in a in a protocol or for a certain decision making new new paradigm we are trying to develop. Uh it can be quite helpful to to inform or have an ongoing conversation to establish rigor because I I always felt rigor is not something

that you you establish it over time. So this this level of quality, this level of um you know references you can make and this is not something that you I I feel that in a team or in a certain community you can switch on once you need rigor and you switch it off once you'd like to have speed. It's rather something that has to be an ongoing practice and sometimes you have to acknowledge and accept that you can't apply the level of rigor you would love to apply. Uh but this can be a conscious decision, right? What we are aiming for is in fact that this is a conscious decision and not by accident. you just

miss all the all the all the work and on all the the proofs uh that have been made about a certain um situation or problem and um yeah so ongoing an ongoing exchange between practitioners and research to to further increase the level of rigor this is something I would love to see for tur Yeah, I mean there's immediately next we need to compile all of the manuscripts that we received and you know we've given ourselves a soft deadline that I'm not going to make public u but you know there will be a

like you know um a proceedings of you know to us at at Khan's 2026 um and I I would like to see that idea um evolve over time whether you know if we have to shoulder it ourselves so be it but I think this is an enormous burden um but I'm I'm really sick of this um feeling that I get when you know I might sit down with someone and they're like did you see such and such a a thing right and un you know it's getting increasingly difficult with the many thousands of projects that there are to remain a breast of what you know of what development looks like and where

the new ideas are. I think it's really beyond you to have something like a peer-reviewed journal where people can either post abstracts for, you know, what their protocol is and does and specifics about how it works. Um, so that there is a kind of newspaper right in crypto for for protocol mechanisms and that sort of thing. And I think that there's an appetite for it. Um and you know it's been really great to you know I sp I've been spending a lot of time in in Switzerland recently and you know EPFL has its own you know blockchain research center and UZ which is supporting TUS here today has its own research center you know the ITLE E in in Austria has its own blockchain division you know it's not like the the

institutions in in academia aren't already there you know there there just needs to be you know something some forum where this is correctly organized And you know that makes it we also need to make it approachable because I think for a lot of projects publishing an academic paper is like the least prioritous thing right especially if you're in startup mode or something like that but to sort of introduce this culture of rigor and I think that this is what uh Angela was was speaking to um I would like to see how that develops and I think that the first step right a kind of market test if you like is going to be how the proceedings is is received but I I think that that could continue as well.

>> Yeah. Any any more wishes outlooks? I mean I will make it quick but like I I think like also like this format is like kind of capacity building like in the end because like I think in the end you always really need to do take the decision like so like you can always consult but in the end it will be you and I think like uh getting just preparing for times where one needs to be quick like kind of I think this format is quite nice because we have time now like and instead of listening to sales pitches we're sitting here so like kind of giving this alternative for like collective capacity building and then I think it's a concrete next step.

I also think like um kind of um what I very much like very open like in inviting researchers also to come here like to this group um another step might be to step out and go to an academic conference which is there like and have a workshop there. So we are briefly discussing it and and this might be something like kind of showing the beauty we had here today like with like a lot of good um submissions in a more academic setup. Maybe we want to adjust those a little bit to the narrative which is sent at this conference there wherever one goes and then like a dialogue could start. I think >> for me I mean we've been talking about research the whole day today but tomorrow we will be actually doing research like the next step is to kind of get into the room and apply the

practice actively. uh which is what everyone invite to do tomorrow with us like starting from 2 p.m. in one of the floor uh rooms on the third or fourth floor but you can find it on the um breakout session agenda and also on the main event agenda. Uh we have the token engineering uh breakout session where we invite you everyone to participate and kind of like do this together or like apply this methodology in practice. Um this is one thing that that I would really much see um as one of the immediate next steps and also just to make sure that we are building something that makes sense because I mean there is a lot of push from the regulatory side on making sure that systems that are being deployed are compliant and are

being you know distributed to the public um in a way where like they are safe regulatory uh but I think that we should also do those things technically or academically. So like we would want to establish a standard where the systems that we deploy they are basically validated and I think that we should um um have the ambition to reach a level where we can say okay this has been approved by experts or like this is a valid mechanism or this has been regardlessly tested and yeah this is what I see as the next steps for the community. Oh, Robert my >> my next steps. I mean, I I pretty much second what Chris just said, right?

Because from the investors perspective, this is for sure um I mean VCs or retail investors, whatever investor cap you hat you want to put on, but this is one of the most lacking parts over the last years where we moved fast, we break things, but typically the price was paid by the the people who bought the tokens basically, right? and invested in these mechanisms and and this is the issue there that yes right it it takes rigor takes time to catch up and to validate things but I guess this is like the process and the the the point of maturity where we are right now um you you have protocols and you have mechanisms out there where you know okay this has been validated this is verified um it just needs more stages or you know

I don't know how you call it but symposiums like this where you know okay I can go there um be it in a journal or wherever I can see what has been valid val validated what's what's not um because yeah this is the bread and butter of is working as sort of like economics person and then a VC setup that you need to go around and say like okay does that make sense is it a scam can that work because it's all new new ideas often right so you always look like okay what has been explored in the past and all that so so it's pretty much to to your point right then you go and look out and you don't have a month to do this, you know, you have days to to make the decision because the team is raising and and so on and so forth. Uh

so that's why um I very much support this initiative to uh uh push this forward as well. Um because yeah, it's for sure something that is not there enough right there. We have these but it's not enough for for the space and it's also if if we as a space want to mature then this will be demanded because no serious you know large investor comes in and so like oh yeah you have an ICO white paper of course yeah I I'm going to spend spend all my money here that I have fiduciary duty to basically invest for a pension fund or or whatever. Um so so that's that point. Um and yeah and maybe uh since we have couple of minutes left um

a for questions from the audience but maybe also for the for the participants today or just general audience or token engineers in the crowd. What would you like to see from a program like TUR? >> Too much pressure I guess to launch everyone. Hey. Um, I just wanted to express uh support for the journal where new ideas are published. I thought that was a really good idea.

>> Thank you very much. Yeah, I I agree. I've been waiting for someone to kind of do it. Um, and you know, I've seen uh like Springer and Elsia, they've kind of got like blockchain themed kind of stuff. Um, and there's, you know, but it it's not quite right, you know, like it's like um too financial or I don't know, a little bit naval gazy or something. It's not really like a you know a report of of what's happened and um or you know it's like deeply deeply theoretical which is fine like obviously I find that stuff very appealing but sometimes I just want someone to say this is the idea this is

why I think it works on sort of like a bulletin um and you know to have a sort of call for for questions or criticism the same way that like letters to an editor used to work you know um but yeah I'm glad that we've got a second a second voice that says that they would find that appealing. I also hope we get to do this again. There there's one more. >> Uh thank you for the all the sessions

and I have a question. Uh unfortunately uh I don't I can't go to the tomorrow's uh hands-on session. So like I'm curious what kind of methodology you will like uh try to we will try to use and like um I wonder if the session is like a live streamed. Yeah. So answer to the second question first uh unfortunately not. So like the breakout sessions are typically not recorded. uh but we will write um a report about it like we will share insights after the fact and to your first question it really depends on the problem um on the problem framing and the team working on the problem so as I

mentioned before it's more a multid-disciplinary approach so the first step is to define the problem the second is to um check in the toolbox of methodologies what is the appropriate setup or the appropriate approach to tackle this problem and then go to first principles thinking and try to arrive at a solution or at least a plan to come up with a solution in further research. So we will see tomorrow what happens. Some of the problems defined are more narrow and more concrete. So we would expect um real solutions to be derived within three hours. But for some problems that

we have, we might end up uh defining a formulation that we will explore for several months afterwards. So um it's basically um we will see and we we will we will see when we get there type of thing. >> Okay. Thank you. Later I want to ask more details. >> Sure. >> Thank you. Um I wanted to ask regarding the research that you do with AI and the slope that it generates. I wanted to understand what is your most grueling your grilling pains when you do research rigor research that takes a long time.

What is the most the thing that takes most of your time that you would like to automate actually using automation frameworks? There are there are some things that I do where I know that the for example the algebra is going to be very easy but tedious right um and it used to be that I would use um I found this really amazing package back in 2020 called simp for Python. I'm sure if there are any mathematicians in the room that know what I'm talking about. And it's basically just um a really convenient way to do tedious tasks in in mathematics. Um and it seems at least to me that chatpt and claude and everything

else already knows about senpai. And so it does that kind of stuff automatically. And so for a for a really tedious but basic you know process in in like mathematical research I find some of those AI tools to be exceedingly useful at you know shrinking down to 5 minutes of of prompt to you know from you know what would usually take me like an hour or two hours or two or something like that manually. Um, so that's that's the kind of stuff where I'm very happy to automate, right? Things where there's literally no intellectual nourishment in that performing that exercise other than maybe proving to yourself that you still

know, you know, how logarithms work or something like that. Um, but when you have a like when you're doing sort of deep mathematics and you're actually trying to drive at a specific insight or something, this is something that I think cannot be shortcut, right? You If the model, the model can only operate, I think, on things that are already known, if that makes sense. Like, it it would just be very strange, I think, if an AI came out with a truly unique, you know, um, never thought of concept and was able to generate that for you despite, you know, how good you might think you are at prompting AI. If it did, I think that that would be really great evidence of actual like, you know, uh, general intelligence or

something like that or creative intelligence. Um but I don't think AI can do that. And I think that often when um people start to become over reliant on AI, their ideas become more and more derivative, right? And simply because the AI is keeping them, you know, uh bottlenecked inside of what's already known. And I think that it was one of the, you know, one of the things I used to tell my PhD students was that really like if you there was an old graphic that someone drew in in a magazine which was like if you draw a circle that represents like all human knowledge, right? Your thesis is just this tiny little bit right on the edge that's going to like bulge that out a little bit. And I, you know, I I think that

that is something that requires, you know, creativity and um and insight um that an AI cannot yet produce. Um and so if you if you are relying on that kind of automation, you're only operating inside that circle. And so you're not doing actual, you know, it's not research in the traditional sense. You might be writing a review or re reorganizing what's already known or something like that. But unless society or civilization now knows something that it didn't before, then your research question wasn't well phrased. If that makes sense.

>> And also I think this very much connects to the conversation about rigor that we had before. So um um ultimately what you can automate is um processes or things that you trust in the sense that you can either replicate so like you can do them yourselves because you trust yourself that you can validate the solution or you can trust the person who did it uh which um again connects back to academia where you basically just reference a paper or reference methodology and say I trust this person many people trust this person because this is something that has been approved and validated and this is um ultimately what you are um allowed to do and if you are within this um framework then you can automate in between. So it's okay to use AI to find proper methodologies or it's okay to

find proper literature and to to do you know research for you in a sense of like collecting information. But then it gets back to a step where you would say now it's my responsibility as the operator of this AI to really make sure that what I am parsing now or what I'm deriving from this suggested solutions is really you know gets my stamp of approval and this is basically the difference whether your research is qualitative or not. I just quickly point out as well that there's been there's actually a really amazing parallel when any time um a system that is good at organizing information is released, there's an extreme moral panic about it, right?

When search engines were released, it's like, "Oh my god, you know, everyone's just using Wikipedia. Now they're all going to become idiots." And we're like, even spellch checkers, right? The moral panic like, "Wow, no one's going to learn how to read." Like, you know, or just normal um you know, word processing stuff. people like, "Yeah, Pete, you know, okay, handwriting probably did get really bad, but literacy has gone nothing but up." Um, I think that um when resources like Scopus in science, right, and um uh Web of Science and Reactis for Chemistry and SciFinder Scholar and all these kinds of things came out, people like, "Wow, the scholarship is going to go down the drain." But the reverse

happened, right? you know that people got really really good at finding, you know, exactly the articles that they wanted. Whereas if you had to go through an entire library shelf looking for stuff, of course you're going to miss it. I think that AI is having that same sort of impact on people now where they're like, "Wow, like, you know, people are going to get so complacent and everyone's going to be a this time in 10 years or something like that." I don't think it's true. Um I I do think it's going to automate the stuff that it can automate, but we're still in that Garpner height cycle thing that I was speaking about before. I think people have dramatically overestimated what AI is good at and have dramatically underestimated how far we are from like that trough of disillusionment is what they call it.

that it's I I I yeah, I think that we will be looking back on, you know, this era of of of AI and laughing about what people were were saying about it the same way that we laugh about those micro computers that people were going to inject into our bloodstreams or something, you know, like it's it's okay to recognize the moral panic about it. I think this the the slop um inference that people make about it is also justified but it is extremely useful in certain contexts and that the trick is finding the context where it works well >> and maybe an observation from our sourcing process uh when we reviewed the applications we received for TUR and um I mean there I would say there's there's

a crowd of maybe I don't know 500 people who are applying to every single crypto conference everywhere and think they have a good good understanding of of the level of quality uh and also the hype topics to submit and for us I guess like everywhere else we had I would say 90% let's say midlevel present everywhere uh content and for us it was not super hard to detect or to um decide and select the the topics of today. I would say due to the all the slop created it's

even easier because slop is can now be handed in by everyone right you don't even these 500 mid-level quality people are not outstanding anymore it's rather I feel the real good stuff is shining even brighter because yeah it's it's uh it differs >> all right yeah it was good closure of of the of the day I would guess. Um so yeah thanks everybody for coming also uh thanks for the speakers today for the shining out uh contributions to today um all the researcher here right if you if you have friends right we talked about the journal um so also also uh for

future editions I guess it would would be good to reach out um all token engineers uh with something to contribute to the toolbox for tomorrow's problems come to the breakout session tomorrow um yeah did I forget anything to announce Angela any last words for the day wrapping it up. >> Just just follow us on talk engineering on Twitter. Um and um stay tuned for the proceedings for the first channel and uh see you see you next time um live or um digital. Thank you so much.

>> Thank you so much. Sorry, I was