FarCon NYC: Community Talks
Aired Date
May 2, 2025
Series
FarCon NYC
Episode
FarConCommunityTalks
Watch on YouTube
https://youtu.be/pKWhH4SFPmgthe community talks were recorded at farcon nyc on 05/02/2025 speakers were voted on by the farcon attendees farcon is farcaster's annual event organized by the community for innovators builders and network contributors interviews with the speakers and gm farcaster followed the main stage talks thank you to all the sponsors who made this conversation possible including media partner gm farcaster media sponsor nounstow and the farcon team of volunteers learn more at farcon.nyc
today i'm gonna talk about diacc philosophy more popularized by vitalik that really means differential defensive democratic and decentralized acceleration and this is actually more of kind of a personal spin of mine through some of my own beliefs and experiences quick about me my name is tayyab i'm a startup generalist i've worked across startups in a bunch of different niches and a bunch of different roles i've been a founder i was a yc alum and i'm fid 66 so i've kind of seen all of farcaster really i'm currently a full stack engineer at icebreaker where we're building the professional network layer on top of farcaster and the badges that most of you guys have around your necks and what this actually is is this is a call to arms the world is headed in a ton of different directions that we all may not want to see and there are steps that we need to take so that we get to the world that we all want to see right about two thirds of the world lives under some form of censorship that's actually over 5,000,000,000 people in the world internet freedom has been in decline for fourteen straight years we all live in some sort of surveillance mechanism and this is not simply true in just china this is actually true in the west the government can get information about you if they so want geopolitically we now live in a fluid global order the word of the year for 2025 is likely to be uncertainty we in the developed world young and old alike face some sort of loneliness mental health anxiety epidemic and the climate continues to rage on in a catastrophe but look i'm a definite optimist i believe that version n plus one solves the problems of civilization n and i'm a believer that this happens not through regulation or degrowth it happens through building and acceleration right we need to get to a place of abundance in the world and the only way we can do that is by solving problems
we look we've been building for a very long time as people we're healthier more connected more advanced healthcare is better child mortality poverty is down we have access to information we have access to intelligence that we've never had before
but acceleration is good right
let's talk about an example from pakistan so in 1947 pakistan and the indian subcontinent was liberated from the british empire and a large minority religious group wanted to separate and have their own nation called pakistan around 15,000,000 people moved from india to pakistan and pakistan to india two million people died in the chaos that ensued there was hostility between the two countries there were land disputes over kashmir where most of the water comes from and pakistan realized that it had to have a powerful military to defend itself and it did so by taking 40% of its annual revenue and putting it directly into the military so much so that we have a famous quote that goes we will eat grass even go hungry but we will get our own nuclear bomb we have no other choice and pakistan got the nuclear bomb it is the only country muslim country in the world to have a nuclear weapon so it got security but it didn't get prosperity 40% of pakistan still lives in poverty today and so what i'm trying to say is that the direction of acceleration matters it's not good enough to simply hit the gas pedal it's also important to guide the steering wheel of civilization's advancements let's take a look at another example probably one that we all care about right centralized ai most of us access ai today through chatgpt cloud gemini amazon hoot and whatnot the most powerful ais are likely going to be built by the by the entities with the most amount of resources which are going to be nation states these nation states are going to make god models that basically follow their principles their values their strategic goals and these are not simply chat bots right these are policy advisers these are military strategists these are the educators right you get a different answer based on the values of the country that runs the compute right whoever owns the compute owns the intelligence and what we need is we need what we all think about which is secure technologies and many of us probably think about decentralized ai we think about decentralized censorship resistant compute we think about local open source models that we can run and be able to have censorship free intelligence which is going to become more and more important in the world so for me i think about these principles of decentralized or defensive technologies that sort of i've come up with right defense favoring technology should as a default setting protect you from the harm that can come from everyday life the classic example is obviously encryption but it's also that stupid freaking medicine cap that you have to open right that protects child's lives basically everywhere resilient infrastructure resilient and antifragile technologies my favorite example of this is open source when we have open source software we're able to review test and improve software as sort of a bottoms up community where you have more eyes than simply the employees that work at whatever company dispersion of authority right classic example is a blockchain concentrated power corrupts or falls in the wrong hands and then they get gain at somebody else's downside and so being able to disperse the ability to make decisions is a really important thing that we will need and this is the one that i think is actually probably the most important we globally face a crisis of trust
and if if there's one thing that we can figure out how to try to solve it's how do we scale trust at an internet level to unlock human cooperation this is the core thing that we're trying to solve at icebreaker and and it's and i believe every one of us would like to live in a high trust society so what can you do i would say you know build and use d act products i think i think building is obvious but using diac products basically signals to the market that you want products that have specific values and that you're willing to pay for those just a quick example of some technologies that some of us use or could use i'm probably i'm not going to go through too much of this but that's all for me thank you so much and accelerate consciously
you guys can hear me anyways okay
and people love that so without further ado please welcome limone to talk how to make mini apps go viral
hello everybody can you hear me cool let's see if the yes it works so i'm limone from rome italy today i'm going to talk about how to make mini apps viral based on my and our experience with farben and a bunch of mini apps that we built before jumping in a quick intro about myself again i came from rome i'm a software engineer but also like community builder so on one side maybe i'm farcaster you know me for shipping products hacking stuff all the time and that's true of course and we do that with builders garden our product studio on the other side i'm also i love community i love building communities and i'm building a community in rome called urbe it we're running it from hosting physical hub the first actually physical web3 app in our country and we're trying to create an ecosystem from scratch but yeah so a couple weeks ago i wrote this article about all my learnings and kind of like trying to lay down some items that i think are important to make sure that your mini app goes viral so after the presentation you can also check it out on paragraph it's even more detailed maybe so first thing there's a bunch of mini apps ecosystems there's like telegram workcoin and there's forkastr of course and the most interesting thing for me about forkastr and when i think about building a mini app on forkastr is that we are inside our social network so every time i even like yesterday when sitting down at the builder day to come up with a new idea to build we were trying to think it in a way that it's multiplayer at first i want people to share it on the feed i want people to play with their friends or use that shared experience with their connection on firecastr and it's extremely simple to use tools like or embed or anything that you can use to pull up profile information or anything related to the firecastr social graph in general and these are some examples so you can see like on farbir for example we have a simple leaderboard that is showing you your ranking based on experience points and we're also showing special leaderboards within your friends so that you can compete and in the middle we have like ponder which i think is like i got a ton of explanation from ponder on how they were showing these profile pictures below every of their predictions because it's interesting how that maybe can make you more engaging with that prediction and of course like your own forecast you can easily invite your friends in it's easy to track referrals as well creating shareable moments i think is probably the most important thing dan this morning was talking about somebody asked the question about discovery and searching mini apps most of it like today happens in the feed and so it's important that you plan your mini app to be something that people are willing to share in the feed you have some actions in the sdk today that you can use to to do that and it's important also to come up with some custom preview images that are that are fun in this case i picked like this example of moshicam howrah mini app that went viral on bringing it to denver if i'm not wrong where after minting your nft here like your profile your new like howrah picture it will prompt you to share it on firecastr design for feed fidelity so making the preview image as engaging as possible including social proof is a trick that we've been using with farbir as well so when scrolling in the feed first of all i want to see like a clear image or even like call to action of what i'm supposed to do what the user is sharing but also i think it's useful to share how many people for example are using that mini app how many people are engaging in that prediction or how many people in the case of farbir are active and playing the game because if it's just me or 10 people maybe it's not fun enough rewarding people you can do that through collectibles badges social status or even just daily streaks or like leaderboard position is something that you know users are willing to share at least in my experience in what we've been building and of course like we've been experimenting with it within farbir in the case of fargate for example you have a daily streak that kind of build on top of your dopamine effect and in the case of warp slot they were trying to reward you to come back every day with five free spins so it's a way to reward users and if they feel rewarded then i'm sure they will be more likely coming back notifications are also another like superpower that broadcaster gives us today so it's i think it's a very powerful tool that we can use to first of all send announcements for example today i received a new announcement for the acid test episode two that it was live for meeting and that brought me back in the app so notifications are definitely another way in which you can bring people back in and you can do that with broadcast notifications so whenever you have something important to announce a new feature or a new event happening you can do that also asynchronously somehow so taking farbir as an example whenever you plant something we are notifying you let's say like few hours later whenever that that crop is ready to harvest and i've been seeing a bunch of my mini apps using some you know like triggered or special event notification according to any actions that you performed in the past and then of course like notify me when some of my friends maybe are interacting with the app because it's more likely that i'm interested into it and yeah like this goes from ponder for example anytime there's a big prediction going on they notify you and then maybe you're willing to go there because it's an interesting one and you end up predicting something and finally planning the launch like a pro so for both our mini apps like farbille and better which is a chain sports betting app we try to build you know some momentum before the launch creating some sort of like preview experiences that maybe you can see better here so before launching the mini app we built some simple use cases so in the case of fireville you could simply tap and plan things in the case of better you could bet on an event so giving you a preview of what you can do giving you a reward if you kind of like predict the event in the correct way so that first of all you're using the app and you're trying we get feedback in the momentum that we built but also whenever you presave the frame you are actually adding it you are enabling us to send you a notification and so whenever the mini app is ready for lunch you can simply notify people hey you like this this is now live and yeah that's pretty much it
so yeah i hope you'll be there but i don't mean you up if you have any ideas or things or questions about this i'm happy to help you hold the time of course thank you
alright next we will have someone who i think is masterful at capturing the spiritual narrative of farcaster i think i consider him and then a group of other users on the network including six flood like there's a few others who are perfect or not i wouldn't call them perfect but craftsman at articulating a lot of the thoughts and the essence of farcaster into words that we then read so no further ado i'd like to introduce jihad
what's up what's up we're just gonna jump right into it what are we building towards if you ask everyone in this room that question you'll probably get a dozen really solid answers and if you ask everyone in crypto that question you'd probably get a bunch of really generic answers things like economic sovereignty decentralized ownership credibly neutral systems whatever buzzword that they can come up with that day
and it's
not that these things aren't true but the goal of this talk is to convince you of three things first that we should all have the same answer to that question what are we building towards second that i think the answer to that question is actually very simple and very in line with farcaster's ethos and finally that i think it has direct impacts on what your job is as a founder or builder before we get into that though let's zoom out we're sort of at a time where there are two really important digital technologies that we're all interacting with every day crypto and ai more specifically we could argue that it's open global financial rails and large language models now if you're here you probably think that these technologies are going to impact our lives in ways that we can only begin to imagine but in my mind there's one thing that's already true the intersection of these technologies is leading to an acceleration of entrepreneurship startup abundance it has never been easier to build a product reach customers and sell to literally anybody in the world anyone can vibe code a mini app and deploy it to farcaster build a saas product on replit no code no fees no middlemen so that's the answer to our first question we are building towards global unrestricted access to entrepreneurship we're closing the gap between imagination and action and many apps are just the beginning the shape of startups is going to completely change over the next few years so whether you're building in defi or cogen or creator tooling or you're building a new blockchain that is the mission that is holding us together that is our north star which brings us to the second part of the conversation farcaster has always been a network that has supported builders from the beginning farcaster was a network of builders and tinkerers and creators on in crypto and i think for that reason farcaster is uniquely positioned to be at the center of this entrepreneurial revolution the social network for startup abundance if you will but but as farcaster grows it's not going to look and feel the same anymore we're going to need a bunch of overlapping networks distinct networks building on top of farcaster that are going to push forward specific visions of the future in line with startup or entrepreneurial abundance so what does that mean for you the founders and builders in the room again i think this has a very direct impact on your job a founder's job is to articulate a specific point of view and then build the network that turns makes that thing a reality
now we could argue that network is a really general term but i think there's three things that make up a network regardless of whether or not they are visible or invisible legible or illegible farcaster is a network but as we'll discuss so is hire so is clanker so is base so is design twitter right networks are made up of three things the people in the network the tools how those people are interacting with one another and the context the orientation of those people their shared lore their shared vibe their shared shared energy and knowledge you could argue in a way that we're building toward an era of hyper locality but it's not necessarily based off of specific geography or a zip code and it's not necessarily based off of network states either these networks are the things that are going to act as the foundation for the builders of the future with the founders that i work with i like to make the distinction between serving versus shaping a network every founder that is building a product that people are using is inevitably serving a network you're listening to people you're adding value a really good example of this that we're all familiar with is the original token bot product farcaster wanted to play with tokens and the clanker team decided that they were going to launch an agent that made it possible to launch a token in feed people got super excited they responded to a need that the community had and it was a big success a non crypto example that has to do with more listening to a community narrative is bandcamp during covid the music industry was you know obviously not doing super well and bandcamp decided to listen to the narrative that its network was telling it and made fees completely free for a day which increased revenue 14 times the normal friday on the other side of things though i believe great founders are going to shape a network not just serve it today i think farcaster is doing a great job of this base is doing a great job of this by putting together a very specific vision of the future and then giving us tools like minikit that allowed us to build towards it really easily regardless of whether or not we are developers clankr obviously we brought up tokenbot as an example but now clankr has become a network in and of itself clanker has its own protocol with clients built on top of it with very specific philosophy around what it means to launch a token and the importance of tokens in the ecosystem and i think all of those things together again the people the tools and the lore the context are the things that make a network and finally i think organizations like splits who have built infrastructure from the beginning that have sort of defined what it means to be an on chain product and are now building specific interfaces on top of that that are defining what it means to be an on chain startup with all of these things you could fork the technology you might be able to copy the tech but you're not going to be able to fork the network so how do we go about building a network how do we be like these great founders that are shaping the future i think first and foremost the playbook has to do with believing in something you have to define what the future you're building towards is and define the philosophy that you're working with without belief you could build a great product but in this era you're not going to be able to build a long term sustainable business all the great teams that we mentioned believe in something but then there are specific tactics and again getting users in this world might be or getting users in this world is is pretty meaningless without getting really granular about what it means to bring those folks together and build trust between them i think there are three things that could help you do that really well first words you gotta write a lot whether it's an internal handbook or externally to align a community around your product writing is not this not only going to articulate your vision it's also fodder for llms you're able to accelerate your own own progress by making it possible for you to use and leverage ai tools second boost i think the single highest leverage thing that you could do as a founder or a builder is to help your customers help your community achieve their goals regardless of whether or not it has anything to do with your product you shouldn't have to depend on them using your saas product in order to help a customer achieve the vision that you guys are all working toward together and finally this is the more specific one but cohorts i think a lot of great founders have done a good job of launching hackathons programs accelerators incubators on very small scales and using that as a sandbox to develop a network of their own get the right people together build trust figure out what it is that you could do to create value and then push that community forward into the broader network to influence it if you do all of these things well it is pretty clear to me that you will win but what does it mean to win how do you measure success in this world again your job is to articulate a specific vision of the future and a specific future that people genuinely want to inhabit and then build a network that makes it real if we're all working toward global unfettered access to entrepreneurship then the success metric isn't daily active users or a specific growth metric it's agency you are helping people achieve the future that they want to achieve in line with the vision and the network that you're building the world is moving faster than ever and the best founders are not just going to build products they're going to build networks your job is not just to capture attention but to cultivate agency in every single person that your product and your business interacts with thank you
okay i before i introduce the next speaker i have a question did anybody hack did anyone submit a project for the hackathon oh my god have you has anyone who here has tried coineroid guys fred wilson gave it a shout out you have to try it you can find it did you graham did you post it can they find it on your profile okay go to mcbane do not coin me though i don't want to be coined right now but try it after the next speaker is every if you were a developer on farcaster before linda joined this was actually your go to person if you had any developer questions despite not being on the farcaster team so i'd like to introduce somebody who i've grown to call a very close friend and someone i look up to a lot samuel huber
welcome everybody guess i'm actually in new york there's a lot of me saying things about the us and i'm actually in new york it's fine we're gonna talk about one year of serving farcaster companies and building a little bit of a farcaster a quick rundown of what that actually means well there's a lot of people in and around the farcaster ecosystem a lot of people with ai no coding but to me there's generally indie hackers who bootstrapped and also just companies making revenue there's a few of those a lot of it to me at least is bootstrappers and just people trying to figure it out and that's also somewhat of what i've basically learned over the last year well what do we do we've it's like long list but tldr is anyone going to market on firecaster or needing tech support that's usually us and this is how it basically feels to build on firecaster this is a little meme this is gonna gonna get bigger because i literally posted this into jetcpt and jetgpt said it's both a joke and a cry for help left in a dramatic style was ecstatic if you're literally spending some time and reading what's up there this is just like two thirds of all the terms that were thrown around last year that generally just mean hey you can develop on top of firecaster you can do cool shit you can build community that's the three things and like the fourth now is with crypto you can earn money and like that's firecaster wrapped into one sentence we use a lot of words to try to describe that so that's fine it's fine we'll figure it out right so the one thing that we all do is we're hanging in there because life is not a race right even though we're in new york everyone is hustling like it's fine we'll all be fine so generally how does that translate to navigating firecaster and i've put a little bunch on the slides that you can read but there's two things in firecaster that you just kinda like have to do and that's just talking to people and the developer calls literally is like a great resource there's also the recordings of those in youtube if you wanna get transcriptions and just read through quickly dylan is also doing a great job at docstack.eath to do summaries every single time that he's in one but it's a community and a platform that's moving fast which if you're expecting something like oh i'm building on top of the internet that's been around for twenty five years you're at the frontier so you'll spend some time just reading through all of that and making sure that you're literally up to date so like attending the dev calls reading the fips and just generally being open minded and contributing to the discussion there's a lot of times something happening which i find quite interesting where dan and varun and the whole like murkla team post their thoughts openly every single time which is literally their thoughts that they have in their head right now and people take that as oh this is the definite thing well no just like actually have a conversation ask questions because literally what's happening is we'll just figure it out together right literally the simple defaults is something that we've talked about and discussed a bunch so we're like should we have a following feed should we just have the home feed how should all of the app work and be set up well you all have input you all are probably in the 1% of users that are gonna actually play with those settings so just remember that a good default is the best to have part of my job is literally to talk to new developers every single day and just figure it out the thing that they appreciate is something to get started right quickly and then they have to adjust but that initial default to just have a successful moment initially is really really valuable like when you're starting into a new ecosystem and most of you probably have been onboarded to farcaster by say dan varun or like a twitter dm or somebody literally doing white glove service to you most people come into the ecosystem and they're like oh i just see great content but i need to figure it out and then going back to like the cheggi petit translation of all of the naming like just sharper them a little bit literally go have conversations help them and realize that for them a good default in holding their hand is what you need to do and also like serving firecast for companies that is what you have to do when you're talking to people just holding their hand and showing them the way is what has worked best for us to really make sure that people understand firecast or to see the value and go there so i've posted a little bit about just some questions people have had about what we're doing and let's start at the bottom with counterintuitive learnings from the past year you don't really need to read all of it it's on the slides for back reference but the tldr is file castor is new so what can you do with it well everything because it's full and open and programmable but then also like everything usually means there's no good default there's no playbook so the thing we learned is just literally show and paint the picture and then walk it together with people and some people will not see that journey and that's fine and we're focusing on the people who want to walk that journey and the one thing that always comes up is what's the expected payoff literally i have calls like thirty minutes sixty minutes where people fall in love with farcaster they become active on the platform but then it's like okay so what is your company that we're doing with farcaster like have you seen the leverage you can get and they're like well how much money is it what are we getting out and farcaster to me is a brand play or a play on building a community if you're in in that sphere if you're out for developers if you wanna make money on firecaster using the wallet is a really really great way to do that we built a mini app with morpho that's now roughly around the hundreds of thousands in like total transaction volume in the like three months that it's out and from offload it's really cool because it's just like actual users like humans in the end just small people using it instead of just institutions that they also have through their coinbase integrations but that's kind of like the thing there and getting back to the show don't tell is even more important on firecaster communication and that's at the top is the tool we have always remember that any can mean where literally there's a lot of shit we have to communicate so breaking it down making it simple showing people the way just like if you watch the mandalorian like this is the way and constantly repeating and reiterating that is really really important and also in there like that this open protocol is there for you you just have to spend some time on it because we're all developing it together and that core understanding is something that i want to drive home today is you can all contribute and you should and nobody should feel left out because they're just like oh should we ask this should we not ask this but that also means everyone tries to figure it out and we have to hand hold each other as well so come to farcaster is still the one page that you all have basically agreed to right you're here so you're already on firecaster but even businesses outside of web three wanna go to firecaster and i'll tell you why because businesses outside of web three have a hard time monetizing audiences and targeting audiences in last november in hamburg i gave a workshop for corporate developers corporate meaning like tens of thousands of employees like billion dollar revenues just traditional businesses how we would call them they're sure like sap of the world and those tech companies but also german manufacturing companies and those businesses would love to have the ability that we have in having all the open data and saying like hey i want somebody interested in a new dishwasher and i can just search for that and target them facebook makes them pay to generally reach that audience on farcaster all the data is open it costs $10 a month on api providers to be like hey i can search this i can directly reach out i have notifications it's also programmable it goes into my back end into my crm well it goes into my back end my crm is something where we need good defaults and we're building those out for those companies to have a product they can use instead of having to build them themselves and we'll get there but this is part of where we currently are so this is a list of things that i think are super undervalued in farcaster the general theme alliance with what the business is not in crypto really like about farcaster and that's the open data all the possibilities with the open data and literally taking that together if you wanna use ai if you don't wanna use ai and building a great experience some of that is embed as an api provider indexing of code's api mainer's api direct messages and literally using all the farcaster tooling there's a blog post detect.visionfarcasterstart where we've just like it's on that little qr web page as well but we just broke all of that down and that to me is the most interesting thing and this is generally the pitch for me is build a mini app because you get to leverage all of that round trip in one thing so if you like mini apps if there's anywhere you think i can hold your hand feel free to talk to me i'm always one message away and thank you so much for attending
okay before i introduce the next speaker can we just give a second to say that every community talk so far their narrative violation is that men in crypto only wear flip flops and gym shorts they've all been dressed impeccably so kudos to that the next speaker was actually my gateway to paragraph he was og paragraph user it's how i learned about the app and he would do full on deep dives around farcaster ecosystem and then different emerging topics in crypto as well and now it's really exciting to see him grow beyond forecaster really kind of capture some mindshare on twitter as well and without further ado i'd love to introduce yb
sweet oh that was loud hey everyone my name is yb and i have been active on forecasters since 2022 currently i'm working on my newsletter terminally on chain full time and today i'm excited to give a brief intro on decentralized ai this is a rabbit hole that i've personally been going down the last couple of months and honestly it's felt as exciting and addicting to me as i first got into crypto five years ago so before we even get into the weeds of things i think it's worth getting on the same page when i first started researching this vertical i thought the question that i was trying to ask was which one wins between closed open source or decentralized ai but it turns out that that's not even the right question to ask see if ai is as big of a shift and a technology shift as internet was then there's more than enough room for all three of these to coexist in fact back in the nineties everyone thought that the next era of computing and internet would be built on the backs of sun and oracle and then it turned out that there was more than enough room for things like windows and open source competitors such as linux and mysql to also coexist so if decentralized ai can even just capture one percent of the total market share and economic impact that ai is gonna have in the next decade then in my opinion i think it's definitely worth paying attention to with that being said where does decentralized ai even fit in so i'm gonna split this talk up into two sections the pre training and post training pre training whenever you hear about the massive nvidia clusters on the news that's pre training it's building the brains of the model and whenever you hear about people thinking chat gpt is being sort of cringe or too agreeable that's posttraining more so the personality or the expertise of the model so with pretraining there's three angles i wanna approach this from one the philosophical side two practical and third the more experimental so going into the philosophy side of things the question that i really wanna address is is open source ai as it stands today truly open source and companies like meta and deepseek and ai too are all pushing open source ai quite a bit and actually their contributions have been so helpful for the developer community to actually partake in all of these ai improvements that are happening but at the end of the day they're the ones that are sponsoring the massive gpu build that comes in and at any given point the reality is is that they can just decide to stop releasing their models as open source and it might not even be because they don't want to it could be that the us or the chinese government say hey we don't want you doing this for geopolitical reasons just as an example so you're probably thinking no big deal i'll just use chatgpt that's what i've been doing anyways and that might be okay for consumer use cases but it turns out that enterprise ai actually cares for them cost at scale or even legal provisions on things like data and customizing their ai infrastructure is actually make or break for their business and the reality is is that there are a lot of teams and developers in brazil and india and australia that are actually very much dependent on these open source companies releasing their next generation of models and staying up to date but they don't have any guarantee of that so that's why it's important that we do have a decentralized state of the art model which isn't bound or dependent for money from any one company or isn't bound by any one legal jurisdiction cool moving on to practical so what's actually happening right now in the space itself i mentioned earlier that pre training process actually requires a ton of capital expenses upfront for things like gpus and the data data centers in fact in the last five years since chat g p two chat g p t launch the scaling laws have showed us that if you just keep on adding more and more gpus and make the data centers bigger then the models will get better which is true but even these centralized warehouses have their physical limitations after a certain point if the amount of operational expenses required to actually spend on these cooling equipments and the real estate and the labor cost that goes into all of this it turns out that it's not justifiable to spend the money when the pre training process efficiency is actually just going down so that's why you have companies like microsoft and google and openai all looking into distributed training as well and the main question here is can we train a state of the art llm over the internet and it's by no means an easy problem you actually have to reduce the network communication between gpus by a couple orders of magnitude but there are efforts being made there's decentralized ai teams like prime intellect and noose research that are taking the pre training process and breaking it down bit by bit and then optimizing different things that they can do so it's very early into the research but even in just this last year it showed the the team showed that there is actually a possibility to make this gpu communication less and the most interesting part to me here was how collaborative the research process is with even teams from deepmind and meta ai i think it proves that pretty much everyone in this space is thinking about distributed training in some form or fashion and that's why you have jack clark who's the cofounder of anthropic and even openai's pre training team all acknowledging that the mother of all models is gonna be built on some kind of decentralized network awesome so going into the last part of pre training which is experimental to me this is the most crypto esque of decentralized ai we all know that bitcoin ethereum and forecaster took the infrastructure layers of money computing and social and made them publicly available and this actually allowed for a new canvas space of different kinds of product experimentation and financial use cases so the question goes is can we do the same thing for intelligence and in order for that to happen we actually make sure we need to make sure that there's a way that no one owns the entire model meaning i or you can't just upload the model to hugging face and call it a day the weights need to be sharded across different nodes who participate in training this network and it turns out that this is actually now possible and there are a ton of teams in the decentralized ai space that are actually productizing this so if this happens then there's a totally new design space of development for different kinds of product building and financial experimentation that we can do with the next era of ai now to wrap up pre training just three points i want all of you to take away one open source ai isn't truly open source when there's a single company actually sponsoring the entire gpu bill two sooner or later everyone is gonna need to care about distributed training as if you wanna get the models bigger and better and then three models as a true commodity actually opens up a whole new market space that we can think about sweet let's go into post training so in a single sentence post training has basically taken different strategies like fine tuning or reinforcement learning and making the models actually better at giving answers that are relevant and more niche in the last six to eight months there's been a huge interest in post training for all of these big tech companies because of the success that openai had with its reasoning model and then back in january i'm sure all of you heard the news about deepseq v one and how they trained a really good model at a fraction of the cost through post training so the thing about it is that we need data for post training and not just any kind of data but really high quality data from different kinds of experts in a wider variety of fields and that's why openai actually spends upwards of $300,000,000 paying mercer and scaleai and upwork to find and source all of these domain experts to fine tune and make their models better now the reality is is that there's not many companies that have that kind of a budget to actually spend and get all these people working for them and it sort of runs back the same story with social media where do we just let five companies end up deciding what kind of information we're consuming on the daily basis for billions and billions of people and it turns out that even the ai ogs and pioneers agree here you have people like andrej karpathy and yann lecun talking about how it's essential that we do have some sort of collaborative and crowd sourced way to do post training strategies for reinforcement learning and data labeling creating evals things like that now the issue is is that this is a non trivial problem it actually requires solving things like coordination bad actors privacy all sorts of things but fortunately for us everyone in this room knows that there is precedence here when you start thinking about the technologies that we're all working with wallets blockchains stable coins micro payments zero knowledge slashing mechanisms then carpentry's vision all of a sudden starts to make a lot more sense you can think of it sort of like a decentralized schools approach where you're trying to get wisdom and knowledge from people on all corners of the world and incentivizing them to actually share their very niche expertise in different parts of what they're good at and you can think about it sort of like a society of minds approach or like a swarm approach instead of just having five giant monolithic models how do you get everyone from all parts of the world sharing their experience to make ai better and the truth is is that the last thirty years of open source have showed us that people do in fact want to contribute if you give them the right incentives and tooling to do so they will come and actually shape this next area of technology that we're trying to go for so to wrap up this talk i just want to say i barely scratched the surface of what's going on in decentralized ai today the ecosystem is growing so fast it's genuinely hard to keep up and if there's anything that i think i would i want all of you to take away is that i believe that we're just a couple of months away from a lot of mind share coming from both the ai side and the crypto side into this vertical so if there was any component of this talk that was interesting to you today i highly recommend to start going down the decentralized ai rabbit hole it's intellectually so interesting and i really do think that it's only a matter of time before everyone realizes that decentralized ai is here to stay thank you
okay one last talk with a person who i think is a legend who's like an icon on farcaster i think i met him at one of the first meetups in venice actually hosted by alok from standard crypto when nobody on the network yet knew that i was a woman and so i showed up and everyone was like wait what you're ted and then i ended up seeing him around venice we actually met up at a brian johnson meet up once the don't die guy and now he's up in the bay area but you know him from many things you know him from bright moments you know him from the books channel and more recently you know him from amps so everyone please give a warm welcome to our final speaker phil
thank you ted
great thank you so today the title of my talk is the farcaster flywheel a little bit about me before we get started the the theme of my career is helping on chain creators make a living and i've done this across a couple domains bright moments where we started irl on chain galleries all over the world recent work with amps which has helped people monetize their audience or engagement market and most recently node which is a non for profit that just announced a $25,000,000 founding grant to help disrupt the art industry and create a generational institution for what we're building here sort of categorically i think if you're here in this room you're an optimist and i think it's really important to be an optimist i think the 37 user meme is amazing because if you're like me and you believe that farcaster is going to be massively influential over the next decade what an opportunity to be involved when there's only 37 users
and but i think the optimism is really important to have some grounding in reality and to me i see firecaster as a leveraged bet on several trends first and the most important trend is crypto and i think this is something that's been going on for nearly fifteen years at this point and it's really important that the principles of crypto continue to be paramount and i think that it's really easy as we sort of focus on building apps to lose those but this idea of cypherpunk this idea of decentralization this idea of self sovereignty and owning your own money to me is the foundational layer of what we're doing here next ethereum is an incredible technology it's this global computer we all know and love and most recently i think you know the ability for base to step in and really power low cost transactions and now there's other l2s catching up as a huge tailwind and so if all of these play out i think it's gonna be massively successful if even a few of them just play out i think we have a really good chance of changing the world and so you know as these trends sort of converge there's two variables and they're actually independent and most people think about them as related and so you have control as a founder over one of them which is your burn rate the second variable which is independent is the growth and the speed of the growth of the firecaster network and so as someone who's considering how to build a company or a product or a business on top of the network you need to think very carefully about your strategy because if you have a high burn rate it's only gonna work when the network has reached sufficient scale you need to make sure that your runway times very well with that and so you need to understand what game you're playing and so you know as a a recovered management consultant i can't help but showing a two by two matrix here if you believe that a network is gonna go very very fast and you have a thesis for that it may make sense to put all your chips on the table go all in try to capture that very quickly with a large team however if you think that the network is gonna grow more slowly it makes sense to extend your runway as long as possible and so understanding what game you're playing is really important sort of for being successful on top of firecaster so sort of with that strategic outline here i wanna talk a little about tactics and i think there's three tactics that are really very valuable and i won't go into sort of the granular level that limon did with apps which i think he did a fantastic job but these are more sort of operational ideas you can take as you're thinking about how to build a business so the first is you know the the merkle team is excellent at communicating their priorities and people sometimes mistake these priorities as ways for them to make product decisions and that's actually not the right framing the right way to think about it is the merkle priorities you should be drafting behind them to inform your go to market strategy those priorities are gonna change over time and so your go to market strategy should draft on top of merkle we've all seen it when there's a merkle priority like usdc rewards it creates a massive influx in the ecosystem that can spur engagement with many apps but if you're building your product on top of somebody else's priority you're always gonna be a step behind so i think go to market is an area where it can sit on top of your product layer it moves a little bit more quickly but fundamentally if merkle decides to go a different direction you're not stranded and you can sort of pivot to make sure that you have a new way to distribute your product to users importantly if you're building on top of this network you need to be solving a daily problem for people there are some businesses where people don't need to come back every single day but often the businesses that are the largest and the most successful solve something called the toothbrush test which is something where you're willing to open the app and use it every single day and as someone who's building on top of a social network as someone who's building inside of an app this distribution is really important the reason you build on top of farcaster is because you have this sort of in feed rich interaction and your app should be something that people are willing to open that you can use notifications to drive behaviors and i i really encourage you if if you're trying to build something simply down to it simplify it down to its most basic level if you're thinking about this incredibly complex sort of rube goldberg device that has layers of tokenomics say what could i ship today how can i strip that down and how can this be something that delivers value to my users every single day and then finally you know i i think that building on top of firecaster and especially if you're someone who's recently experimented with many apps is incredibly helpful but you need to have a plan to graduate at some point if your business is gonna be wildly successful you're gonna hit an asymptote and that's even in the case where farcaster grows very quickly and so you should have an idea of what that asymptote looks like and you should have internal signals that you're looking at to saying hey how big do i think this can get as a function of the network size and are we close yet and that's gonna inform your strategy if you start to approach one of those asymptotes you know that you can sort of jump to that next pond and it's really important to time that this is an art not a science but as you're thinking about these ponds you need to realize that the product you're building today may need to evolve over time to be able to fit some of these adjacent ponds you know i'll just say sort of one point on this because it's salient the world team had some really interesting announcements this week and as i sort of step back and i look at the trend and sort of what's interesting as a mini app builder there are several at scale platforms you know farcaster being the most nascent there's world there's telegram who are supporting this sort of mini app forward strategy and as a developer this is really exciting i love the idea of being able to build a web app that i can get distribution through farcaster jump to an adjacent network that has overlaps world is built on top of another ethereum l2 and then eventually the goal and this is something that dan articulated this morning is to be able to retain those users on a long term basis we're building on top of the web that's really important to remember so forecasters are a great way to go to market but ultimately you won't be able to capture those users and have a direct relationship with your audience and so finally you know just to point to builders as people who are excited about crypto we have a tendency to overcomplicate things great products solve a problem for people and if it's really good they'll be willing to share it through word-of-mouth so it's just helpful to sometimes step back to realize that and understand that farcaster is quite possibly the best place in the world right now one of the most underrated places in the world to get distribution but at the end of the day you need to solve a real problem for people and it needs to be so good that they're willing to tell their friends about it so thank you very much i appreciate your
time
we good all right thank you all right when is adrienne's ready we'll go
you ready gm vargasner gm vargasner we're back and we just finished up with the community talks and so we have a couple of the community talk folks up here with us phil and limon or limonae
limonae limonae limonae
limonae limonae limonae limonae italiano limonae i've been calling you limon
the whole time
so i wanna start with there's a couple things i pulled from your talk but limone why don't you mention the last slide how did you end the talk
can i can i say that
yes absolutely
i mentioned like get get the fuck out of like local loss and and i think it's kind of also aligned with some of of your slide feel
yes
when you know keep keeping things
simple
and just like shipping them out in a while and get the get feedbacks from the users and see what's happened and then you iterate on that i think that's one of the best thing you can do to you know grow your idea or product or and many apps help you a lot with distribution you know so that's why i think you know you just need to get out of localhost get the fuck out of localhost
get in your head yeah
get out of here
and and see what happens it's fun
i think that that sort of philosophy is something i see you doing all the time phil it's like getting it out in front of people right so what like what's your advice around that as well
like yeah i think that the team has done a really great job of creating something that's simple to launch on day one and often the idea is not the hard part it's all of the stuff that comes with it and people try to over complicate and should we incorporate a company and do we need to make a website first and like even now if you go to amps.fun in the web browser it's just like there's not a marketing website yeah and often i've been surprised on firecaster by how often then people will step up to help you i've had people offer to make the marketing website for me which i don't think would happen in any other place on earth
that's a
good point
people are
like hey
i can see you suck at this do you want me to
do it
for you so it's kinda nice yeah
we've we've experienced similar things
has no one looked at the gm forecaster website and why are they not offering to build one for me what i
think they have a time i think they've looked at it and gone oh they need help and we're like well no eric erica like out the gate said that to us yeah so we just haven't taken her up on it yet and then she got busy one of the other things that came from yours phil you said being early without the ability of being patient is the same as being wrong was on your slide and i don't you didn't really talk about that but it jumped out at me and i think that is part of what we see happening a lot and i think it kinda tied into even fred wilson's talk when he was talking about soundcloud and you know like basically they just needed to yeah hang in there a little longer and do something a little different
yeah
can you just like iterate on that yeah what was that thinking
i mean i you know i know a ton of people who have built products on top of firecaster and they've been doing it for a few years yeah and there's a moment where the network growth is sort of very spiky and it accelerates and gets really exciting but then it sort of reverts to the mean and if you're scaling up and you have a ten fifteen 20 person team and the network growth sort of that curve that i showed isn't following the trend that you expect you don't have the ability to be patient and you're not in a financial situation where you can sort of afford to wait now it's different from being an investor where you sort of can make your bet and as long as you don't need the liquidity you can just sit on it right as a proud dgen holder i'm like i'm willing to be patient right
yeah yeah
but as a builder you have to think about runway you have to think about operational costs and so i just think this goes back to make sure that you're sort of aligning your company building journey with the network growth journey if firecaster is your primary go to market mechanism
i think that was just really good advice for us like our ourselves are feeling that as well and i you know okay so i'm a active user of farville which i was just like as i was sitting there i'm like i really should i gotta i gotta harvest my crops i'm really behind they're they're gonna i'm gonna have i'm i've already blown my street
we're so addicted
yeah i know totally so speaking to that like you've i mean being launching something like a game on barcaster in that ecosystem versus somewhere else what would you advise if someone's out there building they wanna build a mini app they wanna build a game why should they come to farcaster to do that like what have you seen what has been beneficial to you
i i think i still need to to discover that like most of that of course most of like me like starting to build it was just to experiment and and have fun i i've never built a game before and it's like totally different from building products that i've i've been building in the past so i'm i'm learning a lot through that but i guess the the the most like the most exciting thing that i'm looking for even for farville talking about scaling i mean you have to something more is around how you know having this game that now it's pretty well you know played how we can use like traditional let's say micro transactions that you are we are used to right to use and play with with other games inside a platform that is a wallet and that should facilitate that and and so i think that that's the biggest opportunity that somehow there's a playbook that has been used for on facebook for example like farmville but other famous games and i think that there are ways in which we can try to learn from that and reproduce it inside verkasten with some pairs and days of like you know having a funded wallet and just making use of it to play the game and buy something around it i think that's the best of all of
the games so that's like the next step is figure out how do you monetize this how do you yeah make it sustainable i mean we talk about monetize but what monetize really means is how do you make this sustainable yeah so you can keep it going
for us for us it's crucial of course like it depends on from game to game from developer to developer but we didn't we decided to not launch a token and so we need to figure out like the the whole way probably but still i know that we have an opportunity and and i'm sure that there's opportunity for ton more other games so i would look into it if i was to start again maybe like even before starting building farbir i already before starting with any game actually i already like trying to plan what type of like mix of transactions will make sense to to integrate if you're not planning to launch a token of course
amazing and then amps has been taking off like crazy you've built multiple apps and things on barcaster what's your advice to somebody brand new they have they have an idea where do they start and how do they find those first few users just to jump in
yeah i think you know the work on any product starts way before you write the first line of code and i've been building a community on top of farcaster for many years now and so when the idea for amps came about it felt like the continuation of a conversation and i think that you know as you think about a network there's contributions that you can make to it and then there's ways where you can receive value back from the community and it's like a it's like a bank account if you're not continually making deposits and you try to withdraw it's gonna be overdrawn and and so for anyone who's thinking about building a product but maybe they don't have the idea yet the best thing you can do is go out there and be helpful what i've been doing this entire week at farcon is sitting down with people who are building apps and saying let me open the app in front of you and show you what it looks like for me
oh it's a
good idea
and that's an incredibly useful whenever people do that for me i learn like 10 things that totally changes the road map for me i did it with jacob yesterday i showed kartik the wallet today anyone who wants to ask me i'll sit down i'd love to go through farville with you
i'll show you
my farville i was
about to ask you
it's very sad i have very few props and i'm sure just the conversation with you hopefully i'll give you ideas you'll explain to me why it's so exciting i get so energized from those conversations and i recommend that it's a a great way to
that is such a gift to users and we know it's funny because we just did that with chintan who was here from base with coinbase wallet like i literally was going in to do something and i'm like oh
this isn't working he's like oh i gotta go fix that so it was like literally like
live feedback while we were filming
from neynar and giving him feedback about using the neynar kit farcaster kit and he is like just maybe and i don't know if it was him or rish actually who asked but he's like would you mind just recording your if you record your screen when you're going through yeah but that's why like the far con's so good just to be in person with people because you can get like yeah you can record things and we can do things remote but just like all the quick intermingling of users and builders and founders like to like have that quick feedback is so valuable this week
absolutely there there's a a piece of software called loom l o o m yeah it's a screen recording device it's very popular in silicon valley and product circles install it download it use people's apps and send them a recording unsolicited you will make a ton of friends and then when the time comes for you to ask for feedback you know it's great it's a give and take varun did that for us for amps like a week and a half ago wow and on one hand i want to pull my hair out because it was all these things i was like no no it's not working and it it quite literally changed the direction of the product and for a week we spent time fixing all the bugs and now it's better because of it and so like it took two minutes out of his day and it totally changed you know our week so
that's huge adrianne any last questions before we let them go get some lunch
no let them eat
alright let's go let's
go jam farcaster
alright thank you alright alright alright alright guys
thank you
alright so good so jam farcaster are back we're with jihad and tiab who both did amazing community talks today you you guys had more of like to me it was more of the like thinking ones if you will it was a little less like practical of like not like the hands on
let's mhmm
you know here's how you build kind of things and more like thought provoking i guess would be the way i would describe it so you really kind of drove into like dacc can you just give a quick what is dacc
cool so there was like a couple years ago maybe a year ago even who knows about time anymore but there was a manifesto by marc andreessen called like the techno techno optimist optimist manifesto and there was this whole thing going around or about acceleration and then vitalik came out with this post called his version of tech optimism and he talks about this idea of like diac and so i come from pakistan and so i have seen what the world looks like when you don't have democracy we have democracy in name but really we don't
right
and
so like the most important thing for me is like how do we get that it's actually the reason i'm in crypto is like how do we give access to people of a of a global financial and legal system and that's why i got into crypto and i read that and it like changed my perspective on acceleration and and acceleration kind of like the thing that we hinam and i always talk about
this which is like okay this is so i did i did pair this well yes
okay good
you wanna accelerate but you want to be very
aware of which direction you're accelerating and the consequences of where you're going like you really do need to know the world that you're trying to build
and like and so and yours is a little bit more push forward accelerate mhmm but really thought provoking in terms of like where we're going what's going on so what wanna do like a quick highlight overview for those
yeah i think to tayeb's point like there's two axes of acceleration that are sort of referenced in my talk there's yeah this idea of entrepreneurial acceleration which like there's more companies are going to get started more projects will be built and then there's this idea of you need to orient yourselves in a specific way so that you could determine what you're accelerating towards and i think that aligns very well with what tayeb is saying where it's like we can't just all be accelerating aimlessly we have to have an idea of where we are going and i think ultimately my argument is that the founder's job is to help its users accelerate toward the same common goal so if we are all building toward idea of you know entrepreneurial abundance a global on chain economy then it is up to the individual founders who are building the products that we are using to orient their users toward that same mission and build tools that allow those users to you know achieve that goal themselves so yeah i i said at the end of the talk i think the the kpi there is if your users are becoming founders that's a good thing and in an era of entrepreneurial acceleration then we want as many people to be building products and projects and whatever it might be as possible
and i think that's where you know farcaster ties in so well to everything that you just talked about was that you know you you're yeah your your users are becoming founders over and over and over again in all different areas which is really fascinating and kinda special in terms of the ecosystem so tying like the ecosystem to both of yours like how would you tie in far like yours is a little more direct like how you tied in farcaster in terms of that aspect when you're thinking about dx
dx where does that i like dx kinda i don't like dx right where does that kinda flow into like the farcaster ecosystem yeah
so i'm i'm a big fan of history and so throughout the world there were many times where you used to work a land but basically you were just a laborer and you worked for somebody
i still do it's just called farville now
yeah we
just had lamonia up here so
and and and you would work the land and then somebody else would reap the benefits right and then we finally got away from serfdom at some point where people that were able to own private property and they were able to dedicate their their resources their risk and their capital into something and that's how we got such a large economy and so we've been in a world where these web two platforms have basically owned us and we're basically doing work on their behalf every single day i think farcaster is an example of like releasing us from the shackles of like digital serfdom
and oh i love that
and that's kind of like why i mean that's why one of the biggest reasons why i've been on farcaster for so long and been trying to push it
by the way nice nice flex at the very beginning of a very beginning of the talk flex after that meth id sixty six thank you i was like oh alright i see okay i see you
i mean i'll i'll be honest like i i do think i did play a big part in the earliest days i hosted a lot of the meet ups that we did in
new york
and i and it's because i believed in kind of like the same things that that many of us do now and i and i knew that i needed to get it out to as many people and it like this community wasn't going to grow if we didn't put that early work in so
and that's why we have such a strong large community of farcaster in new york yeah
that's
part of what you guys did so yeah amazing adrienne what what questions from you she's like i'm just chilling over here
no questions
any other questions before we let them meet
i think we should let people eat what what were we gonna say
no go ahead i have no less time what else
you guys looking forward to today that was what i was gonna say
lunch so you did yeah
lunch for sure and like i'll be honest like i was not nervous since i found out that i had to give a talk and then like an hour before my nerves just popped just like straight just as haywire and i feel like i'm coming down yeah from that nerve right there
the adrenaline's coming out
so i'm like ready to like go outside get some sun get some food
talk to some people
check no longer have to worry about giving a talk and so and i'm excited for the rest of the weekend i know we're like doing talks today we're builders day yesterday but i'm excited to just hang out tomorrow with everyone at brunch at paddle ball at all of the things so i'm just hanging basically
excited about the work i've been working nonstop yeah she's been hacking yeah but i've been working nonstop so i'm like i'm feeling that same thing i'm like i can't wait for tomorrow
yeah just wanna hang out with my
can't wait for brunch
yeah good stuff yeah
how about
for you yeah
lunch first and foremost and then after that same i feel like you know this morning i was chit chatting with people but i was getting pulled one direction doing one thing so second half of the day will just be hanging with the internet friends and we'll go from there
alright well go go have some lunch and we're gonna get our last two up here thank you hey let's see this is
a sweet setup
wait is
this is this mic on by the
way it's
on i like the mirror
oh thanks is
that new longer
i just stopped combing and people liked it more and i liked it longer
just stopped combing it actually made
my life a lot easier
amazing
you know
that mark zuckerberg also gets away with doing it so you can too
that's not in the same position
do you
have a chain yeah i've always had this though yeah
you're you're sporting your zuck look
your your zuck luck it's your zuck luck
hair zuck
copied me on that one
that's so funny alright and we're back with samuel huber and yb so you guys both had amazing talks yours is very different you were more about these decentralized ai which is interesting so that's like a new new thing you've been exploring but one of the things i noticed what you said right up front is like you're doing internally on chain now full time yep like how exciting is that
it's amazing i it's this is now year three so
still going at it crazy
and it's i feel like both of you know this better than anyone but it's tough being a content creator and making it work so yeah figuring that out
but you're doing great amazing
thank you
you're doing it you're doing it
yeah
so why did you decide suddenly to focus on that for for here like for your talk today rather than all the other things that i know you're exploring all the time
i think the last six or so months like slowly i got more and more interested into decentralized ai i think i mentioned at the beginning of the talk but i feel like i went through the same kind of excitement with the rabbit hole when i first got into crypto
mhmm
and honestly it was a good excuse for me to learn like new technical stuff in ai and then the most i feel like the subtle part that people miss is that when you go through crypto you learn a lot of different things and you can map that to what the ai community is talking about
like yeah
who is their this kind of builder or what kind of product is memed in this way and there's a lot of parallels that it's just interesting going through both together so okay
cool who's their boro dutch
oh who's your boro i'm
boro dutch what do
you mean
we're all boro dutch
oh yeah
oh true we're all
we're all warped as admins
wybie you could see i can see like you the reason you chose that talk is because it was like you were interested in it
oh yeah
like it was
definitely
and you're such a natural teacher
thank you
so it's super cool that you take us along yeah
on the journey
yeah i i think it's like i was when i first got into the topic it felt sort of meme y because you just added the word decentralized in front of fill in the blank but then when you start going through it you're like okay like sure as with anything like 90% of the concepts probably aren't gonna work out but that's okay the 10% is really interesting and worth paying attention to so i
love it
how do you actually figure out those 10% because for me one thing i have to bias against is when i hear something overhyped and it's not as ai is part of it ai especially is i count it off as not just all the people hyping but you need to take two layers out and then look deep how do you do that and get to the 10%
specifically with this one was like have they actually shipped a product and like can i use it and two what are the perspectives of the ai community on these projects too because then you get like a pulse check of like is this even like a real thing or not so yeah how do you do that
on the conceptual side because people like especially with twitter and content creation they have the same concept over and over like mobile context protocol for no model context protocol for example it's like oh this really big thing you can have ai agents interact with your shit and everyone's doing it and you're kinda like okay is this real or overhyped and then you have to look one way deeper to come up
i mean it's your own like vibe check at the end of the day right like do you think forecaster's gonna grow or not and then some people in crypto might say no and some might be like over the moon and be here right so i think it's just how you sort of break down the papers
what i'm hearing is your tldr is just don't discount it spend a lot of time
yeah like with the
that kind of is a hard sell in like as a general rule i was trying to hope that you because you break down so much so often you just have a general rule it seems like you just have no
i i think there's really nothing other than just going through all of the papers and trying out the products to be honest yeah
this is so i'm like we can just samuel why aren't you doing your own podcast that's what i'm trying to figure out right now i'm like sitting here going why isn't he doing a podcast great questions and i loved your talk and i feel like you know you've been the the go to sort of ted started off by saying that like before they had you know linda you were linda helping people build helping build people build on farcaster
i love linda like we're doing workshops together she's like now amazing with the media yeah podcast series every friday
i know which is fantastic and i'm so glad to see that so what is your advice somebody's coming in yep they wanna build something they wanna or maybe they're not even on farcaster and they they're building how are what are you advising them in terms of like getting their first users and how to go about the process
okay you kind of did a two part question
i know totally realizing it
the second part is they wanna get users the first part is they're not on firecaster or just they're maybe building something
yeah
if they're building something they need to figure out are they building something for themselves and then all of what we talk about in distribution of it doesn't actually matter at all though it's a hard psychological switch it's like i'm just building this for myself doesn't feel like anyone on twitter talking venture back like at scale millions of users i'm chasing them you know it's like self fulfillment is okay and not comparing to that venture backed
right
dream that you're sold and like if you then think about okay i'm not in that part i actually wanna build something that people wanna use it's like are you at that stage where you want have something to go to market we see like ship fast a lot mini apps really help there
yeah but
then if you're not on firecast you need to get into that and understand is well you've kind of like built a web page probably as like your first thing just wrap that technically my joke nowadays is we have docs for that because i'm writing a lot of docs but it's literally like even if you don't know anything it took me thirty minutes to go into a code base i had no idea on and a framework i had no idea of and make it a mini app so it's just the awareness of that and where do you get that awareness it's well we have talks for that literally but just like yash said it's like google it research it and that's what we have to do a better job at is there in the firecast community with all of the changes we have to at least in that communication have some stability of where people can go and i'm hoping that we're part of that with the content we put out and otherwise it's actually just going up to people like there's a reason i don't have a pay gated calendly because people just need a conversation
yeah that's i i think you're right on there and i think the more we're seeing so much more of that now like you said every friday mini apps and i think that's gonna help so much to stabilize and get people on there who wanna build something and you made a really great point though too i just wanna double click on really quick the building for yourself versus trying to build the thing that you can get somebody to give you millions of dollars for at a vc sometimes if you're building it for yourself that might that might follow later on because it's probably gonna be a better experience
interestingly speaking what
do you think
if you go to people who have done the uber successful thing they solve their own problem yeah but that boils down to psychology of being obsessed with the problem it's like i am in zurich in like a co working space near the eth zurich university it's like one of the most technical places to be in the world and there's a lot of technical people and sometimes it comes up that they have great technology but no sense about who the actual end user is if those people solve their own problem that's really interesting because they know that user then they have a problem of going to market but if you're solving your own problem and then you figure out oh other people just ask you about it that's kind of an inkling of maybe you have some first product market fit type of thing whereas if you're just trying to build venture scale it's a totally different game and you're chasing the idea maze until something sticks well it's way more chill to solve your own problem or have that one thing and if you raise venture you realize you're building a billion dollar company or you're a failure and like that realization is what a lot of people don't have when they are like oh i'm gonna build my idea and i've been asked to invest in this like do you wanna play that game for a lot of people it's a great game to play for other people it's probably not and if you're going back to the psychology if you're obsessed with a problem if it's your own problem it's easy to be obsessed about but right other people are obsessed with other people's problems so it's just boiling down to that i think
yeah that makes a lot of sense adrianne questions for our guests
well just samuel i mean it relates to the thousand true fans from like a content perspective but if you're building right
and
yeah you're building your problem you go niche the internet gives you access to the world yeah so like do you have you know can you get your first thousand customers yeah from farcaster probably it's shocking oh like yeah
yeah you'd be surprised like i was talking to somebody building a mini app i don't wanna leak their finances but why
that's why i'm not that's why
i'm not naming names but they just added a mint button in one of the many mini apps we have and they made like $3,000 within a day and just because that
community i might
need it station from you
on this
yeah no like they just added a mint button into a community and into a tool that people are already using and we're just able to monetize that yeah the question then becomes is that a one off or is can you do that and like phil hit a great point in his talk with the tooth toothbrush test and i would say for monetization is it something that is valuable every week like your newsletter or monthly works as well but then it has to be high value where if it one
time thing okay fine but
targeting based on like literally knowing the contents people post on firecaster heavily underutilized is not a one time thing but the attention to do that and drive that every single day into that sale is harsh
yeah right
it's
but i
was just saying but it's also just on farcaster it's so easy to make like because everyone has a wallet
to make that yes to get the distribution you'd be surprised yeah like a lot of people coming to dtech to like work with us is like either technical problem or most of the time distribution and we now have people coming to us just because i have distribution
right
and they just want they repost and working with us because they've been referred or something but there's no and by the
way we can yes you know we can help with distribution yeah yeah
sarah will
talk we are distributionists
and i'm still due to the gm podcaster mini app the problem i have is when people
you don't have a mini app yet we're getting there
actually you know what it is i asked you to work on a project that's not your problem
that's that's one
of the the other one is you didn't put a deadline on me like i'm a i'm a japanese just in time
you wasn't obsessed enough
yeah so i mean i don't know why you're not obsessed with making a jamf broadcaster mini app i actually do need it now okay that's it's like it's it's urgent
it's urgent
i have a flight back that's seven and a half hours
it's yeah
oh my god you're leaving tonight oh wow
i actually have i actually have the i looked into the youtube api because we were talking about it and i figured out that you can get a pay a webhook from youtube apparently on live streams not sure exactly but latest videos i've now actually written a little script that gives me all the latest videos from a channel so i'm a i have okay some of the pieces
i think robin was doing some work on this too right it's yeah yeah
it's also funny like i'm starting to learn front end development like you know how big ux is and like trying all of the different tools i'm now getting into just writing my first like front end because i've always been in the fortunate position to have other people that we collab with write front ends or being able to pay for it and like things like the gm firecaster mini is why i need it because then in one hour i can do the front end and then the back end is natural to me
mhmm so there wow
alright well let's let these men go eat go build our apps you guys have been amazing yeah go build our apps too
instead there are tools
wolves in you
yeah it's two wolves in
one time are
you doing are you are you doing the one soccer and one one
you're doing concept monetization and strategy and under the dumbass building it yeah
that not dumbass but yeah
minus the dumbass yeah alright well thank you your that song talks were amazing and we'll have them all on gm forecaster coming up hopefully next week we're gonna be packaging them all up and yeah so thank you thanks robert thank you all yeah thanks guys