Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
AI & Decentralization with Ÿ
Recorded: March 27, 2025 Duration: 0:51:30Space Recording
Short Summary
In a dynamic discussion on AI and decentralization, industry leaders explore the transformative potential of decentralized AI in the crypto ecosystem, highlighting trends, project launches, and the importance of user-centric governance.
Full Transcription
Testing 1212.
Thank you so much to everyone that has listened to the space currently.
Just hold on a moment as we wait for all our guests and community members to join. Thank you so much much. Just give us a moment as we get all our guests on stage. Thank you guys so much.
Just give us a moment as we get all our guests on stage.
Thank you again guys.
We're just waiting for one more guest to come up, and then we
can get started with the space. Thank you guys so much for your patience.
All right, ladies and gentlemen, we're going to get started with the panel now as we wait
for everyone to start piling in.
Thank you guys to everyone who's come through today.
Of course, today is episode 32, where we will be discussing AI and decentralization.
We are super excited.
We are joined by some amazing guests today.
We'll give a moment to introduce every single one of them in a moment.
But of course, I would like to remind everyone
that we currently are running a raffle of $50 in my tokens.
All you need to do is reply to the tweet that announced
the space and tell us why decentralization should
be important in AI.
And of course, then you must attend the space until the end.
And finally, follow us and our guests on X.
We will be checking, guys.
Already got quite a few points and answers in the chat, in the comments there, and we really appreciate it.
So, guys, please do keep them coming as the space goes.
But, of course, let us give a moment first to introduce ourselves and introduce myself.
Of course, my name is Synergy.
I am here on behalf of Y to do your AMAs and speak to you guys.
And, of course, if you do not know who WI is, WI is a crypto intelligence platform,
which is harnessing the power of AI and data.
The WI Oracle unveils unparalleled insights
into the crypto market,
seamlessly blending historical wisdom
with real-time market dynamics.
And of course, we are super excited
to have recently announced
some of our largest updates to date
that I did mention in our previous phase.
And of course, that is all centered around the automated trading, which we have just dropped.
And of course, we do have a hyperliquid vault that is currently using our Ypsilon AI to do some
automated trading. So if you want to get on that action, make sure you check out. You can find the
link in our description and things like that. Thank you so much. And of course, the next major
one I want to quickly mention before I invite everyone in of course is that we have announced the zero percent tax that is coming to y super excited
for that and of course more details will be announced around that close to the time but
let's now give a moment to introduce each of our amazing guests and i would love to start with
lava network if you could please come up and tell us a little bit about yourselves and what's cooking.
Hey, I didn't get who is... Lava Network, are you there?
If I can, if it can be myself, I will do.
Hey, everyone.
Super pleasure to be here.
Thank you for having me.
My name is Anna.
Anastasia, if you're speaking, we unfortunately are not able to hear you there.
Really?
I can hear her.
No problem.
Yeah, so, well...
Good to see you here, Anastasia.
Thank you so much. All the time when we have technical problems on the Twitter space,
I just realized how far we are from Web3 and anything else, because technology-wise we need to,
you know, like, fix... Yeah.
Okay, I am getting a thumbs up. Unfortunately, it seems that I'm not getting any volume out
of my site currently. So while I invite, I'm going to invite Anastasia, if you wouldn't mind
just telling us a little bit more about HG. I'm just going to quickly try to sort out my
connection. So if you could just tell us a little bit more about yourselves, and then, of course, we'll give Lava Network and everyone a chance.
And again, so apologies about this issue, guys.
I'll be back in one moment.
Okay, sure.
My name is Anna.
I am Chief Marketing Officer at MedChain.
MedChain is decentralized AI blockchain.
Before, I used to be for four years as CEO of Cointelegraph
Communications so my background is mostly media and before that it also was media
but traditional ones. So basically this is about me.
Awesome I really do appreciate that thank you so much for giving a little bit of an introduction.
Amanda, apologies again.
And I'm not sure, I only just got back now, but did you tell us a little bit more about Matchchain?
Sure.
As I said, Matchchain is a decentralized AI blockchain.
MadChain is decentralized AI blockchain.
So we are voting for securing data sovereignty,
and we are willing to get this power of owning your data to the user.
Because a lot of companies like Facebook,
Meta, and everyone else,
they are gathering our data every day, and they are using
this data to advertise us, and we are not getting any benefits on that. So with our flagship product,
Match ID, which works as universal sign-in and works like your backpack. Like think about this like a backpack. So you can
put inside your backpack everything you want. You can put your data, your degrees,
everything about like your job path and all the documents in the world and in anywhere, like in Web3 world, you can sign in as easy as your Gmail works while staying secure, while having benefits, while exploring your digital value.
So you can understand how valuable you are and your data is.
your data is.
And it doesn't matter how many accounts you binded within the match ID and how many data
you shared, everything will be appreciated and monetized back.
Awesome.
Thank you so much for the introduction.
I really do appreciate it.
And next, I would love to give a moment for Kodaflan to come up and tell us a little bit
more about themselves.
Hi, welcome. Thanks for having me. My name is James Young. I am a founder of Collabland.
We do token gaining and we've been in production for about five years. My roots start from being an Ethereum researcher. I went full-time into crypto
in 2017, seeing the space with all the ups and downs and super excited to be here. We know a
thing or two about bots and I've been heavily investigating and looking at the agent space
And I've been heavily investigating and looking at the agent space for the last two years.
So excited to be here.
Thank you so much, James.
I really appreciate to have you on here.
And of course, we all know CoLab Land.
We've all been dealing with you guys for quite some time if you've been in the Wittory space.
So it's great to have you on here today.
And of course, last but not least, I'd love to give a moment now for Lava Network to come up
and tell us a little bit more about themselves
just before we get started with the space.
Thank you so much.
Super excited to be here.
My name is Yair Klepper.
I'm a co-founder and CEO at Magma Devs.
We are contributors for Lava Network.
And Lava Network is providing access to AI agents, wallets,
for dozens of blockchains out there.
Think about Lava as kind of the on-chain cloud flare,
routing traffic to different providers
and holding them accountable for providing the service.
Yeah, that's in a nutshell.
Since a few months now, when AI,
all the frameworks becoming much more solid and there's more and more development,
there is much more need and demand
for accessing the chains themselves,
whether it's read-write transaction.
And this is what we do.
We support RPC, remote procedure call,
kind of the language of blockchains
and providing access to dozens of blockchains.
Awesome, thank you so much for the introduction.
I do appreciate that.
And of course, ladies and gentlemen,
we're gonna get started with today's session.
And of course, today we'll be talking about AI and decentralization. And of course, these and gentlemen, we're going to get started with today's session. And of course, today we'll be talking about AI and decentralization.
And of course, these are both big buzzwords that often are tossed around in a lot of white
papers and things like that.
And it can be a bit daunting to look at.
You know, I'm sure that at this point, most of us have played around with an AI in some
sorts of chatbots, you know, chat GPT, et cetera.
So we're all slightly quite familiar with AI.
And of course, we will all be quite familiar
with decentralization as we are here.
But it's important to also understand exactly
how does decentralization impact AI,
and how can it actually work?
So what we're going to do now is we're just going to ask
a few questions that just kind of unveil a little bit
of what's going on behind the scenes.
And the first question I would love to pose to our guests would, of course, would be, how do you define decentralization in the context of it?
And I would love to start the question with Lava.
Yeah, for sure. So if you think about decentralization, many people at the beginning were asking us at Lava, are we a decentralized RPC? And the first association for decentralization is maybe it's not so fast, there is low latency, kind of what we see with Ethereum versus Solana.
latency, kind of what we see with Ethereum versus Solana, right?
So, you know, when you think about how do you take in accountability different providers
providing the same service, we think that this is the best way to use the tool of decentralization.
Decentralization is a tool, is not the goal.
And when we started thinking about Lava, we were thinking at the end of the day, everyone
needs to consume blockchain data, read-write transaction.
So even if the end users are using the wallet, they don't know which providers their wallet is using but at the
back scene there are different providers that's running those nodes for that and this goes the
same with AI agents the agents themselves when communicating with the blockchain, they are consuming rewriting transactions, right?
And as more as the
AI revolution gonna turn into
gonna increase, right?
We see that by the end of the year, we believe 80% of the transaction
gonna be consumed by AI agents.
So, I believe the difference for that
is that ai agents don't clips they don't take breaks and the agent infrastructure will be
super crucial to make sure that machines talking to machines right thinks that ai agents are automated script or on-chain bots and they currently read write
those data so the more we give them possibilities of who they use the data from who they use the
data to write this is going to be better for us that later agents are presenting us for the wallets for the exchanges and for every
operations we want to do with them thank you so much i really appreciate i really liked the one
point that you raised about decentralization being a tool not necessarily a goal i can sort of see
where you're going i hope i got that directly directly. And also quite the claim for 80% of, I believe, the transactions to be
by AI agents by interviewer, but I can see that happening. We're definitely seeing quite
a big sort of Cambrian explosion in that area. So I wouldn't go so fast to not see this
happen. I mean, a lot of people would fall into these habits of, how do you say, like, of things being easy for them and having AI agents that can possibly trade, putting
up liquidity, you know, things like that, or even possibly look after a V3 position,
for example, even V4 now, for Uniswap, for example, those things are going to really
take up a lot of space. So I'm very interested to see where we are here as well. So thank
you for that. I would next like to ask James if you would come up for a collab. If you
can just give us a little bit of your views into the question.
Yeah. So when I think about decentralization, it means that it's not centralized. And when
it's not centralized, that means anyone can participate.
So there's some characteristics where it's permissionless
and it's censorship resistant.
And that's how I would define decentralization.
In the context of AI, I think it's going to be important
because the way I look at AI is, it's really just a natural language computer. It
lowers the bar. As a dev, you hear all the hype of Cursor and Windsurf, which are these programming
tools where you just ask the AI in the programming tool to write the code for you, and it does, and it's just getting more and more accurate.
So the way I look at decentralization in the context of AI, specifically with blockchains,
blockchains are just commitment machines in my eyes.
It guarantees a future commitment.
Because a blockchain is decentralized.
There's no gatekeeper there so that it can't necessarily rug you if the code is set correctly.
And so there's a bunch of game theory behind it as well.
But when you write the code, it's expected to run.
So decentralization in the context of AI for
blockchains, I think really means that the user is in control. They can actually have this program,
this bot, this agent, whatever you want to call it, fulfill your commitments, the things that
you want to get done. And what that means across the stack
is that it has to be also decentralized or trust minimized.
So if you are communicating with an agent,
how do you know that behind that agent,
the dev that actually has launched that agent
doesn't actually have the private keys.
So you have to be able to kind of understand the autonomous nature of this agent.
Let's say that you use some type of TEE or account abstraction or key abstraction management system
so that you know that it's only that agent that has access to those keys or funds.
Then you have to also then worry about kind of these longer range attacks when it comes
to model training because, you know, there are abilities for the data itself to nudge
you as a user in a specific way.
So you have to always be able to verify,
at least now that the agent is not hallucinating
and the agent is actually fulfilling the commitments
that you're asking it to make.
So there are different layers in an agent
because it's various technologies that are coming together
that you want to make sure that it's various technologies that are coming together, that you want to make sure that, you know, it's quote unquote decentralized or trust minimized.
So that, you know, as a user that you are in control and that agent is doing things on your behalf in the expected way that you want it to happen.
expected way that you want it to happen. Because these AI agents, they're just natural language,
to me, they're just natural language computers that just lowers the bar to be able to do things.
And I do agree previously with the previous note that I think the majority of transactions on chain
will be programmatic.
You know, you have a lot of bots actually below the surface
that are doing transactions now.
It's just that you're gonna be able to bring that
to the masses because of this natural language computer.
You know, high frequency trading happens programmatically
like in TradFi all the time, but it's very
much gated to a few huge firms and to a select number of people because there's a huge barrier
to entry there.
But with AI, that barrier to entry has decreased dramatically and it's further reducing in
terms of the barrier to entry.
So I do think that in the future, the majority of all transactions are not going to be manual,
they're going to be programmatic.
And I think a lot of that's going to happen in the context of AI.
Thank you, James.
I really do appreciate that.
And to just mention on the last point that you raised, I could definitely see that happening,
especially with a lot more partnerships with mainstream sort of money movers like Mastercard
and Visa.
I'm certainly going to see a lot of transactions, like you said, in the background.
And I guess it's not necessarily a new thing, but rather something that is much more obtainable
to us.
I really like that you called it a commitment machine.
That was a really, very interesting one.
I'm going to remember that.
And, you know, it really, certainly what you said, that data nudges you, and that's very true.
Because when you're using a lot of these, you know, these AI agents at the computers, essentially, you're sort of relying on them in a way to do what you want.
And you sort of automatically trust them sometimes.
You trust them to perform that action, but you need to be able to often actually check
in the background, are those things what's going on correct or not? And that goes down,
for example, like you mentioned with the private keys. And I think it definitely lowers the
bar in a very positive way. I think it's more inclusive and welcoming to more people. I definitely look
forward to seeing what the rest of the next few years has for us in AI. So thank you so much,
James. And of course, I'd love to give a moment now for Matchchain to come up and tell us a little
bit about what they think about the context of decentralization is in AI. Yeah, it's very hard to add something on the top,
what's already been said, especially by James.
I completely agree.
So for me, personal decentralization in AI
is about redistributing control.
So moving away from centralized entities
that monopolize data models and decision-making
and instead ensuring AI is
transparent, user-driven, and censorship-resistant. And at Matching, we see that decentralized AI
as a way to empower individuals to own their digital identity and data while benefiting from AI-driven automation.
So instead of black box algorithms controlled by corporations,
AI models can be trained, validated, deployed through distributed networks,
and users there can have visibility,
governance over how decisions are made and how their data is used.
So for this time, they actually have responsibility.
And the goal here isn't just to make AI more accessible.
It's to ensure that AI serves people rather than the other way around. A lot of conferences I've
participated recently we're talking about misuse of the AI so like all of us we're so blown away
by AI agents and everything what's going on the main question is how to still have a control over that, how to still make AI to serve people.
So I think there is still a long way to go.
And it's super exciting to be at the forefront of this evolution of the Internet, evolution of the technology.
So, yeah, that's pretty much it.
Thank you so much.
It is a very interesting time, Zyvonne.
It does feel like the forefront, the renaissance of a sort of era.
It's reminding me of the industrial revolution,
the next part of technology that comes along
and sort of helps humanity move on in a certain way.
So thank you so much for that.
And very interesting to hearing about these, not spaces, pardon, these conferences that you've been to. I
definitely going to be looking outside of this vent a little bit more into that because
I think there is a lot of space for, you know, misuse of AI agents or AI in general. But
that brings us to the next part, the next question. And
before we get next, start with the next question, I think there's a bit of context. A while
ago I read a little thing that stated that in, I think it was the seventies or something
like that, IBM released basically a memo that said that no computer could be, can ever make
a management decision because a computer cannot be held accountable.
So having that, I definitely would like to ask the panel over here, you know, who do
they think should be responsible for governing decentralized AI networks?
And of course, the question is not necessarily the same as what I posed, but just a nice
little segue.
And I would love to start the...
Yes, please. the same way and i would love to start yes please all right yeah if i may uh uh you know it's it's
really interesting how we um how this user-centric gonna gonna gonna be rewind to um kind of agent
centric right so we delegate a lot of the tasks to the agents and you know i've i've
been in this in this in the space for a few years and everyone knows here on the call how difficult
it is to take and to pass government proposal how participation uh uh looking there right
How do you incentivize all participants to vote on something they don't even know who is this grant going to and what are they voting on?
how do you incentivize all participants to vote on something they don't even know
So I think it's leading.
I don't have the statistic, but my gut feeling is that it's leading us to a very low participation rate, especially on the not day to day basis, but on the more frequent one. So I think this is where it's becoming super interesting
for decentralized AI networks,
how community and network participation
are going to be driven by the governance.
And you see, if you can give the on-chain accountability to define that and delegate that to the agent himself,
kind of representing your principles, the way you think and agree about things,
and maybe even once a week kind of getting a few questions from the agent
to really understand the thoughts about a few topics that the agent to really understand the thoughts about a few topics
that the agent could use in order to participate in this DAO government.
This is something I'm super interesting about.
So with us in Lava, I've seen that the Lava V5 reputation system
is now an example of how our transparency or performance-based accountability can
be delegated and regulated by the network quality.
So for example, when AI agents going to be more
independent and going to make more choices, but still
steer the parameters by the willingness of the
community.
So this is one example of what is actually,
I've seen in a physical example in lava talking about,
you know, economy versus premium tiers
or quality versus speed, et cetera.
Thank you so much.
I really appreciate that.
I think one point I definitely want to say
it is always difficult to get people to participate especially in governance you know. You know
there are many solutions in many ways around this that you can debate for hours over. One
way I even read one time someone was talking about like sort of I am sure everyone remembers
a while ago in sort of 2021 I think think it was, where it was like yield farming, basically DeFi somewhere. And you could just farm away and everything
was great. And people are saying they want to do something like that, where it's sort
of incentivized, instead of incentivizing people to engage with the DeFi platforms via
providing liquidity, for example, they're actually incentivizing them to do a lot of
more governance. But then at the same time, the issue that you have is you also do have a lot of space for, you know, abuse of this system. And we
have heard in this, I'm sure we all have heard in the past of how a bad malicious actor has,
you know, accumulated a lot of the governance token and then used that to, you know, pass
the governance tokens. So we have to be very careful about that.
So, yeah, I really appreciate you bringing that up.
Matt, Shane, I was wondering if you could tell us
a little bit about your thoughts on the subject.
Yeah, so I was thinking while the first speaker was talking
who should be responsible, I think that there is no single answer.
But one thing is clear. AI governance shouldn't be left in the hands of a few centralized players.
Instead, it should be community-driven, transparent, and adaptive. And at Matching,
we believe that governance should be shaped by DIDs,
so verified users and contributors with really reputation-based stakes in the network,
on-chain governance models, so like smart contracts and DAOs,
ensuring decisions are made collectively rather than behind closed doors.
And what else?
Developers, validators, and ethical AI experts.
So multi-stakeholder approach where those building and maintaining AI systems
have aligned incentives to keep them open, fair, and responsible.
So just, you know, I think AI needs checks and balances,
and decentralization is a perfect way to ensure that it happens
without bias or single points of failure.
Sometimes I'm just personally scared that it will be concentrated
just in, you know, like in a single hands.
No, I 100% agree. It is a very scary thought. And it's very interesting to hear that your take on the government's
anything like that. And I think I definitely agree with the VVID.
I think it's VVID, I might be mispronouncing that, but the verified identity
that you were talking about.
And it is always difficult to get people to involve, but there are multiple methods that
people are exploring, and it is great to hear that you guys are explaining, exploring a
different method, so I really do appreciate that.
And of course I would now like to give James an opportunity from Colabland to come up and
give us a little bit of their thoughts on the question as to who was responsible for
governing decentralized AI networks.
Yeah, this is a really interesting question.
I co-wrote the white paper for MolochDAO in 2017.
And naively, I thought if you have a financial incentive, that would reduce coordination.
So my whole take, as I said before, blockchains are commitment machines.
Well, why would people want to commit to something in this trustless manner? It's because they want to coordinate. And you mentioned like governance attacks and coordination
because everyone doesn't have the same incentive in mind.
And so when it comes to AI and now looking back
over the few years in terms of what was kind of unleashed
in the DAO space,
and there are varying degrees of what a DAO is.
There's some DAOs that are very decentralized and there are other what they call Dinos,
decentralized in name only organizations.
And I think that, you know, in the future and the way I think about AI and who can govern
this AI, I think it really has to go to the edge of the network, to the end user, because you want to be empowered and you want to be fully in control.
We just don't have the tools to be able to do that yet.
So it's a very much an open question.
Because if you allow an agent to do proposals and kind of be your proxy, I think you miss a lot when it comes to just human evolution or contact.
Because then the agent is learning and adapting and making these decisions for you.
And, you know, it's going to learn, if you will, rather than you learning and you being able to have that or find
other people of like minds that you can actually be aligned with in the future so that you can
coordinate with them and then continue on with these commitments?
So it's a very open question.
I don't know if I have the answer to it.
I think that the low-hanging fruit when it comes to decentralization,
who can control AI, it's going to have to be categorized
in terms of the realm of just busy work, right?
So AI agents right now in kind of the broader outside of just busy work, right? So the AI agents right now in kind of the broader,
outside of just Web3, booking your flight,
doing all the secretary type of work,
maybe helping you summarize some of your emails
so that you can respond and you're checking it
if you're having complete when you're having the AI,
autocomplete, kind of like autocomplete on steroids, your
email, things like that.
They're still human in the loop.
And I think when it comes to AI in a group setting, and I think this is where DAOs may
come back, there'll be this opportunity for reinforcement learning through human feedback at a group level instead of just at individual level or just at a company that's going to say, hey, trust us.
We'll make sure that we have the AI do the right thing because everyone has their own bias and everyone's motivation and alignment is different.
So it's very much an open question.
And I think that the first approach is we need to do and we need to get that naivete
out of the way.
And that's kind of what I've learned.
So it doesn't mean we stop or slow down.
It means that we make more frequent, smaller experiments so that we can
all kind of learn together. That's why I think AI, like really, from a, it really opens the
door to philosophy, to research, because these are open-ended questions.
Thank you, James. I really do appreciate that and certainly in this question there is no
actual answer to this of course. It is definitely up to every single one of our speakers to
hear sort of a bit about themselves and what they think. So I really appreciate that and
that is very true unfortunately everyone has different incentives whether it is protein
or actin or anything like that. So, and I really like this coordination that you
brought up. And unfortunately we are still going to have biases that bleed into the AI,
you know, more often than not it is very difficult to verify that the data that these models
even trained on are actually free of any, you know, sort of biases in any direction.
It is sometimes difficult. So, something to keep in mind. So, thank you so much for that. And of course, ladies and gentlemen, that brings us to our last question for the space.
And of course, this one is a bit of a broader question, rounding things up in the end.
And of course, the question is, how does decentralized AI fit into the broader Web3
ecosystem? Of course, we first looked at what decentralization is in the context of AI.
Who should be responsible for government? And now we are going to look at it in a web-free system.
Yes, please, I'd love to hear your thoughts.
I think it's a great question, and there are so many philosophical and even technical concepts that you can touch with them some that coming to my mind is uh you know the integration versus uh or actually
the integration with the standards today think about how ai agents gonna speak with each other
you know the on the other the other front you see kind of the ai agents that will be
the other front you see kind of the ai agents that will be only as healthy as the food or the
data sets that they are getting feed from and maybe another thought is about the transparency
and trust right um so the the transparency and trust kind of says that the blockchain or one of the values of that ensures can all see it on-chain, right?
They are verifiable and establishing trust that eventually will be more and more increasing
as the AI agents handle more sophisticated tasks, right?
The other topic, for example, with what kind of food, what kind of data set we are putting into the AI agents.
We've seen at the beginning with ChatGPT that they were really super biased towards left-wing
and right-hand conceptions because they were trained on specific data sets that represent those conceptions.
So how do we make sure that nobody has trained the agent with a malicious code or a backdoor or a Trojan horse or something like that, that in a specific moment of our time, the AI agents will not act according to our values and principle?
So this is why I think decentralized ai fits really much in the
broader web3 ecosystem um and the last topic is something that starting to be closer and closer
to my heart is the next evolution of ai agents everyone gonna have the ai assistance that
representing the transaction
it will make and representing kind of the values.
Think about that as a travel agent
that picks for you the flight,
except if the AI agent is super valid,
100% uptime and doesn't sleep at night, right?
No breaks allowed there.
So I think about that as how my AI agent
is going to speak with my girlfriend AI agents, how my ai agent is going to speak with my girlfriend
ai agents how the ai is going to speak with restaurants that i want to book my calendar
and so forth and this will be feasible only through standards and there is an interesting
concept in the web2 world that's called mcp model context protocol that was
presented by anthropic this is the concept illustrates how standardization might simplify
the the cross-chain integration in the future but it should allow the ai agents for you know autonomous or diverse Web3 application.
So a question about that is how the different contributors
of the standard gonna be incentivized to do so.
And one application and real quick answers
might be with blockchain itself.
So implementing startups kind of like MCP wise
and so forth on the blockchain.
Hope it makes sense.
Thank you so much.
I really do appreciate that.
I'm definitely going to be looking into this model concept protocol after the space.
It's definitely a new concept to me.
So thank you for that.
And I did laugh quite a bit at your girlfriend comment there.
That was quite funny.
Probably a bit funny.
But anyway, I'd love to move on to the next guest's thoughts
on the question.
And I was wondering if Matt Shane would come up
and tell us a little bit about what
their thoughts on the subject are.
Thank you so much.
Yeah, I will tell my answer and unfortunately
need to drop off work calls.
So, yeah, I think that it's perfect marriage.
AI and Web3 are basically made for each other.
AI brings the smarts and automation,
while Web3 makes sure it's transparent, fair,
and actually works for people, not just big companies and everything. We are talking around all this session and in general.
So where decentralized AI is making an impact?
So it's several formulas.
AI plus digital identities.
So instead of logging with Google or Facebook,
the idea allows people to control their online presence.
So AI can enhance this by helping verify identities without exposing personal data.
Is it a future? Yes, it is definitely a change.
Data ownership and monetization.
So today, AI companies profit from user data while individuals get nothing.
And with blockchain, users can own and monetize their data, deciding who can use it and for what purpose.
AI model ownership.
So instead of OpenAI or Google deciding how AI is trained and used, Decentralized AI allows communities to govern
and improve AI models collectively.
So this is about the previous one, about misuse, right?
So while it's in control, it can be in different direction.
So Web3 and AI together mean no more black box algorithms controlling our digital lives.
So instead they create open, transparent, user-first system where AI works for the people, not the other way around.
And actually, the previous speaker like about everyone has AI assistant.
Yeah, it's actually very comfortable.
It makes us very productive while we are saving the time, the precious time of our lives to do something else.
So while we can delegate to AI some routine communication
and AI will use our knowledges for doing that.
We can have a bit more space for something what truly matters.
So, yeah, and as I told, because I need to drop off,
just wanted to say thank you for having me
and for everyone who spent their time. Time is super precious and thank you for having me and for everyone who spent their time.
Time is super precious.
And thank you for all the speakers.
James, I'm your huge fan.
Woohoo.
Have a good day.
Thank you so much.
It was wonderful to have you on the space again.
Thank you so much.
And best of luck for the rest of your day.
I hope you wish best of luck for everything for you.
And definitely time is very precious for everyone. So I appreciate everyone of
our speakers for coming up here. But before we round up this session and give everyone
a moment to sort of let us know what's cooking in their parts of the books, I was wondering,
James, if you could give us a little bit of your thoughts on the subject quickly before
we round things up. Yeah, I'll try to keep this quick as i
mentioned before and just to continue the thread from this space blockchains are commitment machines
and what we're going to find i think is these little pockets of alignment so it's really hard
to align yourselves with a hundred people a,000 people, a million people.
So I think there's going to be a huge fragmentation when it comes to what you want to get done and how you want to coordinate.
And this is why you need Web3 or blockchains in general.
It's because they're commitment machines that allow you to seek out.
It's because they're commitment machines that allow you to seek out, you know, now we have this technology that allows us to not just be able to coordinate and align ourselves with people in our very direct geographic location.
We can find alignment with anyone in the world. So I think that the future of commitments are going to require us to be able to
have all these little alignments. And with that, I think that it's not going to be one agent that
you're going to interact with, but like millions of them and they're going to, or not millions,
but maybe hundreds on a personal basis.
And you're going to have this multi-agent ecosystem.
And I think in that way, you're not going to have one frontier model.
Going back to the original question, that's going to like have this ability to suddenly
have AGI. I think that the blockchain is needed to help these multi-agents also coordinate.
And in this multi-agent ecosystem, I think that's where AGI will be birthed from.
So blockchains are going to be that ecosystem because it's too hard to find alignment with one company because this company they all
have their incentives so you're going to as a user find alignment with others you'll probably govern
in your little small pocket on the internet multiple ai agents these ai agents in this
multi-agent ecosystem will coordinate with one another and
find each other as well. And you're going to have this graph of agents. And I think that's why you
need a blockchain, ultimately, because it's a commitment machine. And it's an alignment machine.
So I think it pairs really well that way.
And that's how you get decentralization overall or in general.
And you're going to have these agents that are going to be very tuned to what you want it to do.
But if it's highly tuned, it can't, its action space will be very limited.
And so how do you have agent or agents that can do a wide variety of things? not just humans, is going to be this kind of next kind of problem set that we have.
As we know, this is the worst that AI is going to be today.
And exponentially, it's just going to get better.
And so I'm excited.
I don't know where the future is exactly going to go.
But I do think that you need coordination amongst humans and with
agents and i don't see any other technology outside of the blockchain that can help with that type of
alignment thank you so much james i really do appreciate it. And of course, ladies and gentlemen, that does bring us to the end of our session.
I'll give James and Klepper,
if I hope I pronounce it correctly,
each a moment to tell us a little bit more
about what's coming for their amazing products.
And I look forward to hearing more from them.
And thank you again to all our amazing community members,
those that are live with us now,
those listening to us after the fact.
Thank you so much for your time.
James, is there anything you'd like to us after the fact thank you so much for your time james is
there anything you'd like to share with the group before you leave uh yeah we're doing a lot of
exciting things at collab land uh it first starts with onboarding um people uh and so we have a
whole account abstraction uh infrastructure that we've been working on under the hood for about two years now that we're
finally going to release so that you can have a smart account, you're self-sovereign, and you can
engage in chat. So chat, group chat is a great way to be able to communicate with other people online but it's also the ui or the ux for ai agents so we're going to
be moving forward with the ability in collab land for you to communicate with agents without having
to leave chat and we're going to be starting to launch that very soon in q2 this year
very soon in Q2 this year.
Awesome.
That sounds very exciting.
I very much look forward to it.
I've been following you guys since you guys really started.
It's been amazing.
So it would be great to have you on here, James.
I wish you guys the best of luck with everything that's coming up.
And now I'd love to give a moment for Larva Network to come up and to tell us a little
bit more about what's cooking in their books
and what we can expect from them in the coming months so you know i think now than ever access
to blockchain different chain is more the most important thing right how are you gonna
trust the agents if some of the nodes are not running? How do you gonna trust the information that the agents getting feed from
if this information is not the latest block and so forth?
So I think infrastructure has to evolve
and provide this reliable, scalable access to the AI agents.
So we are doubling down on supporting AI agents, wallets,
the same as we did before, but with a greater speed now.
There's going to be a couple of announcements
where they're really big and massive users
from the enterprise world, different frameworks
that are going to use the Lava plugin
in order to serve their AI agents framework.
And yeah, that's just the gist of what we are going on.
And last but not least, obviously, we're going to present soon the roadmap for 2025.
Well, that sounds very exciting, guys.
Make sure you check out Lava Network.
I'm sure they have a lot that's cooking,
and I'm sure the roadmap is definitely going to help.
Let us know what's coming up.
And again, thank you guys so much for everyone
that has come through to you,
for our amazing guests and community members.
I hope you all have a wonderful rest of your day,
and I will see you all for our next episode next week.
Goodbye. Thank you.
