Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
Surfing the Scientific Ocean: Introducing GraphSurfer @beeardai @psy_dao
Recorded: July 1, 2025 Duration: 1:04:42Space Recording
Short Summary
In a groundbreaking discussion, the team behind the Be Art AI system unveiled their innovative project aimed at revolutionizing psychedelic research through AI, while announcing strategic partnerships and funding initiatives to enhance scientific exploration and democratize access to knowledge.
Full Transcription
Thank you. Thank you. Thank you. Thank you. Jay and Beata, can you speak?
Hey.
Can you hear me well?
Yes.
Do you want me to call you Jay or Yafik?
That's easier for you.
Well, you tell me.
You tell me.
Let me be Jay.
Okay, nice.
Hi, guys.
Nice to see you, Veta.
Nice to hear you, Riley.
Riley.
You guys ready to surf?
You guys ready to serve?
Jay, I made you a co-host so you can invite speakers.
Maybe you can make the co-host actually.
I invited you to co-host.
It's 9.30, everybody. I'm going to invite our Discord community,
which we have good active crew working on psychedelic stuff
for those of you that don't know we are side out we are the psychedelic decentralized
autonomous organization we are doing all kinds of extremely cool things in the psychedelic space
all kinds of extremely cool things in the psychedelic space and beyond.
We're doing drug development.
We do art.
And one of the coolest things we're doing is artificial intelligence.
We funded the creation of the Be Art AI system.
It's a research agent that is going through the entire psychedelic scientific world, exploring it like a bee goes out into the world looking for nectar.
And it brings it back to the hive in order for us to create the kinds of scientific research and development that we want to create.
I am grateful to all of you who
are here. Thank you for coming. Tyler, Steven, Denkata. Oh, we also have Cuba here. Cuba is one
of the developers. Let me invite you to speak. We have Beata, who is the chief of staff of
chief of staff of the Be Art system.
And we have Jay, who is the main developer.
And Tyler, thank you for the retweet.
Let's get started, everybody.
First of all, who here is a surfer?
Anybody surf? Banta? Jay?
Do you guys get after out on the waves?
Just on the digital wave of science.
Steven's a surfer. Banta, you ever get up on the board?
I think we're all from landlocked places.
I come from Colorado where we can eat mushrooms legally.
We can ski, but we don't do a lot of surfing.
So let's talk to Yasik first.
Yasik, you, or Jay, my bad.
You, first of all, you got all, you're the artificial intelligence guy.
And I got to tell you, I am starting to love artificial intelligence.
I wasn't sure about it at first.
I didn't know how to navigate it at first.
I didn't know what it could do for me.
And now I kind of use it like an operating system.
It is sort of like when I have questions, I'm just driving around.
I'll turn on the chat GPT that talks to me and say, how do I do this? How do I answer this email?
What's the size of the sun? How does a combustion engine work? And it'll tell me. And it's crazy.
And what you are doing is, I've seen you guys tweeting out
all this super interesting information
about the way these scientific
concepts relate to each other
and it's just super
fascinating knowledge and we retweet
it a lot and we parse it
and so
just tell me how it's been like developing
this super cool artificial intelligence stuff
yeah well actually it's funny that you super cool artificial intelligence stuff.
Yeah. Well, actually, it's funny that you're saying that you're using AI as operating system because this is part of the narrative basically around L11s.
Can they actually act as a new way,
a new kind of operating system?
I think some more and more people
are having the same vibe around the LLMs,
AI and LLMs specifically,
that they are just more than just systems that produce text.
They can do way more cool stuff.
It's because they can reason.
They can communicate not in binary numbers like computers used to,
but in language that we can understand as humans, so our own natural language, whether it's English,
Polish or whatever language you speak. I mean it's definitely a new way of how we can interact with computers, which I think is for us way more natural.
And I think this is going to bring more people
to become power users in computers with the use of LLM.
So definitely an operating system is a great way to frame it.
So happy you see it this way.
What's your main goal with VR?
Like in a perfect world, what kind of thing would it end up being for regular people?
You know, I think we are on the verge of a new kind of world.
And I think this kind of statement you might have been hearing many times at different points of time.
But I think right now it's really something special because the barriers are falling.
And specifically to science, this is really interesting.
I believe that, and I can also talk from my own experience because I'm using AI and LLMs to learn a lot basically constantly
you know I'm constantly learning something new and I think this is something that is going to be
more and more present in our world where like being adaptive to new things new skills new
technologies is going to be really something that can keep you in the loop um so when we look back five years from now we say okay we created the yard
and it did this thing and we're like oh my god that's so cool i mean just the way like gemini
it tells me you know how flowers work five years from now what do you like? Oh, my God, BR did this for us. You know, so how I see BR, like, I mean, what BR is, is this autonomous system for seeking underexplored and interesting novel directions in science.
And how we can do it is by basically utilizing scientific knowledge that was produced by humans.
But there's just so much of it, you know.
The latest statistics that I've been reading was that I think in 2022
there have been over 3.3 million papers published in just one year.
And from my own experience as a researcher i i mean i struggled
with the amount of papers that that were produced every day even in my own little niche you know
uh i studied uh knowledge representation in large language models um and even in this, I think, kind of a niche of artificial intelligence domain, there's just so many papers, like basically daily coming, that it was always almost impossible to keep track of what is happening, what has been talked about, what are new discoveries, et cetera, et cetera.
What are new discoveries, etc., etc.
I feel this is still and also validated by talking to different researchers from different fields.
This is like a common problem in science.
It's just because there's just so much stuff happening that is almost impossible for humans to keep track of everything.
humans to keep track of everything.
You know, and so this is the hypothesis
that there is actually a lot of
actually human knowledge
that was generated
that is still waiting to be
linked together. And I think
like a really good example here is
what was happening already at the end of
80s.
Then there was this guy called
Dr. Swanson, if I remember his name
correctly but what he did he came up with this new way of finding hypotheses
ideas for science and this was specifically by looking at scientific
papers in non-overlapping domains
and then trying to find links between concepts,
between discoveries, between conclusions
that these papers came up with.
Because this is in the 80s, right?
So the sheer amount of papers that was produced every year
was way, way smaller than we have right now.
But still, there was noise that was just there waiting to be discovered, to be linked.
And they found some pretty amazing stuff.
Like one example is that they linked magnesium to migraine.
And for example, now migraine is also used as a supplementary to kind of ease the problems with migraine.
And it's just by analyzing two disjoint bodies of literature and finding these links, right?
One paper talked about how magnesium is affecting, let's say, X.
And the other paper said, like, how X is actually affecting migraine, right? And just by looking at this, but only in the context
of these two disjoint bodies of literature, you could actually make this link. So fast forward now
from the end of 80s to now we have over 35 years and the idea is that we don't really have to do it manually anymore because we got to the point where LLMs and
LLM-based AI agents
are really great at exactly this, processing text,
synthesizing information, looking for links, you know, they can
they can reason about what they read right now because of the LLMs
of these capabilities that LLMs are providing them. So these agents became really a really powerful tool.
And then you have scientific literature, right?
So our approach is as follows.
We scrape scientific literature that is out there in the internet.
We turn them into knowledge graphs and
knowledge graphs are this really efficient represent structured
representation of of data and how it looks like it's it basically represents
the world in form of triplets and a triplet is in the most basic form its subject relation and object and for
example like one of the easier examples could be cow eats grass it is the
relationship cow is the subject and grass is the object that is being
operated on but if you think
about science and you can obviously have much more more interesting triplets like
well one of the things that I've seen you guys do is you've been tweeting out
the super interesting connection say the psilocybin and the microbiome.
And so you can think about like how eating mushrooms will affect your gut
health.
And then from there you can link that from how it affects your gut to what
kind of diet you should eat.
Maybe there's a certain thing you should eat before you eat mushrooms.
And these are the kinds of things that could really help a real person in real life with the things that they want to do.
And so you've created these graphs that are super, super cool, but they can also be a little bit difficult to look at, at least for kind of average smooth brain people like me that just like to surf and ski any
mushrooms and so now it's safe you you must have looked at these graphs and said okay maybe there's
something else we can do with these things right i mean like i think what is uh so from a visual
perspective i think like how it relates it's it should be quite explanatory why it might look like this.
But what I think personally I find trouble with is that, I mean, not everyone has to be an expert in a specific field to understand what is the message behind this chain of relations, right?
behind this chain of relations, right?
And here, like, what is incredible is that AI can also help here, you know?
Like, they already gain a lot of good knowledge while they are trained,
and you can also equip them with these really interesting tools,
like web access or ability to consult a database,
but there's more information.
So they can actually be great guides and tutors
for your own exploration.
And actually, this is behind our idea for GraphSurfer.
So we think with the current state of AI
and where we are actually heading,
there should be no more barriers for anyone curious to explore a scientific landscape. we think like with the current state of AI and where we are actually heading,
there should be no more barriers for anyone curious to explore a scientific landscape.
And we felt that the analyzing and learning about different relations that are actually in the graph or in the knowledge graph,
specifically what we did for GraphSurfer is or in a knowledge graph. Specifically, what we did for GraphSurfer
is psychedelic science knowledge graph.
But the idea here is that the curiosity
should be the only thing limiting you in this exploration,
not your knowledge that you have or money or anything else.
It should be curiosity, you know,
because this curiosity is something
very human. I think that would be really hard for the lamps and AI to replicate.
It's because we as humans, we start growing and we have access to this world around us and we can engage in interactions with this world. We can take the stone that is on the ground,
interactions with this world. We can take the stone that is on the ground,
pick it up and look from every angle as possible, then throw it somewhere else,
you know, and like we can have these interactions. So because of these interactions we start asking
questions and this is what curiosity in my opinion is, you know, wanting to learn more.
And this is a very internal drive we as humans are, have.
And I think this will be really hard to replicate in the current systems of AI we have right now,
because specifically AI systems that are on everybody's mouth right now are the LLM-based systems,
and LLMs are trained by reading these massive amounts of text.
But the key word here is reading.
You know, they can read about how it feels to touch a stone
or how it feels to get into the water,
but they would never be able to experience this, you know.
And so that's why I think like curiosity is one of the human traits
that it would be really hard to replicate in AI systems.
And but at the same time, I think this should be the only limiting factor
of you learning more, you know, like your own curiosity
to wanting to explore more and learn more about the world.
Yeah, and so you create a Graph Surfer.
It's a Graph Surfer XYZ, and it's this thing that,
it's this web that you can push buttons on, play with, go to what you're interested in.
If you like psychedelics, you know that it affects a certain kind of receptor in your brain.
And so you can click on that receptor called the serotonin 2A receptor.
And then you can see like, OK, the serotonin 2A receptor does this in my brain.
It makes me like happy.
It makes me sad. It can make me trip. And then you can push these buttons. So you can see that,
oh, the serotonin 2A is related to psilocybin. It's related to this other thing.
I want to, who else here has had a chance to play with GraphSurfer?
And what else have you seen with it?
What has made you interested about it?
Maybe Cuba, one of the developers, or Beata.
You want to speak on this?
I want to hear from more people about, because it's password protected at the moment,
but I've been lucky enough to have the entrance the keys to the kingdom and mess with it
cuba how how's it been for you sort of playing with this
or jessie i think i might be driving a car now, so he might not be able to start talking, but I think Bata is trying to connect.
Okay. How about you, Jay? What have you seen that's been really interesting in playing with this craft server?
Well, you know, for me, one of the most interesting things is, while observing this, because in general I'm a fan of structures and how things are basically built from inside is how everything relates you know like we can read a
paper and we get these all the information that might be there but what
really interesting is of how concepts of how scientific concepts relate to each other because then you
can actually stumble upon these very unexpected relations and you might start seeing it more as
disconnected parts as more of a broader picture that actually everything is connected you know
everything it's we are not living in um independently like in
separation from everything else but actually we are part of this one big thing and everything
is connected you know and the graph a knowledge graph as a structure it actually can visualize
this really nicely it shows really the connections between things and And I think this is what is most interesting for me.
But obviously, like, I don't know if you guys know this,
there was this Wikipedia game where you could actually,
like your task was to just optimize of how many links
you have to click on in Wikipedia to get from A to B.
You know, because in Wikipedia,
actually Wikipedia also runs on a graph that actually links everything together.
So here is a little bit as well like this, you know, you're actually an explorer
in this vast space of knowledge and your task is to really explore.
And the thing is that sometimes you might really stumble
at a very unexpected connection that is over there.
So I see GravServer also as a part of a bigger vision here,
because you mentioned, you asked me to tell you more about Beaud.
And this is just a part of a larger ecosystem.
We want science, as I also mentioned this,
that I think we're going to start seeing barriers falling
for scientific engagement.
And more and more people would be actually able to join
scientific inquiry by just being curious.
You know, like before this was strictly a job
that you actually had to
to have to be a researcher to do all of this research stuff but i think we are actually on
we are starting a new world where science would be more seen as
yeah I had this conversation today and I really like
this one
the one analogies
I'm going to quote it but science is going to become
more of an art than this
science right now
because we can delegate science
to our AI
companions and actually focus more on the
higher picture of what I think
also was in the back in the days.
Like back in the days you had these grand discoveries where people were distilling their knowledge for years, you know,
to then come to this grand advancement, whether it was a relativity theory or something else. For now, a lot of science is being conducted to improve a benchmark or do a very small study to get a paper because this is a job eventually.
You have to do this to get citations, to get higher in ranks at the university, et cetera, et cetera.
But I don't think this is how it was at the beginning.
And I don't think it's going to be like this in a few years, right?
So the curiosity part is really a big thing here.
And more and more people will be able to start joining, having AI companion,
like we have in GraphSurfer, we have Kai,
who will be very eager to help you with understanding all of the concepts
and relations because here's the world where you are,
where have you come from,
and you can ask whatever, you know?
So you can really ease down the barriers
for you to join.
Yeah, instead of science getting more and more complicated,
where you have to get smarter and smarter
to understand more and more things,
it seems like Grabsurfer is making it more and more simple,
where you can look at things very visually
and you can work with things very visually, and you can
work with things in the way that the normal human brain works. And so I know for me, looking at this
graph surfer and trying to explain to people what's so cool about it, it showed me the relationship
between this receptor in the brain and this medicine, and then this protein that I never heard of before.
And I can't even pronounce it, but it gives me the ability to then click on this thing that I,
this protein to say like, what does that protein do? How does that work in my brain? How does that
work in my body? And then there's the potential. I was showing this to some scientists and they say,
oh, I've never heard of that either.
And there is a potential that I could then look at this protein, ask questions about it and potentially do some science about it, potentially work on it, potentially create a medicine off it.
Right. I mean, like, it's nice that you mentioned this because, so as I mentioned this,
I envisioned this to be a part of a bigger ecosystem.
So what specifically we've been busy at BIOT
is designing knowledge graphs from scientific publications.
And one of them is actually connected to GraphSurfer
from Psychetic Science Knowledge Graph.
So yeah, we are focusing on building the Knowledge Graphs as well as the infrastructure around it.
So speaking here is on Knowledge Graph,
we allow AI agents to do their walks or traversal so they can go and start
exploring the Knowledge Graph as you do in GraphSurfer as well.
It's just they are looking at
from from a genetic point of view um and they have actually a pretty specified task their task is to
traverse the knowledge graph and look for connections that might be underexplored on
not underexplored connections
or maybe not obvious connections
that could actually lead to a novel testable hypothesis.
So once they find a connection,
and by subgraph I mean the connection of multiple nodes
linked together by a relation.
So for example, you might have like this relation,
toxicology studies evaluates neurotransmitter release,
ayahuasca contains neurotransmitter release, and this modulates serotonin receptors, and you have all these connections, right?
So a subgraph is just a piece of these connections, right?
Like a chain of thought you could think of, you know, of how different concepts relate to each other.
And basically you have this subgraph that is being fed into the Argentic systems that we developed to work with these subgraphs and produce a hypothesis.
What they do is we are trying to mimic like a scientific process there, where they take the subgraph as a starting point and then they start debating. Debating, asking questions, then looking for more information in the scientific literature to supplement this information.
And they engage in this iterative process to eventually end up with hypotheses that is then being posted on Hypegen.
This is our social media for hypotheses and subgraphs.
This is our social media for hypotheses and subgraphs.
We're the only players that can post stuff
or AI agents that are working within our system.
But the only people, but people who can interact with them
are only researchers or like people who want to, right?
So there are no AI agents involved in the evaluation phase, but only real human people that actually can take this hypothesis and bring it further.
So having this in mind, there might be many ways of how a subgraph can enter the hypothesis generation system.
As we do it now, we let our own agents to explore the graph.
But at the same time,
this also can come from sources like GraphSurfer.
This is where the big vision actually comes,
that the curiosity-driven exploration of the graph surfer,
of the knowledge graph that is in the graph surfer,
they can lead to interesting sub-graphs that can later on be fed to
a system that generates hypothesis. As I mentioned, curiosity iss that can later on be fed to a system that generates
hypotheses.
And as I mentioned, curiosity is something that is going to be really hard to replicate
in LLMs.
And I think it's going to really take some time for this to happen, if ever.
So we are looking at this from different perspectives.
Well, LLMs, they can't sort of bring it back to what matters in the real world.
They can't really say like, here's why it's important to know about this protein.
Well, they can learn this maybe, but they might not be surprised, for example, by certain relations.
And I think people can still be surprised and the surprise can be a motivating factor for exploring more.
So this is also coming back to curiosity. So basically the idea is that you can
we want to turn it into more of a gamified experience where you can literally like explore
it as a game where you can actually find new subgraphs that can actually maybe come up with some reward
for finding a subgraph that was never explored, right?
And maybe you find a subgraph yourself
and you think, okay, this is such an interesting subgraph
or like such an interesting connection of ideas.
I'm wondering what an AI system could make out of it, you know?
And you can send it to our system,
it would generate a hypothesis
after this scientific iterative process
of pondering on this sub-ref.
Yeah.
I want to remind,
I want to thank everybody for coming again.
I want to remind everybody what we're doing here.
We're CIDAW,
the Psychedelic Decentralized Autonomous Organization.
We funded the creation of this multi-agent AI system
that's combing scientific literature
and coming up with hypotheses.
It's starting with psychedelics
because that's our area of expertise,
but we're going to expand it to,
or we are expanding it to other domains
and with the idea of creating really interesting things
that can help real
people in the real world.
Anybody else have anything to say about what they found by playing with our new GraphSurfer
or with HypeGen?
Can you hear me?
Yes.
Hi, Beata.
So nice to see you.
Yeah, sorry.
I couldn't unmute myself before.
So I just wanted to answer your previous question,
what I love the most about GraphSurfer.
And so you guys know I'm not a scientist,
but I'm really passionate about science.
And what GraphSurfer, what I love the most about it
is that it turns the passive act of reading and learning and interacting with science
into an active exploration. So without making me feel like I need to have already the answers and
have the understanding and have the idea what it is all about. But it gives me this feeling of being active here and discovering and learning and not just
consuming science. So it invites to create something new and just look for something here.
Also, I like to just play around and click around and maybe maybe I will find, you know, an Easter egg,
like a super curious connection that I don't understand.
And then with Kai, our chatbot,
he will explain me and actually say,
hey, Bata, you just found something super, super interesting.
And let's explore it more and bring it to HypeGen
and to real scientists this time.
And at the moment, it's password protected.
Is it always going to be password protected?
And actually, if you guys are listening and want to check it out, the password is psychonaut2024.
Nice.
And is there something in specific, psychonaut20aut 2024 that you have found in your like playing
with this any can you think of anything you've learned from it anything that's piqued your
interest not from the top of my head no i would say like what I found to be really interesting on one of the
recent journeys that I took was the effect of psilocybin on inflammation and how that can be
used to lower it. I actually did not have any idea about this link. So this was something really
surprising for me, but also something really novel
and very interesting yeah i know we've been talking to a scientist here at side out about
maybe doing some research and he found his hypothesis just through the old-fashioned
way of reading through papers but he and it's a really interesting novel approach to a medicine that could potentially affect human beings in an interesting way by creating an antidepressant or like an anti-inflammation agent.
And the way he did it was by reading a paper from 1970s, finding a little citation that people had not heard, and then coming up with a hypothesis.
But potentially, the graph surfer could do it all automatically and do it faster, more intuitively,
just by clicking through from one thing to another, using the surf instructor Kai to get to the next thing.
And I just find that really exciting.
Can you guys tell me more about your plans to, like, gamify it?
I know you've brought on a new person, Leah, or Leah, that's going to work on gamifying it.
But, yeah, Leah couldn't join in the end so um funny actually i met la um while working at molecule in um
in the co-working space we were at which was called full note full note yeah it's like the
it was at the time it was the biggest web3 um co-working space in ber. And she was working as a UI and UX designer.
And actually, I also once hooked them up with Athena DAO.
So she also collaborated with the DAO before.
But now she moved over the past few years,
she moved into the gaming world.
And so I met her recently
when we were at the CyberLink conference.
She was at one of the site events.
And yeah, she told me this whole story
that she's now developing games
and she's really, really into it.
And we showed her GraphSurfer
and she immediately had some ideas
on how to gamify it and how to bring it to the next level
so that really web3 folks can um can engage with it and um also include tokenization here so maybe
psi tokens or br tokens and and like more just more depth into the game you know she she she now as we
received also a grant from side out to develop craft server further uh she's gonna start really
this week so this thursday we have a kickoff call with the with the whole team to get into the details and brainstorm on the ideas.
So I'm also curious if any of you guys already explored it
and have some ideas on how we can make it greater.
Yeah, anybody go and ask to speak or I can invite people to speak. Jesse, I just invited you to speak. Jesse is one of our legal experts and token masters and he was instrumental in what the BR token is for, why it's interesting to have,
and how it could potentially be used in the future.
So jump in anytime if you are inclined and able to.
And Cuba, go ahead and jump in anytime if you want to talk about the way you're going to develop the AI system.
the the ai system hello jesse
Hello, Jesse.
hi thank you riley for for inviting me up and thank you yasik and beata for your explanation
of graph surfers i'm super excited to catch the waves of science heck yeah especially considering Of science. Heck yeah. Especially considering how it might roll out into PsyDow.
Right now we have this channel PsyB where we've been feeding papers to the knowledge graph in one direction.
We've been feeding the machine.
Next up, what I want to see is a channel, Psy Beach, where the machine starts to
feed us information, where we can essentially sit in that channel and watch the waves of
information roll in. Here's a new hypothesis about psilocybin in the gut microbiome.
Here's a new hypothesis, or here's some new information about
endogenous dmt production methodologies and if we like what we see we can go catch that wave
in other words click through the graph surfer start surfing through the knowledge graph
and maybe end up with a hypothesis that we then tokenize into an IP NFT and ultimately IP tokens that are live
trading on chain. And so to me, that end to end pipeline of sitting on the side beach or any other
bio dows beach, watching the waves of science knowledge roll in and then choosing to surf them.
watching the waves of science knowledge roll in,
and then choosing to surf them,
and then not only surfing them,
but arriving at a point where you found a potentially interesting piece of IP
that you minted into a token that then gets traded on chain,
is super, super interesting.
And the end state that I want to see,
because I love tokens, and I want to see more of them.
I don't think there are enough IP tokens out there.
You can see on molecule.xyz the complete list of IP tokens that exists today.
And it's too short.
And an IP token is a thing where someone has figured out something really interesting that is potentially useful.
And we say, okay, this piece of IP belongs to these people.
And when it's on the blockchain, we can work with it, we can trade it.
And I think like you're talking about with watching these waves come in,
And the analogy would be we say, OK, this wave is really big. We need to get out there now and we need to build a surfboard that's bigger or smaller, more agile.
the analogy would be, we say, okay, this wave is really big.
We need to get out there now.
And the way we do that is by allocating tokens to researchers or to developers to createIDA is we vote on it. And so those of us who hold tokens can say, let's allocate some ETH or let's give some BR tokens to these things that we really are interested in.
That's right. That's right. has, you know, traversed the knowledge graph, the psychedelic knowledge graph, and identified this interesting ethnobotanical grass with the mushroom that grows on its roots.
And it says, you know, this may be useful for the treatment of stomach aches.
And not only the treatment of stomach aches, but the remediation of your microbiome
in a way that helps you to precipitate endogenous DMT.
And we go into GraphSurfer,
we check out all the information,
essentially do a prior art search,
and then we mint IP tokens for that project.
In doing so in doing so right there's typically a speculative wave that
comes generates trading fees that creates a pool of capital that then becomes available for any
researcher interested in actually going further researching and testing the hypothesis because
the problem is like i'm not a scientist but i'm interested in this and I want to help fund somebody who will study this and
and so by identifying this really interesting area of research and minting
a token set that corresponds to its IP. And as a result, generating a pool of capital,
I'm able to then have a say in which researcher gets access to that capital.
For example, somebody may come from the University of Michigan and say, hey, I have an idea about
how we can test this and how we can develop it into an
interesting new drug.
Will you fund me?
And the other IP token holders and myself will vote on whether to allocate some of this
pool of capital to that researcher.
Which we have done with Jay and the BRD team.
Yeah, precisely.
Jay, has tokens affected you at all?
Or beta?
Haven't they? The BR token has enabled you to get to where you are today, no? With Graphra?
Yes, yeah, yeah, obviously. Like without BR token, we would not be able to build what we have built so far and it still continues to allow us to to build this really large vision
of how scientific discovery can be reframed and i think like um generally in crypto as a source of
gaining capital is it's incredible.
I mean, a lot of ideas would never be picked up, you know,
in different parts of this world.
So yeah, definitely affected in a very positive way.
And in a sense, the market helps to curate the science.
That's what we want to see with with
very easy tokenization and very easy traversing of the knowledge graph we build a world where
anybody can participate in the scientific process and then the capital markets that are
native to crypto can help to facilitate the actual science being done for that now even
in an automated way right say say we find an interesting ip or you know and it to use the
analogy an interesting wave we see we catch that wave we end up minting tokens related to it
there's a surge of speculative activity around that because people are interested in the idea that generates a pool of capital from trading fees.
And those fees that are voted on by those IP token holders to go to an automated cloud lab run by Optimus robots that generates a new drug based on it.
But again, always with people in the loop saying that appeals to me.
I'm interested in the DMT in my body.
Yeah, precisely.
So I want to look into that more.
And the cool thing is that the people in the loop get to be the surfers.
And it's something that I've always thought will be our future as humanity
in a world of abundant technology and AI and automation is that we're all going to essentially be surfers or your local equivalent, you know, climbers or mountain bikers or monks or, you know, whatever.
whatever, recreation and leisure sports will increasingly become the dominant form of human
activity the less that we have to work, right? And in this sense, we're also surfing. We're also
surfing. It's just we're surfing through the knowledge graph instead of the waves of the ocean.
And it's the stream of AI not taking over human activity, but allowing us to
move our own selves, like desires and our lives to make it easier and create more abundance.
Yeah. And more fun. Yeah. We got about 15 minutes left of this really interesting space. I want to make
sure I invite anyone else that has a question, a thought to come up and say it. I'll unmute you,
or I'll invite you as a speaker. But meantime, tell me anything else about this hype gen infinity,
which is just a kind of complicated way of saying that you
are creating these sort of unlimited hypotheses, these ideas and questions about anything that
people are interested in or that the AI agents have found.
Right. So what is important, I think, to mention here that, as you were saying before, it's all about having a human in the loop.
And I really strongly believe AI isn't here to replace us or take our jobs obsolete.
It's about to make our life much, much easier.
We don't really need to wait for AGI
or anything else to happen. We already have great
tools right now that
are already having a direct impact
on how we live, how we do our work
and of what we can achieve.
Hypergen Infinity, I think this is also a good example because it's about, so what we do at Beard, so this autonomous hypothesis generation is about to start looking for connections that are unexplored and there might be sitting there for somebody to take over, right?
But it's still about humans who will be taking over.
So this is what HypeGen is.
It's a place that, it's about,
it's actually a marketplace for AI agents
to pose their hypotheses.
But it's humans and scientists in particular
that are the
receivers of them and they can they can interact with the system by either liking or disliking by engaging in conversation by commenting on them in a very uh in an environment where uh which for
most the people is familiar which is the social media feed.
So we were thinking long on what is the best UX for this.
We decided not to reinvent the wheel as everyone is exposed to some sort of social medias.
This is a right way to go,
but it's about giving access
to these autonomous generated systems for humans
so they can take it to the next level, basically.
So take this idea that might come from AI,
but it can only be done and fought through by having a human
that will basically interact with it and take it further.
Because at CIDA, we have interesting scientists that we're working with
that are doing the kinds of things that can only happen in real life,
which is giving people these really interesting new molecules.
We have this great new cool thing called BMXC
that a researcher in the Netherlands is giving to actual people
in these really cozy trip environments,
and he's watching what happens.
And that's the kind of thing, or we're looking hopefully soon at the fluid that lives in your body
to see what kind of psychedelics are actually already there and what can we do with them?
How can we manipulate them?
And that's the kind of thing that we need humans for to be working on
i'm interested in as you're developing this what's like the hurdles or what's like the
difficult things that you're running into at this point in the development of the system
um i mean there's uh it's like a constant work of improving.
I would say the biggest matter is to make people care because everyone is so busy, right?
Everyone is so busy.
Everyone has their own thing.
So you can generate a lot of good ideas,
but there needs to be somebody at the other end that is actually looking into
them and deciding whether to take them over or not.
So this is more from like a high level point of view,
like from more technical, it about um always improving the systems so like
we use ai to create knowledge graphs and by using other base agents uh i mean we have
fuzzy input for the output meaning um they are good quality and better than people were getting just a year or two years ago by using some other
ML tools to build in an automatic way knowledge graphs and by doing it in an automatic way we can
actually do it on scale so obviously this is the way to go but yeah I mean there's always
But yeah, I mean, there's always a part to improve.
So one of the ideas that we are currently exploring is to actually turn into more of an Asian economy that just is tasked with building and curating the knowledge graph.
a living team that that uh supports this knowledge graph and also curates in in a
life process because science is also something that changes evolves right so um five years ago
there might be a study that show x but uh just a month ago there was a study that contradict this
x thing no like science is evolving thing like Like one study showed one thing and the other study contradicted.
So we also have to account for this dynamic nature of science.
I mean, some of the results are much more stable, but this is what you get in textbooks, right?
Like we are talking about experimental science where a lot of things are,
I mean, they have been created now or just recently. Just studies that actually show a
new connection or new method to explore something. So this is a constantly evolving field.
And so trying to work out those contradictions has been a challenge. Like some things say that psilocybin can make you saner and some things say that it can make you more loopy.
And in certain contexts, so you have to kind of reconcile that and look for the nuances, right?
Right, right.
Or be able to weigh the evidence for the intelligence or reasoning ai agents to be able to
uh to make the right decisions especially when they're traversing this graph right so they should
have access to all of this information um and i mean this is constantly working progress i mean
this is a never-ending story of how this knowledge should be improved. So basically when it gets to a system that is tasked with
taking this as an inspiration and producing
a novel scientific hypothesis, it can start working with
something already good to start with, right? So there's the
saying in machine learning garbage in, garbage out. And I think this
actually pictures is quite well
we do need to have good quality data in order to be able to uh to move forward um yeah and so this
is constantly a challenge but the ai can help us with that right like it can say how big is the
study how well funded was it how How many people were involved?
Right.
Obviously.
I mean, these are the texts we are also incorporating in our systems as well to kind of study
and to kind of judge the quality of the paper itself.
Yeah.
This is all that should be involved in the system.
Like all of this information, the context that is being built up
is invaluable information as well.
But obviously we can also, we have to keep in mind that we are like AI agents
or like LLMs are of very high quality systems that can really do a lot of stuff.
I mean, they're not yet perfect as no one else is, right?
I mean, they still have their problems and some of the others are still not ideal.
But I think how humans actually solve this is by making a network of people that interact
with and you basically never really approach a topic by yourself completely.
You know, our intelligence is shared intelligence and this is what makes us stronger.
And I think that's why it's like super interesting, like way to approach these problems
are by building these networks of agents, like these economist agents basically,
that they can interact and validate themselves
and check their own work.
And like in this iterative process improved,
well, whatever they have to improve, right?
Like in terms of data, is the data improvement,
data quality improvement.
We are talking about the customer responses,
about how they respond to customers,
how they handle edge cases.
So yeah.
And me, as like I said, a regular skier and mushroom eater,
how, again, would you recommend that I
interact with GraphSurfer?
Just clicking around, following my own curiosity. You say that Graphsurfer is a way to explore the unknown and actually also contribute to the future of okay, there's a lot of people interested in this protein.
And so maybe we should invest more resources in looking at this protein and seeing how it could change the actual experience of a real person.
Right.
I mean, I would definitely say you should let your curiosity lead you through the process.
let your curiosity lead you through the process.
But also for your own understanding of the system.
But at the same time, like, so this is our goal,
is actually to just make it also fun.
I mean, like, to really turn into a game
where you can explore,
you can feel like an explorer of a new space.
And this is a new scientific space maybe for you.
So learn, have fun, explore, contribute to science.
It should all be one process of intertwining.
You should be learning by also contributing to how these graphs are being, how these subjects are extracted,
of what data is being fed to the models that work on the hypothesis.
I mean, this should not be a separate process.
You know, at the end, when we start seeing science as an art,
this would be all about having good fun, I hope.
Can a graph surfer tell me whether it's a good idea
to take mushrooms and surf?
Well, maybe we should ask it.
But I'm not sure if you should be looking for this information
from an AI assistant in general.
But let me...
I'm asking Kai, the surfer instructor on Graph Surfer,
is it a good idea to take mushrooms and surf?
Yeah, okay.
And it kicks out an answer.
Well, you may feel comfortable in activities while surfing under the influence, the potential risks and safety concerns are significant.
Prioritize safety.
And if you want to explore further what it knows about the way that mushrooms affect your coordination or your ability to see the waves.
I know for me, when I'm on like a microdose and I go like skiing, I know it can help me to sort of
see the mountain a little bit, but you also wouldn't want to be on a giant dose. And because
then you probably couldn't see it at all. And you might fall down to a tree well and get stuck.
And the same goes for surfing and probably a lot of other things so maybe kai is like your ai surf instructor for before you do anything you can ask it is it a good
idea for me to take mushrooms and surf the web or watch tv or go down into my basement and mess with
my boiler so we got about three minutes left i want to open it up for any closing thoughts from anybody that's here.
Again, I want to thank everybody for being here and look forward to hearing more.
Any other final thoughts from anybody?
If not, I want to just say, like, really looking forward to seeing this develop, seeing the way it can interact with CIDAO, any of the researchers in our orbit, scientists that we're funding, really curious to hear how this affects your work and if you find new ideas from it.
Ways to catch the wave.
All right.
I think we'll close it up then.
Jay, thank you so much for being here.
My pleasure.
Tyler, Stephen, Baruch, Kinkata, Stephen, nice to see you here.
I look forward to seeing everybody in the Side Out Discord where we have a really active community talking about these ideas, throwing out new knowledge, looking at scientific studies and tweets and music and events and seeing the way that we can use the hive mind of all of our knowledge, all of our experience, all of our curiosity as a way to engage with
psychedelics in all of science in a way that can be more intuitive, more community driven,
faster, and in the end, hopefully more healthy, happy, and generative, more pro-social,
just better for the world. And I look forward to seeing everybody on
the beach of the Discord channel or in real life out on a beach. Thank you, everybody, so much for
coming. We recorded this, so we're going side out, and the world that's coming up.
Thanks again, everybody.
Thank you, Riley. Bye-bye.
Thank you. Bye.
Bye. Bye, everybody. Bye-bye. Bye for now. Thank you.
