Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
RAC FM🦝 talking DeSci w/Josh from DeSciWorld🧐 P/by Alphagrowth
Recorded: July 9, 2025 Duration: 1:04:18Space Recording
Short Summary
In a recent discussion, the founder of DeSci World unveiled exciting developments including the launch of their Bonfires protocol, a unique fundraising model, and strategic partnerships aimed at enhancing decentralized governance. The conversation highlighted the growing trend of decentralized science, showcasing its potential to transform research funding and collaboration in the crypto ecosystem.
Full Transcription
gm red eye let's get you that mic or that co-host brother all right red eye if you can grab that
co-host son it would be uh much appreciated and we'll get the mics out all right josh we'll just
get the room built up a little bit bro we'll the co-hosts uh up and running and we'll get the mic sent out and everything good morning red eye gm fam
gm gm
how are you doing brother you okay i'm doing i'm super pumped for this interview i have been
watching lots of uh news prior to this so i'm gonna keep the news in and uh seen so many good things peter teal is i've been laughing at
him all morning just as an aside um raw it's dsi is absolutely exploding beyond belief right even
since like the few months since we interviewed dr pepe which was back in uh uh late february or early march right
like even if you just go and look now the wealth of information that's out there is absolutely
phenomenal and something you know i'll tell you about uh dc world that i noticed yeah
not many projects that i would like go to do a bit of research for a show
not as many projects like actually have given me a headache as much as dc world because they've just
got that much out there like that much content even just going through the twitter stuff if uh both both the official account and Josh down there yeah I was like my goodness how do you
accumulate and kind of you know how do you accumulate this and kind of like shorten it
down to like get everything within an hour do you know what I mean yeah oh I do I mean
DCI has been something I've been interested in pretty much since I found crypto. Like my educational and pretty much most of my, almost all of my educational background is in the hard sciences. And so I've been trying to figure out how to marry these two concepts for a while. And it's awesome to see that people have already been chugging away at this uh sort of problem so excited to see uh what sort
of unique stuff they've got under the hood oh bro they're they're like partnerships their
collaborations like they're so uh like far and wide and varied i was like oh my goodness like i
had no idea to be honest when i asked them on the show i had no idea that these guys i mean these guys
are like the real real deal right i've just been reading some of the stuff about what they were
doing in mexico and stuff earlier and i was just like wow this is like see this has got like serious
depth for a project and that makes me excited for an interview you know right what we'll do
red i will just say hello to josh there i see we've got a speaker now and then we'll do the rack fm intro and then we'll get down to business right the
room will build up yeah uh hello josh have you got the mic there it's nice to speak to you again yeah
yes hello guys how we doing oh lovely mate it was great there touching touching grass with you
yesterday and just uh getting to know you a little bit before today but wowzers you've got some project on your hands right there i'll tell you that much
my goodness we do but yeah as you are as you said it is you know communicating it in a way that
people truly understand has been our biggest issue and so that's something that we still need to work
on clearly well actually that's that's right fm's 40. uh red eye i'll tell you it's kind of we
always see it the guests you know do you want to come on people are saying like what's the style
what's the format we always stick to the same thing we like campfire chats dude you know not
only do you get to know the project you get to know the people a little bit but we have a nice
like relaxed style over here at the same time as making sure we get like all like critical info across to the audience you know if there's anything
we don't understand we're thinking well maybe the audience some of the audience might not understand
so we are inquisitive and i will say this josh you know if there's anything that we mention i
doubt there will be but if there's anything we ask or something you know maybe you don't want
to go into it because of ip or anything like this just tell us like just say
no comment you know and we'll move on to another question don't be afraid to stump us on anything
if there's some anything you don't want to talk about brother but what i'll do josh i'll quickly
do the beginning uh for the because we record and archive these and put them on uh all streaming platforms yeah so
when finn gets his studio back up and running our our archivers just moved from uh oregon to montana
so he's busy setting up the whole studio again you know but we always do a beginning
and the recorded podcast version will always start at this point so without further ado good
will always start at this point so without further ado good morning rat fm and it is a beautiful
morning people it is wednesday the 9th of july 2025 and here we are guys bringing you another
dc project uh dc world this is going to be an absolutely fascinating hour because there's so
much to go through these guys have got so much
cooking under the hood and without further ado let's just uh welcome the main man of the moment
josh hi josh thanks for coming on today hey thank you so much for having me i'm really excited to
chat with you guys yeah man it was great getting to know you a little bit yesterday josh and
yesterday i said to you oh well let's find out a little bit about your backstory it was unrecorded i think we only
had like six of us in the room and i wasn't gonna do this today but i went away and i thought you
know what his backstory is quite fascinating and we actually should get it on the recorded
archive show because like i was really interested in your your little backstory very quickly a sort
of close question for you are you the founder of dc world or are you a co-founder of the project
just to start i am the founder of dc world yeah i thought as much i thought as much so just give us
very quickly summarize like your your background
you know have you always been a crypto punk have you always been out of the box right just give us
a little bit of background up to the bit where you decided to create the the project a sci world
okay um and to be clear i am the founder but we have co-founders now. There's people who have made Deciworld their home also,
so I will talk about that later.
But okay, so myself, yeah, I've always been a little bit
against the grain of authority,
as most people that come into this space
and do creative things have been for my whole life.
In my yearbook, it says most likely to to everybody. And my most likely to was
most likely to talk my way into and out of trouble. And that's because I've kind of generally always
had an issue with taking orders from people and the systems of authority that we live in. So
that's just me. Then I moved to China in 2018 so that I could learn Mandarin and I was there for three, three and a half years.
I got into crypto around that time just because of cross-border payments and I then used crypto to
find some things on the dark web and the standard early start in crypto stuff. And then as I was
living in China, got some really interesting perspectives on sort of like social control mechanisms and such things.
Then really went because of that deep into crypto as a sort of technology for ideology, technology for the future of humanity sort of vibes.
And then COVID happened and I was in China and I left China just luckily at the point of lockdown.
at the point of lockdown.
And all kinds of prevailing narratives around COVID
and sickness and how to be healthy
and all these things were happening
that seemed some of them to be a little bit weird to me.
And I had opinions and started doing my own research
and corroborated with scientists that I know
who were working directly on these things.
And it became clear actually
that the scientific establishment is not open
to many countervailing
ideas and generally science which is the pursuit of truth kind of is no longer that it sort of was
quite clearly something more corporate more controlled more institutionalized and it just
sort of led me naturally to creating DeSci World which was originally called DeSci. This was at the end of 2021. After the bull run,
I saw all sorts of obscene wealth going into things that had very little meaning. And I thought,
let's do something but with science, created DeSci as a project, realized that DeSci was a
small movement that had just been starting. So called it DeSci World. And then that was the
beginning. Unreal, unreal and and I'll tell you
something it's exploding beyond belief you know people were talking about RWAs it's like then you
know we've had all of these buzzwords you know DP and RWAs and I'm of the mindset right now that
I'm thinking actually DCI is the one sort of real kind of frontier that the where we're going to see some real
major changes and we'll go into the corporate overlords and stuff uh a little bit later there's
a few examples i'll tell you what i'm going to do josh i'm just going to send uh brian he's he's
actually uh he runs the company that sponsors us alpha growth right uh but brian's really big into dsi
and he's really clued up so brian if you want to grab a mic you know the score you know the drill
right josh before we get into it i know red eyes got a load of stuff you mentioned the team and
you mentioned you know now we've got co-founders right so when i was looking and i tried to go
through as much information as i could
you've obviously got core team members right and then you've got like also collaborators as well where they might be doing other projects but they're collaborating with you guys right and
then obviously you've got partnerships and things so talk to us about the core team the people that's
actually kind of maybe on the payroll or or whatever what's your core team
look like this setup yeah uh so we've had some change over the last four years of course people
in people out um and basically i would put the core team into two strong verticals which would
be development and non-development um the the dsy world team is building bonfires which is also in
this chat bonfires is a protocol by us built by DSiWorld.
So the development resources at DSiWorld are also building bonfires.
But the initial team was firstly myself for a few months and a friend who was a developer
who I sort of subcontracted.
Then in May of 2022, I brought in the first cohort of teams.
And there was two people in particular Jelani and Carolina
both doctors uh sorry both scientists Jelani's a doctor who became early co-founders and supported
at Desai World until now Jelani has actually moved on about a month about a year ago to Amino Chain
which is the probably largest Desai project in terms of investment that A16Z backed them some
years ago he's their chief
business development officer, but they're still friends, homies, advisors of DSI World.
We've added marketers, we've added mostly science-related folks, people who have degrees
or PhDs and such, mostly all on the communication side because DSI World is mostly about communicating
and flag bearing for Desai.
And then about two years ago, I met my now other co-founder, Chief AI Officer, Carlos,
who's in Ecuador. He is a crazy AI-minded chap who basically brainchilded what is now Bonfires.
And as the last two years have gone on, I've also managed to rope in one of my best friends,
Dan, as our CTO, who previously ran his own boutique development shop and built a number of really high quality on-chain DeFi protocols. I managed to rope him into being our other co-founder.
So him as CTO, Carlos as chief AI officer, and me as CEO. We are now the sort of three co-founding
team. But many of the other team members are still here or still support us from the
sidelines and we have now got 14 full-timers and two part-timers in the team across the
verticals of marketing development uh internals and bd but mostly i do bd are you talking our
language bro this is this is a perfect show for us man so i'm gonna ask the question it's kind of the same question we'll ask everybody
in this situation you know uh did you guys do any fundraising i mean are you self-funded have
you went out and done any raises or anything like that yeah so you know as true to this space as we
possibly can making everything up as we go along uh i was fortunate enough to take some good risks early 2020 and made a decent bank
during the bull run. Not very good at selling, so thankfully I was selling to fund Desai World
because I probably wouldn't have much left now if I didn't sell my funds for Desai World. So I did
that and funded it for the first two, two and a half years by myself. Then it became something
awesome and Desai was clearly a vertical that was going to happen.
So I raised a small $150,000 round from a few angels at 2023. Then we survived on that until
basically halfway through last year. And towards the end of last year, if you remember,
the DeSci boom happened with BioProtocol being invested in by Binance. I was actually in the hotel room
at the time when Vitalik and CZ sort of had this private meeting with the BioProtocol and their
investors. I was in that room and left and thought, right, well, now's the time. We raised another
$600,000 from 50 angels in November of last year. And that allowed us to expand the team on the
development side and really push the sort us to expand the team on the development
side and really push the sort of scaling of the product on the infrastructure so we did that and
we're just at the end of our raise like our raised funds kind of now uh sort of right near the end of
the runway and we are currently conducting right now in private but it's going public next week
uh an ico sale which is actually an nft sale so you purchase the nft and it gives you rights to
50 of the token when it goes live in about two months so that's our fundraising journey and
happy to dive into that because some we did some pretty innovative stuff very very grassroots
basically everything non-institutional took a lot of risk all like relationship trust based
um in my opinion it's one of the sort of like truest crypto raises of the last like few years and would love to tell that story actually in some way so that people
can get excited about it. Cause for us with all of the VC sort of influences and token
captures high FDV, low floats, uh, market making shenanigans, we feel very, I feel very
proud anyways, to have raised in such a grassroots way.
That's music. There are is that, and It sounds like you've got a pretty lean team.
I mean, it's not like you've kind of squandered millions upon millions of dollars.
It seems like you guys are running quite a tight ship,
which we always like the teams to be as lean as they possibly can
and stretch every dollar as far as they can.
Red Eye, I'm going to welcome Brian. Anything just to chip in on the funding side of things or anything we talked about there? and you know stretch every dollar as far as they can hey red eye i'm gonna welcome brian anything
just to chip in on the the funding side of things or anything we talked about there
no go ahead and bring brian and i've got a i got a few questions stage up but i'll ask him once
brian gets in here all right guys he's still on my side of the world i believe uh good evening
yes sir 9 p.m um yeah dude look check look, check it out, D-Site World.
It's really interesting.
What part of the stack are you tackling?
Are you tackling IP, or is this kind of like a suite of apps
that are tackling different parts of D-Site?
Oh, oh, oh, oh, oh.
I'll tell you what, Brian, keep that question for a second.
You know the softballs, the lead-ins a little bit, right?
I was about to just say something.
So D-Sci's exploding.
And before we actually got into the product, Brian, that's your question the minute we get there.
Why the need, Josh, for D-Sci?
And why the need for it now?
So, like, just explain to us the existing problem
that because there's many dsi applications there's many projects going on right
explain the initial sort of core problem that all of these dsi projects are trying to sort of
you know overcome for example what's the core problem for a project like you why do we need you
oh god i mean look i could talk for hours on this so I'm just going to try and distill it.
There's the official answer which is the following. Institutionalized science has
brought many benefits but also many drawbacks just as with institutionalized everything and
so the core issues that we identify generally as a space and as a DSCI world is publishing,
the ability to publish scientific journals or scientific papers and to not have to pay
for it and to be able to be paid for it.
And that includes peer review, which is a completely broken system.
You can dive deep on any of these.
Funding of research itself, the incentive mechanisms for researchers and for funders
and for people participating in trials, also a big problem um generally organization and collaboration and
knowledge sharing big problem just siloed institutions again incentives are bad um the
technology actually is really old-fashioned and poor some some universities are still using like
windows 98 and shit like that right it really is as bad as it sounds um and the last one is data massive problem data generally um the access to
data the the ability for you to earn money from your data that gets used privacy over data uh some
of the dangers that you have around that uh the reproducibility of science obviously which requires
access to the data um new movements being spun up all the time around data. There's one thing called FAIR,
F-A-I-R, which is a standard for high quality open source data. It's getting pushed, but,
you know, not very well. So those sort of four core pillars. Now, final bit on that is the
open science as a movement has been trying to tackle some of
these problems for decades but to little success because they don't have the technology embedded
in their methods to really be powerful right they don't have the the monetary systems the asset
control systems they don't have the incentive mechanisms uh to make it something that scales up
above the clearly entrenched corporate interests of
the science today so that's like the official reason and then i have my personal reason but i'll
open the floor to feedback on those things first right okay so brian yeah you were mentioning about the stack and you were mentioning about the the ip i guess you're talking about the fact that
obviously scientists don't get to kind of own their own data or whatever right brian what
are you going to say there mate yeah potentially i mean so so the open kind of um open source data
thing is cool i want to really begin and check that out uh it's funding and research totally
broken i i would say like what i've seen in the industry as the biggest problems is that just the cost of science
and most importantly, like my healthcare science,
and that's what a lot of like de-science is around,
but you could take it and extrapolate it to many other things.
My question would be is like,
how are you guys solving like capital formation funding issues
and what's your new,
that's what I mostly work on is capital formation,
liquidity funding types of mechanisms.
I'm curious what you guys are doing there.
Yeah.
So DeSci has basically mostly as a space tackled funding first.
Research Hub is doing things in publishing, but it's not that innovative in terms of revolutionary publishing stuff.
in terms of revolutionary publishing stuff.
Generally, funding is the thing that has been tackled.
and it's because funding is basically using DeFi tooling,
which has been developed quite well over the last five years.
So people use decentralized fundraising platforms
and then tokenomics or token representations of the fundraise,
DAOs, for example, to do novel fundraisers for sites.
And so if you've been following DeSight for the last year,
most of what you've seen, on the bio protocol the bio dow side um are basically capital
raises for communities around specific disease areas who then issue another capital raise for an
ip nft which is an on-chain representation of an ip share which is then given to the people that
contribute to the capital raise.
So that's basically what Desai as a space has been tackling. And we are also tackling that with
our protocol bonfires. But we're doing it in a slightly different way. And I do still have a
problem with some of the way that Desai funding mechanisms have been built. We'll talk about that
as well. But specifically, our protocol Bonfire is what it does is it allows
you to deploy your own Bonfire, which is essentially an AI powered coordination protocol
around a particular task, or if you will, a particular research question. So you deploy a
Bonfire, you say what your research question is, people can raise to the Bonfire, they have co
ownership of the Bonfire. So if you contribute to the bonfire, you get an NFT, which is your vote. And then basically that treasury is leveraged by the
community, assisted with the project management, is assisted by the AI systems that we've built.
And then you use that capital to go and fund the operations to achieve your research questions. So
it's truly grassroots, like Kickstarter, essentially, but dows for anything. Bonfires is a general purpose,
but you can use it for research questions. And that's why we built it to allow
this capital formation,
self-organization around particular research tasks.
And I haven't, so I haven't, I also haven't touched the knowledge economy, which is actually
the thing that underpins all of this, which I can just quickly do if you want.
So going on the site, it seems like you got like a, like an NFT mint around kind of like
a couple of different topics and that unleashes an AI swarm to then go find and fund.
And then does, is it, I'm sorry, I got to double click it in there.
Is it AI driven the way the funding works
or is it like a DAO vote?
How does that mechanism work?
Yeah, so a bonfire is literally like an ICM,
like a capital race,
like a fund for Kickstarter, for example.
You say, right, I'm launching a bonfire
because I want to research this rare disease.
And that's my research question. So the bonfire goes live. People contribute to the bonfire because I want to research this rare disease. And that's my research question.
So the bonfire goes live.
People contribute to the bonfire, right?
They say, right, I want to back this.
I'm going to give you $200 and that's going to give me one vote.
So each vote is worth $200.
The NFT is worth $200.
This is decided by the person that launches the bonfire, right?
So I'm going to launch it and I'm going to set membership at $200 per membership.
Someone comes along and says, yeah, I support you. They put their money in, they get the amount of NFTs out that they bought. And those are votes. What the AI does
is the AI helps your bonfire to coordinate around whatever your objective is, right?
So for example, it's to research this rare disease, some therapies for this rare disease.
The AI is going to say,
okay, what is the disease? Like, give me some documentation, give me some context. You basically feed it the context and using this AI system helps to structure your context and use it as a project
management to achieve that research question. So it says, okay, now you need to go out and gather
this data. You can speak to these people. We need to use this money to pay for this lab equipment
access, et cetera, et etc etc and the ai helps
to facilitate all those actions but it's still the humans at the end of the day that are signing off
on the votes that the agent puts forward uh deploying the capital where it needs to be and
doing the fundraising etc this is this is awesome dude highly highly i i gotta like dig in there's there's a lot going on
here yeah i mean so we're actually soft launching this week right so like there's a lot of uh
information yet to come out like over the next week two three weeks we're uh we've got docs we
haven't released yet we've got lots of big partners we've partnered with uh asi which is
singularity net we work we're doing a twitter spaces on the app Ethereum Twitter account in like two
weeks time, uh, with the EF we've been invited to go and talk about our product.
There lots of really great stuff happening. Um, so more data, more information,
more, more documentation will be out like soon.
Yeah, Brian, you missed it at the beginning. I said,
this is one of the very few projects that we've went to do a deep dive research
on. And literally we just came away with a headache there's just that much information it's it's
quite phenomenal uh i don't think i've actually come across a project like this uh red eye i do
i have a couple of questions about sort of uh protocol separation and things but you want to
do you want to pick the mic up, Red Eye?
Yeah, sure. I'll ask. Since we've been talking about Bonfire's AI, I have a question related to that. So we talked about some of the, I guess, intro level tasks that you guys
or would be potentially expected from Bonfire. And I guess I'm curious, is there like some, let's say, ultimate task that you're hoping
that Bonfires AI will be able to help with, where you're like hoping to especially train
this for the help of all these individuals that contribute to it?
Or is this more meant to be
a more generalized agent to be able to help people with any part of the I guess the research stack
yeah okay so not to abstract things even more but okay so firstly bonfires is an unopinionated
protocol anyone who can launch their own bonfire for any research question or like job description
or business idea.
It's like an intent.
You say this bonfire exists for this reason.
Anyone can do that for anything.
And every bonfire has its own agent as well.
So it's not one agent.
It's each bonfire has its own agent and your agent becomes a specialist in whatever your
bonfire is trying to do.
So firstly, unopinionated. But the real purpose of bonfires is to give people a tool
so that they can coordinate themselves independently and effectively around whatever
it is they're trying to achieve. And in the process of doing that, what actually happens
is the bonfire trains its own local model, right? And so it actually basically
makes it cheaper and continuously more powerful to run. So the more your bonfire runs, the more
powerful it becomes. And at the same time, that data that it's using to sort of coordinate the
knowledge that it's using, we actually extract that into what we call an Ngram, which is an
on-chain knowledge unit. And we share that on a global open source knowledge network
with all of the other bonfires.
And it's actually an open source knowledge network
for any AI agent.
So even your allies and your virtuals could tap into it.
And this open source knowledge network is a vector store.
So it's only readable by AI.
And it basically will, if done correctly,
if bonfires takes off,
should be the most powerful open source data set of them all because it's an
AI powered system, but it's open source and the attributions of the data are attributed to the
people that created it. And so if it's retrieved, if someone else in the network, another bonfire,
another agent uses your knowledge, you're actually paid for that in the knowledge token,
which is the fundamental underpinning of this knowledge-backed economy that we're trying to build so bonfires is actually a tool for coordination to begin with for anything
but the intention is that it allows people to contribute to knowledge-backed economy which if
done correctly should allow them to just by doing the things they care about just by researching
the thing they want to research earn an income in the knowledge token that's awesome yeah and that kind of answers the second follow-up question
i had was that you know is it possible to earn from any additional training that you as a researcher
or a contributor uh to a particular bonfire or i guess down the line that knowledge you give to
the bonfire gets used up by other people like is there a way for you to earn from your contributions as a whole and it sounds like you can which is awesome
both those things you've got it exactly so both those things firstly uh the the obvious sort of
overt one is your knowledge that you use in your bonfire gets shared with other bonfires and there's
an omission of the token there but actually the training of your local model on your bonfire
is producing data and
also a highly tuned specialist model, which you could turn into a service and pay, you
know, ask people to pay for usage of, for example, if you make a research question around
a particular rare disease, and then your agent becomes a specialist on that rare disease,
maybe you could even offer it as a service to healthcare providers or something, which
would also allow you to generate revenue and the system tracks contributions to the bonfire so it can be distributed back to you
for your contributions yeah this is awesome i you know i have a few friends who are working on doing
some small localized um ai model stuff and they're wanting to do highly specific uh training uh for
models of like really niche information, just assuming that some of these
more generalized large language models won't be able to assist in the way you want them to,
because you're asking for something extremely niche. And I know that there's not, he is,
you know, my friend, and there's many other individuals there, they are not the only ones
thinking about this. So if you do go through this crazy level of deep diving to provide some level of additional information,
that's cool that you can be able to not only share this,
educate people asking the same questions, but you can also earn from this.
I think that's really, really cool.
Amazing. You totally got it.
We did get a question from one of our really good friends and good supporters uh
he basically said is there sort of anything in place that prevents like two of the same
bonfires from existing sort of parallel or is there not i mean each bonfire has like a uid
essentially so they would be separate deployments, but they could be identical questions.
And you could, after launching it, get exactly the same amount of people to join, the same people to join, and the same knowledge and context to be shared with your agent.
So you could make it exactly the same, but it is still an LLM-based product.
And so basically every time you're calling the LLM, it's going to produce
something slightly different than the other call, because that's how these things are working,
right? Like it's a, it's not a, it's not a, it's a statistical retrieval, right? Like, so it's not
a rule-based retrieval. So we're going to, basically you could try and replicate it, but
it will always be slightly different, which is fine. And whenever the knowledge is shared, so if
my bonfire and your bonfire share something on exactly
the same topic they will look slightly different and it's down to the retrieval the vector retrieval
to decide which one is most relevant and like that's just how the system works it will like
our thesis is that by tracking retrievals of the of the knowledge and knowledge network
you'll get this metric which says this piece of information is like useful, like X amount of usefulness. Um, and over time you should see
a divergence between high quality and low quality information, especially on the same subjects.
Superb, superb. And I'm just something that really got me, you know, when you were talking
about when you started the project and if we look off, go, Red Eye, mate. I'll jump in in a minute. Yeah, go on.
I was going to say, this makes me think about how to potentially prevent,
let's say like bad data from being uploaded.
Like let's say, you know, I wanted to contribute some information that I believe is truth-seeking and then someone else submits information maybe to the same bonfire
that is opposite, we'll say in as many objective ways as possible, of the information that I shared.
How would the, I guess, LLM or how would the bonfire be able to handle that sort of mismatch of accounts um and is maybe there a way for like better
knowledge to kind of rise to the top relative to let's say uh additions of information that
are false or have problems yeah so there's two aspects to this and there's like four answers so
the two aspect is uh locally to the bonfire and globally on the knowledge network. Like if you, if you submit bad data to the bonfire, uh, that the effect of that would basically be, um, you'd have a situation where two, two disagreeing pieces of information are in there and the bonfire will recognize that and say, look guys these things are in conflict you need to resolve the conflict and so it identifies the conflict and if it can it will resolve it otherwise
it puts the conflict to a governance vote so the people inside the bonfire vote on which is canonical
knowledge like guys this is a fundamental disagreement like what do we think is true
and then you can do that internally and also users the individual members can have have their
own identity this is not ready at launch but this is like part of the individual members can have their own identity.
This is not ready at launch, but this is like part of the plan.
Individual users have their own identity.
So they have a rating and every bonfire has its own rating and every unit of knowledge, which is shared globally, will also have its own rating.
So that's like reputation systems, which are also important based on user feedback and such.
But the global knowledge network is where it's really interesting. So our thesis is that if you have this system where all the information can be
uniformly shared, like it's essentially a normalized data structure. So it's one vector
store with the same structure and all this different information is stored in it. If you
don't put too much opinion on the retrieval, instead you just let the vector search do its thing,
right, which is you say,
okay, I want to run a research experiment
and it goes into the vector store
and finds via similarity search,
some research structure,
some methodology for doing a research task.
It will bring those into your bonfire.
Now, what we anticipate is that over a large enough network
and a large enough scale and
enough queries, that mechanism of retrieval will produce a pretty decent summarization of what is
good knowledge and what is bad knowledge based on the vector search. And every time something is
retrieved, you give it one point per retrieval, right? And so that basically becomes a metric of
how useful the AI thinks this knowledge is
for the network.
And then you can use that
as a reputation mechanism as well.
And we're actually really excited to trial that
at scale across many different vectors,
sorry, across many different queries,
because we believe that
just based on the semantic search itself,
which is basically vector search,
you'll get quite a large distribution of like
sorry quite a positive correlation between what is good knowledge and what is bad knowledge based
on the usefulness and and this sorry to ramble but this kind of underpins what dc world is about
which is your knowledge and my knowledge might be different and one might be scientifically verified
but my one might also still be useful to some people. And maybe it
could be verified, but just isn't. And so you don't really want to stop knowledge that hasn't
been officially verified from entering into the scientific ledger, because that's how you end up
with dogma, which is where we're at right now. You had one last question, Bobo, and then I'll
hand it back over to you. You had mentioned sort of this like decentralized id for people who are contributing uh on dside world or contributing to bonfires is that something
you guys are hoping to build in-house are you working with other sort of like did providers
uh to like work with uh bonfires of the various products you guys are going to be building
yeah i mean like we're gonna we're gonna create dsw id which will be an id for the dside world ecosystem including bonfires um
but we're not going to build all that in for ourselves we have a lot of very good friends
in the did space the identity space the zk space who already have tools for example we're very
close with the team at human dot tech uh they recently bought gitcoin passport and they've been
doing they're like a primary competitor to world but they're just not like super dystopian it's like
we'll use some of that stuff as underlying infra bunch of friends in the identity space uh and and
really we're trying to build as less as possible internally uh by using as much of like the
interoperability stack as possible uh but what we built at Bonpheise is all in-house because no one else has built it.
But when it comes to like running node infrastructure,
the decentralized computer stuff,
we're working through partners and Akash, Aether, Naphtha,
all these sort of protocols that have already been doing a lot of great work
because that's how this space should be run.
Appreciate it.
Yeah, that's awesome to hear.
You know, you guys specialize in what you guys are specializing and not afraid to lean on uh you know your friends and folks that are specialized
in another thing so making the best use of your time and money love to hear it red eye don't worry
about me son have you got anything else on the menu come on you know what it's like over here
well um you know i'm gonna go i'm gonna go a different route in a minute, you say.
So anything else you want to add right now?
Yeah, I guess I'm really interested in access control for particular data or maybe it's for research or like some sort of analysis that's been done.
or research or like some sort of analysis that's been done.
How are you guys envisioning access control being a part of the user experience?
Like where does access control come into play for DSiWorld?
Is it mainly people that are, say, researchers contributing data?
Are they able to control access to maybe the
raw data or are you guys am i thinking about access control in the wrong way um related to
dside world no you're not you're asking the questions that need to be asked and we also have
like a lot of these questions internally some disagreements with different people on the team have different views on this.
Generally, the premise is this is if you want to participate in the knowledge economy, which
is where your knowledge is shared and incentivized if it's retrieved.
If you want to participate in knowledge economy, then your knowledge needs to be considered
open source.
And that's sort of generally the framework, which is don't put anything into a bonfire that is public that you don't want the network to know.
That's the first thing. And also generally, we believe that the future of privacy is like, no matter how hard you try, your stuff's not private anymore.
So if you want to say something that's going to get you in trouble, just say it in private to your friends.
Like just that's the world we live in, right um they had to do that during the prohibition period like we're gonna have to do
it now and so those are just the general underpinnings but of course we want people to have
access to things they'd be able to use this but with with the privacy they desire so um at at sort
of stage one is do you decide to connect your bonfire to the public network if you do you
participate in knowledge economy you earn knowledge incentives that's how you should run it if you don't you
can just run it locally and it's useful for you as a tool but you don't participate in this global
knowledge sharing so that's the first thing um then you move very quickly towards things like
you know trusted execution environments locally hosted data locally hosted compute and that way
you can actually selectively choose what knowledge gets taken and put out. That requires a bit more development work, but we have all of the sort of framework for that ready.
And then the kicker is that when you share your knowledge with the knowledge network,
again, you assume that it's going to be public. But what actually is happening is in the process
of achieving your task in your bonfire, for example, a research question,
as you're achieving that research question, you're developing almost a methodology to reach that
conclusion. So it's like, okay, what is the subject? It's rare disease. Does it require lab
work? Yes. Where is the lab? This, this, and this. So your local context is it's a rare disease
project. It requires local lab space. The local local context is it's a rare disease project.
It requires local lab space. The local lab spaces in London, like that's your identifying local
information. What's actually happening when you're sharing it with knowledge network is you're not
sharing that localized context. You're sharing the decision tree that allows you to build that,
that research question, that research question. So the methodology, which is, what is the research question? Does it require
lab space? Is it online? And that question is what we share on the knowledge network,
so that other people in the future, if they do their own research question, instead of starting
from scratch and building their own context, they can just import this decision tree from another
bonfire. The bonfire that created it gets gets paid and you get this pre-made context
which speeds up your process but it's not with the identifying information right so it's like
public knowledge or public insights and like local data
that's really cool yeah and i agree uh about the you know in in terms of science, I agree that all of your data and anything related to your research questions should be open sourced.
I mean, in order to reproduce data, you need to know exactly what someone did previously to be able to verify, you know, what they received was correct.
received was correct um so i'm fully on board there i've just been i've been toying with a
So I'm fully on board there.
few different ideas related to uh some decentralized science some more like a
decentralized history if that's even a a niche uh sort of like decentralized academia
no 100 or like archival networks or something yes yes that's that's a much better description of what I'm actually thinking of.
But anyways, I don't want to go too far down those routes.
I'm just really interested in what you guys are building,
and I'm always fascinated by what you guys are thinking about.
I guess which types of users and researchers
and people who are seeking information you guys are trying to help out.
Yeah, amazing. And look, guys, the platform, we're soft launching it this week seeking information you guys are uh trying to help out yeah amazing and look guys like you know
the platform we're soft launching it this week with some strategic inside partners um next week
more partners and maybe by the end of next week the week after we'll have like more availability
to give to some newer partners if you guys wanted your own bonfire you could definitely have one and
you could give a research question that is relevant to RAC and then just use it.
And, you know, the price of using it would be a, you would mint one of our NFTs because that's how you mint a bonfire.
The NFTs are linked in our bio.
We haven't gone proper public with it yet, but we just have given private access to some peeps.
And B, you just talk about it online.
Like if you just said, hey, look, this is a cool thing and this is why it's cool and help us build community around it,
that would be the only thing we'd ask.
We'd even run the compute for you for the first few months.
I'll tell you what, Red Eye, it's not often that like Rack FM,
like publicly on a space is going to say,
like what an unbelievably fascinating project.
Really, guys, just go out there and just do a little bit of research a little bit
of work and you'll you'll you'll just get to this rabbit hole and you'll never find the bottom of
it i mean i did a couple of hours today right about two and a half three hours and i was just
like oh my goodness gracious me how the hell am i going how the hell are we going to conduct this
interview but they always do work out you know and you know red eye you know what they say every good band
right every great band needs a good front man and josh i just want to say mind you rep your project
so well dude you are so like well read so well spoken it's actually great to speak to founders
like this mind we really do appreciate it dude i've got a couple of questions, but I did want to just ask something that I was curious about.
And you founded this in 2021, right?
And I was thinking, well, you know, AI wasn't really sort of like publicly available or, you know, it wasn't out there in the masses that we see now, you know.
The project where it is today, I'm guessing it didn't start out that way or did you
have like this idea from the beginning and or have you had to like pivot and can you can you
just tell me about this like like how you got from where you were when you found it to where you are
now if that makes sense yeah uh in a word carlos uh the guy from from Ecuador, that's the main reason, but I'll tell you the story.
So when I started this in 2021, I started this because I was trying to push back against the transhumanist future that we're heading towards.
And for those that don't know, transhumanism is basically the application of technology to move humanity to post-humanity.
So it can be body augmentations, brain machine interfaces,
you know, living forever, gene therapies, etc. Now, whilst there's a lot of great progress there,
and it's doing a lot of great things, I'm just very concerned about the speed and the direction
and the leaders of those movements. And I don't think people are properly informed, especially
some of the people operating and investing in the space. So started it with the intention of like,
let's do more open knowledge exchange. Let's do more ways for people operating and investing in the space. So I started it with the intention of like, let's do more open knowledge exchange.
Let's do more ways for people to be participants in science
in a decentralized way.
So I always had this as my, that was the reason I started this.
But because of my sort of concerns around transhumanism,
I was also greatly and still am greatly concerned with AI
as just a technology.
And for me, I didn't really understand it at all. I had
basically no knowledge. I basically refused to use it for a long time. Started Decylworld,
wanted to make decentralized funding and decentralized knowledge sharing, but didn't
really know what the mechanisms were. And so started building the company. Then in 2023,
some guy turned up in the Decylworld discord and said, hey, I'm looking to do a collaborative research project,
and I have this idea with this AI knowledge graph system.
I really didn't know anything at all about AI,
but I spent like four hours a day for the next like three weeks
chatting with this guy whose face I didn't see for six months
about all of these systems and all his crazy ideas.
And I thought, wow, dude, this alignment-wise in terms of open source knowledge sharing
and collaborative decentralized groups in research was the perfect alignment, but at
the center of it all was his AI system.
So I was kind of a little bit unsure, certainly uninformed on what ai is over the last two and a half years obviously
i brought him into the team said let's build it together he's been working for us since now he's
co-founder role um over the last two and a half years just absorbed as much from him and the
ecosystem and from doing bd with other ai systems about ai so now I feel very fortunate to be as informed I am as I am about AI just
through this role. I will say, um, whilst I'm like, before I was like, I refuse to use AI because
fuck that. Now I'm like, well, the inevitability of the thing is here. And it's so powerful that
if you don't use it, you will be fall behind. And whilst I appreciate all those things, I am still no less concerned about the adoption of AI for the future.
But I feel like being informed is definitely the right place to be.
And the idea here is that if we can build systems
using the same technology as the establishment is using,
then we have a fighting chance at making something meaningful come out of it.
And whilst I, every time I say that that I hear in the back of my head, the road to
hell is paved with good intentions.
Um, I still think it's the best effort that we can do is actually to try and
build systems that give more ownership, more incentives, more transparency of
this tool to people, uh, cause it's going to be used against us no matter what.
That's what I said about the privacy, right? Like privacy is is gone no matter what so try and do it on your own terms
as much as possible yeah we're not fans of the transhumanism movement i mean we're not exactly
alex jones but we're kind of not fans on it at all i like what you're saying about ai it reminds
me a little bit of like you don't know the horse and cart versus when the the motor car was created you know and people might have looked at it and said you know I'm
never driving one of those blah blah blah but it was inevitable you know it was always going to
happen that people were going to convert and mechanize transport right hey dude there's a lot
of stuff to honestly I'm going to have to listen to this interview red eye you know what we're like
we go back and we listen right and we unpick these interviews because when you're hosting
you can't you can't enjoy just being a listener you know dude you've got some ridiculous sort
of collaborations and i saw i want to talk about this stuff out of mexico and stuff but
i saw that you guys uh have done some stuff with q blockchain so we are well i am personally a bit
i'm a big dow guy you see we've been running deals with our music and things for quite a while we
we were early to the dow game uh probably mid 2022 we were starting to utilize uh dowdows back end uh
tech uh so you you mentioned about the bonfires and you mentioned about the
sort of voting and i believe it's operating obviously on ethereum uh can you just talk to
us because i'll tell you why i'm asking this i did see is it the france way he made that uh tweet
about the like dow red flag right uh can you just talk to us a little bit of kind of how you might have pre-empted
traditional like sort of doubt issues uh the governance did you just have to sit down and
really brainstorm how you were going to have governance for these bonfires or can you just
talk us a little bit into the relationship with q blockchain maybe as well yeah i mean so q uh
were some of our earliest like official partners They're really great guys who were super interested in DeSci from early on.
If you don't know, they built basically a governance blockchain,
one of those cool systems where they were like,
let's build a system that is both compliant in the real world
and useful in the blockchain world
so that more real world companies can use the blockchain.
I think probably actually they were a bit ahead of their time.
And in 2022 2023
no one like no one was using daos so i'm not too sure on the status as of like this year but i've
met the team they're all really nice guys super excited about dsai and they supported us from
early they gave us a small grant which we never used so it's still sat in a treasury uh to deploy
a dao on queue and they had some really powerful um sort of infrastructure for casting votes and
and outsourcing any uh arbitration uh um arbitration to the ecc which is the european
commercial courts something like this basically they had this integrations with existing dispute
mechanisms and it was cool but it was mostly an experiment uh we didn't have enough traction and
the blockchain didn't have enough traction
for us to really push it.
So that was just the Q thing,
but we did a couple of press releases
because it's good outreach.
But specifically to your question
around the DAO mechanisms,
I've been in like a dozen DAOs,
some proper DAOs, some DAOs in name only.
All of them have been really poorly organized.
Deciworld was never a DAOo but we have always been decentralized uh again poorly organized even even just with like
me leading it and that's that's my fault really but so one of the things that we really know is
that if we're going to work in this space we need to be better coordinated so bonfires is at its
heart a coordination protocol but at times you, when it comes to deciding on conflicts or deciding on big decisions, you need to go to a governance vote.
And so we have some innovations in governance, but mostly just the automization of the governance proposals, which is the big innovation.
It saves a lot of time.
But when it comes to like novel governance mechanisms and voting mechanisms, we're working really closely with some of the
biggest brains in the space to integrate these into bonfires. So there's things called community
currencies, there's obviously quadratic voting and other voting mechanisms, all of this stuff,
which we're basically very excited to use bonfires as a sandbox, right? Because as an infrastructure,
what it does is it's a sense-making protocol. So it makes sense of all of the craziness of
the communications of a DAO, and then use that as a foundation to go and do governance on top of.
So I mean I've got loads of ideas on governance but that's the high level.
Let me just get this right though, you're building your own governance
systems like into your protocol, am I right in thinking that you haven't you're not utilizing sort of somebody else's governance systems which we've built the let's call it the social consensus
mechanism right which is how like we've got 20 of us in this call what do we all agree on right
what are we trying to work on like you submit transcripts from all of your meetings and all of
your spaces i submit from mine i have documents i have ideas we're talking in a telegram channel
like what what is the consensus from all of this like the ai system that we've built will find
those consensuses and then post governance proposals so that you can all vote on them and
execute on them that for those for the posting of the governance we use snapshot and we'll use any
of the other well-built protocols as they come out and are built out but what's difficult and what we
have built proprietarily is that sense making of all of these people in this decentralized protocol doing things
and like sharing information. Like what does it all mean? What are we trying to achieve? Like what
are the conflicts? That's what we've built. Fascinating. Absolutely fascinating. I'm going
to look at that a little bit more as well. I didn't get to spend enough time on that because things like uh governance uh fascinates
us and and you know we've we've done a lot of work ourselves on doubts i mean we've got on our own
particular day with the rack fm deal we've got like i think about 120 uh past props up on there
well not all of them passed obviously but we've done it we actually did a beats by governance
believe it or not uh i was telling you last night we released music you know we actually did a thing where we gave the community who were holders of uh an
original uh album nft right we kind of give them like the voting rights and uh every week we gave
them like three beats to choose from and the selected like you know whichever beat one we
actually decided to write a track to that and
then put it in an ep so the ep was actually called beats by governance and it was 100 decided by the
community nobody could have ever and it was just the beats you know we had to go well finn had to
go and write the track to the beat after but do governance fascinates us i'll stop talking because
red eyes got his hand up and we've still got some more questions. Go on, Red Eye.
Well, I guess, you know, I've had a bunch of ideas these things that I would be interested in potentially using DCI World and Bonfires to help with.
But I'm curious, how will DCI World or myself as a creator of a bonfire help with attracting additional contributors to whatever cause that I'm working on. So let's say I wanted to do sort of,
I wanted to do like water runoff testing around factory farm areas.
And it doesn't matter, it could be globally.
I live in the US,
so I'm more interested in US-based factory farming.
So being able to do water runoff testing there, there, you know, it's highly restricted to do any sort of recording or
testing inside these facilities. Or like, if I wanted to do look at collect data changes
to water composition, like the minerality or the, you know, the chemical
composition of water after natural disasters or composition of trash and plastics in local
waterways or something like that. If I wanted to try and do something like this and create
data and research and analyze the data and try and produce additional information,
the data and try and produce additional information how would the site world or myself as a creator of
that bonfire dedicated to this get additional contributors to help with this endeavor or
provide additional data yeah so it's a great question and like the discoverability uh and
the incentive for joining is sort of a core reason that we build this which is you don't know what
you don't know and also maybe you don't know how useful you are to somebody else. And so what
we're trying to build with this system is, okay, on the surface, you have a bonfire, which goes live.
People can contribute to it because it's a good system for organizing around a task. There's
accountability towards the goal. So the AI agent that is attached to a bonfire actually uses the bonfire's research
question as its anchor. So when you say, I want to raise $10,000 to test water, and then you go and
say, I want to buy a car for $5,000, the agent's going to be like, well, why would you do that?
And then you say, oh, I want to buy the car so I can travel to the water plant. And it's going to
start throwing up all these questions to you and ensuring that it's aligned with the goals
of the bonfire, which is an accountability metric and then of course you still
have to pass the governance vote and the governance vote is held by the people that contribute the
money so as an incentive to contribute funding it's accountability for the goal which is good
and no nothing in crypto really has that anymore it's mostly just like raise the funds and get the
token and hope that the guy
behind it does the thing. And so that's cool, like full accountability. You could also share
the context in a sense of like, you can contribute your own research and have that make an impact,
which is cool. And your information is also shareable in this global knowledge network,
which means more people can see your work and you can get incentivized for the usage of it.
But then also
with the bonfire system by being connected to knowledge network like people can query any of
the agents any of the bonfires or even through a separate agent system via connecting through api
to the knowledge network they can say like oh i'm interested in doing water treatment stuff like
who's working on that and it will say these guys over here. And it becomes something that's a discoverable through the vector search,
which is also really, really exciting because it's searching data is very difficult,
but through vector search is very easy.
So we basically use these systems to help the discoverability of your work.
And that's just as a permissionless, unopinionated thing that the system does itself.
And then of course, if you're working on dope stuff like that,
and we're Desai World, the founding company of Bonfires,
we're going to push that through our network as well,
which is very strong, and we know everybody in Desai.
So basically, we'd say, hey, guys, come support here.
We'd help you handing gloves to support you to find Desai people
and people building decentralized testing and verification stuff
and that sort of thing.
That's awesome. That's really awesome. I appreciate the context there.
Guys, I'm really sorry. I have to go. I've got another call now. But what I will say is firstly,
thank you so much for this. This has been really awesome and definitely invite me back on at some
point. We'd love to chat, especially as we forward with our our sort of releases um anybody who's interested please follow me dc world and
bonfires uh all in this call right now um follow us whatever reach out on the bonfires twitter
there is a telegram channel in the bio join the telegram channel say hello and this is for
everybody in this call i know i don't know brian's still here brian just left uh you know feel free to support us by joining that telegram
channel share it with your friends we are soft launching the protocol this week and next week
there is actually a mint button in the bio as well which we haven't publicly announced we just put it
in our bio feel free to go there that's the ico uh launching at 10 million fdv in september the nfts are the ticket to the ico
all this information is coming out next week but go support us and tell your friends and all that
good stuff oh josh i can't thank you enough and let me tell you uh from the team at rackfm you've
got some really ardent supporters in us we we absolutely love this project and what i'll
probably do is i'll probably uh retweet this
out probably tomorrow the space and i'll put up some time stamps uh will definitely help like you
know do our little bit to help generate some interest and get some eyes on you guys from from
our side 100 i really appreciate your time today josh been a fascinating insight to your project
okay it's been awesome i really appreciate your support guys and
like as you as you know as we grow and as we we all grow let's keep chatting and it's been it's
been fun no worries mate go to the other meeting i know you're a very busy guy and we appreciate
your time but take care good night and god bless okay cheers team bye bye wow red eye that was absolutely fascinating wasn't it holy shit dude yeah
this Mary's kind of my my two current primary interests in you know for
professional stuff science and crypto and sounds like they've got something
really really interesting here so did you see the tweet I put in the nest dude
did you this is one of the only last things I didn't get to ask him nor did
you see the tweet in the nest or where they're like Mexico head right that's
cool did you but yeah like dude UNESCO like Mexican senators and stuff like
that like this is not easy for like a blockchain based project though even to
be able to get a foot in the door to be able to even kind of talk to the people of that level
never mind like helping out on projects and stuff absolutely like these guys are about as legit as
legit comes do you know and what i picked up on as well there was a real like differentiator
in what they're doing compared to like some other like you know flat what i would call fly by night
project you know what i mean people heard deep in and they're like oh we're gonna go we're gonna be
a deep in project you know people heard like uh rwa's oh well we're gonna be an rwa project you
know the vision kind of from day one was there about
what the guys wanted to do yeah really cool really cool i've really enjoyed this hour yeah
yep i saw finn just pop in very briefly and then yes shall we shall we close it down and i might
spin up a little half an hour after party or not are you busy or not oh yeah i'm down for a half hour after party okay because i know b-bans is going to be traveling i
think she's on a trip so she might want to jump in and you know which is like she gives us every day
right yeah i'm totally down to do that well we might get finning as well right i see us are all
there foxy hey kit shorty we'll spin up a little unrecorded by all
means jumping guys you're more than welcome uh we will end the show here but i'll spin up straight
away once again this has been rack fm you know we make shows that we want to listen to we will
put this up tomorrow with some time stamps and a bit of you know information etc but we have thoroughly
of you know information etc but we have thoroughly thoroughly enjoyed tonight and if you guys know
rackfm we we don't we don't buzz like this that often about a particular project or a particular
interview you know we have good interviews and we have great and then we have like tier one
this was a tier one interview today really really on the money but this has been rackfm guys it is
wednesday the 9th of july 2025
and this is rack fm over and out good night and god bless take care see you on the after party
bye bye
Speaker
bryancolligan
from
AG Product
DeSciWorld
from
DeSciWorld
