RAAS Rewards as a Service Building a Next gen Rewards Engine to Incentivize DAOs Hart Lambur

[Applause] so um james has i have better graphics than james um this is the first time i'm giving a presentation from something i made from my ipad um i did this once uh internally for our team but we'll see how you guys like it uh so it's all hand-drawn um i got a thumbs up cool um the concept here that we're gonna be talking about is just rewards uh and incentive programs and how we can generalize them and there's a new concept that we've developed internally at uma that we call the optimistic rewarder that we're going to introduce here and just to warn you guys this is a fresh concept we haven't fully fleshed out every aspect of it but in the spirit of eth denver we wanted to introduce to you guys and get some feedback so i think there should also be lots of time for questions at the end um for background i'm heart lamber i'm one of the co-founders of uma we've done a lot of things over our existence from really before d5 existed what we are really focused on is this thing called our optimistic oracle and it's an oracle system you can ask any verifiable question to and get a truthful answer from and we can use it to build all kinds of crazy things this being one of them so um my core thesis and i don't think this is super controversial uh rewards uh and incentive programs are a really really big part of crypto um i actually think crypt all of crypto is basically a giant incentive structure um but specifically we've seen how yield farming or different sort of incentive programs have led to the very quick and rapid adoption of a lot of the innovations in defy and web 3. um and i frankly think it's one of the coolest things we can work on and play with and the core design pattern here what all these programs have uh in common is just hey we take an action and we get a reward that's it action reward action reward um and you know to go through my james you like my graphics to go through my great graphics here we'll go back to uh the og of yield farming arguably um synthetics with their staker contract some of you in the space might know they have this standard piece of code that people have used all over the place uh to run an incentive program and the contract's really simple you put in some rewards you stake an asset a token and you earn some admission rate super simple worked great this is the basis for a lot of yield farming innovation but again action reward um we keep going here we'll fast forward to our our friends at olympus what's the concept here here we're gonna my mental model for how this works we're gonna stake an lp share a uniswap lp share that's like om eth we're gonna stake it we're gonna wait f that's the action we take we're gonna wait five days and then in five days we get some reward back that's the structure worked pretty well until maybe it didn't but it's super interesting curve our friends at curve and they're ve economics i'm still not sure how to pronounce this actually but they're vote escrow token nomics what's the action here i'm going to go and i'm going to take a token i'm going to stake it for a period of time maybe four years i get four times the voting power action reward um and then to plug some of the stuff we've been working on for a second uma has this concept of kpi options that we've been playing with and a kpi option is a pretty simple idea where you have a token really a synthetic token that'll pay out some amount of rewards if a given kpi or metric gets hit and the cool thing about this is it can be a kpr metric that's observable in the real world did your tvl hit a number did you get so many discord users whatever it might be but again here there's an action and a reward we get so my big claim my big thesis here is that reward structures are really what's been responsible for making dows and defy and web 3 what they are today without these types of structures and these incentive programs we just wouldn't have had the growth we have and it's like growth hacking on steroids and it's super fun um but there are problems so it's a lot of work to program one of these incentive structures um i think this is again one of the most powerful things crypto's brought us um but there still aren't that many design patterns out there because it's a lot of work to come and code them up and write them in solidity and we all know how hard it is to find a solidity engineer and all that kind of stuff um and once you code them up they're brittle and hard to change um and so what we've seen in effect is instead of inventing or modifying these schemes people fork them they just copy them and change the parameters i don't think that's a good thing i think would be better if instead of people forking them they actually tweaked them and improved them and made them you know different or newer or ran more experiments the other other point here it's a bit specific to kind of my view of the world is that most of these incentive structures we went through they hinge on real world events uh sorry they can't hinge on real-world events there's no way to build an incentive structure that's triggered by a real world kpi like if your project is beating your competition do you get more rewards or stuff like that with uma's kpi options we can do things like that but a lot of these other programs don't so what's the idea here what we're looking at with this optimistic order concept that i'll dive into a second in a second is if we can generalize these incentive structures can we figure out a broader framework that we can use to do this action reward pattern one that's super flexible and lets people tweak and experiment and the the sort of the ideal here the the blue sky scenario is if we could reach get to a point where this is actually like a no code structure so that people could come and just invent the rules of their incentive program and it would just work and i think that's a pretty cool dream we're not there yet by the way but i think we could get there um and so uh frankly in a lot of time around um lisbon and youth lisbon we came up with this idea of an optimistic reward or contract that we have gone ahead and written it's actually been audited too so what is the optimistic rewarder well the idea here is we still follow this action reward pattern and we tweak it you do an action you get an nft the mft is not meant to be like a picture pretty thing although i certainly could which i think is kind of cool but the nft is just meant to be a receipt that says hey you user xyz did this action and it's unique to you and we can track your history of all the actions you do in the future that are associated with that nft so i now have a non-fungible representation of all the actions you've done then we can use that nft and associate it through a set of rules through an arbitrary rule set and this is the part that's kind of important totally arbitrary rule set an arbitrary rule set that says that nft deserves x amount amount of rewards and however we calculate those rewards is just purely defined by this rule set we specify somewhere so like wave hands and be magic about how we specify that rule set but that's the idea and so to kind of look at this visually the idea is like like let's just pretend we got this contract that's our optimistic order contract we load it full of rewards um we have uh the code of the contract says hey i'm gonna take an action and i'm gonna get an nft i get my nft back whatever that action might be go on with my day and then i have some set of rules that i'm going to be able to use that nft to claim those rewards so what is this rule set and how does it work our rule set is any any for these rewards can be any set of verifiable logic anything that i can actually that we could objectively go and look at and say hey that's going to get me this many rewards this rule set we actually enforce off-chain which means we don't need to program it in a smart contract and we enforce it the way we enforce it off-chain is optimistically and here's where i got to do a little sidebar into what does optimistic enforcement mean what does it look like how is it relevant to what we do and this ties back to what we call our optimistic oracle and the concept here is pretty simple the optimistic oracle is a structure where i can ask a question or in this case propose and say hey here's here's how many rewards i'm owed and as long as nobody disputes it it gets it gets set to be true it's valid so i can use this optimistic um uh optimistic oracle to enforce this arbitrary rule set i've written out like literally i could just write it um on ipfs or something so i can say hey i want to withdraw this many rewards i have a challenge period of let's call it an hour and as long as nobody says hey you're wrong during that challenge period uh my my my my claim gets set to be true and i'm allowed to withdraw those rewards so this whole thing hinges on this concept of a dispute and a dispute would get triggered to an additional stage of verification that we call our data verification mechanism and basically um token holders vote on whether the outcome's right or not we've used this optimistic oracle in production for a couple years now hundreds of millions of value flowing through it and it actually works surprisingly well where people actually will dispute invalid data up to this other system and so now we can use this to again enforce optimistically this arbitrary set of rules um so sort of toy stupid examples but let's define the set of rules is hey i'm going to stake some uniswap lp shares get an nft and my uh my rule set says in five days i can claim that i can burn that nft to claim x tokens cool i just programmed like an ohm fork that way it's kind of cool kind of interesting i could write a second rule set that says hey i'm going to lock my voting token i'm going to take my token i'm going to deposit in this optional order contract and i get this nft that i can't do anything with for four years let's make that be my rule set and then in four years time i can burn it and get my tokens back but in the interim i get four times the voting power that that's cool we just programmed like curves vote escrow economics um so the concept here what i'm trying to highlight is this design pattern is extremely flexible where we could program a lot of the existing reward structures that have worked well as well as tweaking them so okay my point the design pattern my handwriting is awful here the design pattern the design space is huge there's a really big space here that i think we want to explore i think the space is relevant you know not only to d5 projects web 3 but increasingly in the real world too we invent better incentive structures and this optimistic rewarder structure is designed to be super flexible um so let's let's like try to invent something maybe new here uh just to showcase what this might look like and let's pretend that we have a new dow that wants to distribute a token to its holders this dow has some sort of staking mechanism associated with it so we say hey day one or day zero we're gonna have this initial community and we're gonna distribute to all the early community members we're gonna give them an nft and this nft says you're an og and once you have this nft you can now go and the protocol gets turned on and there's some staking function where you stake to earn some api some let's call it standard yield farming but this og nft gives them a 2x multiplier on their apy for being the original guy so it's kind of cool right there's an incentive to be the original person i get this multiplier but then let's just i'm just making this up by the way guys so let's just then say that in order for that user to claim their assets they actually have to burn their og nft and i lose my 2x multiplier going forward well okay now i've kind of created a little bit of an incentive where i want to not i want to huddle i want to stay invest in the protocol i don't want to actually claim my rewards because if i do i actually lose my my multiplier my og status cool kind of interesting right kind of a cool way that we might have uh incentivized early users to be supporters of the protocol and to stick it in the long run and i'm just putting this out as like an example of something that we could invent or create with this arbitrary rule set and um some of you in this room might know that risk labs the foundation behind uma has also supported or launched this l2 bridge called a cross it's at a crosstalk too you guys should all check it out because it's actually really awesome super fast lows fees handles most volume but we made a promise when we launched this that we're going to distribute these tokens whatever across token com is going to distribute the community and so we're quite excited to use concepts maybe like the one i just described throw them out in the community and say hey would this be a good way for us to distribute our initial tokens um is that something that'd be interesting so if you guys are curious about that maybe come join the across discord um because we are literally building a community from scratch um okay i am going to leave lots of room for questions and i'll wrap this up in a second um i i guess i want to summarize by saying like okay what's my point here rewards are powerful and impressive the reward systems and reward programs have done lots of useful things for d5 but they're hard to tweak uh this optimus rewarder is our idea for a super flexible way to build an incentive program that's easy to tweak um you can design in this program any incentive structure that you can write down rules for so if you can write down the logic you could actually build that incentive structure to go back to something i said earlier the dream here is we actually get this to a no code structure where you could literally write down logic or maybe there's like a gui to drag and drop rules and it it it like visually programs the incentive structure you want to enforce um and then everything here this rule set for your rewards program is then enforced optimistically using uma's optimistic oracle um so you guys can find me thanks thanks for my talk thanks for coming to talk you guys can find me on twitter at um it's also my mom's birthday today i really doubt she's listening but happy birthday mom we have f hey we have actually four minutes for questions i will optimistically enforce four minutes for questions i think that's great so go