Tokencraft with Eddy Lazzarin

[Music] I'm really excited for this talk but I want to start with a quick question Who here May issue a token at some point okay that's a lot of people so if half of if at least half of that group does not ask me questions you're insane like you need to ask stuff there's so much in this talk that I have omitted or glossed over or simplified because there's tons to talk about regarding tokens they're Complicated I talk about them basically every day with our entire portfolio I've been thinking about them every day for years

at this point and I have no idea what I'm saying so if you if I don't you need to have questions and typically when I have this type of conversation with our portfolio companies it's in the form of a conversation a dialogue where they have some thoughts they have some opinions things about their protocol things about their product and those Things help inform what I can say back and forth so I have a bunch to share I'm going to stop at a bunch of different points and say are there questions what do you think if

you want interrupt at any point there's mics going around and they're going to Jump Right In really quick so let's make this as conversational as possible I honestly I think it has to be to be as fruitful as it can be so first what are we talking about we're talking about tokens a Specific type of token which I'll get to first I'm not a lawyer so obviously none of this is legal advice this is just my thoughts as an engineer and as someone who's seen this a bunch of times since a token is just a

balance in a crypto program that's really all it really is there's many ways to do and to think about tokens many that haven't even been invented yet I hope you'll invent a few of them and but for this talk we're going to focus on just a few So first let's let's start with meme coins since they're back in Vogue everybody's talking about them again it's a little funny because memecoins have been in crypto since the very beginning in fact even before there was ethereum there were memecoins and the way people would do them they'd actually

take the original Bitcoin code base Fork it make a small change give it a name and put it out there and if you go and look at like Wayback machine and Archive.org and look at uh you know captures of coin market cap from 2013 you'll see the top 100 littered with them right they're not a new idea that being said I don't think I I hate to say I don't think they're particularly interesting I hope that nobody here makes one and we'll get into and we'll get into specific reasons why the quick way to think

about what a memec coin is I I I I think there's so much more to say but they explicitly have no purpose And that's actually a critical feature they have to have no purpose because if they have more of a purpose well then it's something more interesting than a mcoin already but also you start to get into protocol design you start to get into legal questions the fact that it has no purpose is what makes it a purely speculative vehicle and it being a purely speculative thing is what makes it just gambling which makes it

legal because gambling is legal right so those Those things have to kind of hold together obviously because it's purely speculative there's extreme price volatility which is kind of the feature of the thing it's not a it's not a bug and and we're saying I I'll mention later in this talk a mem coin example for a specific purpose but mem coins can be decentralized but they could also be a scam right that's when I say a scam I mean there can be deep a symmetries of information in the meme coin right where Some people have kind

of arranged in advance they're going to dump like this or share the information like this or you know allow uh maybe like there's hidden bugs in the code it looks like a regular meme coin but there's actually a back door or a mint or something a mint vulnerability something like that in the ideal case you know you could mint the meme coin and distribute it across a whole bunch of people and you could say it's decentralized and I'll think about It more in that setting the other setting is really more of a scam but anyway

not going to say too much more about that a second key token category that we're also not going to talk about is stable coins right maybe some of you will use them in your protocols I hope you do because I think they have a lot of promise and I think they're critical for mainstream adoption uh they've been taking off as you can see uh chiefly as collateral on chain uh but uh and not so Much for payment but I think that that's likely to change even though that's been something people have been saying in crypto

for a long time hasn't really happened but we can talk more about that later a third category that people don't really think about as much it has a bunch of bunch of names I like to call them arcade tokens some people call them soft tokens sometimes they're the second token in a two token model this is a token that's on chain and transferable But it's typically price dampened in some way usually as a result of the issuance and Redemption mechanism so it could have an unlimited functionally unlimited Supply think of it like an arcade where

you know you can get little tokens in an arcade to play arcade games and they have a monetary value you had to pay for them and they have a utility at the arcade but it wouldn't really be reasonable to buy them uh to expect price appreciation because you know that The way they're being issued is very loose and very easy and you know they have limited usage in terms of redeeming from maybe a single counterparty the final category which is the most interesting category and the category the rest of this talk is about for some

reason people just call token but this is Network token you could call it a protocol token right this is the token that represents a decentralized protocol right it is the core the it's deeply Integrated with it in some way potentially has high volatility but it's part of an economic model and I'm going to talk a little bit more about that economic model in Big Picture examples you you know the examples because they're the biggest crypto tokens ethereum maker Unis swap uni compounds these are all classic examp more developed examples that I'll refer to in the

talk if you don't know how they work you you should and we'll we'll talk more About that a critical thing is that these tokens represent a larger system you can't just make them in isolation their issuance and Redemption and creation and destruction and their value capture and their controls have to be a part of a bigger thing you don't really design a token you design a protocol and having balances and assets is part of the way protocol work and this is this is an example of that so any are there any questions about this before

we we go On yeah would you consider something like fly a token no yeah yeah eventually maybe right now because it's offchain and because there no transferability it's not yet that's just a point you know it's uh a balance in a database but eventually if it was transferable and if it was on chain and it was part of a larger system absolutely like I and now it may uh you know I'm although I do talk with the Blackbird team a bunch and I think that're you know they're they're Thinking about it really deeply it could

also end up being in the category of arcade tokens right it could actually be a great example of that where you go to restaurants you could imagine paying with fly and you imagine they can do extra work to ensure that the fly value Redemption value at a restaurant is more or less stable and people can kind of take it for granted and reason about it but that it could still have transferability and there's a known Onchain process or at least Grand economic process for how it's issued and how it's redeemed so it could be in

the arcade token I guess one followup is like they released as points centralized not on Shan yes is this something you're seeing more teams doing yes and I think it's a good idea with some caveats I'll say more about that later I may refer to this a little bit in in conversations like uh you know Hillmar knows this so well in fact one of the The coolest slides I've seen was uh I saw like the Quant aggregations of the faucets and sinks for some T really incredibly cool slide but this is just a very simple

mental model I think everybody should have in their mind when they think about the token it's the faucet here is your ability to print the token when you make a new token there is a single power that you have this is probably the most interesting power single power that a token has Beyond Programmability is you can determine the conditions of its issuance that's an unbelievably power powerful ability right because it means like take ethereum as a simple example it means that you can always guarantee that validators will be paid because there's no fixed pool we're

pulling from we will print the tokens they will get paid right now that's that doesn't mean that you should have infinite issuance and massive massive Inflation necessarily it's just we should we shouldn't take for granted the power of this faucet and I'll say more about what I think are good things for this faucet to come from but when you print more tokens and you increase the floating token Supply great hopefully the protocol is paying for things that are very very valuable and will expand the supply and quality of the goods that the protocol is supposed

to make available but you should also keep in Mind the sync and these are the things that eventually destroy or retire or lock up or remove from circulation the tokens that we're creating and the reason why you need that balance is because you want to preserve ve the price of the token so that in the future when it pays for things it's paying for them with a valuable asset right it's it's about creating sustainability in the end and creating a degree of equilibrium that's hard to make that Perfect there's tons of caveats I have to

say here but just always keep that in mind when you're designing the system that your fundamental power is issuance but you need to always take into consideration the sync and the sync should be tied to the good that is being created by the system like the thing that is fundamentally valuable by the system another quick thing I'll say here is it's it's not uncommon to think of a protocol as a Marketplace like think of It as having a supply and a demand side so there's the faucet there's the snc and then there's also the supply

that's the parties or mechanism that is creating the thing that is of that is a good that is that people want and then there's the demand side of the market which is where people pay to consume that so the pattern match match this against eth very quickly eth we pay the security budget to validators that's the inflationary Reward the sync is when people pay for compute a portion of that is paid to validators a portion of that is destroyed and the EIP 1559 burn uh the so the demand side is demand for compute and storage

and verifiable computation the supply side is network capacity block space that type of thing and do you have any sort of mental models to kind of reason with you could say the inflationary and deflationary of Economics of Supply [Music] or uh the inflationary and deflationary sort of Economics of the kind of uh Supply and how that impacts the protocol do you have any sort of mental models there or any ideas around there so making a system that can achieve a perfect equilibrium is incredibly difficult so I wouldn't Endeavor to do that especially from the start

I don't think equilibria are like real things in nature that are sustained always I would Say that uh I tend to think having a degree of inflation is good as opposed to a strictly fixed Supply because that inflation means you can make a guaranteed payment for something and another quick one is the fee where you extract from the system should be at the point that the valuable thing is consumed right so don't try to put fees or extract value from the system on like a weird part of the supply side or some Like random like

Corner over here take it when people actually consume the valuable thing so for UNIS swap that's they're consuming liquidity right they're actually paying for a trade that's like the the good for maker it's paying for the leverage that they are consuming the risk that they're creating in the system they pay for that uh for ethereum it's block space computation uh those are very valuable things so force people to pay for the Valuable thing is does that sort of answer your question sort of um yeah I was also kind of uh sort of alluding to you

could say you know back in the day used to have things like bonding curves and stuff like that that kind of played around with it so so do you have any sort of uh um suggestions there like whether to use a bonding curve not not per se but mechanisms of that sort yeah yeah there's a whole toolkit of them I mean like bonding curves are good when You want to create a market where there's very limited liquidity and you're okay with the price schedule being some sort of fixed curve that might not actually reflect market

demand right like we can talk about those maybe maybe we we'll spend some time on it yes when you're measuring the impact of syns and sync design hey thank you um there are obviously the official syncs that you just discussed and obviously in like ethereum you can have Unofficial syncs contracts that send to dead or those things how do you measure that ecosystem wise very hard to measure I think of those as like explicit and implicit right explicit is like you know you build the fee in it's paid it's burned whatever implicit is people want

to float some people trust it they want to hold it uh you know they uh they use it as collateral in some other thing right there's all kinds of syns like that I'd Be weary about designing a protocol where you rely on the implicit syns just because they're harder to quantify of course they can be very powerful but let them appear in the backdrop right try to measure them but like let them appear in the back drop the system should be hold together without you relying on those implicit syns those can be very dangerous and

some things that seem like syncs are not people will often say like I'm going to say in a few slides like For payments oh people use it to pay then it's a sync because they have a little bit in their wallet maybe but how do you measure that should your system rely on this you know also it's a liability in a way because that sync could be depleted like it could back up immediately right it's not real Double Entry bookkeeping at that point yeah yeah exactly so here's some good and bad reasons so these are

really quick Cuts I could talk about these Forever but just okay so here here's one incentives this one drives me drives me nuts sometime sometimes people will come up and say Eddie I want to make a token uh you know I'm really thinking hard about it I like love to make a token and then I say okay great what does it do and they're like well it's gonna pay for this that is terrible that is not a real that that's just printing money and then like throwing it somewhere what you really it's it's so easy

to get caught Up in the idea that you can make a token you can pay for a thing it's like having an infinite ad budget that you didn't pay for this is not a real protocol design and worse it will create huge issues for your reasoning about the system right your goal as an entrepreneur is to try to find product Market fit you're making a product and you want to ensure that it's a thing people actually care about that they actually want to use that they're Willing to pay for because you're creating more value than

you took as inputs if you just spend it on the wrong incentive then you can end up distorting your ability to read the product Market fit it's like uh if you if someone says yeah look I have like five million in revenues you know I'm feeling really great about things and you check underneath the hood and then their ad budget was 20 million right that would you would just know that there's no There's nothing real underneath there because they're paying for it now thankfully in that case in a traditional company it would be on the

balance sheet so you could see but in a in uh crypto if the token exists somewhere else it can be very very hard to measure you'd be surprised how little work people do to try to quantify how much they're actually spending on customer acquisition costs indirectly through some sort of token incentive plan so I'd Say it's not that you shouldn't use the token to pay for things you should use it and a good protocol will use a token to pay for the right things but if you just think of it as a way to blast

money in a specific Direction you are not going to learn the most things and you're not going to create something of longlasting value that outlasts a single hype cycle so there's so much more to say there but instead what you want to do is You want to pay for Network effects you want to grow the network effects of the system right now it's hard to make sure that what you're paying for is valuable that the supply that you are hopefully bootstrapping the development of is real Supply that you need more of and that is good

and works but there are ways to do that you know the first uh yield Farm as far as I know was compound back in 2020 and uh they would pay for Supply by just giving people a little bit of token when They deposited one of several collateral types into the protocol that's a pretty simple one to verify because you know what the collateral is and you've handpicked what collateral types are being used so you can know that it's useful collateral and has value and you can pay for it accordingly and that's all the system needs

but if you think about your protocol what are the what are the things that are critical to developing a network effect what is the Supply side what is the demand side and how can we boost high quality Supply right and as I said before subsidizing demand can make things a little bit more complicated so be very very very careful with that right you what you really want is uh the things that lead to the accumulation of a network effect another thing you want to pay for which doesn't apply to all protocols but it applies to

a lot more than people think is security you want to pay for security now when I Say security I mean in a very very general sense I actually thought about trying to pick a different word for this but it's hard ethereum proof of stake pays for security right there's a token issuance that pays validators to do the work to make sure that the block space Supply is high quality that is what is happening they are all making sure that the block space is good that the compute in the block space was paid for is correct

and Is replicated across the system and verified that's at the infrastructure layer but you could also think about this at the application layer this is what's happening with compound and with maker when they pay regularly for security in the case of compound what security is uh for those for those who don't know because it's everything is layered in into a single pool there is always a risk that a collateral asset that is dangerous or Low quality gets added as a collateral asset to the pool and that's very very risky because if a garbage collateral is

added someone could deposit a bunch of garbage and take all the things that are valuable so the protocol is specifically designed to protect the deposits in the pool that's what it's doing at all times the compound token holders are voting carefully on the admission of new collateral assets on the debt ceilings of collateral assets and their goal is To ensure that suppliers never lose their collateral uh or the deposits and that uh borrowers pay their fees and pay their debt so it's making sure that the thing is secure and it does exactly what it promises

that the ability to deposit and to get leverage out of the system is very very very secured and that it behaves as guaranteed this is a kind of security and the token exists to ensure that that security is delivered the same is true with maker and I think I was Talking a little bit about story protocol earlier without going into the details of of of how it might work if you had a network where there are various intellectual properties that you could commercialize or make a derivative works from you could imagine a very critical concept

of security would be that the intellectual property that I'm commercializing is a legitimate intellectual property that's a very hard thing to measure in the real world you Have to do a lot of legal work and a lot of diligence to ensure that that is a real legitimate intellectual property and the person who says they own it actually own it but if you could take that for granted if the protocol ensured that this degree of legitimacy was the was true for all the assets in this network then that would be a very high quality product and

I could say that about any application in crypto the product is guarantees verification Credible neutrality and if you can pay for that with your token then that is a high very high likelihood of being a longlasting and valuable protocol I can say a lot more I know there a lot but are there any questions about that or any thoughts yeah let's like dream up what your like protocols like perfect world and all the things that your protocol needs and then you know budget it out in your like Supply and then try to go pay for

those things and you know That's your budget for uh all the hard things your protocol needs yeah but I'm saying like what is your protocol selling that's like the question like what is your protocol selling what is the thing when someone comes up to it they're going to say but is this is this is this good like is this true is this high quality and pay to ensure that it's high quality always yeah yeah and that it's ver and high quality is my colloquial way of saying verify Verification like verification is a boring word but

it's that's what's so exciting yeah okay thank you was there another question yeah do you think security the way you define it is inherent in the value of network tokens say more um the way you're defining these Network tokens specifically it kind of seems like security is actually a key piece like you you can't have a network token where You're not paying for security you can't so I'll give give another example my next example doesn't have it but the the the point is that the Network token should pay for security the network token's value comes

when people pay for that thing that is secured they're consuming it the token should capture some of the value from that to sustain its own price so it can keep the system going it's all about making the thing perpetually sustainable and then also There some systems create risk like with compound with maker they they have systemic risks and those risks instead of being spread across the users of the protocol can be internalized into the token holders so they absorb the risks and the nice thing about pushing risks in there is you can pay people to

do it right people should people should be paid to take risk and when they take that risk and they can do work to mitigate it and ensure that it continues To work now some systems don't require paying for security continuously right so for example immutable systems can have very very low security costs so take UNIS is an example I hope like people have been thinking a little bit wait a sec this is this describes some of these but this doesn't describe Unis swap right it doesn't uniswap is totally immutable each version is separate it does

not have an ongoing security budget and it doesn't need that and that's Because just a property of Unis swaps design is that it externalizes security concerns each pool each uniswap pool is separate and isolated and the trader who goes and swaps something in the pool is the one who accepts the risk of the pool right and because of that you don't need to pay continuously for securing there's no systemic risk that needs to be carefully monitored and paid attention to but not all protocols can be designed such that they're Immutable for example you could not

do do compound without with immutability because it needs to be able to decide which collateral assets are safe because of the way that uh liquidity is shared across assets and so if you can make the system immutable then you save a significant cost and you may also be able to minimize or eliminate governance altogether and if that's the case I think that's a great idea so here's Another bad reason people make tokens they say it's for voting this is like a governance token who wants to vote like what is the point of that like I

I I like to me voting it's not that you shouldn't have voting in the system voting is a last resort this is true politically in a way as well right not to get too crazy but voting is a last resort in other words we don't have a way for society to organize itself Properly at the appropriate organizational level so we need to do the most painful conceivable thing that is expensive difficult to analyze low signal like it's it's it's very very voting is very very very cumbersome we go and we ask everybody to weigh into

a potentially incredibly nuanced complex multi-dimensional question you should only do that if you really really really need to and the evidence so far is that token governance has relatively low Participation it has high opportunity costs uh it's difficult to motivate experts to do the research it's not that it's totally dysfunctional it's just that it's a cumbersome slow thing and you should apply it only when necessary as a last resort and everything that can be automated and built into the system or delegated to a market mechanism or something like that ought to be and and worst

of all if if that's the story for why it's valuable like you're Telling people well why would people want this token oh they want to vote well you're just deluding yourself no one wants to vote the Voting is part of the responsibility of being a token holder what you really want is you want to align the network you want all the stakeholders of the system to be incentivized and to think toward the same goal there's a ton to say here I think this is an incredibly powerful property that tokens have it's very very Difficult to

achieve with Equity but uh the fundamental idea is that you have adjacencies like these are Goods that are next to your product that are important for the functioning of of of this protocol you have suppliers you have users you have uh technical Partners all these types of people you can make deals with them but you know what's even better than making a small simple cash deal is they have tokens and now they want the protocol to succeed Too and you can create links between all these different adjacencies that otherwise might be competing and the thing

together is now much much much more potent than it would have been otherwise right there's all kinds of tools you want to deploy in this setting vesting Milestones locking I'll say more later but you can imagine some of these me methods being automatic and some of them being manual like in the early days manual being like BD Deals Partnerships conditional token swaps made out of your foundation things like that they can be automatic like people deposit collateral they lock it they have to promise to do some verifiable work they get paid if that works done

they're the what they earned unlocks over time Etc we can be very creative with this I think people really underestimate how potent this is and the effects we see because a lot of the most pronounced effects are subtle right They're just people agreeing with each other or people not competing or people sharing open source code or people promoting each other's products on social media right which is like a very important part of of how crypto has developed so happy to say more about this this one's kind of on the fence for me and I just

mentioned it because it it comes up a lot people bring it up a lot is is payments like what my token is for is for payments remember I'm talking About Network tokens I'm not talking about Arcade tokens I'm not talking about stable coins right I'm specifically talking about the protocol token the evidence so far is that people don't want to pay with protocol tokens in fact there's friction in doing so they have to get a balance of it at least the way networks work today they have to secure some they also have to reason about

the risks of holding a potentially volatile asset people don't Like paying with a volatile asset even if they think that it's if they think it's valuable they may not want to want to sell it and if they think it's not valuable they probably don't want to hold it and have a whole stock of it available for payment when they need to it's not that I think it's guaranteed not going to work I think that it could just we need a lot of interesting evolution in how to think about these tokens for it to work um

so it's a Little fuzzy and then like last point is that if a payment is not a clean sync now taking a fee out of a payment can be a sync right but the payment itself requiring people use the token for payment is not a direct sync so happy to say more about that any questions so far about some of these high level good and bad reasons yeah thanks uh how do you see the potential of governance to actually help with security so like in in in the case of um Compound that was kind of

clear right um so so that's not necessarily like a a bad thing to have governance if you know that decision like the decision how to secure the network isn't can cannot be automated but has to have maybe a human in the loop like okay we're going to decide that this token pair is um safe enough for us to put us into the collateral pool so it's it's true obviously that's what's happening the token is for voting for Security I think it can achieve that purpose and there's evidence that it can work right maker does that

compound does that a does that Bunches of def5 protocols do that there's also security councils literally right where there are people with a special power to freeze the system in the case of a exploit or something like that so it can definitely be the case it's just very sensitive you're introducing a point of failure you're introducing a political risk Right there's you have to really think about how you're designing that uh compound and a and Moro have all thought all lending protocols uh from a you know with an overall similar purpose they have spent a

ton of energy thinking about how to get experts to vet their protocols with every change they to pay for it they have to ensure that there's at least some number of experts that they weigh in publicly that they're constantly being evaluated that their Token holders are evaluating them I'm not saying it's impossible it's just this is a pain can you do without it can you minimize it right the to the extent to which you can minimize it you're making the system more robust right that's that's that's that's the frame and of course um I don't

think it can always be removed in all cases but yeah does that answer your question yeah no totally totally yeah so sorry if I'm oversimplifying it but like uh it sounds Like a lot of it is like bad incentives bad good incentives good right like um so like yeah how do you turn that from something bad money bad good money good yeah I but how do you turn it from something that's kind of more subjective to something that's more objective right because if you're saying hey like maybe I'm spending 20 million on marketing today but

I'm only bringing five in then that might grow over time like is is there any way you suggest to like kind Of measure that more objectively yeah so it it depends on the thing so like in the marketing example if you're spending 20 million on marketing for 5 million in revenues then the question is what's my CAC LTV ratio for example right like I'm paying for customers and then I'm getting this many revenues from customers if that ratio is underwater in other words you're paying more than you're receiving then you know something's broken with your

cxs and you Need to lower your customer acquisition cost that's like the sort of basic like marketing Playbook now in protocol design you have to think about what you're what you're paying for so I'll give you a very crude example from like earlier in crypto history that people may or may not know I don't know if anybody remembers steam stem from like a very early I love it yeah so that that was around that was kind of an early social Network it wasn't decentralized the social network but what was the the kind of critical conceit

was that people were paid for comments so like you made a comment and then you would get like paid in some token automatically for that comment and uh you know if people liked your comment you would get more payment that's it it was just that simple so if you're thinking like wow that seems really broken and I could easily exploit That you're right right and uh and what ended up happening is that there was a huge amount of comments the system was constantly going down because it was flooded with comments and the comments were garbage

because what you know you're not we're paying for comments not good ones right so in that system the issue would be that the thing that you're incentivizing is not the thing that you wanted right so that's why a critical Piece is verifying the quality of the supply that sounds like oh maybe that sounds easy maybe that sounds hard it's hard it's incredibly important and if you can do that then you can feel safer about subsidizing it right so to answer your question is like how do you get concrete on like what you're incentivizing incentivize things

that are you can verify to some extent AO as automatically as possible their quality And that those things directly lead to highquality network effects as opposed to vanity metrics like a bad thing to incentivize people making random ass transactions on my network that that doesn't help anybody there's no point to that right what's a nice thing to incentivize high quality supply of uh you know high quality conversations that's harder to measure high quality Capital deposits that's it's easier to measure Right does that kind of answer your question yeah so like just the last basically combining

simple marketing okay uh simple marketing metrics with what you're trying to achieve with your actual product yeah but it's but yes but also it's not just marketing right it's security it's liquidity what whatever the good is that's being created by our system so it may may or may not be marketing yeah thank you I'm curious uh if you Think it's a like should you design a protocol around the incentive or like the thing that you want to pay for and uh like how do you think about designing around that incentive changing over time um in

terms of the application and the use of it and like what you're trying to incentivize the incentive changing over time or the thing that you are wanting the thing you want to grow more of changing over time exactly yeah and like just how you think about yeah uh yeah You got it yeah yeah so this is why I really like manual programs as well like doing an automatic program is critical if you want it to be truly decentralized and sustainable because the go the goal as I'll say more about in the path is decentralization part

is you want to be able to walk away from the system and it still works that's like a great little test so obviously that needs automatic payments for to sustain the thing but in the earliest stages you should Aggressively use Manual allocations of tokens right to bring on the right Partners the right contributors whatever and the things you incentivize in that case ought to change over time because you're learning more the point is you're trying to find product Market fit you're getting signal from the market continuously you're learning about what's working and you're you're realizing

okay this is a new bottleneck this thing solved here's my new issue This thing solved here's my new issue you should definitely be ad adapting what you're incentivizing to the extent that you can verify it as well yeah yeah uh one thing that we haven't touched on is Dows um and so the Playbook we see a lot in our Niche is people use a token as a money printer and then justify that as voting so they basically do all the things that said not to do exactly um and the Voting is for part of a

dow what do you vote on what can it do very Dubious can Dows be done correctly like do you see okay yeah yeah they they they can be I mean how much this say now I mean I guess um well one thing that's helpful like when do do correctly there's like a couple parts of do correctly one is legalistically yeah if you haven't looked into it or learned about it yet I I think it's going to be in the curriculum is our Duna framework that that that we're recommending people take With Dows this is like

a legal entity in the US you you don't have to be in the US to take advantage of it that uh can serve as a very much more appropriate alternative to a corporation right for Dow members and what it does is it allows them to vote and to have like Express opinions that are legalistically binding and real you know real and protect them from liability in such votes and do things like pay taxes and be compliant with the law without There being like managerial controllers of the system or being like a formal Corporation where people

are hired and fired and so on so first that that that helps a lot with that leg of doing a dow a lot of the Dows that are trying to do their thing are not doing it right because they're completely avoiding that angle and taking a lot of unnecessary risk the other side is what is the purpose of the thing well like I said just paying for Stuff and then voting is like ridiculous that's like not that's not a thing at all in my opinion a DA should be a group of stakeholders of varied interests

who represent who who are the represent and humans associated with a protocol right so the story is what's the protocol do right right for UNIS swap which is simpler because it's not a it doesn't have to pay Ono for security but it does collect fees these are the people who hold tokens use Unis swap build apps on Unis swap are all around its growing ecosystem and they have shared interests that they would like the protocol and the foundation to represent and to carry out and that makes a ton of sense to me as a dow

now whether they're doing it exactly perfectly right or not like it's hard to say but they should be able to allocate from their treasury they should be able to support specific programs they should be able to do things like uh make specific protocol Upgrades and so on and so forth right that makes sense and I guess in terms of is there a way to defend the da from voice influence being tied solely to how early you were is that something you should try to avoid against oh yeah that's a that's a great that's a great

and more subtle question I think naive I I had like one line in here about about that which is like naive token governance like simple token governance yeah it's just like it's not that great Just like okay you have more you have these tokens you vote now okay there's just like there's way more we can do that's much more interesting you can have like different houses of votes you can have automatic committees you can have you can put civil resistance mechanisms and then do like right where people are randomly select I mean there's so many

creative interesting things I could talk all day about ways to appropriately I think just Taking how much token people have is there's advantages to it obviously you want I mean I'm not like some radical where I think that that's bad like people who have more stake in the game should potentially have more voice right but the goal is to try to get high quality signal about the entire ecosystem and that's a complex and very subtle endeavor yep so yeah thank you yeah sure I'm not sure if I can express this properly so I'm going to

give it my Shot right we were talking about um you know the value of tokens um are what is securing the network uh the protocol in this case so in a case like maker D where they have two separate tokens one for utility and one for governance like they're they're obviously decoupled right but there's also the idea of hey you could is there a need to have those be decoupled in the first place Place could those uh could the people who are staking just use the Native currency themselves to stake rather than have two separate

things well in maker there's the there's two tokens because one is the product and one is the network token right right and I would say it's not a utility token I would say it's a stable coin well if you're a user of it it's a stable coin if you produced it because you made took leverage out of the system which is one of the products then it's a debt coupon that you need to purchase back to repay And reclaim your collateral right so it's kind of this sort of Dual Purpose token sort of there's subtleties

in it the the ultimately the idea and this is a really nice product design decision that I think was great from the maker team is that you could imagine they have like the die holders vote you know you could imagine the die holders have some responsibilities but I think what they did was they said let's simplify the social contract for the die holders they Can assume that the die will be stable that's it what are you getting when you get die you get a stable token that is worth a dollar each and that's all you

need to think about and if there's a problem that's maker token holders fault they're the ones who will get burned if something goes wrong you don't have to worry about it right and I think that's really important because if you're just using it as a stable coin you don't want to think like oh it's a stable coin Except for these complex caveats where I have to like track some obscure like debt rate or what you know insane things so that's more of a product design decision I think and I think is was a good one

cool so a little bit about the path to decentralization this this little chart okay the vertical axis is decentralization with more decentralization at the bottom and the horizontal axis is protocol functionality you could you know with More on the right so our goal of course we want to get to the bottom right corner why do we want to get to we want to be really decentralized and we want a protocol that really like does its job like it it has rich and high quality functionality and now should I motivate more like why we want to

be decentralized is that like I'm feeling like maybe I should should I is it obvious why like who do raise your hand if it's really obvious why we want to be Decentralized wow okay 60% 70% okay okay okay okay well so well okay then I'll say a little bit about why we want to be decentralized decentralization is part of the product I know people say oh people don't care as much about decentralization totally I totally get it I totally get it but that's end users don't care as much about decentralization this is like end users

don't care about the specific idiosyncratic reasons for monetary Policy with their dollars of course they don't care about that why would they care about that they do care though that their dollars hold value they care that their currency doesn't inflate away they care about you know that their bank doesn't steal their money like all these things they people actually care about a lot of these things things and they make those choices and those choices are reflected in their Market choices like what they choose to use and what they Choose not to use people like decentralization

because they like there to be a lot of high quality products and if the thing that you you you make is very decentralized many people will build on it because it's the most solid foundation I can talk more about this at length I have like another talk about this but I'm presuming that decentralization is a critical goal and if you want to issue a network token it needs to be decentralized that's the Goal we want to we want it to be decentralized and we want robust product fun protocol functionality so our recommendation and this is

kind of you've heard about Progressive decentralization is that you start in this corner you start at the top right corner so this is the centralized but functional corner now it's not safe to release a token yet so I'm going to say so what what we want to do is we started our Good start and then the things on the left these are like not an exhaustive set but like examples of things that as they happen we are probably becoming more decentralized right we have diverse stakeholders we have more economic activity from more people there's alternative

clients the system's permissionless we've minimized single points of failure like these are a constellation of properties that are starting to describe a decentralized Thing and as we become more decentralized then we can get closer and closer to having a transferable token so at the beginning when we're centralized if you release a transferable token you're taking extreme legal risk extreme legal risk and but as it becomes more and more decentralized those risks diminish slowly I'll have a little graph about it in a second and you can do things like points points are points are great why

are points great because you Can incentivize things without risking people being burned when I say being burned I mean someone goes in they buy a ton of a token the price tanks because you tweeted the wrong thing now they're going to sue you right they're going to and the SEC is going to have a lot to say and uh that's obviously a scenario we want to avoid so the way we can avoid that well we make the token non-transferable now people can earn a thing yeah doesn't have a market cap They can't reason about how

much exactly it's worth yet but they know they're accumulating something and people actually do respond there's significant empirical evidence that people do respond to these types of incentives once becomes more decentralized maybe there's a point where we can make it transferable maybe we have to keep the team and the foundation and you know investors locked up we probably should long lockups are Good and uh but then we can allow people who received it in an airdrop or people who received it for doing some good work it can become transferable and a market price can form

yeah what about a test so likeing all of these like full functionality with transferability and so on you know claiming it's a test net um and then like once you're ready like okay now we're doing the same thing and um kind of transferring the balances onto the Main net yeah well if you're transferring exactly the same balance then was it just main net all along probably yeah don't be don't be too cute right the the the cuter you are like the more risk you're taking in a way right because if if if it if it

was secretly main net then people can still be burned markets can still form the this is like a very subjective and hard thing to convey um and remember I'm not a lawyer but a lot of the what it comes Down to is like how does it look I don't mean that to be glib right I I mean like really like were you facilitating the formation of a very dangerous high-risk Market where you promised people money and you took their money and they lost it if that's what happened they're going to crush you it doesn't matter

how cute you were being about no but I did said them legal magic words at one point it doesn't matter if you really were trying to do the right thing and you're Earnestly progressing in the correct direction and you're making that clear and you're being honest and you're not being crazy and you know telling people nuts things and the facts basically show like look business is hard making a making a business that works is hard enough making a crypto protocol that's decentralized and works is even harder then the facts will paint that and even if

you were in some gray areas at some points as long as the intentions Were extremely clearly good and you went out of your way to try to avoid harming people then you're probably in a safer spot and and I'm not suggesting like the laws don't matter I'm just saying like I'm actually saying the spirit of the laws is very important so like I guess what you see in um networks launching today is that you have test net where like everything is functional and andoss and transferable yeah um and then you you might not have you

know like Onetoone transfer of the balances to the main net but you do have airdrops which are you know highly correlated with what happened on the test Nets right so it's it's like an indirect way if you're in if the thing that people have to do is like real work to earn that air drop in the future and you're saying like look here's your points you did some real work well that's not an investment of money that that's uh you know they're you know it's not you know they're doing Legitimate things to help build the

network that's totally clear if at The Other Extreme the way you're giving points or whatever is you say hey transfer me a shitload of money basically make an investment I promise I'll give you like these bunch of valuable points later that just sounds like a sale right you see how like it's not the it's it's it's the full scene being painted that is what matters here's the simple test test for Decentralization the simple test is if you disappear if your team disappears if the token issuer disappears whoever the the key party disappears what happens to

the Token price if the token price would go to zero if you disappeared you're not decentralized right if the token PR if the token price would go to zero with a single failure that thing is unlikely to be Decentralized now if there's nuance and it depends how people think about the token how people I said the Optics really matter right so if the if the way people think about the token is not as some is not as an investment if they think of it in a more diminished way and there's good reasons why they might

think of that then you know there's there's there's a gray area but it's a this is a simple little too reductionistic in certain settings but a Really a valuable test that I think you should consider and so you might say like Eddie why why are we going from the top right to the bottom right why don't we just start at the bottom left and I've actually been asked this question many many times like can't I just launch a meme coin meme coins are legal right like they're they're just gambling but they're legal like uh you

know with caveats they they they can be they can Be legal why can't they just release the token and then just go right add layer on and add functionality later the reason is a little bit complex to explain but I think it's really really clear the first is that this is how much legal risk approximately you're taking so this is like red at the top with gray in the middle and then some green kind of at the bottom right this is very roughly how much legal risk you're taking with a token if you have a

token At these stages if you're centralized very very very bad if you are more decentralized it gets better and what's even better is if there's a rich protocol functionality and you're decentralized now this is the difficulty in finding product Market fit this is a really critical point okay if you're totally decentralized how are you going to find product Market fit you have like 95% of the people who use your system to convince to do to try that one Experiment that may or may not work that's just not realistic it's basically I would say it's basically

impossible if you're decentralized to find product Market fit finding product Market fit is easier when you're centralized because you're in control like this is this is how Entre this is what entrepreneurialism is all about is you're in control you can continue experimenting you can keep moving things around until you feel the pull of the Market and it tells you like this works this is a valuable thing that we want so if you kind of over play these on top of each other a little bit right you can see there's a clearer path right where you

can go from finding product Market fit while centralized and go down while maintaining product functionality which you decentralize one at a time because you found the functionality that would had a product Market fit and then Trend toward decentralization with the mem Coin example the interesting thing here is if you launch that Meme coin and it's a legal meme coin and there's no you know you did everything right you might say okay can I just wait a little while and then later like try to try to hook them up and find product Market fit the minute

that you benefit from that Meme coin like selling it in bulk or in um you know or making a promise like I'm going to make a protocol that uses this meme coin as its main as its protocol Token as its Network token you are moving it way back up the decentralization access so you may have literally moved from a safe Zone back back into a danger zone it is not a one-way passage you don't go and become decentralized in your commodity and guarantee that you'll never be a security again you can rescure you can become

a security again right so there's a lot more Nuance to say here about like Um cute schemes where you convert meme coins into Network tokens but I think that they're really really really challenging and there's a reason why we recommend this path I'm just curious like a lot of this talk right is around you know how we design a token and the approach and all that but when you're building on a product uh like when is the most successful like examples of when people start to actually consider doing this um like is there like a

a Point where you look when to consider issuing the token like when you start to actually consider token design and issuing a token is there like a point where you look at companies and you're like this is a really good time for you to now start to actually cons consider these things and move into this category the earlier the better yeah in my mind you're really designing a protocol first yeah and then you're seeing where the role of a token is Necessary and the truth is I think with no exceptions when you complete a full

thought of a protocol design and this is how the system works it's obvious where a token ought to be in fact it may not work without one there is a pattern where you can make a protocol that's here and then you can have a centralized app and business here there's no reason you have to like only do the protocol in fact like Something that Chris Dixon likes to say like a little me little pattern I think is is useful to consider imagine if coinbase made Bitcoin right they could have made Bitcoin obviously they didn't but

you could imagine the first exchange saying you know hey I'd love like a digital asset that people could blah blah blah they they come up with Bitcoin and it is Bitcoin as it exists today like it's totally permissionless it is exact like Open source all the things that we like and maybe they're one of the early Miners and they like get a get a good amount of it but it really belongs to everybody and it just goes out there and it's a truly decentralized protocol but then separately they have a business that's a traditional real

business that is a client of that protocol right so your business is an adjacency right think of how like they're compliments together they do better together I think It's a very compelling and reasonable Playbook to say hey I have a company I'd really like but I really need a protocol or I have a protocol I like well I but I can think of a company that I would love to be a client of that protocol and to develop them together push them out let the protocol become its own thing that controls its own destiny like

with the unisoft foundation which is like authentically independent of labs unisoft labs which developed the Protocol originally and they may go in different directions they may even compete someday like who knows right that's good that's like the crypto way so I'll say that any other questions about this that's that's definitely a path that we're trying to go down um so curious yeah who besides like uni and maybe like dydx have you guys seen that have done that well and obviously bod super well that like they both came out of a6z but um curious how like

who else Has done that well in terms of having than a centralized business that can you know have Revenue that drives value to the equity and if people ask you oh is the value going to the equity or the token um how do you you know answer that in you know in a way that makes sense and is really both yeah so there are some other early examples we'll talk about we can talk about offline but in terms of answering that that second question just because I I'm know I'm Running out of time answering the

second question the reason why when we invest we prefer both right the token warrant and the equity in most cases there is exceptions but in most cases is not because like we want we just like want everything it's because we want to preserve alignment with you and you should make that decision later that's it shouldn't be a constraint on the entrepreneur like you're going to figure that out there's many cases where they Thought it was going to go one way you know we thought we were just going to make a protocol and do a token

and the equity was like a vehicle to make this thing and the company goes away later and then they decide like no actually there's like an adjacent business that we would really like to run and vice versa right where they decided eh the equity turns out not that interesting like we want to be one of many companies that are building on this protocol right So the in the early stages I'd say leave your doors open and figure out what works finding product Market fit is hard enough don't burden yourself with some commitment in advance let

yourself wander around and you're going to learn as it develops so I'm going to kind of rush through this last section here and if there's any questions let me know so first this is a hypothetical token launch schedule there's a lot of info in this but this is just say like the way It's set up it takes at least at a minimum I think this is a pretty aggressive schedule takes 6 months to launch a token like why does it take so long besides protocol design and the network selection and implementation of various protocol code

which should happen far before the six months before the token hopefully that's happened for a long time maybe a year or two years three more right who knows but in the last six months before token generation And launch you have conversations with exchanges to have security audits to do conversations with institutional custodians you have to worry about employee custody team custody there's Foundation setup governance setup the Duna filing there's tons of work to do maybe conversations with market makers analysis for like airdrops and other manual token distribution events there's a lot of work you can

take a look at this it's obviously just a hypothetical No two have been the same and yours will will will vary but this is just to get in your mind that issuing a token is it's not just you just Deploy on Main net and you're done there's a whole bunch of stuff to it thankfully it's becoming a lot clearer and we can you know I can help answer some of your questions on that um but anyway I'll leave that at that and then I don't have enough time to go into extreme detail here but there's

only four categories of Ways to distribute tokens that I can tell maybe I'm wrong I don't know so tell me if I missed something but there's only four buckets the first is token sales selling tokens to people selling them privately to investors no problem right the law is very clear that that's totally okay selling to the public big problem don't sell your tokens to the public don't do it it's one of the single riskiest factors you can do and It just don't do it placement so this is I say I you see I have a

little warning like ex you know warning Mark or whatever when you issue a token especially if your token is part of a protocol that's supposed to be functioning people will need access to that token like people will want to be able to buy and sell it obviously people who are performing the work for your system will want to be able to sell it potentially to pay for their operational Costs if there are some and obviously people who want to be future contributors to the system or be a part of the thing will need to buy

tokens so they're going to go to a market and you want there to be a reasonably mature Market a market that has no liquidity is a volatile crazy market and if it's volatile and crazy you're telling a very bad story so I'm condensing this story into its minimum time frame but um it's it's worthy worthy to keep In your mind when normal companies mature and they go through the IPO process that is a long time period where they are vetted in many many many ways and they need to survive long enough to get to an

IPO right they've reached a degree of maturity where the market and Underwriters have deemed them far enough along in their Journey that it's worth iping generally right obviously they can still be extremely risky but they're further down the risk curve is the point People don't see how startups are brutal a real startup is a roller coaster where the where the valuation of the company is constantly going up or down based on the whims of the next deal so if people could see it they would see that a normal startup is like incredibly turbulent chaos but

over time hopefully is maturing and then maybe reaches a point where it's a little bit steadier and still fluctuations but a little bit steadier when a token is released early In the history of a protocol everyone gets to see that roller coaster and worse the market generally gets to participate in it which means that the token price is now a spectacle that everyone can participate in and what's funny is if you normally if you think about it if you went to a Robin Hood or whatever and you look at the price of a stock if

that stock price goes into the floor it's pretty likely that that token that that that company is not a token That company with a stock had a catastrophic event right like why would the price just like Plum it into the floor something horrible happen to that company whereas in reality if you look at a token price for an early protocol if it dropped into the floor oh maybe it was just Sunday like that's normal but people don't think of it that way they still think well wow the token price dropped that thing must be dead

oh the token Price is high this project must be ripping and like growth is incredible neither are necessarily true it's basically a low source of information that people are accustomed thinking of is a high source of information so I think the extent to which you can smooth that token price and Ure that the market behaves in a less crazy way is supportive because you don't want people getting the wrong idea this is an early protocol it's developing we're Trying to make it more mature and we're doing our best that's the idea not like this is

a roller coaster and you should get on the third bucket of ways to distribute tokens is automatic programs I already alluded to these security budget like paying for verifiable high quality work paying for the supply side all these things these are basically ways that tokens are automatically issued by the mechanism in order to reward the types Of things that you want and then the last program which is very interesting and I wish people thought more carefully about at maturing very quickly is these manual programs air drop everybody knows aird drops air drops have some significant

advantages and risks but also some advantages Retros right retro pgfs like optimism has popularized Grant programs prizes Partnerships these are basically ways that you in the early days and maybe a foundation later on or Maybe like the Dow members as things develop can find ways to take tokens do some work out of band allocate them in clever ways to potential partners and I think if I could leave you one idea on this because I'm out of time is that uh we often think of airdrops just because of the way his crypto history has developed as

something you give directly to users to end users but I actually think this is like this is a little wrong the Unis swap airdrop was although It was to anyone who swapped and was an LP and you know whatever with Unis swap just because of how early Unis swap was in its history those are primarily developers those are just crypto native people who were likely writing smart contracts building stuff just probabilistically they were not a random sample of people so there were there were some like regular people but it was a very skewed set of

people and that is part of what makes it so impactful and Also recall that that was in the early days before the extreme professionalization of airdrop farming Advanced so I'd say when you think about manual programs don't think as much about or maybe maybe discount slightly your ability to distribute it to regular people it's very hard to do that in an effective way that is most valuable for them most valuable for the protocol think more about how to get it in the hands of long-term stakeholders who are Going to build on top of your protocol

right people who are leveraged people who are going to raise their own money people are going to build their own business people are going to bring users because that's all they think about those are the people who are the most valuable to concentrate your scarce token toward for the long-term success so I'm going to leave it there thank you very much [Music]