Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
ETHCC Recap: What’s the lowdown on Cannes?
Recorded: July 17, 2025 Duration: 0:47:49Space Recording
Short Summary
In a recent discussion, key trends in the crypto space were highlighted, including the integration of AI agents in DeFi, the launch of Robinhood's support for real-world assets, and the declining presence of gaming projects. The conversation also touched on the potential for growth through asset tokenization and the need for legal frameworks to support innovative financial technologies.
Full Transcription
Thank you. Jesus, sorry. Yeah, I forgot to mute my, or turn the audio off on my computer. How's everybody doing? Hopefully everybody's okay. We're going to wait for a few of our speakers to come up and we'll bring them up.
up to the stage. I also see a request from Sora Chain. I'm not sure who they are, but Jeff,
when you come up, you'll let me know. And if they are our guests for today, happy to bring them up too.
In the meantime, if you'd like to request to speak, please hit the little microphone button and we'll bring you up.
Hey Jeff, how are you doing, buddy?
I am pretty good.
I'm glad to hear that.
What makes today very good for you are pretty good.
Oh, do you know, it's been a long week and it's coming to the end. I think that's as good as I've got this week, aren't you?
So it's a long week
You're busy, you're working
Probably a lot of family stuff
Exactly, but it's Thursday evening
Tomorrow's Friday, famously a nice
Poets day for me, so I'll probably finish around
Three o'clock in the afternoon
Is a good tradition in the UK
I love that
Have to start living that lifestyle You need the poets in the UK. I love that.
I have to start living that lifestyle.
You need the poets
in the US, right?
I mean,
I think Australia took it on.
US should keep it.
What is it?
Poet?
Poets.
Piss off early
tomorrow, Saturday.
Nope,
never heard about that.
we do. Us older people in the uk keep it alive we go down
to the pub have a couple of beers with people relax oh yeah jeff if i was where you are i would
go meet meet up with you and poets through uh through saturday it's perfect i probably said
that wrong. Cool.
We do have a Sora chain.
Are they the guests that you invited to come up to speak?
Yeah, that's correct.
Awesome.
Here we go.
Bringing them up.
And I see a familiar face in the audience.
GM, sir.
Hilarious from the audience.
Hope you're doing well as well.
Cool. We have Barnavas up here joining well as well. Cool.
We have Barnavas up here joining us as well.
And I think SoraChange should have already connected, though.
It's a little slow to show on my side.
Yeah, I can hear you all well.
Awesome.
Well, I'm just going to set the stage stage and then we can kind of kick it off.
But GM, everybody, to another Ontology Spaces event here.
Today's topic is going to be a little different.
We're not going to go with our usual programming, which is either going to be on DID and privacy and kind of exploring it with our partners or in one of the more novel applications
that we're doing a bit of R&D with, which recently has been on-chain gaming. We're going to recap
ECC because our very own Jeff was our correspondent from Khan. And well, you gave the inside scoop to us in one of our internal calls.
But I'd love to hear from you kind of some of the topics that you heard were kind of causing some
waves, really kind of stirring it up over in France, and maybe some things that we've already
known about, but maybe there was something new, aka Robinhood, while you were out there.
Yeah, absolutely.
And so we met on the event we were sponsoring down there on day one, pretty much, I think it was, right?
So we had a good chat about AI agents and World ID.
I expressed some opinions on those things.
And so we had a really good chat.
And AI agents was pretty much quite central,
I think, to the entire event and AI in general,
whether it be open source AI, decentralized AI,
all these different things, which was really good to hear.
Because obviously on our roadmap is this entire
thing around the need for identity within AI agents, within AI in general, and AI modeling.
Myself and Humpty has been speaking about this for quite a while. We spoke about it as well,
so I was up during the event about that need for legalities and different things on there.
And then I went to a couple of
more events looking at AI agents and how they're going to be used. I was surprised how much DeFi
dominated that conversation. And I'll bring you in as well on this, Swayam, as well. I think you
went to a similar event as me following those things. And I'm sure you went to quite a few of
the AI events as well. But I found there was quite a lot of talk around AI agents
and DeFi as well, and what they can do for DeFi infrastructure
and allowing people to make more yield as well.
And I wonder if you found that same thing on there as well.
Yeah, first of all, thank you so much for the divide.
I think we had one of the best conversations.
Like, it was one of the highlights from the invite. I think we had one of the best conversations. It was one of the highlights from the cons.
Yeah, I mean, there are a few buckets for AI in general
and DeFi is big because in general,
because it's on-chain activities
and AI agents has become the UX
to carry those deals or carry those trades.
Yeah, it is predominant at this stage, but will eventually span out.
On the application layer, agents, especially for DeFi, is dominating at this stage because
I think in Web3, the most easy thing to build is a layer on top of the existing infrastructure, which does the intent-based design.
So, yeah.
Yeah, and then, so we saw that quite dominating Humpty on there as well.
But something I know we've been interested in, and we also spoke about at Cannes as well, was that requirement for good data for AI agents. And I think this is
something, sorry, Shane, might be looking at as well, those personalized agents and understanding
your own personal data better. And of course, that comes with a whole host of questions around
data security, identity, how do you share that? How you keep it safe the need for potentially zk proofs
in there as well and then also that legality around what your personalized agents can do so
um obviously i should give you the chance to introduce our chain as well to people and
explain what you are working on and how that fits in with those themes of data and personalized agents and things as well yeah of
course so in essence what we are building is an ai training infrastructure so the best one best
model here is to think bit tensor but something that is optimized for edge devices and where data
the raw data never gets shared. So what essentially we are building
is a layer for coordination
and layer to scale edge training.
And that's where the model comes to the data.
So like if you take the current scenario
of AI training in general,
the data is centralized
or you aggregate the data
and get it to the cloud
or somewhere wherever the model is being trained.
Usually this is cloud. And that's where all the data compliance is
and different regulations across the globe comes in.
But if we bring model to the data, we perhaps solve the core problem of moving the data.
But solving the more difficult problem that AI is facing these days is the bottleneck on data itself.
And not just data or any data, a good high quality data on which you could perhaps actually build some useful, meaningful agents.
And that's the infrastructure that we are building at Zora.
Yeah, which is really cool. And Hanji, I want to give you the opportunity to come in as well on
this, of course, otherwise I'll just gas on forever. But I think this theme of our AI agents,
the need for data, the need for personal data access as well, is really relevant to the things
we've discussed as well, I think.
Yeah, I mean, I think we've talked about, first of all, for a long time, the need for kind of these privacy systems in order to be able to act in a way that is, I guess,
opt-in in a very public space. And what I mean by that is the blockchain is very public.
Everything you do is transparent.
People can see what you're transacting,
how much you're transacting for.
Anything that lives on chain is,
and for the most part, permanent too.
So to be able to introduce these systems, primitives, if you will, so that you can
with confidence use the platform while not necessarily revealing everything about yourself,
I think is very important. I think many of those things that we've discussed here at length,
and obviously this is what ontology specializes in, are things like decentralized identifiers and verifiable credentials, ZK proofs, and even with what Ontology is building with Orange, which is ZKTLS.
I think all of these are different ways in order to be able to provide a level of privacy and freedom, really, to people to navigate this space more comfortably and confidently as well.
Yeah, so this was a theme through a lot of the discussions I was having with people in terms of, it actually started in discussions when I was at Prague, at ETH Prague, starting talking about the need
for privacy within blockchain.
Because as you just said,
blockchain is very transparent.
Solved a centralization problem,
introduced a transparency problem,
as well as transparency being good
in some respects.
And now we get ZK starting to solve
that transparency problem.
And then we get this huge advent of AI agents and AI in general being built.
And just to be clear from all the conversations I had in CAN and every I've spoken to, AI is not going away, certainly not in Web3.
going away, certainly not in Web3. There's lots of reasons for it to be in Web3, even some good
cost reduction reasons as well, as well as the data privacy reasons and all these different things.
And so that was probably my big first takeaway was AI is here, AI is being built, it's here to
stay, everything's happening on that. My second takeaway, I guess, was less optimistic because it is part of our roadmap.
And that was gaming projects are few and far between at the moment at these events.
I don't know if they've sort of run out of runway, if the environment's just not there for them.
the environment's just not there for them um but but i didn't come across a lot of active gaming
projects that were really pushing any narrative out in there which was
not not a massive surprise but slightly disappointed that they weren't there i think
they're probably protecting their funds and and can was not cheap it's not a cheap place to be
it's not a cheap place to exist and it's it's it
certainly knows how to charge for things there's a lot of yachts and lots of beautiful people
wandering around um so gaming not massively present which was slightly disappointing but
then we we get to things like rwa's real world assets. And of course, we have the Robin Hood announcements about them supporting and bringing them on chain.
And I've actually spoken to a couple of people, hopefully somebody we can get on the show in a few weeks,
something on that as well, who are bringing RWA support and are keen to talk to us about what that could look like for Ontology as well within Onto Wallet.
So that could be really interesting.
So RWA is coming big.
I think this is an American-led thing.
I think the changing and sentiment in America has really made everybody stand up and take
notice and everybody start thinking about things.
And we see different places around the world now starting to ease regulation
whether that be in Hong Kong or in the UK or in other places just starting to allow
more discussions about the tokenization of assets and and I think that then becomes really interesting
so because going back to AI if we bring R's online, then your agents can interact with those sorts of things and interact with purchasing RWA's for you, even if they're on chain and do different things.
And so, Swayam, I just wanted to come back to you on this.
I mean, we spoke a little bit about the potential legal hurdles for agents acting on your behalf.
legal hurdles for agents acting on your behalf.
And I wonder if that's something you've thought about with Sorachain in terms of,
do you get to a position where these AI agents can make purchases or do things on your behalf?
And what does that look like in terms of the legalities?
And where does things like identity and things drop into that, I guess?
Yeah, I mean, that's a very deep and critical question on what we are building this stage with AI in general.
agents and there needs to be grounding from where the agents are getting the metadata or
where they're taking any actions or defined tasks for the agents. That's why blockchain is really
good where it like we can do the provenance and all different kinds of traceability. But the problem starts when an agent can create another agent
and that agent can do another agent.
And that's where I think, let's say if I,
or anyone develops an application using all this web new tools,
wipe coding, let's say Noable or any of these good platforms,
who owns that IP?
Like, do I own the IP?
Or is it Lovable that owns the IP?
And if, let's say, if I build an agent
that builds the software,
which, let's say, is Lovable
or something similar to Lovable
that builds the software,
how do we define the ownership? Like, who owns the piece of software? let's say is Lovable or something similar to Lovable that builds the software,
how do we define the ownership?
Like who owns the piece of software?
Where does the IP belong to?
And I think software may be something they can probably link with online identity
or GitHub accounts
or probably some other ways of doing the IDs.
But the question is like,
if agents are capable and they would be to use crypto wallets or any of the
payment, now Robinhood is coming in.
So we have payment rules which are being integrated into agents.
payment rules which are being integrated into agents.
Now, what if,
Now, what if?
What if it could mimic a real person identity
and buy a property somewhere?
Like they can pay, right?
And they can buy a property.
And entire government structures
or the systems have been designed
or the government itself came in
to protect one's property.
And that's where we run into a lot of different nuances
of these problems and having a ground truth and identity is one of the core pillar to do it
now we can do it in different manners we can use all different privacy tech
but essentially like it needs to fall down to a ground hard truth which i personally think
is where the biometric data comes in it's where like one is to one
is like even if you're fingerprinting or having your face recognition on your own apple device
or any of your smartphones,
that is one of the core biometric data that one needs to link to,
to prove the identity.
And I think,
we need a tree of identities,
like we cannot have biometric data for everything,
but let's say my LinkedIn, Google,
or any of my online identities
must be linked to one ground truth.
And that's where I think the challenge is
on how to design this
and how to make sure that this is embedded
in the agents itself.
Yeah, I totally agree.
And obviously we had some discussions around the pros and cons of
biometrics and but just coming back to the ip actually hunty because i know you spoke quite
heavily about creators and ip and and we've we've had people on the show as well who i know you know
very well talking about ip and i i wonder just in general, what are your views on that in terms of that role of
identity around agents and who owns the IP? Because as creators, we'll use AI more and more
in their creativity as well. What happens to IP on that? How do you maintain ownership? How do you
keep that traceability in that log of who owns what as agents create agents who create agents who create IP
as a creator, where did that leave you in that respect?
Yeah, I mean, that's pretty hard, right? I mean, one of the things to consider is this
data set that the AI is being trained on, right? Like, where does it come from? Is this open source data or is this really just scraping the internet for work that others
might have created and are not getting any kind of reward back from?
And I think, I'm not sure how this will work, but I think that there's going to need to be some guardrails put in place on these platforms where we share our content to, first of all, protect our works.
But second of all, to be able to claim some reward for the work that we're doing.
claim some reward for the work that we're doing.
And whether that is just something that's built into the application
for direct monetization on the platform,
or in terms of maybe it's encoded in the software,
where it recognizes that similar to Google indexing the website,
there's an agent crawling the data,
and that it actually
tracks the ID of that agent so that if that is reshared, it's almost like a bit of provenance
from where did that information get sourced? How did it get repurposed? How is it getting
monetized and then be able to like redistribute rewards based on its usage in the future?
and then be able to like redistribute rewards based on its usage in the future.
Because, you know, IP law is there for a reason, it's to protect creators,
but also it's there to ensure that there are, you know,
fair usage of different types of IP based on the copyright of it.
based on the copyright of it.
And so I think all of these things
And so I think all of these things are going to need to be encoded
are going to need to be encoded
and in some way traced back by the platform
and the agents that are doing the scraping.
One cool thing that I saw,
which might connect here in some way,
is I did go to the BASE event yesterday
that was happening in Los Angeles.
And the big announcement yesterday,
aside from just generally talking
about BASE being a much more unified ecosystem, is that the BASE app, which was previously known
as Coinbase Wallet, now allows for any creator who posts on that app to directly monetize that post
based on the interest of the people who are engaging with
that content. So for instance, currently, I think the standard on most platforms,
including here, X are to comment, repost, or like, that's basically your three standard
engagement points for the content that you create on these platforms.
There's no way to monetize the content. And so beyond you taking this content and maybe
having someone go and collect it on another platform, or you name it, maybe subscribe to
my Patreon and all that, there's no way to have any direct incentive for posting
things on a platform. With this new app, you're able to now have people not just comment, repost,
and like something, but actually buy it because every post is coined. And I think it's a new
behavior. It'll be interesting to see how this is received by, you know, kind of a growing audience because they're slowly rolling this out. It's still not available for everyone. But again, it's these new systems that are being put in place so that a creator is able to profit on the value that they're creating on a platform, right?
What's missing still, of course, is anyone then could still scrape that and repurpose that in
some way to train an agent. And so I think it's beyond creating these systems that enable the proper incentives for creators to continue sharing kind of their works on these platforms is how can these platforms then protect the value that these creators are creating?
And how can they then, if, for instance, like I've already mentioned, this gets repurposed, how can they
reap some of those rewards? And I think that there should be, and I think there will be,
identity systems that are built for agents that allow for this kind of like traceability of their work and also their reputability or accountability
for its success in the marketplace, which is really just the internet.
I saw the other day actually, Humpty, thinking about this poem-based system of posts and
creation. I was just having a quick look
to see if I could remember the name of a guy.
I follow a guy I followed for a couple of years now
who's been trying to work heavily in music,
NFTs and Web3 music, decentralized music,
whatever we're calling it.
And he's literally last week, I think,
he posted just saying they're moving away
from trying to use NFTs for music and actually trying to make each individual track have its own coin and sort of coinifying, if that's the phrase, these music tracks.
And I thought that was a really interesting approach in terms of, you know, rather than trying to, because we've seen people, you know, mint an NFT of this or of this blog or something similar to what we've seen across lots of mirror.
I think you can you can mint the blog and things like that as an NFT.
I think that's still a valid approach.
But this move towards maybe a coin based approach for creation is quite interesting.
is quite interesting and potentially given what we've seen with some of the meme point success
and things like that is potentially something that could work really really well um i guess
we'll find out and i guess i guess on the ai agents as well uh swayam i guess you're thinking
about what does uh monetization for creators and that marketplace look like for that
as well. And I guess proof of ownership of agents also becomes important in that in terms of who
owns that agent, how do they monetize it, how do they make sure that they are seeing some benefit
from it. And I just wondered how far along you were in trying to figure out what that approach
looks like for you. Yeah, I mean, first of all,
I don't think that we're going to slow down AI.
I feel like we're at a point
where it's just going to continue to accelerate.
So this idea that we can prevent it
from scraping the internet
and kind of repurposing all this content
and making whatever shows up on the internet appear real
when it really isn't,
I think that that's just going to become harder and harder.
Yeah, I mean, if you just see the research from Epocare,
which is one of the leading AI research firm,
like what they have published is all the publicly available web data
is almost crawled and it is expected that the models would run out
of available public data by end of next year so like there is a big hypothesis with all this
ai in general like if you have any problem throw enough data and enough compute and you will be
able to solve that problem now the biggest bottleneck is going to be on data because compute is scaling and compute like we
have enough compute and different energy sources which are coming up to power that for now but
data is one of the biggest need of the day in which like for anyone may it be GPT or Anthropic
or Gemini
so far they have been competing on
the benchmarking
on some critical tests
and they are able to do
perform well just because they have
good data. Grok4
is basically competing well
because they have all Twitter and all different
kind of data that perhaps other platforms don't have.
So at very core,
like even if we are monetizing data or creating different data marketplaces
or the social media platform from base,
I mean, we're already running out of data
and almost everything is crawled.
Now, the only essence is how we bring more quality data
to these models and how would the next good model
will look like.
And our core hypothesis of SORA chain
is about going towards more specialized models,
SLMs, or small language models,
or even specialized language models.
Because let's take an instance.
Let's say if you broke your needle and you have an X-ray,
and if you want to upload that X-ray
and get some consultation or insights from AI,
it's not possible.
Because in general, the models have not been trained on the data.
Now, to train such data or to train such models, you need data and loads of data.
Now, there are three ways to get this data.
One, get all the publicly available data, which they have already done.
Second is going towards synthetic data.
So creating data out of AI,
but it is as good as the base model.
It is garbage in, garbage out problem.
So you can have probably 30 to 40% of your data synthetic,
but you still need a good high quality data
such that you can reduce the bias in the system.
And the third is propriety data,
which is basically trapped in edge devices
or institution, hospitals, doctors,
or even patients like you and me.
Now, this is where a lot of value is
on how we unlock this data.
But data itself, like so far,
we have seen Ocean Protocol
and so many different versions of monetizing data.
But the core problem is how do you value the data?
Like how do you create and identify the value or the price discovery for the data?
And that's where, at least from my perspective,
we are not really seeing a product market fit there.
And one of the things that we have a big hypothesis on is,
what if, what if, if we bring model to the data
and you're essentially like as a human
or as your core job eventually would be to train the intelligence,
like how we train anything else. Like your core job is to train the intelligence like how we train anything else
like your core job is to make the AI better
and the intelligence better with all your data
now this could be any data
like even if you're walking, sleeping
all that data that you're capturing
is super high valuable for the research
and for this model is to do better
and this is iterative and this is repetitive
like it needs to do better. And this is iterative and this is repetitive. Like it
needs to be better each iteration
and the more you do it, the more
better the models get.
And yeah, that's where we are building
the infrastructure and
if models are good, the agents
would be good. And
identity is going to be a very
critical layer that
we are thinking about in our system
or perhaps any AI system would have to think about identity
because at the end you need to prevent
and you need to think global in terms of the entire system
because this is global and it's going to be agent to it.
and it's going to be agent to it.
I think the most complicated part is going to be agent to agent.
I think the most complicated part is going to be agent to agent.
Yeah. So I don't know if I really answered your question.
No, no, no, you did. You did. That was absolutely fine.
And it just made me think as well, actually,
going back to the ECC recap point of this is it really made me think that
one of the things that was kind of implicit in a lot of the conversations I had was the underlying importance of some of the deep in things being built.
Yeah, exactly.
So I think one of the biggest category of projects that I saw there was data, especially aggregation.
Most of them were deep in and doing different categories of data aggregation.
No, exactly that.
So, you know, I spoke to a project.
I can't remember the name.
I'm getting too old.
But I spoke to a project, and they were looking at mobile data, actually,
providing solutions for deep in solutions for mobile data
so people could host data hosting on their Android devices.
And actually, then also, you could use that as your
storage device instead of Google Drive or Google Photos and things like that. So you've got this
decentralized solution. And I think those data solutions in a deep-end way will become more and
more important to AI agents and AI providers, because as you were just saying, one of the
things we need is personalized
data that you can take the agent to, if you like. So you can take the agent to your device and say,
here's all my data, or you can take it to your data deep in Vault and say, here's all my data,
learn something about me. Because we will, again, as you were just saying, we'll very quickly run
out of publicly available data and global data. And we need that much more nuanced, personalized data
to really make these agents applicable to us, if that makes sense.
Yeah, of course it does.
And there's so many good projects which I've seen some project building data doors
for Tesla data, which is basically all the driving data on Tesla,
that you can put that to data DAOs
and then they have created the entire system around it.
But then I've also seen something very similar to Google Maps
and Apple Maps in how they capture the data with the cars
moving around the city and people walking the streets
with those big cameras and the entire system
to capture the live feed.
And I've also seen a couple of projects doing that.
So in essence, I think we need this data at all different levels.
Deepin is one of a very good way to do it and monetize
and create the structures to get those data.
But then again, we are not really reaching the edge devices and where like you cannot really
share the data and you would perhaps never want to share that data and that's where we are
positioning ourselves like going for the data that we can enable but we can like we are we see
ourselves as more like an aggregator like we're also aggregating the private data, but as well as the data from Deepin and all this data does.
And you can train any models with it.
So we are not limiting ourselves to one approach,
but just opening the entire arena on what a problem requires,
what kind of data does it require.
And they should be getting access to that
while respecting all
different privacy by default zk fhe everything is additional like by default data should not
leave your device or your premise that's the core of what we're doing yeah now i'd like that can i
just check hopefully i'm not the only person who's not 100% sure,
because otherwise you're just answering this question from me.
But when you talk about edge devices, what exactly does that cover?
Edge devices could be anything.
So your smartphone, your laptops,
and now the form factor is changing how we do AI, right?
Now, that's a meta as a smart glasses.
Now, they come with inbuilt
model which is running which is capable for the inference. So like one of the core thing is
happening is the hardware tailwind is super capable. Intel and everyone are shipping the devices with
the capability to train the models locally and to at least run the model locally. And these are small models, right? Like 1GB, 2GB, 3GB models.
Like you can already do that with Lama and Gamma running on your device,
but they are local. Like they are not learning globally.
They are limited to your own fragmentation.
Yeah, so edge devices could be any form of device. It could be even machines.
It could be DP devices. It could be deep-in devices.
It could be, let's say, aircraft.
It could be anything.
Perfect.
I get that now.
Improve my understanding.
I love it when I learn something new.
So, Hamti, just to bring you back in here, for me, a lot of these conversations were around these topics.
You know, the advent of RWA's
come in via the Robin Hubbard announcement, the need for data, the need for data storage
and data privacy, underpinning all that, lots of Deepin conversations around what Deepin
is doing to provide that. AI agents across different types of approaches.
I went to one event and saw a really cool presentation by Bagel. I don't know if you were there, Swayam.
Bagel, I thought, gave a really nice presentation on open source monetization, actually.
I think it was mainly around monetization of open source models and open source platforms,
which is applicable to many, many things, of course.
And so that was quite cool to watch.
Lots of DeFi things, DeFi AI things, however we pronounce that one.
And then, of course, underpinning everything in some respects was this growing interest in stable coins.
And again, I think this is probably dropping out of the US
with different bills going through.
But stable coins were central to a lot of the discussions I was having.
What stable coins are available?
How they earn yield?
How AI can help them to earn yield?
How they're collateralized. Actually, that conversation
was less than you'd expect. It was much more a bit around how do you make them earn yield and
what are they going to do for you and how can they interact with AI and get the
market beating returns on there. And I must admit, I saw some on the surface, what looked like
phenomenal AI platforms to do your trading for you.
Some of them were black boxes. Some of them gave you some input.
So depending on how much you trust AI depends on what you could do with them.
But we're talking things that at times outformed the market by 20% APY, which is incredible in reality.
APY, which is incredible in reality. That's a very quick 40-minute roundup of CAN.
I think what's really interesting from my position is I think AI is going to unlock lots of this
technology we've been talking about being important from identity to D- to privacy to zk's um to to defy even demystifying defy
in so many respects and i think he's going to create so many opportunities in web3 that we
could start seeing real use cases come to fruition in a way that we've been talking about for so long
and so i got a real positive vibe around identity and data and privacy from CAN
for that reason, actually,
because AI kind of demands so much of so many things.
It's a hungry beast to do things
and also can be problematic if not managed properly.
And so that management becomes really important.
Anything else you want to just check on on there, Hunt?
Yeah, that was a quick 40-minute roundup.
So just throwing it back at you now.
That was a great roundup.
Thank you so much.
Obviously, I think it sounded like a great event.
I know that there was some murmuring as usual before a conference
that it's not the right place to do it but it really does
sound like there were a few standout projects and topics uh ai being one of them of course defi
really cool to see defi taking front uh you know the the front of the what does it got the front
stage again probably saying that wrong um but this also goes back to something you mentioned, I think, at the top of the hour, which was, you know, you didn't see a lot consumer driven, or even maybe some narratives that are being developed and pushed by some of where we are at in the consumer space, where we are going with kind of
new applications that present, you know, this to a new audience, potentially. And of course, like you
said, regulations, things changing in the US, I feel like people are excited to talk about like, how
crypto facilitates a, you know, that that opportunity for everyone to be able to make money on chain.
So I think I'm not surprised actually to see DeFi taking front stage and also, of course, with AI kind of converging the two.
I think that personally, this is from what I've been seeing, I do think that the future of DeFi will look very different, both because of AI and social.
I think AI will help accelerate it.
Or should I say first, social will help increase the discovery of it, and then AI will help accelerate it.
Meaning more people will understand what is DeFi, and then they'll be able to leverage these tools, like you said, depending on how much you trust them,
to be able to go and create these different mechanics
that work for them to be able to maximize
what they can do with it with, of course,
whatever risk tolerance they have.
But it's really exciting.
I think this self-sovereign money conversation
that we had last week, I think carried over even to some degree this week.
So I would love to, I'm excited actually to put this together in another newsletter.
For anybody who's serious or interested, I am recapping all of these conversations now through my own newsletter at news.cryptosapiens.xyz. I'm using that as a way
to syndicate the content that we created on Ontology and then loop it back into our blog.
So the last three articles, I believe, are mostly Ontology. And go check them out if you prefer to
read and not re-listen to this conversation to get the highlights and some of the major talking points.
Aside from that, you know, I think I'm excited to see the conclusion of OnChain Gaming next
week.
I did talk to a lot of my friends who are in the OnChain Gaming space who were at the
base event yesterday with me, and many of them signed on.
So we should have an exciting conclusion to that four-part event
that had a very long break in between
because of an interest in ZK
and kind of the updates from Orange on ZKTLS
and now with our very own Jeff being at Khan.
So that's going to be a fun one.
And Jeff, I know you have a lot to say in that as a hardcore gamer.
If anybody here is interested in that, please do come back next week on Thursday.
It is 3 p.m. EST or Eastern Time.
I'd go ahead and do the math on where you are, but we'd love to see you then.
Jeff, anything else before we close out?
Only that you just reminded me of something, Humpty.
There was a really grumpy old man who obviously I had an affinity to listening to him talk.
I was about to say, were you looking in a mirror by chance?
One of those things that really resonated with me,
as he said, he said this phrase,
he said, Web3 has never had a problem
with attracting people who can build amazing infrastructure,
but we've always had a problem with people who can build amazing infrastructure. But we've always had a problem with people
who can build amazing consumer front ends and apps.
And I think that still holds true to some extent.
I think it's becoming less true, actually.
But I still think it holds true to some extent
in that we still, and I know it's becoming sad so often,
it's becoming a bit tiresome to keep saying it,
but it's still something we need to get better off.
Yeah, personally, I think that's social. If you look at what X added onto their platform, which was Polymarket, it's a prediction market built on Solana.
This is now it has a consumer front end, right? Before you had to go to this very funky crypto first
platform, not so much anymore. I think mobile first, web first applications that target consumer
experiences, you know, using very familiar and fun front ends with very robust, you know,
and kind of sophisticated backends that are powered by crypto.
That's the future. And personally, I'm very excited. I actually tweeted something out just
moments before launching this space where I talked about the article that I wrote last week,
or I should say earlier this week, that recap last week's space that, you know, was talking
about the introduction of DeFi and social.
And now, of course, seeing that become a reality
with a front end that looks just like a social app.
So yeah, this is really fun.
And in general, I think AI agents would become the front end
for most of the complex things that we do with blockchain,
and especially DeFi-like products.
I think that's where there's a huge scope
to build intent-based design with the UX of an AI agent
where it just works and it does what you need
without bridging and doing all the complex stuff of blockchain.
So I think, yeah, I think we are getting that.
And like ai might be
the one which provides the pmf for blockchain and for the technologies that we have been working
last decade yeah i think you could absolutely be correct on that's way and just to say
quick thank you for turning up i know i invited you quite late on in in the day and gave
you very little context i really appreciate you joining us today and let's keep in touch and keep
talking and keep us up to date with your progress on everything as well so really looking forward
to seeing how that goes yeah perfect thank you so much for having me uh it was pleasure to be here
and yeah i think identity is at the core on where we are heading uh yeah we're waiting for
some cool stuff from your end yeah and we'll talk and i'll explain why um um vampire attack
injection attacks are a problem for biometrics and we can have a long discussion over that as well
hey of course awesome and yeah thank you so much for coming up, Jeff. Thanks for, you know, recognizing that, of course,
and for inviting Sora Chain to come and speak with us.
Who knows, maybe the next four-part miniseries
will be dedicated fully to AI, right?
And we'd love to have you back
because we're always trying to explore these different verticals
or rabbit holes, really, because we're going down them
that we can explore. So we'd love to have you back then.
Yeah, of course. Happy to be here and for the next ones as well.
Excellent. All right, everybody, thank you so much.
Hope everybody has a great weekend.
And as our friend Polaris who's been in the audience would say,
make sure you stay hydrated, drink lots of water.
We miss you, buddy. All right. Bye, everybody.
Bye-bye. Thank you.
