Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
How AI Could Revolutionize Blockchain Technology
Recorded: March 24, 2025 Duration: 0:17:17Space Recording
Short Summary
The discussion highlights significant developments in the intersection of AI and blockchain, including innovative projects like Oasis's blockchain with a confidential VM, and the potential for decentralized AI. Trends such as the commoditization of data, compute, and bandwidth, and the rapid scaling of startups are noted. The conversation also touches on the synergy between Bitcoin mining and AI compute, and the growing importance of privacy in AI interactions.
Full Transcription
Welcome to Nasak Trade Talks, where we meet with the top thought leaders and strategists in emerging
technologies, digital assets, the regulatory landscape, and capital markets.
I'm your host, Jill Malendrino, and joining me this afternoon. We have Sanjay Gupta,
Chief Strategy Officer of Oredine, Juan Lopez, General Partner of Vanek Ventures, and Marco Stokic,
head of AI at Oasis, we're here to discuss AI and how it could revolutionize blockchain technology
and why there's still so much confusion about their intersection, certainly a key topic for us on trade talks.
It's great to have all of you with us. Welcome to the show. Juan, let's kick it off with you.
Give us a brief overview of where you sit within the AI and blockchain space.
Thanks for having us, Joe.
Yeah, my name is Juan Lopez.
I'm a general partner of Vanek Ventures.
We're a $40 million early stage vehicle,
really part of the broader Vanek umbrella,
but primarily focused on investing
at the intersection of FinTech, crypto, and AI.
For those that I'm familiar with Vanek,
we're a global investment manager.
We specialize in ETFs and mutual funds.
The firm was founded around 1955 but continues to be family owned.
And we're increasingly excited about the opportunities in crypto and AI and have made a few investments.
So super excited to be here and chat more about that.
All right, you got it.
And Sanjay, tell us about the work at Oridine.
Sure, Jill, again, delighted to be here.
I'm the chief strategy of O'Rodine based in Silicon Valley, a high company,
which is focusing upon technology infrastructure for both blockchain and AI.
And so this is a topic very dear to my heart.
We're actually focusing on power in the future of both.
And we provide all the infrastructure which drives the future of both blockchain and AI.
All right. And Marco, Oasis tells us.
Thank you, Jill. It's a pleasure to be here.
Oasis, we have actually built our own blockchain quite a while ago,
and we built one of the roll-ups on top of it.
It's a confidential VM, but we don't need to go too deeply into it.
But what we realize with AI companies is that there is a big need for privacy
or rather confidential compute, and that's what we're assisting projects with.
So Marco, let's talk more about that here because AI and blockchain, we hear a lot about these emerging technologies.
Sometimes the names are used synonymously or we're not understanding exactly what they do and where they fit within the ecosystem.
So why does AI and blockchain matter to each other when it comes to security, data-driven, decision-making and so forth?
I mean, there's great synergies between them.
The latest kind of pitch that I see and it really clicked for me why this is so beneficial to both is, and it's a quote or actually a statement by Sri Ram, one of the founders of Eigenlay.
And he said, crypto has an app's problem.
We have built so much infrastructure, but no one is building on top.
I mean, I don't know if you've ever used the crypto app.
I very rarely recommend it to non-crypto people.
And trust or AI on the other hand has a big trust problem.
And this is starting to become kind of obvious to many because we are again handling all of this data into these big companies like Open AI and we are trusting them with this data.
And the intersection between those promises kind of the
perfect symbiosis like crypto or AI can become the applications on blockchains and blockchain on the
other hand can help AI companies establish trust with their users with things like confidential
compute or decentralizing some stack of the AI infrastructure. Yeah. And Sanjay, I think Marco makes
an excellent point, right? Because when you think about blockchain, it helps to make AI more
transparent data integrity, which is at the backbone of, you know,
all companies these days, I feel as if talking about security and governance around data was a
conversation that almost happened after the fact, right? And that it wasn't almost thought about
security by design when these technologies started to roll out. What are your thoughts around that
when it comes to data integrity and what AI means for that?
Thanks, Jenner. I think makes an excellent the future of AI. AI, there's tremendous applications,
but there's also a question upon areas like deepfakes, the data integrity.
There's also going to be the future of AI which is driven by agentic workflows where you're not even taking a direct derivative of AI, but you're looking at AI agents communicating with each other and taking critical decisions.
In fact, there are some statistics put out by Forbes, one of the articles, saying that AI agents could be over $125 billion industry in the next five years.
So if you're looking at ensuring that there is security, there is, for instance, you know,
there's technologies in blockchain where you can guarantee security and authenticity.
of these transactions, you can put it out on a public ledger.
There's a strong intersection point between applications for providing security to the
AI application layer, which I think is a very critical overlap between the two areas.
Besides, there's an overlap on the infrastructure layer that we can get into down the line.
But I think on the application layer and the secure application layer, there's some strong overlap.
Yeah. And one, I'm curious to get this from the next perspective when you think about AI-driven trading and market insights. How is that being leveraged? How is it changing the way you think about investing and strategy selection?
Yeah, totally. I think every conversation when it comes to crypto and AI starts with sort of two core inflections.
And I think we've accurate it to Sanjay and Marco. It's really about trust and it's about ownership.
And when we sort of zoom out and think about that, especially in a world that has increased skepticism to sort of what's real and what's not,
as well as the collective sort of demand from society to participate and have more governance or have more of an economic say and
what happens with these systems, we see a really strong sort of setup for how AI could be co-owned,
perhaps through protocols to sort of manage these agents, or perhaps through the data itself that's being leveraged to train some of these AIs.
So when we sort of think about our strategy, we're a venture fund and we want to make sort of direction a long term.
That's we really think about founders thinking deeply about these inflections.
and looking to leverage them in ways that are both adding additional governance and additional attribution to the systems that are managing AIs,
but also thinking more deeply about how they could be mutually owned by a,
by other communities, so as opposed to just a hyper-scalers.
Well, let's take that conversation a bit further.
But in another way, when you think about AI and Bitcoin or crypto miners competing for those energy resources and being able to triage that demand and electricity, I mean, it's going to unprecedented levels.
What are your thoughts around that in terms of, you know, who's getting what and how they're, you know, able to triage this kind of demand?
Yeah.
Totally. Yeah, I mean, I think the super interesting thing that we've learned from Bitcoin is its ability to create this sort of industry when there's an incentive. And it begs the question. So when there's a valuable resource and I think we have a new set of commodities coming down the pipe, whether that's data, compute, bandwidth.
if given the right incentives, could we bootstrap those resources at scale and how could we monetize them,
how could we trade them in a way that is not just through sort of the bilateral agreements that are in place today?
So when you think about who, you know, who wins and what sort of setup enables that incentive,
I think it's exactly what many of the protocols and teams that are working on today.
It's quite early stages, but we're quite excited to see how different teams can take models from
or the perspective from Bitcoin and apply it to deepen in order to bootstrap new resources that can
benefit the AI economy. Again, things like data, compute bandwidth that are now starting to be
seen more as commodities, but can be traded on open blockchains.
Yeah, and Marco, I see you're shaking your head in agreement there. I'd like to get your
perspective as well when you're thinking about developing new models and applications around
the blockchain.
Yeah, I think this is the perfect kind of transition because this is something that
blockchains have been great at historically incentivizing users or kind of building an incentive
system for common good. Still, Bitcoin is the largest computer network out there, larger than all
the AWS data centers, etc. Of course, those are not very usable for AI compute, but it just shows
that it is possible to incentivize users. And that's for me kind of the one of the big selling
points, why decentralized AI makes sense because it's been proven.
historically like hey it is possible to aggregate this much compute now we just need to focus
it on something like decentralized training for new models yeah and sanchi i want to go back to the
conversation for a moment when you bring up agentic ai and how it's continuing to evolve beyond
automation um does this affect the blockchain in terms of um safety and security as it becomes
more autonomous and moves to more you know defy infrastructure yeah
The notion about blockchain is decentralization and it's about an overall element, well, it's trustless, but there's a notion of trust which is inbuilt into the way that blockchain is created.
When you think about agentic AI workflows that are going to come up, I mean, think about tomorrow.
It's not just going to be about you and me interacting.
It can be about your agent, a Jill agent, interacting with a Sanjay agent or somebody else on this panel agent.
So in that particular case, you're talking about decentralized applications even further.
There's going to be a need for trust and security that it's built into that application layer,
which is going to become all the more critical.
And think about now these agents multiplied in the thousands and millions,
as it keeps to any kind of customer interaction, customer service interaction,
in critical decisions is going to be driven more and more by these AI agents,
where security is going to become even more critical.
Yeah, and I think that's encouraging to hear because again, my initial point in the beginning
was security and transparency was almost in Afro thought because the technology was evolving
so quick.
And it seems as if security by design, governance by design is becoming more of a forethought.
And I think that's encouraging for the space as this infrastructure continues to develop,
especially as more tokenization and real world assets come on chain, that's going to be the
source of liquidity, right?
As you think about the capital markets and how they're becoming digitized.
Sure. And, you know, Jill, can I just touch upon a different topic?
I just want to get back to the infrastructure, the energy, the equation.
Because you know, at Auditian, we actually designed cutting-edge semiconductor silicon chips,
which power both blockchain and AI.
We actually build a full system stack on the full software stack.
And we, you know, one point about energy, right?
Because we know that Bitcoin mining in particular and any other proof of work type protocol,
they consume energy, AI consume significant mon energies.
But one of the parts that is often overlooked is the potential synergy between you.
So as you look at areas like Bitcoin mining, it provides a tremendous ability to balance supply and demand of electrical grids because you can rapidly cut, you know, increase or decrease the amount of Bitcoin mining that you do.
In the case of AI compute, you cannot do that.
You need a steady state AI compute.
But you have AI data centers that have visited all around the country, which are putting up, let's say, a one gigawatt plant.
They don't consume the one gigawatt completely.
And the ability to use a mining operation for, let's say, Bitcoin mining, to balance that demand of electricity can become very relevant.
So a lot of people are looking at these joint data centers where you have both AI compute and Bitcoin compute on the infrastructure layer, which are competing
Your specialized cooling technologies which are used in blockchain compute, which can be leveraged
AI compute as well to drive more energy efficiency and therefore provide a more sustainable
AI infrastructure for the future.
Yeah, that's a great point that you bring up.
And I think that we should, you know, shift that conversation when it comes to sustainability.
But I'm going to talk about the global ABA blockchain race for a moment here, Juan, between the U.S. and China in terms of where investments could be made as we're in this race for innovation and certainly a competitive landscape.
How are you thinking about that from an investment perspective?
It's an interesting question for sure.
And zooming out, I don't think anyone can say with confidence that there's a competitive
edge at the model layer.
It's clear that all the labs are investing heavy resources and some of the smartest people
in the world are all sort of working to build some of the best models everywhere in the
world.
And I think we're starting to see that from multiple fronts.
When we think about where to compete and where to invest, we think much more closer to the
application layer.
So we think that's where the largest arbitrage exists from some of the sort of SaaS businesses and services businesses that exist today.
And where AI could come and essentially build a protocol where it can be essentially services as software and start to productize that.
And basically can...
do the work of an entire country through different verticals that sell different services.
So we've made an investment in a company called FitChat that's doing this for investment research,
and we're excited to sort of zoom out and think about other applications that can greatly benefit
from the advances in LLM today and start to productize that and sell that to customers today.
I think you've seen the story from Stripe and from White Combinator that we're now at the
one of the most interesting times to build a startup and scale it because you're seeing companies with 5 to 10 employees get to 10, so even hundreds of millions of AR at an unprecedented pace.
So I think that's specifically where we're most excited to invest is at the application layer because they can always expand vertically.
Whereas the models and the infrastructure layer itself, it's just a lot more capital intensive and it's getting commoditized much where rapidly we think.
Yeah, which also makes the MNA space interesting around what's happening in digital asset infrastructure as well.
And I think that's another theme that will certainly see a lot of focus, even, you know, digital asset companies that are looking for traditional IPO routes, right, to, you know, get more liquidity and acquire more capital.
You're starting to hear that.
as well, but I still think at the end of the day, one, when we think about it from the capital markets perspective and the technology perspective, it's still early innings. And even though the technology is evolving so quickly, our conversation could look very different six months to we're up for now.
Totally, totally. And that's why we're a C-Sage Focus Fund, right? I think we're at the point where we're simply looking for some of the best founders and teams that are going to come and navigate this environment because it's very much rapidly changing. And it's less clear once you're looking to make investments at a growth stage where, again, value could shift quite rapidly.
Yeah, and Mark, I want to hear the same from you as well, where we could be in the near term, where you think we are in terms of the evolution of the AI and blockchain environment.
And what this could look like longer term, it might even be a very immature question to ask because, you know, this conversation, again, could look very different in just months from now.
We are very early.
And as one said, I think the application layer is the most interesting one, because currently
everyone is just kind of trying to build things, check out new features.
But since it's evolving so quickly, you kind of have to reiterate quickly.
And it's, you have very few teams that are actually building companies.
Everyone is now building products and shipping them and just kind of thinking about the fun
side of things.
But what's still missing and this has been a crypto problem for a while is projects actually
thinking in a like three to five year horizon to build something sustainable out of it.
But obviously, that's the difficult part and few people want to build companies.
It's way more fun to solve complex problems.
Yeah. And when you think about projects that could come on chain, what areas, what sectors
do you think holds the most promise? Where would you like to see more work being done as it relates to projects coming on chain?
The decentralized training space is incredibly exciting, but this is not the application layer and that's super complex and I'm happy to kind of leave it to teams like Prime intellect and Jensen who are building new decentralized training methods to allow these foundational models to be built on a decentralized compute network.
What is exciting to me is all of those that kind of improve on the privacy level, private chatbots, for example. Just I mean,
I try to remind people daily.
If there is ever an open AI hack and there is a leak of your email with all of your
prompts that you ever did with an open AI model, horrible.
I mean, honestly, I would be wasted because I use it for everything daily, like personal
things, health problems, everything.
So just solving this, I would give up a bit of the quality of the model just to have
private conversations with LLMs.
Yeah, that's a great point that you made because I was talking to co-pilot this weekend about my allergies.
Isn't it too early to start up and it's only March?
So I leverage it in the same way that you do.
I virtually don't even use search anymore.
So Sanjay, to wrap this up here as we're nearly out of time, where do you see projects having the most promise or how would you solve for those security issues?
What do you think the next evolution of this?
Where should the focus be?
Yeah, so clearly there's security, and I'll actually take a slightly different kind of that, you know, because I know Huan and Marco talked about the application there.
There's work that needs to be done in the infrastructure layer for not just security.
Okay, all right.
Let's wrap that up there.
It seems as if we have lost Sanjay there, but appreciate everyone's insight.
It was great to have you on the show.
Thanks for joining us on trade talks.
I'm Jill Malendrino, Global Marketing Reporter.
I'm for the future of both of them beyond what has been done at the application layer.
