Hello everybody, welcome to influence.
How is everyone doing today?
Hello, hello, good to be on here, nice to see you, Masha.
Hey, what's going on everybody, fantastic.
Yevgeny and how are you today?
I'm good, wonderful. Thank you. Hey everyone.
Awesome. We're here to discuss a very interesting topic, the AID apps built for the next hundred million of users.
And at that I'm going to pass the mic over to you, Tom, to kickstart our conversation today. And thanks everybody for joining.
Terrific, it's great to be here. Thank you, Masha. Thanks everyone for joining. This is going to be an interesting one because it's something that is certainly not my area of expertise.
And I don't think it's our projects real area of expertise influences building the cloudless compute network. And this is something very different but we always want to learn more about the sectors that are adjacent to ours and I think AI is certainly
adjacent to anything cloudless. And I think that a lot of what Fluence does will enable real AI in a decentralized manner which I think is an entrusted and verifiable manner which I think is super important.
So I think there'll certainly be some overlaps here. And I think we're really interested to hear ourselves, the type of apps that are coming. And so thrilled to have these speakers talk about what they're seeing and share their expertise with us to help
advance our knowledge. So who wants to go first?
Any speaker want to kick it off?
To introduce yourself and let us know kind of what you're doing in the space amplify or who wants to go first?
Yeah, I'll go ahead first. This is a crypto grad. So hey, what's going on, everybody? My name is Parmit. I'm the head of communications here at crypto grad. And obviously, the topic of discussion is AI, which is super exciting within web three.
So some of the things that we're working on here when it comes to AI related stuff inside the crypto spaces, we're trying to more focus on the educational side of things with crypto and AI itself, because we are firm believers here that know the next wave that's going to really help the
you know, the newest trader or the more professional trader is helping them take advantage of AI by using those type of backbone structures to help make more solidified trade and get better insights on different trades that you might be trying to devote into with some
you know, some more backup detail from AI itself. So that's kind of our big thing that we're trying to build education into AI and build, you know, a community focused around helping traders were amongst another and also using the power of AI to kind of grow their trading skills in a nutshell.
Super interesting. And out of curiosity, help us understand more about how that's working. When you say AI, that means a lot of different things are using one model, your own model, are you building off of some general base model? Like, how does it help us get a little deeper into that?
Yeah, so with it, we are obviously using a model that's pre built, but we're obviously taking advantage of it, because most people I don't believe truly have taken advantage and understanding that they can use trading, or get better insights on trading through AI, because AI is a tool that's very powerful on providing insight, data, and things like that, that most people may, especially as a new trader, may not know what to kind of, you know, do or how to do it.
AI itself can pretty much give you the step by step process on what to do. So that way you can become a lot more of a professional trader, because especially when you're new, you're starting out, you may not understand know, know what you know, pips are different things of that nature, you may not understand how to read graphs, charts, and these are all, you know, obviously defining factors that obviously can help you really have success in trading. So we pretty much know, give people the tools, we pretty much tell them what they can type in, how they can look up details, how they can
get your insight on their trades through AI, so that when they're obviously going out there and they're ready to execute that trade, they feel a lot more confident because they went through AI, they typed it in, they typed in a question like, hey, what price points should I get into this coin at? What are some of the back history of this coin? How long is this coin been now? And obviously, they can see all those finer details as they're getting ready to divulge into it.
It's super interesting. But my question is, are you using, what's powering that? Is that your proprietary model? Is it just a front-end on chat GPT? Is it, what's the actual engine behind what you're talking about?
Yeah, so it is a engine of behind a chat GPT. We're using them as our backbone platform for AI.
Gotcha, super interesting. And love to learn more about the opportunities and risks there. Who's up next?
I'll jump in. I was getting rugged before slightly, so I think it's working now.
And yeah, so basically, amplify what we're doing is we're using AI and machine learning combined to basically build more of a decentralized automated yield aggregator, right?
And the premise of what we're trying to do is we looked at how these lending borrowing protocols are operating at the moment, especially when it comes to centralized entities like Celsius and BlockFi and Voyager,
and how ultimately it's the human element within that that cause so many problems for the users there.
And we've seen FTX as well blow up when people get greedy when times are exuberant, and then they also don't make great decisions when times are bad as well.
So we're trying to sort of prove out, I guess, that we can use AI and ML to power a protocol to deliver long-term sustainable growth, right?
Something that is rooted in sort of logic and long-term thinking rather than short-term gains.
And so we're basically building that out in a number of different ways, and there's a number of areas we'll be focused on to basically automate that entire process while it's also de-risking actually compared to centralized entities.
So yeah, so it's just the alpha release, which has gone quite well.
And then we're going to obviously look to do a beta release, and then beyond that, we'll flesh out a much more comprehensive platform.
Well, I've got a couple of questions for you, but first, same question I asked before, is this your proprietary model you're using, or are you front-end on another, on a different, more kind of general purpose model?
Yeah, so one of our senior devs, it's his own proprietary tech. So yeah, we're building everything out ourselves because there's no real roadmap for this sort of thing, right?
And also, obviously, it's a machine-learning model, and over time, the decision-making will get better as it sort of analyzes on-chain data, what sort of work previously, what hasn't worked previously.
So yeah, so it is proprietary in that regard.
That's definitely interesting. And you also just also mentioned quickly decentralized. I'm not sure, it doesn't sound, it sounds like it's a proprietary model, so just curious, what's the decentralized component here?
Yeah, so basically, the way it works is we basically wanted to operate in the sense that the model will basically look for opportunities, perhaps it will look at on-chain data and look for new protocols or new liquidity pools that we could potentially use within the overall strategy.
And when I say me, I mean the overall system.
What will happen then is that the system will then propose it via governance model to the community, and they will vote whether they want to integrate this new pool or this new protocol into the overall strategy.
And if it's approved, then ultimately, it will start to balance yields across additional pools to what it currently has.
So the idea here is that we're not manually making the strategy.
There are some other competitors of ours who came to be decentralized, but they have, let's say, for example, five validators, all five validators belong as part of the team, and they all reach consensus among themselves.
So not really decentralized. What we want to do is we want to have a community-powered platform that is ultimately powered by AI and ML under the hood.
Interesting. All right. I still have more, but I want to move on to another speaker to introduce themselves. Who's next?
I just joined. I might be next. Hey, everyone.
Great to be here. Thank you so much for the opportunity here, and thanks for everyone for joining in as well.
I'm Jesper. I'm the co-founder of SciFinance, PSI. We're live at scifi.app.
But what we see at scifi is a huge UX problem in crypto, especially for onboarding new users.
But even for existing users, even experts like all of us here, they're just general issues, and we really believe that AI can also help play a huge role in this on many different fronts, both educational,
but also just in terms of, in our case, like fraud detection in terms of financial transactions help only unlock certain transfers when it makes sense, things like that.
We're already playing around with some very kind of crude prototypes on how to prevent hacking from the wellness that we're building.
And I think obviously AI will have a huge impact on the entire world, obviously.
But I think it goes really hand in hand with crypto as well. So we're excited to integrate and I'm excited to be here. Thank you.
Really interesting. Yeah, I mean, I think having those AI tools to help prevent fraud are interesting.
I can't although I can't help but wonder if some of the tools the banks use to hold up our transfers are also AI generated.
It caused all of our frustration. So I don't know what point these two worlds come right back around and meet in the middle.
That's my that's a little bit of my concern.
Anyone, anyone else? I guess we've heard who had their go, or do you want to go ahead? Yes, sir. Thank you.
So hello, everybody. This is Adam from rain at work. I'm global ambassador for rain at work.
So rain at work is a public blockchain that is compatible with EVM and it's designed to be lightweight, easy to use and developer friendly.
So we're actually trying to make our platform a suitable place for building the apps and the idea is also one of them.
So I say this is EVM compatible. So we were able to leverage the benefits of the cost system.
So we're trying to address the common issues like, you know, high cost, love efficiency.
So we have we also we don't also we don't only have the apps in the ecosystem.
We also have game fi NFTs. So our platform is designed to facilitate actually the rapid migration of existing the apps to the rain at work.
And yes, sir, that's my take.
Terrific. Interesting. Can you give a little bit more detail on that? Because is there other apps live on the system now?
Like, how does the integration work?
First of all, we think that there are some essential problems in existing public blockchains.
Like, for example, there are very high fees. The interfaces are not user friendly.
They are not they don't have a support for developers.
So our project is actually on the blockchain for solving these problems.
And for our topic AI right now, we only we already have a lot of the apps, but especially in the recent years after the AIs, AI being more popular,
we're thinking to get more AI powered the apps into our ecosystem.
And right now we're also thinking about designing and editing or interface to make it more developer friendly as a wish bus or Mango at the end of the day.
So, yeah, that's actually what can I say. If you have any question, I can answer.
Well, you know, I think I think what will happen, I'd love to go into next.
But first, I want to make sure we've got all our intros is some examples of some of the related apps you mentioned, because I think that's what this space is about.
I'm kind of curious how you can talk to you about that. But before before we get into that, any have we missed?
I'm just trying to think. I think we missed anybody.
Evgeny, do you want to you want to give an intro as well?
Yeah, sure. Hey, you're getting here.
Co-founder of Fluence together with Tom and Dmitry.
We're working on, as Tom said, Cloudless computing platform, which is basically like decentralized and serverless computing platform capable to host any kind of workloads, including AI.
Like we plan to run AI models on it.
So like should be useful for any related applications and provides full tolerance.
Cheap prices compared to centralized cloud security and, you know, a lot of developers features out of the box.
So we believe that actually like deep in platforms and decentralized compute platforms going to eventually replace centralized platforms as they are more secure and transparent and can evolve quickly.
One question I think I'd like to get into a little bit is on these apps.
First, you know, maybe turn it right back to Erdem because I would love to hear people talk about a little more specifics on some examples.
And obviously, a couple of you have financially specific applications.
I think are super interesting. We've heard about those.
But I think it might make sense to go back to Erdem to talk about any if you have any off the top of your head examples of AI related apps.
Either you're on your platform or not or things you'd like to see anything like that you want to get into.
Yes, yes, for sure. Right now we actually so if you basically want to be in our ecosystem, there's a process.
So right now I cannot tell a lot of examples of I.D. apps in our ecosystem.
But we know the importance of AI. We know the importance of I.D. apps.
And we always try and expand our partnerships and we're always trying to expand the projects in our ecosystem.
So to give you some use cases, for example, I think AI powered landing platforms.
They can assess borrowers or they can assess credit burdens more accurately or leading.
I think it can lead to more efficient loan approvals and lower default rates.
Like there are some trading bots. You know, I think if you build some AI driven trading bots,
you can analyze market trends and execute trades with more accurately.
And essentially you can optimize users investment strategies and you can maximize returns.
You know, also, I think AI integration, I said we have also some gaming projects in our ecosystem.
I think AI integration in decentralized gaming and entertainment, the apps also can enhance user engagements and immersion.
So, you know, there are some algorithms and these algorithms can enable dynamic content generation
or it can develop the gaming experience for gamers and it can provide adaptive gameplay mechanics.
So I also have to talk about the chat bots, you know, like chat GBT
or there are a lot of chat bots right now that developed by different companies.
So I think AI powered virtual chat bots, they can provide users with real time support,
which I think is really important in any project because your community is always seeking for information
and sometimes even your moderators can help them.
So I think AI powered chat bots are really important for any projects in the future.
So, yeah, I think these are the use cases that I can say the AI for the apps.
Well, listen, it's interesting and I think, you know, I guess we might as well dig into this financial one
because there are so many of you guys, those speakers on the call that have financial exposure applications one way or another.
Do we enter a world, or let me rephrase the question differently,
how long before the world is just AI bots versus AI bots in terms of trading?
Like, I know we're not there yet, we got to get there pretty soon.
And then how does that world look like and who wins?
So talking on who wins, if it's a battle of AI versus AI,
I think it's ultimately going to come down to a few things.
One, experience people that have obviously that will end up staying in the field for the longest,
that end up garnishing the most amount of knowledge of, you know, that during that time to kind of figure out what works and what doesn't work
and what's best for the user because ultimately we are building most of these systems to kind of better entail
and help the user get the best experience when trading or whatever the application may be used for.
For Cryptograd, our entire process is focused behind the idea of helping that end goal user get better
and obviously have a better confidence when it comes to trading.
So I think it's going to be more focused on if you're here for the longest, you get the better understanding
and you obviously understand what the user needs, you'll have the most success
compared to someone that's just coming in and just trying to focus, doing one thing
and hasn't gone through trial and error.
Trial and error is the best way for success and I think ultimately the best person that sits there for the longest
will know what the user needs and they'll come out on top.
No one's going to say it's the number, the amount of GPUs thrown at the problem
or the amount of dollars thrown at the problem?
Nope, nobody wants to say that because that's also an answer, I suppose.
Well, maybe you have to tell the name directly, maybe it's going to be more promoting.
Nobody is asking a question, maybe you have to say like,
Jesper, it's your turn, maybe it's going to be more promoting.
Yeah, fair enough. I mean, I'll call people out, definitely.
I mean, I guess, go back to amplify. Any comments here?
So, I mean, this definitely isn't my area of expertise at all
when it comes to sort of trading bots if we're talking about specifically
but I imagine, you know, regards to its AI, you know, there's still,
I suppose, human intention behind it, right?
So, you might have, you know, AI bots that are trading for like a day trader, right?
10% gains here or 3% gains there and, you know, you're battling longs versus shorts
but then you'll still have the human element, right?
Trying to chuck in whatever amount of money it is to get a 100X or get a 50X
and, you know, you can always tell the AI like,
listen, don't sell my position until it reaches this amount
and if you get more volume going in than it is that's selling,
then obviously the price will still, you know, go up
and I just think it's just going to be an easier way for people
to basically manage their positions.
It'll be sort of more of a backseat approach.
You know, I was working with a team called Octavia.
Previously to this, people might have heard of it.
You know, Web3 Assistant, AI based, again, using sort of chat GPT in the background
but you can ask it to do certain things like, you know,
buy into this presale at 2 a.m. in the morning when I'm in bed
and, you know, use one Ethereum, here's the contract address
and it'll do it for you and then, you know, sell at 3X
or sell at 5X or whatever it is.
You no longer have to connect your wallet, you know,
make sure you've got good connection, you know, watch the launch of a project.
You can just basically take a passive approach to all of this
and so I think AI will help improve that side of things for sure
but I wouldn't know probably the intricacies of the sort of trading bots,
you know, algorithmically or whatever for small gains, yeah.
Listen, that's a great answer.
That is terrific, showing how AI can help affect a strategy
that just saves labor effectively, which makes a ton of sense.
And I also would, I would re-characterize your earlier comment
or rephrase it as there are different time horizons
and, you know, AI may be very impactful in some time horizons
and less impactful in others.
That's also a, you know, real possibility as well
when you're talking about, you know, you mentioned sort of gains
but it's probably more time horizon based potentially.
Let's see, where else should we go with this?
Are there any other non-trading applications
that we should talk about with regard?
And we're trying here to talk about what's bringing the next 100 million users here.
So what, I mean, trading is a big one but are there other applications,
AI related that we think are worth having on the radar,
whether anyone here is involved with them or not?
Yeah, I think like as a gaming becomes a big part of crypto
and AI is widely used in gaming,
this might be one of the interesting use cases
for sort of crypto games, web tree games to use AI
to, you know, make the gamer experience better
and to make games more interesting.
And on the other side, basically, of course,
people gonna use AI to automate their gameplay,
their actions to, you know, get more points,
more like rewards, more tokens, higher ratings and so on and so forth.
So like people gonna leverage actually AI to get
And especially this is gonna happen to crypto.
We saw it before with, what was this name?
Like Axie, Axie Infinity, but I think it was like mainly people
But like, I think in the future, we'll see a lot of AI agents
in these games trying to get like maximum value out of this.
Yeah, and that's actually a really interesting point
because then what happens next is that make it uninteresting
for human agents to play or for humans to play
if AI agents which are presumably are better.
And then if that's the case, you end up having proof,
you need to have authentication and proof of human
I mean, you start to, it starts to get to an interesting place
where there may be games that are dominated by AI bots
and others that require constant human proving potentially.
Yeah, and actually with permissionlessness of crypto,
it's actually more challenging than in Web2.
In Web2, you can try KYC as much as possible
with credit cards and stuff like that.
And then crypto, like ideally if you wanna play,
you just connect the wallet and you play,
but then any AI agent can connect the wallet.
So it's gonna be challenging how crypto, like Web3 games
are gonna onboard users and at the same time defend these bots.
Yeah, I was just gonna say like on the other side of that coin,
I'm also very excited about that future where the gaming world
are going to be so intelligent and engaging at a whole new level,
like 10x, 100x more than today.
You're gonna make friends.
You're gonna make actual friends in a game
that are not humans necessarily, right?
Like you're gonna go on quests that are,
maybe the game developers themselves didn't even know, right?
So they're just building kind of the start
and then the AI engines in there and agents and whatnot
just kind of evolve this world into something
that even the developers didn't even see.
You could even see games be developed within games.
It can really go in so many exciting ways in my opinion
and be way more engaging than anything we see now.
I think that's pretty thrilling to be honest.
I can definitely, definitely see that happening.
No doubt and not to mention that some of the games could from the get-go
be developed by AI agents, right?
Like games could be fully developed by that, right?
I think that's something that's extremely exciting.
And then of course, you know,
blockchain just plays such an important role.
You know, I almost see it like as an underlying,
almost call it like a security infrastructure for AI, if you will.
So just like humans, you know,
we can see ourselves as a very like advanced form of AI, right?
Obviously where the A is less important there,
like it's not so artificial.
But my point is that we can't even hack the blockchain, right?
Like all considered and AI system is starting to catch up with us,
of course, and in many ways smarter depending on what the task is,
But my point is, you know,
for some of these systems and I'm just talking in general now,
not necessarily just about games,
but we're going to need some sort of like security layer to run this on,
And that's also why blockchain is important for the world in general.
We don't need a financial system that runs based on trust.
We need it run on blockchain.
Of course, we need everything run on blockchain,
but the AI systems need to be constrained as well to blockchain or else
like, you know, they can essentially run wild just like humans do.
So I think it's going to be absolutely crucial.
But what I was saying in terms of the games too is that,
of course, as we all know here is like blockchain is going to be exciting
just in terms of digital property, right?
Digital ownership, things like that.
I really am excited for when we start truly seeing kind of cross
universe asset transfers, right?
So you own something in Counter Strike and you can move it over to
Minecraft or something like that or whatever have you, right?
I think that's going to be exciting.
It's going to bring a lot more people into gaming in general and bring
a lot more money into it, a lot more value into it as well.
Yeah, listen, I think that is interesting.
And I think, you know, it's very reasonable,
a bigger point that gaming has real potential here to drive a
lot of use here, no doubt an AI app,
whether that is an AI developed game,
whether it's engines that help people play games or monetize games,
that seems like an inevitability,
just a question of when and how.
Are there any other verticals that we should be thinking about,
whether you guys are in it or not?
Cryptograd, anything on your side here?
Amplify, you want to dive in?
Yeah, I was doing a sort of a mentor session with a bunch of
university students here a couple of days ago,
it was actually in partnership with Coinbase,
and we were looking basically how to get to the grassroots
and make people aware of the benefits of blockchain and,
you know, remove misconceptions.
And I was working with a team of young women there,
and they were basically saying,
one of the major issues that they face even today,
which I didn't even know about,
is you have a lot of these bots or a lot of these malicious
actors who actually, you know,
take the faces of these women on their public profiles,
and then they attach them to like pornography, right?
So like naked bodies, whatever else,
and there's nothing that these women can do.
They can report to the police,
but there's no law to protect them in any way,
The police can't do anything about it.
And even if one gets reported and taken down,
And it's gotten to the stage where they just sort of accept
that it's going to happen, which is pretty terrible.
But that got us thinking about, well,
if we had like a social media platform on the blockchain
where you had verified users in the form of like,
you know, a verified NFT using ZK technology
then every user would have a verified profile.
you can use AI that sits on top of the,
basically the social media as a moderating tool
that can look to remove any sort of malicious content
extremely quickly, right?
So you can program an AI to say like,
if there's over a certain percentage of nudity
or if there are certain images that look racy,
and here's a previous example of 10 million images
that would be considered racy,
remove it immediately and block that account.
And that's something I think that social media
apps at the moment really struggle with is
moderation of spam and malicious content
and bad actors and all the rest of it
because it's just too much for them to do
and the current technology that they use doesn't really,
it's not really adequate enough.
So we really built this idea
and sort of presented it to Coinbase
and it did particularly well.
but it just got me thinking of how AI
can actually revolutionize that side of things as well
and make it a much safer place for people to interact.
I think if you had a social media platform like that,
then that would encourage a lot of, you know,
positivity and I suppose good or good intention people
to come together and remove malicious actors from the space.
Listen, I think that is a terrific concept.
I think though you've also highlighted another problem
with AI, which is it actually helps accelerate
the generation of the exact images
and things that you have concern about.
And you could easily see a program to create,
you know, a thousand of those a minute a second.
So I think you're right that there's a real,
there's also a protection on that
and if you interacted with blockchain and verifiability,
no doubt, but we're also seeing,
I think you sort of touched on maybe inadvertently
some of the AI applications
that could be pretty negative and harmful.
Yeah, for sure. Absolutely.
Like it's like everything.
There's always the good with the bad, right?
I suppose if we have something that's verifiable,
that means people won't necessarily want to
be on that platform because as we all know,
the blockchain records everything that's on it.
So that includes, you know, potential messages,
potential images and whatever else.
So in order to have a verified user on chain,
then you have a system where you could report that user,
I guess, for malicious behavior.
And so the idea there might be that social media
has to move towards this new system
where everyone has to be verified
in order to be able to interact with each other.
Now, is that possible for exploitation?
Of course, but I think it depends
on the policing policies overall.
But just like anything else,
of course, there's the good with the bad.
And yeah, we don't really know, do we?
We don't know until it actually happens
how good or bad that's actually going to be.
And there's always the issue that potentially
more rights taken away from you in order to offer you
more safety or more protection,
which is what people are sort of worried about.
But, you know, it's a hard sort of question to answer
until we're sort of living in that world.
Totally agree, totally agree.
But I sort of phrase it is the ability for AI
to create negative content quickly actually leads
to help to accelerate the world you're talking about,
which is an AI-enabled world with more protections
So you kind of it kind of pushes you into the world
you're talking about a little bit.
But crypto-grad, you were about to say something
and then we head off in this direction.
No, I'm actually kind of excited to see
if we headed off in this direction
because I think this is a great topic to bring up
regarding, you know, I guess, impersonating,
you know, things of that nature,
but not having that verification
to kind of make sure we avoid these issues.
I think, you know, even though as AI progresses forward
and, you know, we see AI taking over in the future
with like, you know, different ways to do things,
I think there still needs to be the accountability factor
with a bit of a check and balance,
with humans still kind of remaining in the mix of this
to be able to check and verify anything that is going on.
You know, same thing with AI being used
in different applications such as trading, gaming, et cetera.
You know, there needs to be a mixture and there's a check
on how we're handling information, you know, data
and things of that nature because obviously
that can get out of hand pretty quickly.
And as long as we have, you know, a human
behind the mix of it, you know, kind of, you know,
watching how the data is being used
and how, you know, all these are being played out,
you know, I think we'll always have that check
and balance, never have that fear of, hey, you know,
our information is being misused or, you know,
this person is not the person who they say they are.
And it's also creating that trust and accountability.
And I think that's a big factor too with communities
and different projects is having trust and accountability
behind the people in the project knowing
that they're actually going to be doing what they're saying
they're going to be doing and not, you know,
Listen, completely, completely agreed,
and all those are important issues.
I guess I've got, you know, another question here, I guess,
and we sort of vaguely touched on this,
anyone have ideas on ways that, you know,
AI adapt developers, I guess that even goes beyond
AI related, but can include token economics
and incentive mechanisms that encourage
participation and engagement.
And the challenge, which I guess we sort of touched on
already, is the extent that we do that,
you run the risk of back to the AI engines
and bots, you know, exploiting those token economics, right?
So is there a way to do that?
Erdem, do you want to dig in?
Well, maybe they can be kind of a reward system,
like you can implement token rewards
and maybe it can be incentivizing for engagement,
like think that users could earn tokens
which is participating in AI training processes
or providing feedback on the platform.
And I think these rewards can be distributed
based on the quality and the quantity of contributions,
and maybe it would encourage users to actively participate.
Or it can also be used in the sense of staking and governance,
like you can introduce staking mechanisms
for users to hold and stake tokens,
which that doesn't only contribute to network,
but also grants them voting rights and governance decisions,
like your users can reward with additional tokens
or voting power based on the duration
and amount of tokens to take.
And so essentially, it's going to foster a sense of ownership
and alignment of interest,
or you can increase the utility of your tokens
that serve specific functions in the AI-D apps ecosystem.
For example, tokens could be used to access premium features,
like unlock advanced analytics tool
or purchase data insights generated by AI algorithms.
So if you offer valuable utilities,
I think developers can drive demand for tokens
and encourage users to participate actively.
You know, when I started in my blockchain career,
I was a community manager.
I think community is the backbone of any project.
So establishing community-driven incentive programs,
they can also foster a sense of belonging
and collaboration among users.
So developers can reward users
who contribute to community forums
or participate in discussions.
And I think that can take the form of tokens,
or like maybe some kind of exclusive privilege in the community.
So also, you can adopt an unlocking mechanism
like users could earn increasing rewards over time
based on their participation.
maybe you can motivate your users to stay engaged
and active on the platform.
So ultimately, they're going to know
that the reforms will be rewarded proportionally.
Listen, I get it, and it makes sense.
Just don't forget that AI engines
are perfectly capable of using tokens as well.
And actually, you know, this is something we've thought about
because Fluence is governed by a DAO.
And DAO is basically obviously on-chain voting
that governs everything with regard to Fluence's
going forward from treasury to roadmap to everything.
And we've actually written in that, you know,
AI engine owning a token is a valid DAO participant and member.
And so that is something that we've kind of already foresee
as a AI agent effectively being a governance participant,
which I think is interesting.
And, you know, I think we'll see more and more.
Let's see, who, anyone else, want to get into kind of the token-related
incentives related to AI applications?
We can talk about either specifically how to use them
or also ways to, they might be misused.
I guess that's also a possibility.
Yeah, I think it's going to be interesting to see how this is,
because, like, again, if you think of humans as, like, you know,
also a very advanced, quote-unquote, AI system,
you know, we can kind of draw some parallels.
But, you know, in the real world, we also have incentives, right,
through money and, you know, you can go to jail or, you know,
you can, like, unfortunately die if you operate in this world
and you do something wrong, right, like, in many different ways.
But in the same way, like, that's how you can help
constrain people to certain actions.
And the same with AI systems, right, like, you need to,
you know, they need to have some sort of sense of,
like, what will kind of drive them to do something better, right?
And, like, do they have a sense of good and things like that?
I think for all, at least biological systems,
survival is certainly rooted in many things we do.
So in the same way, like, how will that look in the AI world,
I think, could be some interesting things, right?
You know, it's like if you tell an AI bot,
hey, you can make more money, you know, they'll need to,
of course, be able to use that money for something,
but maybe they can, just like in the real world,
they can use that to improve themselves or to get something
that elevates them further into some other state
or something like that, right?
As long as that makes sense for them,
then they want to get more money.
And then the point is, then you can incentivize them
smaller, probably, to what we have now in the real world.
And there's a lot of chaos,
there's a lot of bad things happening,
but, like, largely speaking, over 8 billion people,
you know, it's, like, largely okay.
Again, definitely room for a ton of improvement.
But it's not like we're all killing each other,
and it's definitely better than it was, like,
500 years ago, in general, in general.
Yes, yes, no, you're right, you're right.
Looking at that time span, way better.
Cryptograd, what's your thoughts here?
So a couple things I want to touch base on.
One, we talked about incentives,
and incentives are a big key factor.
I think incentives, obviously, are a driven factor
to kind of get people more actively involved.
But then you also talked about the misuse of that as well.
I want to first talk about how, you know, in Cryptograd,
we do offer incentives as well.
And obviously, there has to be a check and balance
to make sure that whatever we're incentivizing
to obviously get people to do
is not going to be AI-powered
and not going to be a misuse of it.
So, obviously, over here at Cryptograd,
we offer a reward system for people
that give us feedback on, obviously, AI,
and, obviously, the general feedback of the AI
and how it's working for us on our side.
And, obviously, we also offer
a general reward system for people
that obviously referral people into it.
But we also have an authentication system
that obviously does a bit of a check and balance
to make sure people are not misusing the system
just to have a large gain in the process.
And on top of that, we also put auditors in place.
And I think that's a big thing that people got to start
looking to do down the road is put auditors in place,
in which they check every data that comes in
to make sure it's not AI-powered
and it's authentically generated by a human
who typed it off on their end.
So, I mean, the feedback is obviously human-based
and everything and so forth.
So, and it's an autonomous system.
So, obviously, it kind of has a round-the-clock system
that can constantly have the check and balance.
Certainly, certainly interesting.
I'm trying to think of an example.
Can you get a little more detail on exactly that
because that's actually, I haven't heard that.
So, how do you verify and how do you validate
that it's actually user-generated?
Yeah, so the way we validate if it's user-generated,
so a human, once the feedback is obviously received,
a human that's from our team over here, a cryptograph,
obviously verify the info is legitimate.
They'll obviously do a back-check on the info
and see where it came from
and obviously check the data
and see what's the origin
and make sure it's not just from like a,
you know, because obviously, AI, if you build it,
it's almost difficult to, you know,
it's difficult to say where the, you know,
it's difficult to use AI to say,
hey, I posted this information from this location.
You know, typically, you know, most of the time,
AI is, you know, it's pretty much a software program
or just, you know, it's just coding, you know.
This, what is human side of things,
we can obviously backtrack all that data from location,
how active the person has been,
how often are they logging into their account
and what type of feedback are they providing
and how, you know, particulate is that feedback
on what they're providing for us, you know.
So, and obviously, it's rigorous, you know, constant,
you know, digging through that information to make sure
it's not going to be something that's going to be abused
to obviously just, you know, have a large gain
for their factor, for their site.
Well, that's a lot of work on your side.
That's a lot of manual work.
To scale that's going to take some effort, huh?
No, definitely it's going to,
but obviously our goal with the project is long-term,
it's not short-term, so we're obviously going to be adapting,
you know, upgrading and trying to find different ways
to make Cryptograd stronger and better for the community.
Let's see, I'm trying to think about the best place to go here.
Are there any topics that related to these applications
that I haven't talked about that we could?
I mean, one thing we haven't talked too much about
is interoperability, and kind of curious if that's something
people have thoughts on or ideas on.
Anything related to interoperability of either AI systems
or AI and blockchain or, I mean,
even between different AI models.
I think we've got no thoughts.
Oh, Jesper's got thoughts.
No, I mean, I think that AI systems will be very,
you know, useful to kind of find different great opportunities.
I think some sort of yield optimization was mentioned earlier.
But it could be other things too.
So just, I think at the end of the day,
what we want is like abstraction,
just like with the internet, right?
It's not like we're sitting there, you know,
and thinking about TPS, IP, and SSL, things like that
when we navigate the internet.
But all this stuff is going on underneath us.
In the same way, we need more and more abstraction
in crypto as well as the space matures
so that at the end of the day,
we don't really care what chain we operate on.
We just interact with the apps and whatnot.
That's going to be, you know, at least closer to the end state.
So something similar to the internet today.
But my point is that I think AI definitely can help
with that abstraction as well,
kind of routing things to where they need to go,
finding the best path for certain things,
improving the speed of your experience as well.
So just constantly be in the background
and just making sure your experience is the best as possible
and interpreting what you want to do
in the best way possible, basically.
It's almost like having these live agents
kind of as a middle layer between us and the blockchain.
And they're just all these agents running around,
helping improve everything.
And, you know, like it's like an Amazon warehouse,
but with like AI agents that make sure
that all the stuff goes in the right places.
But importantly, makes our experience
extremely seamless and easy.
I don't know if it's going that direction.
I just, I can see that there could definitely be a use case.
Yeah, I can see it, but even expand it up
and you actually just jogged a thought of mine,
which is related to, I can't remember,
it's called predictive markets.
There's actually a very technical term for it
where you effectively auction off
any type of service that needs to happen.
I've got to get from X location to Y location.
And that is just put out there with anyone
having the ability to bid on achieving,
on delivering you the package,
whatever it is from that location to the other, right?
And so that then opens up a sort of,
it expands way beyond just like an Uber
to a whole number of different type of agents, right?
And so that's kind of the type of things
you're talking about, right?
There are moving into the physical world, right?
You're abstracting even the physical world.
And yeah, I think it's similar to like intent-based
kind of architecture, right?
And then you could imagine because you could imagine
like what is the real problem?
Like even the real world,
let's say you brought up Uber, for example,
it's like the problem is not really,
I need Uber to go from A to B.
The problem is I need to go from A to B, right?
And then so whether it's Uber,
whether it's Lyft or any other app doesn't really,
I just need to go from A to B.
Of course, I have other constraints.
Like I don't want to go with certain transportation methods.
Like I don't want to go by boat or swim, obviously, right?
So there are certain rates you can constrain it.
But at the end of the day,
the point is that yeah, similarly for blockchain,
I think the more useful to us really at the end of the day.
I think my point is, and your point as well,
that AI is a big part to enable that type of application.
And that could be a very scalable application,
could onboard lots of users in all kinds of ways.
You'd be like, I want to make,
this is what's in my fridge, what I want to make,
There's tons of things along those lines as well.
But listen, we're coming up,
I've got about five minutes left,
and I'm actually going to have to run,
if not a couple of minutes early.
Evgeny, do you, anything you want to take away here?
Any questions you want to bring up Evgeny to anybody?
I think the only thought I had just been thinking about
the more comprehensive AI models we have,
the more widespread they are,
the better we need to think when we design tokenomics.
Because, you know, every flaw in,
you know, a token like a crypto-economic system
or a new project can be attacked,
can be used by AI agents even more efficiently
and can be discovered faster by the AI agents
because they just have infinite resources.
Basically, they're limited only by computational resources.
So they're going to try to, you know,
find all types of attacks to protocols
and utilize them to, you know,
maximize revenue of their creator.
Yeah, so this is just something to think about
when we build new web tree projects.
And Evgeny is even more right about that than he knows
because the computational elements that they're using,
to find the exploits, to find the exploits,
We're providing the compute that they may use
to exploit the systems, right?
To be a little flippant about it.
But it's something we can definitely see.
Any closing thoughts here?
Amplify, any closing thoughts?
I think it's, first of all, my sort of thinking is,
obviously, this is going to revolutionize the world
And yes, with every positive, you know, step forward,
it's going to be a negative step forward as well.
I just think ultimately, you know,
with the inclusion of blockchain
and blockchain systems running throughout the world globally,
which will eventually happen for airplane tickets,
concert tickets, you know, your car service history,
your house service history,
everything will be on the blockchain.
I think ultimately having a traceability across several industries
or across multiple industries or the entire economy
will help drive transparency
and that transparency will automatically reduce
the level of malicious actors that are out there
because it becomes easier to get caught.
And I think ultimately, I think that combined with AI
will have a much more positive effect on society
than the negative sides could have.
But I think without blockchain technology,
without that, you know, immutability, transparency,
I think ultimately you could be in a very dystopian situation,
but I think ultimately the good will win out in the long run.
Jesper, Erdem, anything else?
So actually, for the last years,
everybody's talking about regulating the blockchain.
I think in the next years, we can talk about regulating AI.
All right, should we close the space off now, then?
Thank you guys all for joining.
It's been terrific having everybody on.
Definitely learned a bunch about some of these financial applications
and others I wasn't aware of.
So thanks, everyone, for participating,
and hopefully you'll join a future one of ours
next week and the week after.
Thank you so much, everybody.