Panel Talk - Can Small Teams Still Build on Ethereum?

All right. Thank you so much everyone for joining us and special thanks to Mark, to Cheeky, Gavin here and Chris for taking the time to talk to us a little bit. uh this panel is about uh whether small teams can still build on Ethereum and I think this is quite an important question uh these days because I know we've been speaking with a lot of builders and around the challenges like I'll be super honest a lot of the challenges related to AI and as well security so I'd like to you know like maybe start the debate it's mostly going to be like a debate we don't want to go on a yes ma'am type of discussion here specifically not okay and uh yeah I think probably I'm I'm just going to like start with the first question and then feel free to take it away you guys so what changed for small teams uh building on Ethereum from you know back in the days let's say even like five years ago or more even three uh compared to nowadays

funding

let's let's Just

sure

funding I think I think it's I think the the landscape for how funding is acquired is much I think it seems to be much more difficult. Um that's just my impression. Does anyone want to debate that?

Like if you want to say a little bit more like what do you mean more difficult because I know like it's an overall but yeah let's get into it.

Yeah sure. when I first came into the space that was it was quite a long time ago the there were you know the beginning of ICOs and so you know one of the cool things that Ethereum solved was this ability or I don't know if solved is the right term but like it provided a solution for um for for crowdfunding so it meant that people could u pull their money toward something uh and then gradually over time that became like uh you know in uh regulatory risk. So a lot of teams shifted to private sales. And so then uh a whole bunch of VC capital came in and started uh betting wildly on uh just about anything. And so like you know money just poured in and that seemed like teams could raise off of you know uh very surprising um you know uh materials amount of material or yeah uh but even that now it seems like it's it's drying up to a certain extent like the bar is rise like raising but like um yeah if the bar raises then like how do teams get that funding.

But maybe we can double click on that a bit and and can I ask you I mean what would you attribute this uh drying up of funding to? I mean there are definitely multiple reasons for that right so like I mean one of them is like generally market sentiment right there is not so much um capital available or like bare market whatever right the other is um capital has shifted to other um focus points right like people don't invest in in crypto they invest in AI for example or um the the other extreme would be abundance of projects right because like there is like like the bar has risen with respect to what what you need to provide in order to get funding because of kind of like how easy it is to generate something right that was maybe previously um already um sufficient to get from it. So like do you have an opinion on that? Do you know like Yeah,

I think a lot of the I think reality has set in. Like I think that um I think that we've largely been in an extractive uh dynamic where teams uh raise a whole bunch of money and then you know the goalposts shift and they just you know they use the funding up. Uh it doesn't you know we don't have like failure conditions that were committed to in advance. So like the money there's no like credible path to recover any of the funds if the if the you know if the thesis fails if there's no traction so the money just gets used up and it leaves uh there are other ways like theft um of course but I think a lot of it is just like uh poorly managed and you know there are other elements too like just I guess gambling on tokens people aren't there to see you know some sort meaningful traction. They're just there to like try to get a little bit more money and leave.

How much time do we have on this panel?

Okay. In that case, I'm going to start with a little bit of context setting. I want to give you guys like an introduction to myself so you know where I'm coming from. So, I'm I'm Cheeky Gorilla and I work on Protocol Guild, which is a uh a nonprofit foundation that tries to support financially Ethereum's core protocol developers and researchers. Um, given that context, if you tell me that you cannot build on Ethereum as a small team, that is extremely worrying to me, right?

Like Ethereum, the doors at Ethereum are wide open for anything you want to do. Um, if you disagree, that's a fundamental problem that me in my role as being very close to Ethereum's core pro protocol development, I would love to hear more about. Now, before I was Cheeky Gorilla, I was an entrepreneur. Um and I spent seven years at a startup um which I co-founded um and when the startup was co-founded um it had five founders, five co-founders um and that was the entirety of the of the of the startup. No no no other um staff involved.

when we sold and luckily we managed to sell the startup um whatever eight years later, seven years later um there were two full-time employees in this company and so um this is my context. I believe that there is a way to bootstrap effectively. I believe that there are tools out there that can make people more efficient such that downsizing actually makes sense. I understand that downsizing can be a very uh negative thing, right? Uh it's been kind of personalized with like the chainsaw approach.

Um that's obviously not great. um the pursuit of higher profits by firing people to replace them with AI. These are all like bad things. Um but I I do believe that like the the entrepreneur entrepreneurial bone that you can potentially have. Um is a very very valuable thing, but not everyone has it, right?

And then this brings me to the funding issue. um entrepreneurship and by that I'm I'm talking about like building as a small team is not meant to be easy, right? Um there was easy money in the past and that was a net bad for Ethereum for blockchains in general. Easy money does not result in highquality projects. There might be some exceptions.

Um, but this would kind of be the stance that I take is that, you know, if you don't have a fully fleshed out business model, if you don't have a go to market, if you don't have like a product even, which unfortunately was the case in some of these blockchain pro projects, um, you deserve to fail, right? Does that mean that it's harder to build on Ethereum today? Potentially, if you think that like hard means easy access to money, but I would argue that's a healthy thing. Um, secondary to that is the fact that we are in a bare market. There's a looming AI apocalypse.

Some of these things change the variables a little bit, but purely from a funding perspective, um, I think we are not above or below parody for the rest of the world, right? Any entrepreneur that wants to build something, you you're not worse or better off in blockchain, I think the odds are more or less the same. Um, it's your job as an entrepreneur, as a small team, to be able to prove yourselves. And if you can do that, there can be a lot of success for you. If not, then maybe entrepreneurship is not for you.

So, you're saying the infinite garden needs to be weeded.

Damn. Um, the infinite garden is a beautiful metaphor. Um, but like the infinite garden also has some like dark pockets of just like extraction and casino. But that's okay. I don't I don't want to sound judgmental about that.

Um, you know, you can use Ethereum to buy heroin. you can use Ethereum to buy surprise birthday gifts for people, right? Um, so like it's it's not our decision to to to, you know, decide what people should do on Ethereum or not, right? But for you as a team who wants to build on Ethereum, Ethereum truly is an infinite garden, right? But if you want an exit, if you want to make like a lasting business out of it, um, it's not that black and white, right?

Ethereum, yes or no? There are so many other considerations that I'll stop monopolizing. So if we're speaking about funding, I think it's important to first recognize that that's a loaded term and it means different things depending on context. So and it can be a very personal definition as well. So if we think about for example the ICO paradigm which is something that um has come up in the conversation already which led to things like the you know um the IEO and IDO paradigms and so on the um let's say the the yard stick of success was always in in in some sense still today the market capitalization of the token that represents your project.

And this has created an extremely distorted view of what it means to have a successful project. Cuz by and large and still today, some of the most, you know, popular successful cryptocurrency projects are still um have never turned a a profit in their life and may never turn a profit in their life. Um but that's okay because the business model isn't to create a successful business. It's to successfully sell tokens. um and that has been the um the the winning strategy for a very long time.

Now I think that obviously this is unsustainable. Um but it's important to recognize that the people that took that kind of mentality seriously which by the way is not restricted to crypto. This is also true of things like Spotify and Tesla, right? And, you know, Uber, right? The the um sort of new understanding of what it means to finance a company is how quickly can you get to a ridiculous valuation that could never possibly be justified.

And we're also seeing this now with OpenAI and Anthropic as well, which are looking to go public this year at a $1 trillion plus valuation. Um, so this is like the era that we now live in. Um, so when we say teams, small teams have complaints or specific complaints that they don't get access to funding, there's actually two sides to this coin. One is what do they mean by ample funding, right? Because if it's that they they need to raise a hund00 million so that they can get a market cap token up to a billion dollar valuation, for example, then sure, that might be more difficult now because that well has long since been poisoned, right?

Um, but if you don't really need to raise that amount of money and you're not really wondering about how liquid your token is going to be on exchanges or, you know, centralized or decentralized or otherwise, then you probably don't really need to raise that much money. You know, you can have a small developer team um that produces a product that covers the salary of the people that are developing it. You actually have a successful business. you actually can take that coffee shop mentality of what it means to run a successful business and bring that to a blockchain and run a product that you know clears its operating costs um including the salary of the developers and the founders and things and have a totally successful business that's actually more successful than any of the projects in the top 10 on coin market cap today which is a first a really haunting realization to have um but secondly should also be I I think um encouraging that if you're not holding yourselves to these specific benchmarks of what it means to have a successful project, the challenge actually isn't so great, right? Like the to to run a flourishing business as a small team now seems much more achievable because you don't have to be, you know, of polka dot sort of uh size um to be considered successful.

Um now that separates I think um you know entrepreneurial teams into two categories because there are those that are like yeah really I really do just want to make my salary every year and work for myself in the same way that small business owners do and then there are the more let's say Silicon Valley mindset people that are like no no no I'm going to try and try again until I get a company up to a billion dollar valuation and take my exit. And if you're in that second category, the let's say honeymoon period for crypto is long since passed, right? Getting new projects up to that kind of um that kind of critical threshold that you can afford to buy a small island or something like that. That's probably not a realistic expectation to have anymore. Um, but if you're in that first category where it's like actually I do just want to work and I do like this stuff, then I'd say it's actually more achievable now than it was before because investors are smarter now.

Um, and you know, just like the, you know, degenerates that have been um, yield farming and liquidity mining and things for the last decade um, have learned not to trust these kinds of processes. The VCs have also learned not to trust it and they're looking for these kind of smaller, more sustainable business models that actually make sense. And sure, the ticket size is much smaller than it used to be, but it also doesn't need to be as high as it ever was. So, I think that there's an important distinction here. Funding does have these two um you know, contrasting definitions.

And there's a sensible one and there's an in an insane one. And the insane one, sure, that's getting more difficult, might be impossible, but the sensible one is actually getting more and more achievable every day. Yeah, thank you very much. So, so we're basically proposing a contrarian um position. We're saying can big teams build on Ethereum right now, right?

Because uh kind of it sounds like it's uh now the time to build as a small team because this is what you can do. You can do this with the low operational expenses and you can get quite far with the tools and uh the setups that you have available. um you can advance um quite sophisticated projects with the slim approach and then the question is um do you get to convince um users and investors to double down on that and and follow through with business cases. So um I think um it's it's easier now to to get from idea to prototype to kind of like try to gain traction in the space rather than um raise big and and set up uh proper operations and try to to to build something re revolutionary. So quite the opposite right um sounds to me

yeah I mean I think it's I think it's doable I think it's just that the standard has changed. I think we started out in this I mean I love the infinite garden like you know uh imagery I but I think that like if you're trying to build a product then that's sustainable and that grows I think it needs to uh rather than saying look at all of the things it could potentially do you need to find the thing you need to find that one thing and really uh show that it's it's got potential for traction um And yeah, otherwise I think yeah, the rose is off the bloom.

Yeah.

So I think for big teams um it's not that it's going to be harder for them, but there is a much more narrow corridor that they have to explore. Um and it's going to be a requirement for big teams to do this. Um especially teams that are bringing their own capital. So for example um as the stable coin space continues to grow um there is a a necessary and unavoidable component to that that you may need three or four different law firms for example representing you in different jurisdictions. Um and you need to have the authority um to you know be able to uh negotiate and sign off on um the centralized banking infrastructure required right to to support that stable coin.

So if you want to implement something like make a DAO or something like like the early version of Make a Dow where you've got um you know a simple collateralized back um algorithmic stable coin like um LUSD and and liquidity right in in Switzerland did um that's fine and a very small team can do that and you know the original make a team was quite small and liquidity is like what three people I think um but if you're doing something of the scale of tether or USDC um that needs a large organization I think um and a necessarily large bureaucracy to help secure it. Um and also to lobby governments for example to make sure that the oncoming legislation isn't at ends with the business objectives even at a sensible level. So I don't mean necessarily malicious lobbying but you know that kind of back and forth that people have with governments. So stable coin projects are by and large going to be completely annexed by extremely large organizations. Um, and we're already seeing that the stable coin, you know, um, this that whole stable coin category of crypto is largely being adopted as kind of gift cards.

And like that I choose to think of it as a kind of gift card economy. So the reason Amazon has gift cards is because it's like restricted money, right? If someone buys Amazon gift cards, the only place you can use them is in Amazon. Um, and that means that the time that they sold it to you, they can immediately start putting that into an interest bearing account and that sort of thing. Um, so it wasn't long before Tether figured out how to do that, but with entire blockchain sort of things.

And if you'd never thought of USDT as being a crypto gift card, maybe you should. Um, and the same for USDC. But you can see that now Poly Market is also doing the same thing where if you're going to be using US dollars with Poly Market, you have to buy the Poly Market stable coin to do it or a Poly Market gift card in order to use their platform. Um, and this will increasingly keep happening. Um and I think if any small team thought that they had any chance of um you know going after those kinds organizations of those size with their own stable coin paradigm I think that that's probably quite naive um simply because they haven't considered the overhead costs and things required to do that.

So there will be parts where you know large organizations are necessarily better suited to specific applications on you know on blockchains in certain um in certain primitives but you know that's I think that these are few and far between um similarly you don't see you know Google starting you know small businesses either right if you want to start something like you need a large organization to absorb you know the risk and R&D cost to make something like that available or if you want to start something like you know the Google Maps application again very difficult for a small developer team to do that but that doesn't mean that a small developer team couldn't have come up with Instagram and they did right like there it's it's really a um you know these these businesses and teams are going to specialize in the domains that they are well suited to and I think that that's okay and hopefully not controversial

on on the funding topic Um, you mentioned stable coins. I think another niche we can talk about is public goods projects. Um, and this is unfortunately this is obviously near and dear to my heart. Protocol Guild is one of the largest funer of public goods. Um, and if you started off in the Ethereum space wanted to build a public goods project, um, odds are you may have done so a few years ago because you were promised uh, sustainable funding.

And this is the one place where Ethereum has absolutely positively dropped the ball. Um because it's a bloody difficult problem that we thought we could solve and we failed to so far. Um specifically there were projects like Optimism RPGF um who promised to have like recurring rounds. You know a portion of their transactions fees would go uh consistently to public goods builders every single quarter um promising promising a lot of reliability. um that's been paused, right?

Gitcoin is a fraction of what it used to be in terms of the amount of funding that they have. Um even protocol guild, you know, the amount of funding that we distributed uh two years ago was 10x the amount that it is this year. Um and so unfortunately, Ethereum, we had such amazing ambitions when it came to public goods funding and we have failed most of those ambitions. Um, and so now is actually a very bad time to start a public goods project, which is ideal for a small team. Um, but again, and this kind of goes back to the point I made uh at the beginning, which was that that's also normal, right?

Like um the free market is alive and well in Ethereum. And there's a reason why public goods funding um is a is a such such a a long-standing issue and why governments end up funding most public goods projects, right? Um, so if you are a small team and you want to build build a public goods project, I I might discourage you to do so unless you're you really have a brilliant idea that you can uh get a ton of grants from, but otherwise um you really probably do need a business plan. I mean, look at the EF the way they're spinning everyone out and saying, "Okay, now figure out how to make money." Um, it is tough times for public goods projects specifically.

I think it's a good time for like ICOs. I think I think if it's done well, I think there are ways to do it well. I think that we need to give token holders like real uh power and to find a good way to be able to execute as like as a operations team uh while also uh ensuring that ultimately the accountability is with the token holders. U yeah ICOs anyone? I mean ICOs are the main way that that Protocol Guild makes money.

Um basically if if projects stop creating tokens then that's what we're asking them to donate, right? So uh we would love for ICOs to come back but uh there needs to be a the token needs to have a purpose, right? It has needs to have utility of some sort. I firmly believe. If you disagree with me, we can litigate that afterwards.

Um and you need to have a business model, right? you need to know you're building something of substance. That's the only caveat I would say.

Yeah, definitely. I'd have to agree with that. Either way, I mean, for a while now, we kept on saying like generally, correct me if I'm wrong, like, okay, just token launching for the sake of token launching is not a good idea. And nowadays, even more so. And just coming back to what you guys were saying regarding um being a good time for small teams to build to build on Ethereum, right?

I think I think you're right. Uh it is a it is a good time for the teams to build, but also let's keep in mind this aspect of you know like AI coming in and making it really easy for people to build. And if you keep into account the idea of fail fast, fail often. I think we're going to have like a multitude of founders that just, you know, like put out, you know, like ideas and projects out there. So how do how do we, you know, like how do we skim through that if I can call it like that, right?

fail fast and fail often is a very dangerous um this is something that like is true in most situations but in a blockchain context you really can't apply that. Um and you're right and I was just having a conversation before this panel where we were stipulating that things are about to get really really bad. Like if you think hacks in DeFi have been bad up until now, just wait for next year cuz we you we're now getting teams coming through that really don't understand blockchain at all. Um you know can barely read or write code um who are getting clawed or chat GPT to code the entire small contract base for them and using an agent to deploy it. Um, and you might think that, oh, that's better.

Like the AI is going to make less mistakes than the humans, but I promise you all of these models are trained on the existing, you know, the existing database that it has for it from which the model comes. Um, it's going to get really bad. Um, so yeah, this this move fast and break things kind of I've been critical of that mentality in DeFi for as long as I've been in the industry. I think that you should move slowly and with caution and remember I I don't come from a financial or computer background. I used to be an organic synthetic chemist.

So my training is in pharmaceuticals where you absolutely cannot move fast and break things um because you're killing people in doing that. And DeFi should take a similar mentality that you do not put the thing in a you know in a customer's hands until you're 100% convinced that they're not going to kill themselves with it. Um, and the Silicon Valley mentality is, oh, that's okay. We'll just release some and then patch it later. Um, is exactly what brought, you know, cryptocurrency to its knees.

And I remind you that we're still in the wake of um, you know, a fairly highprofile hack that affected one of the largest DeFi projects, you know, in the world. um that's now really like threatening um the the the continuation of of the DeFi paradigm, you know. Um, and it's not the first. You know, we we've seen major hacks against pretty much every um every corner of DeFi. And um the fact that AI is going to make it easier to deploy um you know, unreed code by small teams that have no idea what that code means or does is actually a terrifying prospect.

Um, and yeah, I don't mean to sort of pour cold water on it too much, but I'm much more pessimistic about the future of DeFi after AI. Like I think that we would be much better off without it. So, so is this a call for regulation or is this a call for introducing a body that sets standards and requirements or uh I mean what is the solution here because obviously I mean things get out of hand right if we just build system critical infrastructure and don't do it properly but um in the end I mean consumers or users need to be protected and the question is How do we raise the bar on those standards? Because ultimately this is what will kind of like um clear the wheat of projects that succeed or not because on in the short run you can get give them funding. can see them maybe you know prospering for one year or so what whatever like there's hype going going to build but uh um the question whether they will stay or fail or whether they can become uh actually used infrastructure is if they um if they are established based on minimal requirements or minimal standards that we all agree on um how do we introduce them I don't know um is it through funding?

Is it through VCs kind of like making the selection uh whom to give money or is it through users or is it kind of like a natural selection process where we have uh people just failing failing miserably and being hacked and kind of like uh uh clearing themselves out of the market. Um I don't see those as optimistic paths towards uh positive evolution of the space. So it's definitely something that would require, you know, a more grounded conversation on how to guide this development rather than just like let it happen over the next couple of years, I guess.

Thanks. I I guess everyone is was I don't know probably asked it by someone that uh for example, hey, my telegram was hacked. Do you know how to find support in Telegram? And no one is replying like and probably hacked account is left forever. Yeah.

So I have same feelings in when we see new projects. Uh Mark is right. I feel like the whole industry of web 2 is caring about web three because like we position themselves like oh we can get money on the launching tokens and memecoins and etc. And they actually don't feel that it's technology real use cases is like laundering money or etc. Yeah.

So my question is what if let's say ecosystem itself like Ethereum they will set up I would say some label mark on a project whereever how it related to Ethereum uh let's say like okay they have this minimum I would say regulations yeah by Ethereum not by regulations by word because still each country is fighting what is right what is not and saying okay if you actually communicate with your users you should use not but let's say signal if you actually um having a security you should have one console from aum foundation or let's say operation support by one of these providers and uh so like set up something as a minimum and if you you have label like probably quick start and fail label and second label should be like startups and product and etc. What do you think if probably ecosystem itself become those regulator and uh uh while actually trying to invest money and build u more products around maybe invest money to themsel and build requirements for products and then someone who wants to I don't know join this product or maybe interact with this institutional web to they will say okay this ecosystem is the most regulated and maybe most trusted by by us.

Yeah, I think um so certainly you know I I made the distinction between you know crypto and pharmaceutical sciences um before and certainly the the regulation is not just um you know implied. it's explicit in pharmaceuticals and also expected like consumers demand regulation right and even you know even after that regulatory process is in place they're still critical of it and I support this you know um in crypto the problem is cultural there is a sort of anarcho libertarian um mindset that has drawn a lot of people to crypto in the first place where to escape you know the regulatory scaffolding of traditional finance they found in Ethereum, you know, a specific answer to that criticism and for like for example anonymous developers um to you know bring about products that um you know that they consider to be an act of free speech, you know, and we saw the the um tornado cash sanctions really sort of drive these issues out into the into the spotlight. um and at least in the United States has been has been concluded that it is an act of free speech and you know deploying smart contracts is protected speech. Now, after that, if you've got communication software and things like, you know, Telegram and Signal and and that kind of thing, and then we've got some rubber stamping process from the Ethereum Foundation that says, you know, we agree with or you know, this project is specifically in compliance with or something like that, you're now bringing back a regulatory entity into um you know, into a process that thrived simply because there was no regulatory entities or at least that's the you know, the the vision I think that Ethereum was um presenting itself with or if not Ethereum then at least Bitcoin from which Ethereum you know um arrived. So uh I think it's not a terrible idea and there could even be an informal process where it's not necessarily a requirement but you know um there could be a sort of decentralized standard the same way that we have other standards you know in other platforms where you sort of volunteer that you are compliant with such and such a you know a protocol.

Um but I think that if we made it sort of um a hard requirement that all projects have to follow this thing to a certain tea um it would be corruptible in two different ways. One is that forcing anyone to do anything like I said is it flies in the face of the anarch libertarian mindset and then the second one is that you would immediately find companies like telegram petitioning right to be admitted to that standard. Um and this is the same problem that regulatory agencies have today where um whatever the standard is might necessarily be to the benefit of a specific company or a specific individual um and once that becomes an incentive then the regulatory um platform becomes untrustworthy. So it's um it's an unsolvable problem but it's not one that is isolated to Ethereum and blockchains and everything else. this is a saturating societal issue wherever there is a, you know, a compliance or quality assurance problem that we have to deal with and no one's figured it out yet in any of those other contexts.

So, I'm not sure we're going to figure it out in this one. I'd be interested to see how like if security firms kind of band together to like come up with some sort of agreed upon like operational standards or something like that.

Yeah. But what I see is like different expectation. No. Okay. Uh thank you.

Probably I'm too fat to stand up all the time. Um actually yeah. What what I see like different expectation. Yes. So we expect uh being involved in uh in crypto maybe one stuff uh end users like my parents etc they expect another stuff uh countries another and there is like is there any chance that we will at the same line in terms of expectation but yeah Mark is right that basically blockchain and ethereum exist because many ideas escape what from those regulators and etc.

You're right. probably u but yeah how then actually not leave this technology as just technology but make it practical uh generally and yeah I saw message recently in Twitter one girl um she was writing a message like now uh running a product goes from $100,000 just to $1. It means that you can use CI to run something. Uh but probably it's just to create but not to survive or actually make it real as a use case because if even at the beginning US co has time to build something but then you will actually fall down into operation and building new stuff that's all. So what is small projects?

I don't know. Five people. Six. One. So, we have labels that you mentioned.

I don't like labels, but I like education. And labels could be like a form of educating, right? Um I would also say that the requirement for labels for education is also a reflection of the fact that we are very early right and all of these issues that we've been describing all this doom and gloom are teething issues and teething issues are not the reason why Ethereum is going to fail or blockchain is going to fail. teething issues are to the entrepreneur, to the small team, your million or billion dollar opportunity, right? There are no standards today or at least not enough.

Um, all of the wallets that we have, almost all of the wallets that we have are garbage, right? They don't do enough to protect users, neither from themselves nor from the developers. These are all opportunities. Um, forming together, banding together, um, and trying to educate everyone is is a natural step that every single industry has to go through and we will get there. And it's it's a it's it's an important thing to do.

The Ethereum Foundation can do part of that. Other other projects, other, you know, stakeholders need to need to participate in doing this. It can all be the EF. Um but fundamentally this is also where the opportunity is for for for new projects to to to I I guess develop. Um and so you know bringing it back to the the the AI argument um this is not novel thinking.

We we all know that like AI is is right now like being used to discover um crazy vulnerabilities. Um but give it a month and the AI will be used to harden the code, right? like it it goes both ways, right? And so these tools that right now like this month has been ins sorry um this month has been insane not just for blockchain but like look at like Linux development all the kernel issues it's insane like every week I'm writing my my Windows friends which unfortunately I still have friends using Windows saying like update your goddamn computer immediately you know and like also people on like it's it's mental and it's clearly correlated with AI right but that will swing in our favor eventually right the the these small teams they will have all those tools they will have the mythos in their pocket and like whatever the the V3 version of that is. Um, and so the all these things that we're talking about that like yeah are generally negative sentiments are also massive opportunities but it's hard you know but this goes back to the fact that entrepreneurship is risky.

If you can solve a real problem you deserve reward but you need to figure out how to solve that real problem and that's hard.

That was

Yeah. Yeah. Let me just

Yeah. So just just to go just to go back to what you were saying uh regarding AI that it's you know like the problem and the solution at the same time. Uh yeah I I pretty much agree on it and to lean even more into the educational aspect of things. Uh I think there's never been a more important or a more crucial time to focus on education because nowadays I mean you know everybody like discoverability is pretty much straightforward. you have AI tools, you have everything you need in order to discover and start to build something, but then educate yourself into building correctly and adequately.

That's a bit tougher. So, of course, I'm a community builder myself and I'm obviously going to lean hard into this aspect and that's why I think communities are important in order to build education. But I think we need to do a better job in putting those educative educ educational materials out there and structuring them in uh I don't know like more clearly so that people can tap into this a bit more easier. That's that's my take on it. Sorry, Gavin.

No, don't be. I'm I'm glad I thought you were just passing it to me. Um yeah, I I think that the I mean we're a small team. Um, I, you know, we're building our MA and, you know, we're like six people. Uh, we don't get paid yet.

Uh, we'll crowdfund and then when we do get paid, we'll get paid a lot less than the industry standard. And that's okay. Uh, that's kind of what we're signing up for. We're able to do a whole lot more with a whole lot less because we have these like automation, this AI tooling. uh just being aware of what the sort of boundaries like what where there be dragons or not is an important part.

Um but yeah like I think actually everyone's going to be a small team like I I I think that it I mean having the the you know the kind of leverage that you get from this sort of tooling and as it gets better uh you know it'll you can just do a whole lot more with a lot whole lot fewer people which means I think more teams striking out to to build things that they couldn't build before. We I I honestly I mean we started uh there were some you know just context but we started like maybe four months ago and uh we're very close to having a working product and you know a crowdfund so uh that was not possible before and uh you know we've we've even had we've had our work audited we've used this automation to build uh production stuff and we've had professional audits that you you know that so far anyways seems like we're like it's working but you know there be dragons it's uh you know we I think you just have to be a little bit aware of what's uh where the dragons are

actually you touched upon a subject that it's quite interesting and partially I know like the struggles struggles and so on because I'm on your discord channel and I see like the work that you guys are doing and the heart and soul that you pour into this and it's really it's really cool to see such level of dedication but just like uh security reasons right like okay nowadays you do a lot of audit audits right some of them are quite expensive right and um obviously one of the concerns on everyone's mind I think at least or I see this a lot in like teams that build is like okay I'm doing this audit how am I able to prove like it's absolutely you know bulletproof in an area of of AI Right. This is like a massive challenge and I think it's pretty much every small team probably the bigger teams as well and also you know like smaller teams or small teams these days can be even I think it's more and more often just like one person teams and AI agents the rest you know maybe in the future we're going to have like AI agents solely you know like uh yeah it's it's evolving for sure. So how how do we tackle this? Like how can you be sure that or absolutely bulletproof that your audit is working it's fine and you're uh pretty much like proof like bulletproof to to security breaches right

I I don't think you can be I think it's a just it's a measured risk I think that you don't go you don't try to be bigger than you really are. uh people use like preser perverse in incentives to try to like you know go for the to be that billion-dollar unicorn or whatever as fast as possible. But um I think that anything that's worth growing starts small and uh just kind of knowing your place and earning your way. I think that you know there are things that you can accelerate and take shortcuts probably but some things you just can't.

Yeah, definitely the mentality of build fast and securely doesn't necessarily. No.

Yeah. Actually that there's like people like look around about like what about this competition? What about such and such project? But actually like the projects that your your so-called competitors, they're their own biggest risk. So many projects just even if like even if there wasn't any competition they would just die anyway.

So it's almost not worth paying attention in the most like for the most part. So obviously it depends on the um the specific place that your product lives in like the the specific product category that it belongs to. Um, I work in, you know, DeFi. So, when I think about exploits in DeFi, I'm thinking about all the exchange protocols and lending protocol exploits that I've seen before. And if you have, and I I disagree that you can't know, you absolutely can, right?

If you're careful enough, you can design a protocol that is unexploitable and provably unexloitable. Um, the problem is is that most developers aren't smart enough to do it. Um, and AI makes them feel smart enough, right? Um but you know if like I consider um protocol development in DeFi to be a part of a specific part of computer science called embedded systems engineering and this requires you to really understand how fixpoint arithmetic works in in something like Ethereum because you don't have access to like floatingoint libraries and that kind of thing and if you have a look at you know um let's say the bunny exploit from September of last year on unis swap v4 or the balancer exploit you know um towards the was that the end of last year um these are all effectively rounding errors in how the protocol behaves and absolutely can be designed around if you know how to design around those things. Um and once you get to a kernel um that has um the rounding driven specifically in the direction that cannot be exploited um then it's very quick to prove that that protocol cannot be exploited.

Um now you might have to make some sacrifices along the way. For example, um demonstrabably it will mean that um traders who are trading against an exchange that's using a defensive programming mandate will pay more for the tokens that they are trading or not because the gas is higher because you literally have to force the price curve to round um in the upwards direction. And so a lot of protocol designers like no no we don't want that because it's going to hurt our volume. But make no mistake, they knew that and did it anyway. And so they deliberately built the exploits into their protocols knowing that these things arrived.

And I've had those conversations with these developers. You know, I've had to, you know, mentor and consult um with new developer teams who are like, "No, no, no. We're it's not that we're taking a shortcut through stuff, but we have a specific view of how the protocol should operate." And in order to compete with, you know, such and such a protocol, whoever they thought their competitor was at the time, they deliberately took a slightly weaker um security perspective so as to get the performance out of the protocol that they wanted. Um but they knew that it was exploitable when they did this.

Um, and this is true of um, you know, I I really worry about all of the hook protocols that are being built on top of um, you know, hookup supported uh, decentralized exchange architectures that don't have um, some sort of code review component to them. So there's two different models, right? There's things like algebra where there is a specific process whereby your hook is implemented inside their protocol which means that it's like very permissioned um, if you want to deploy a specific process on top of it. And then you've got things like unis swap v4 where anyone can just launch a hook and that kind of thing. And by and large those protocols are all getting exploited.

Um because unis swap and algebra both give you a prescribed grid architecture. Um and we call these ticks usually in the in the industry and 99.9% of exploits on decentralized exchanges happen at tick boundaries. And this is where you necessarily have to make the distinction between rounding up and rounding down. And most of these protocols don't have a sense of directionality.

So it's like you know there is a prescribed numer using you know if you'll excuse the um you know the um finance jargon but the the point that I wanted to push back on is that even before AI we knew how to do this and you know I think that AI is only as good or is only as skilled as the skills of the operator who was using it. Um, and what's a shame is that as we move into this paradigm where we just assume that the AI knows everything, um, that's going to erode that skill base. Um, the anecdote that I wanted to provide just briefly is that have you ever wondered why, you know, if computers and AI and all these things are so popular and so important, why there's really only four semiconductor companies, right? And most of them don't even make the semiconductors. We're really basically just down to TSMC.

And it's because it's impossible to develop the skills now to make semiconductors if TSMC goes down, which by the way is based in Taiwan, which if you know anything about the oncoming conflict, right, is potential, you know, a potential way to completely wipe out the um, you know, the AI boom before it even gets started if China was so inclined. But if it goes down, there is no other reservoir of skills in semiconducting manufacturing that would allow that to happen. Um and that was a choice that was made early on by companies like Nvidia um that they would instead of build their own chipsets they would outsource them for TSMC manufacturing. That's effectively what's happening with the AI now where after a generation or two because people have continued to outsource development to AI tools the skills required to make this kind of determination about you know fix point errors and and how to control them will just be gone. And at that point, I will agree with you that it's now impossible because there will be no one left with the requisite skills anymore to to perform these tasks.

But at the moment, we're not there. We can actually prove it now. Maybe not for much longer.

Uh we can wrap it up, I think, right now, and just like draw final conclusions. Uh I I don't want to I can say from an educational perspective obviously I'm going to lean hard into that bit that it's definitely education and coming back to what you were saying Mark uh regarding having the skills to you know like actually do a good job obviously education and documenting stuff is very important continuing to use like AI and all the tools we have in a sensible way right and yeah that's my that's my main conclusion from this but um I don't want to monop ize. Let's just maybe do around the room and you know see uh what you guys think

just in general

uh to conclude this like can can small teams still build and what's

they can it's going to be exciting like they have to reinvent themselves but we have to push forward. I need to run. Thank you so much.

Thank you.

Yeah. I mean I said we're a team of six earlier but um you know we're a team of so many more. It's really fuzzy what the team is and that's really what we want. We want we have a really strong community uh who've been with us in a previous project and uh it's just amazing. So we're a much bigger team than than six.

Um but I think that's maybe the way to do it is if you can if you can identify who it is that you can attract, you know, people with common values and give them opportunity and yeah uh you can you can have a much bigger team not just with AI. So yeah, community focused as well.

Yeah, absolutely.

Totally important. That's down my lane. Definitely.

What about you, G?

So in Ethereum, we have this nice coping mechanism during bare markets where we say it's not a bare market, it's a builder market.

And god damn it, now is the best time to build on Ethereum because once you ship your products, we'll be in a bull market and things are going to be a lot easier. But you know as each of us have made like a separate point that's worth double clicking on like know your limits right understand if you want a mom and pop business or if you want to be like the next whatever unicorn um the approaches will be obviously very different um and for the love of God like if you're relying on an AI too much to create your business plan you you've probably already lost right like creating having a million dollar idea is worth a million dollars because they don't grow on trees, right? Ethereum is open for business for anyone, any idea, right? So, go nuts. But bear in mind that self- sovereignty is very important.

Uh, and you are telling people to trust their self- sovereignty with you. So, take that responsibility seriously because today no one's taking it seriously enough and we all need to take it more seriously. So, be thoughtful in your design. If you have a million-dollar idea, you will succeed. Point blank.

You will succeed. Um but timing is everything and now is the best timing. So build build.

Yeah, I think um yeah, I'll I'll I'll wrap up my um perspective on the on the AI thing like this. Um before the age of silicon, right? Um there were people called calculators, right? That that was actually their profession and they were very good at doing, you know, fast arithmetic in their head very accurately. things like, you know, they some of the most talented of these individuals who commanded very high salaries could also do things like square roots and things very quickly.

Um, if not in their head, then with a piece of paper, but that was their that was their career. They earned a salary crunching numbers every day. Um, and when um a $2.80, an 80 cent you know calculator piece of silicon um you know released by by Casio was available um those jobs disappeared and that entire business prospect disappeared right the companies that hired and trained calculators right were no longer profitable and the reason is is because crunching numbers became commoditized and so if you think that being able to um you know turn an idea into a program in sort of the web two paradigm um is not being commoditized by this you might have another thing coming because things like Facebook aren't really that innovative or interesting um what was difficult was that it required um a unique synthesis of skills that you know even Mark Zuckerberg by himself and he is like however you feel about Mark Zuckerberg he actually is a prodigy at what he does um and even then with his you know um you know coding polymer math skill set still required the help of a few other people you know from around the Harvard dormitories to help him realize that if you had the Facebook idea today you could move it into a website in an afternoon it's it's a fact right it's really just not that difficult I know people that um run as a coding exercise in like a boot camp converting all of Facebook to Rust for example using AI tools or something like The the problem here is that Facebook's success is no longer contingent on the skills of the people that produced it. And that means that at that time, if thousands and thousands of people all had the Facebook idea, they all would have tried to realize it at the same time.

Um, and it's not clear whether or not Facebook would have flourished in that kind of environment. Um similarly you know the the DEAX paradigm from 2017 from which Bangor was born you know that's no longer that difficult to do with with an AI system even things like AB and make a DAO and you know hyperlquid right these are all relatively easy to implement and so your your million-dollar idea needs to be like in order to be evaluated that way needs to be so out of left field right so unique so undiscoverable that only you have it, right? And I think if you're going for the, you know, the unicorn style um evaluation, that's the mindset that you need to have. It needs to be so non-obvious because all of the obvious stuff is now just going to get churned out, right? Like, you know, plastic Easter eggs out of Chinese manufacturing companies.

Yeah.

Yeah, definitely. And to emphasize on that, I think creativity, you know, giving like the era of AI is upon us. Uh creativity will be probably paramount on building anything. So the more creative you are, the more you're going to actually, you know, like nail it as a small team, as an individual, or maybe even as a bigger company. Right.

Right. Thank you so much, guys. Hopefully I've done a fairly decent job.