Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
The Hedera Technical Community Call is live on March 27th, 2025, at 12PM ET. Join us to discuss HIPs PR-1103 & PR-1068 and AI Tools with speakers from Hashgraph and SEALCOIN. #Hedera #HIPs #AI #Web3
Recorded: March 27, 2025 Duration: 1:01:50Space Recording
Short Summary
Exciting developments are underway in the Hedera ecosystem, with new project launches focusing on AI integration, post-quantum cryptography, and innovative fee structures. The upcoming AI hackathon and the introduction of HIP 991 highlight Hedera's commitment to enhancing security and user experience in the blockchain space.
Full Transcription
Thank you. Hi, everyone. Thanks for joining. This is an exciting call. So I'm glad to have you guys here.
Today, we're going to talk about a couple of hips, and then somebody from our team is going to talk about our AI hackathon.
Before we get started, I would like to bring up a particular thing that I would invite you guys to participate in, and that is our EVM survey and tooling survey.
There's going to be some links on the screen that you can
check out and uh participate in that survey on top of that we are having an ai hackathon
coming in the future and there's going to be some information on the screen as well
so you can get some information on that um and before i introduce the speakers i'd like to
put up this disclaimer about hips that are early in the stage
of the life cycle. And this is basically speaking to the fact that there's no confirmed estimated
time of arrival or delivery for the deployment of these hips on mainnet. And just keep that in mind.
So I'm going to keep that on the screen for a little bit, allow you guys to read it.
And that's great.
So now I'd like to introduce our speakers, our wonderful speakers.
First up, we have Jeremy from SealCoin, which is a.
It's a decentralized IoT platform and embedded software,
and they develop embedded software for IoT devices.
And he's the CTO there.
So he's going to provide some good insight on this HIP
that they've brought forward
on post-quantum cryptographic algorithm on Hedera.
And of course, we have our own Nadine,
who is a developer advocate here with us,
who will support that and answer some questions
during the Q&A of that.
And then after that, we have a sponsored VZIP,
which I'm particularly excited about.
It's a neat opportunity for developers
and application apps to make money using hedera by setting their
own or uh sorry sorry i'm thinking about it differently this is a way for um applications
to have a solution to where users won't have to care about using hbar to transact and and the
application can have an account to sponsor the fees for them.
Greg will speak more on that. And then finally, we'll wrap it up with our own Jake Hall, who is
a developer relations engineer. I'm sure a lot of you know him. He's been all over Twitter about AI
and AI agents, and he's done a lot of work in that space. He's very knowledgeable and he'll be joining
about 15 minutes or so into this call.
And anyway, without further ado,
I would like to just invite Jeremy,
would you like to come up and talk
about the quantum cryptographic algorithm, HIP?
algorithm hip sure thank you michael so uh okay can you share the screen thank you yeah hi everyone
Sure.
Thank you, Michael.
my name is jeremy pensier and i'm glad to present a new header improvement proposal
Proposal titled Support for Post-quandem Cryptographic Algorithm.
titled support for post-quandem cryptographic algorithm
In this session I'll walk you through the motivation behind this HIP,
the algorithm we are proposing to integrate, and how this benefits the Hedera ecosystem.
Okay, so go next, please. Thank you.
Why this heap matters?
So, as many of you know, quantum computing is evolving rapidly.
While we are not there yet, a powerful enough quantum computer could eventually break current cryptographic standards including ECGSA and EG25519, which are currently supported by Edera. The risk isn't just theoretical. A known attack strategy called Harvest Now, Decrypt Later involves bad actors collecting encrypted data today
with the expectation of decrypting it in the future using quantum capabilities.
While this primarily concerns encryption and decryption,
digital signatures are also at risk due to the well-known Shor's algorithm. At the same time, regulators
including NIST and the European Union are pushing for adoption of quantum-safe cryptography.
This HIP is about positioning HEDERA ahead of that curve by integrating a post-quantum algorithm.
ahead of that curve by integrating a post-quantum algorithm.
So, yeah, we evaluated several post-quantum cryptographic candidates
and selected MLGSA, formerly known as Dilution 5.
It's based on a lattice problem called module learning with errors, which has strong theoretical security against quantum analytics.
It has been officially selected by NIST for standardization.
NIST stands for National Institute of Standards and Technology.
It's a US government agency that develops and promotes measurement standards.
In other words, they identify the most robust algorithms and advocate for their adoption.
And we also looked at the practicality of this algorithm. Signature size is about 4.6 kilobytes which is manageable.
It also has solid Java support notably through the Ponzi Castle
cryptography library which make integration much easier.
which make integration much easier.
Among the available post-quantum algorithms, we selected MLGSA 87 for its balance of
practicality and security, but we also consider Falcon, Sphinx Plus and SQI-Sign.
considered Falcon, Sphinx Plus and SQI Sign.
MLGSA is quantum safe, but signature size is larger than classical signatures.
Falcon is great in terms of signatures, but it requires floating point arithmetic
arithmetic that is vulnerable to side tunnel attacks.
that is vulnerable to side tunnel attacks.
And it requires a bit more computational power
compared to MLGSA.
Sphinx Plus is extremely secure,
but with signature size reaching 17 kilobytes,
it's too heavy for most DLT use cases.
SQI Sign offers very compact signatures,
but it's still early stage and competitionally demanding.
So MLDSAs track the best balance,
secured, standardised and practical to implement.
So what exactly will MLDSC enable for Hedera? Here are the
core capabilities we are proposing to implement. Users will be able to generate
new content-safe key pairs, sign transactions, and Hedera nodes will be able to verify those signatures.
We also extend the Hedera SDKs and APIs to support this, so developers can adopt it without changing the way they interact with the network.
One of the key challenges is the size of the signatures.
One of the key challenges is the size of the signatures.
4.6 kilobytes is significantly larger than eCDSA.
To address this, we propose an adaptive fee model.
Classical transactions will continue under the current fee model,
while transactions using MLGSA
would have higher fees.
Even though the current fee model already accounts for signatures or transaction size,
we would still need to consider potential increases in computational costs.
Okay, so with this heap, we're not replacing existing algorithms.
We're adding MLGSA as an additional option.
Talking about security, we're improving it against future attacks while staying secure against present threats. There are no breaking changes. Developers
and users can continue using ECGSA or ED25519 if they prefer, but for those who want quantum is there. So implementation plan. We have outlined a four-phase implementation plan.
First, we'll enable key generation, signing and signature verification. Then we'll add MLGSA
support to the Hedera SDKs and API.
Then we'll implement the FeeModel.
And finally, we'll monitor performance and optimize based on FeeMpack.
Okay, but adoption won't happen unless we empower developers.
That's why we'd need to deliver updated documentation
and API references,
step-by-step tutorials for integration,
webinars to explain the importance
of post-quantum cryptography,
and workshops to work developers
to walk developers through the integration process.
through the integration process.
And for the next steps, to the best of my knowledge,
this heap requires council approval.
From there, we'll gather community feedback to refine the proposal
before implementation and testing.
And then we would think about an adoption strategy
to promote post-quantum transactions.
And that's it.
Thank you, Jeremy.
Yeah, I appreciate that.
It's very insightful.
If you could, just for our audience,
could you talk about a little bit about SealCoin
and speak a little bit to this HIP
and how it will enable SealCoin
in your embedded IoT devices?
Yeah, so SealCoin is creating an ecosystem
for the for the iot devices and to do that we we are using a token
built on top of hedera because of the of the quantum threats that we are aware of because we are closely working with cybersecurity companies.
We are really attached to improve the security
of the devices.
And yeah, by improving Hedera,
we are also improving the token
that we are using for Silcoin.
Awesome thank you yeah I appreciate that answer. I'd like to invite Nadine up if she'd like to Hi, Michael. Yes, please. We're good. First, amazing. Let me say how much we appreciate all the effort and all those diverse perspectives on this topic, because post-quantum cryptography is super important for the better. And even though hopefully or probably we won't face a large scale quantum threat
issue for several years, it's fantastic to see everyone in the community already thinking
and planning ahead. So absolutely love to sit and please everyone share your feedback
on it and tell us what possible standards you would like to see so we can open up this conversation
and then in the end implement what is wanted.
Now, I just wanted to talk a bit
how we discussed about this in Hashgraph.
So we've been thinking of other possible standards
and while the lithium is certainly super secure,
it has, as Jeremy has mentioned,
quite a large signature, so around 4.6 kilobytes
um so we were thinking also of something different which is falcon um which provides pretty much the
same quantum security but with much smaller signatures um so only 1.3 kilobytes uh this
could help to preserve hedera's advantages of having this high throughput and also
the low fees you're actually known for the issue here is that currently falcon has no standard
which is fully accepted by nists but we do plan that this will come out pretty soon probably
within the next half year and we can already develop and test bits and pieces and then
maybe switch over when we have the final standard there was um there was something mentioned about
side channel vulnerabilities on the slides and i think it's also very important to stress there that
this is a very valid point but everyone who's working in the industry and knows cryptography is aware of this
and it can be mitigated with careful coding so there's a multitude of potential cryptography
standards that we can look into and thus please again if you have one that you think is very much
needed for your project comment on on the hip, reach out,
let us know your needs and questions.
And yeah, that's it from my side.
Thank you so much, guys.
Thank you, Nadine.
Thank you so much.
We actually have a couple of questions here from the audience.
So that's great to see.
It's kind of a two-part question.
The first I can answer,
and then maybe you guys can speak a little bit to it at color. And then the second one I'll give to you guys. The first is, if we could put it on the
screen here, is there a timetable for quantum capabilities to be achieved? And now, of course,
at the beginning of this call, we put up a disclaimer saying that we for these hips that are early in their life cycle
we don't have estimated time for this uh feature to really uh reach mainnet um but that being said
we encourage people to engage in the conversation um where we have the uh github link up there for the pull request please go there
and and start engaging um we like to see that um and if if anyone would like to add color to that
answer please feel free and if not uh there is a a second part to this question um that maybe maybe we put it up on the screen is
does Hedera have an advantage over blockchains to combat the quantum threat?
If so, why?
Sure. I mean, to the second question, because we're S&T and Pheasantian fault tolerant.
So that's definitely a massive plus there.
And then when it comes with the implementation,
as said, we're currently exploring different possibilities.
The HIP that Seedcoin has mentioned is definitely a great start.
And engineering is talking with Lehman.
So we're on the ball.
Now, there's really no rush because it's a couple of years
until this really would be an issue. So instead of implementing something right now, it's actually good that we take
the time and see what the fully-fledged standards are going to be, which are hopefully coming
out in a couple of months, and then take it from there.
Perfect. Jeremy, anything you'd like to add? I would say the same to answer to the first part of the question.
As Nadine said, there is no threat before a couple of years.
Even if with the involvement of AI, maybe it could happen sooner.
I think we have some time
before being threatened.
Cool. Perfect.
All right.
I guess that brings us to the conclusion
of the Quantum Hip.
I would like to next up talk
about sponsored fees. This is a really very interesting hip and our one and only Greg Scullard
will be speaking to this. Greg, for those of you who don't know, he's been with us since the launch
of Hedera. He's been an incredible advocate for us and his Twitter handle
is HederaKid. So I'll let you go ahead and take it away. I don't need to further butcher this
introduction, Greg. Go ahead and take it away. Thanks. Thank you, Michael. And hello, everyone, and thank you for taking the time to participate in this discussion.
So, yes, sponsored fees is a HIP that is in progress.
If we can move to the next slide, please.
So, oh, OK.
So I guess in terms of why this HIP matters, one of the things that came to my mind was
if you were ever as a child given pocket money, suddenly my parents would usually say don't buy
sweets and guess what I would go and do? I would go and buy some sweets. So this is to try and kind of put this into context on to the next slide.
So approvals that exist today on Hedera and, you know, most of the networks that have an EVM where you can do approvals through smart contracts.
Alice can approve HBAR to Bob so that Bob can spend it with an exchange, send it to a friend,
send it to a smart contract.
But Bob cannot use the approval to pay for transaction fees or token custom fees.
And one of the challenges for Alice here is that if she approves Bob to receive 100 HBAR
or to use 100 of her HBAR to spend on smart contracts,
for example, she has no way to control or to ensure
that that's what Bob will do with those HBARs.
Bob may send them to an exchange and cash them out and go and buy some sweets.
So with this HIP, we're giving Alice a little bit more control. Next slide.
And more also importantly with the current approval solutions, you cannot use them to pay for transaction fees.
You can only use them to send in crypto transfers or to send to a contract.
So other solutions for Alice to be able to, or for Bob to be able to spend some of Alice's HBAR
would be for Bob to create a transaction
and designate Alice as the payer for the transaction fees,
and then send the transaction to Alice
for her to countersign and essentially approve
potentially approve that Bob spends HBAR on her behalf.
that Bob spends HBAR on her behalf.
Or they can use schedule transactions
where either Alice or Bob can send a transaction
to the network and then the other party has to
countersign this schedule transaction so it's executed.
But that means that Alice has to always check what Bob
is doing with those hash bars that she's allowing him to spend.
And he could sneak in a crafty transaction
to buy some sweets, which if Alice doesn't catch it,
then, yeah, obviously, he'd have a bag full of sweets.
So on the next slide, what if
there was another way? What if we could increase the scope of approvals to cater, for example,
for instances where you would want to grant someone some HBAR, but have some control over
how those HBARs can be spent? And this is very much the crux of this particular HIP.
So on to the next slide.
And they're not super clear.
We can see that Alice can grant allowances on the left hand side to Bob
for HBAR tokens and NFTs.
With this particular HIP, we're adding scopes to the HBAR allowance.
It's not necessarily changing the token allowance or NFT allowances so that when Alice grants
an allowance to Bob, she can say here's a 100 HBAR that you can spend on crypto transfers but nothing else. Or here's a
50 HBAR you can spend on gas fees and transaction fees and or token custom
fees and which can be either tokens or HBAR. As you know custom fees on tokens
can include payments in tokens like USDC, for example.
So the idea with this HIP really is to increase the scope of allowances
so that Alice has far more granular control over how those allowances that she's giving away are being given.
Allowances can be unlimited or limited in quantity.
So I mentioned allowing 100 HBAR,
but Alice could just give unlimited access to her wallet to Bob
for transaction fees, for example.
And of course, she's able to revoke those at any time
if Bob is spending too much.
We'll also be looking at enabling the management of allowances via smart contracts
so that a contract can allow or approve spending for another account.
And there will be changes, of course, to mirror node queries so that the query response indicates the scope for approvals.
And there will be changes, of course, to Meronode queries
On to the next slide.
So from a security and backwards compatibility point of view, as I said, Alice can revoke allowances at any time.
So even though we're adding those scopes, Alice still retains full control over the allowances.
And any allowances that have existed prior to the HIP being deployed will default to what they are
today, which is crypto transfer.
So their purpose will be unaffected by the change.
If a scope isn't specified, it will default to crypto transfer,
which is the current behavior.
So no major backwards compatibility concerns.
So in conclusion, the HIP opens up a bunch of additional use cases with those
additional controls. For example, and I think that's what Michael was alluding to in the
introduction, a gaming company can approve HBAR to their players who no longer have to
acquire HBAR to play the game. And this is a, pun intended game changer for onboarding new users who are not
mostly familiar or wanting to interact with an exchange to acquire cryptocurrency. But the gaming
company in this case retains full control over their own HBAR and they can entirely control or
choose how to make those available to their users. It could be even given or done on a temporary basis.
You know, here's a, you know, a coupon in a way for playing the game for, you know,
up to 100 or 1000 HB.
And when you're done, then you have to go and purchase your app.
Foundations and investors are also very interested in this.
They can grant projects and be sure that the granted amounts are used for the appropriate purpose.
You know, if I grant some hedge bar to a project and there is a tacit understanding or contractual understanding that this is for transaction fees,
but the project does a rug pull on me as a foundation and sends them to an exchange
to buy some sweets, then, you know, I'm in a lot of pain to try and recover those funds legally or
otherwise. So, you know, in an event that a project doesn't deliver, the foundations or investors are able to keep control over their HBAR.
We also have potential use cases where an intermediary can hold HBAR on behalf of an
enterprise who doesn't want to hold cryptocurrency on their books and the intermediaries can charge
the enterprise a monthly fee for the service in fiat. And again, this is a good step forward for
adoption. So the HIP 1068 is in draft. We welcome your comments before it is submitted to HIRO
and the council for approval and implementation. However, I should say that there is a competing HIP
which looks to implement something that we call,
or the HIP refers to as hooks or Lambda functions
that you could attach to your account.
And then those functions could be set up
could be set up to essentially manage approvals.
to essentially manage approvals.
So it's a far more, I would say flexible,
far more programmable way of implementing
something similar to this delegation
or approval of funds.
So we're considering whether this new HIP is essentially a suitable replacement for 1068 that I just discussed today.
So inviting you to comment, contribute as per the other HIP and to look at this other one on the hooks as well.
Yeah, we want you to tell us what you think.
Thank you very much.
Thanks, Craig.
Just one thing that I wanted to ask.
I didn't see this in the comments, and maybe it's probably not possible, but we get a lot of questions about
people who are developing dApps where their users don't want to care, not just about not
having HBAR, but not even having a wallet. Is there a way that dApps could integrate some
sort of mechanism where they wouldn't, the users wouldn't even have to care about a wallet. They
could just click an approve button and then, you know, they can complete
the transaction, whatever they wanted to do with whatever application it is.
Yeah, I mean, you know, today your DApp could be completely walletless, generate transactions,
generate transactions, send them to a backend,
send them to a backend, which would sign them and execute them.
which would sign them and execute them.
But, you know, I think the notion of a DApp or Web3
is that you are custodians of your assets, right?
If it's NFTs in a game or something like that,
they belong to you.
Otherwise, you're more into something that's more akin to Web2
where you never own anything because, you know,
it's all being done by accounts that are controlled by someone else.
So I think this chip is very much more in the sense that
a wallet could be created for you, you know,
but you don't have to,
and you would have to approve transactions potentially.
You could see things appear in your wallet in terms of NFTs and so on and so forth.
And this could be a great way for people to get to learn and to understand
what decentralization and self-sovereignty of assets means in the context of Web3,
but without having to learn and understand
how to purchase HBAR.
And I think that's a big hurdle for most people.
Most people I speak to in my social life
who know nothing about cryptocurrencies,
they've not necessarily bought shares on the stock exchange.
And that's essentially what we're asking them to do.
Register with the central exchange,
transfer some funds, get some cryptocurrency.
And this is like, whoa, hold on.
I can't be bothered.
You know, sign me up to Xbox Live.
So yeah, potentially this HIP or the other one on Hooks
could certainly help to smooth the path to adoption.
Yeah, mass adoption.
Awesome. Yeah, great.
Thank you so much for that presentation.
That is a very exciting HIP.
And just like Greg said, please,
here's the link on the screen.
If you could go to that and contribute to the conversation.
We love to see that.
It helps us to gauge the engagement and also is we consider it in our roadmap. So please contribute to those discussions.
Okay. Next up, we have Jake Hall, who many of you know, he's in the AI space. He's developed a lot of tools around AI.
I don't need to give much of an introduction.
You guys know him.
He's up and coming.
Let's hear it for Jake.
No, you're on mute, Jake.
Just come off mute.
Test, test, test, mic test. Yeah, there you are sorry guys yeah no i just had to i saw my hadera hat
so i thought i had to put it on i should have had it on uh before but uh yeah no thanks for
having me on i've got a lot to talk about thanks for joining
should we get straight into it should we yeah let's it? Yeah, let's do it. Well, in that case, I'm gonna share my screen then.
I was trying this earlier, the setup's a bit dodgy.
Let me give it a go now.
So I'll go first share my screen and there we go.
And now I'm hoping if I click play,
I can't now see the screen, but can you see this presentation?
Yeah.
And I'm going to be talking about.
Basically, everything AI on Hedera now, not quite everything.
We're still missing out a really big chunk of Hedera's involvement in AI, and that's going to be the stuff which people
like Prove AI and Equity Labs are doing on the data integrity side. I'm focusing more so on what
you can do as a normal developer on Hedera. How can you use AI to develop on Hedera? I'm going
to showcase some of the tools that we have.
But also if you're interested in AI agents,
what you can do with AI agents on Hedera.
And there's a lot to speak about here.
I think I've only got about 20 minutes in this slot
and I spent one minute of it not talking
because my mic wasn't plugged in.
So I'm gonna try and get through into it.
But definitely at the end, if you've got questions,
please ask and I'll also be monitoring the comments.
Now we'll break it down into two parts, using AI to develop on Hedera and we're
going to show some of the tools that you can use in order to do that as well as AI agents what the
different AI agent frameworks we have what are the toolkits what are the different protocols and what
are the different uh the recent improvements that we've made to the actual Hedera consensus layer
which will which will be really suitable for AI agents.
And then finally, I'm going to give a quick plug to our AI agent hackathon,
which is starting in a couple of weeks.
So definitely, hopefully these can sort of give you a few ideas
as to maybe what you can do if you enter it or just the awareness that is happening.
So let's start off with using AI to develop on Hedera.
Now we've got these custom GPTs.
Obviously you can use generative AI as you wish,
as you probably are doing so if you're a developer.
If you're not, I'd recommend you do it
because it's gonna speed up definitely
some of the more simple tasks.
But actually if you've got ChatGPT,
you can go over to the custom GPTs
and you can search for Hedera Hivemind.
And I quickly prompted Hedera Hivemind just before this,
I asked it what knowledge does it know?
It knows pretty much everything that's in our docs
and it's got a lot of the most important repos.
So for example, general Hedera knowledge,
it's got all of the documentation in its backend
and it's a vector database, smart contracts as well.
It's got loads of smart contracts
from the smart contracts repo.
It knows about all of the latest hits.
Local node setup.
If you're looking to start running your own node
when it comes to testing,
often you'll find it's much quicker.
I definitely recommend it.
You can use Hedera Hivemind
to help you start on that journey.
Solo as well.
If you don't know what Solo is, Solo is
kind of similar to Hedera Local Node, but it's like a more robust version where you literally
have your own Hedera Node from the Hiro source code. You have your own Hiro Node running on your
local machine. You can also use Hedera Hive Mine to do that. And then of course, SDK usage,
the stuff which is the bread and butter.
So if you are looking for something
which is domain specific,
you're more likely to get less hallucinations
if you use Hedera Hivemind.
Definitely recommend that.
If you're not a chat GPT guy,
you can also use Ask an AI on our Hedera Discord.
So I think maybe we can plug the Discord up here now.
It's hedera.com slash discord.
That will take you to it if you're not in it already.
But essentially, there's a channel on here
which is linked up with Kappa.
Kappa's another AI vendor which we use.
And similarly to the Hedera Hivemind on ChatGPT,
it's got all of the access to the latest documentation.
You can query it on here and you can ask on it.
So that's just a bit of background
as to some of the tools that you can be using as a developer.
Obviously keep it in with your existing
generative AI, chat GPT, LLM workflows,
but just letting you guys know that there's something
which is very specific to Hedera.
If you want to do something which is domain specific
and there's a risk of hallucination,
these are probably the tools
which you want to be heading over to.
As well as of course the team,
I'm always here to help as are most people on these calls.
I know Michael will be happy to assist too.
Cool.
Let's now go over to AI agent stuff.
Now, the first thing which I want to just give a quick overview about
is the Hedera Agent Kit.
Now, the way that you can think about this Hedera Agent Kit,
it's an open source repo.
You can contribute to it.
It's an open source repo,
which essentially acts as the mastermind,
as the go-to library if you want to integrate Hedera
into an existing AI agent framework
or an existing custom-made LLM flow.
The Hedera Agent kit is really the kit
which you wanna be using.
And I'm gonna go into that in a bit more detail
in the following slides.
But just while we're talking about the Hedera agent kit,
what can you do?
And I wanna make these slides as actionable as possible
just so that it's not just listening and consuming stuff
so that we can actually know what exactly we can do
as developers, as community members,
to help contribute to Hedera
and help build out the use cases on our network. So the first thing exactly we can do as developers, as community members to help contribute to Hedera and help build out the use cases on our network.
So the first thing which you can do is you can have a look
at the open source repo.
I've shared this link so hopefully we can get it up
or we can communicate it after this.
But this you can find on our GitHub, totally open source,
have a look at it, clone it, but also raise some pull requests.
And this is something which we're really looking to do.
Currently, it covers the basic Hedera functionality,
things like creating tokens, transferring tokens,
checking balances, a few mirror node queries.
But actually, to extend it out further,
we can do things like adding in decentralized exchange swaps
or adding in the ability to stake and unstake
and other sort of protocol integrations.
If you're listening from the community here, you've got your own protocol, please reach out and we can talk about how can we integrate your specific functionality into our agent kit so that it helps and what I'm about to discuss, which is first and foremost, our Langchain or Langgraph tools.
and foremost, our LangChain or LangGraph tools.
And what this is, is it's a tool which is compatible
with the LangChain LangGraph AI agent framework,
which allows you to build AI agents using LangChain.
This again, all of it's open source.
You can play this out yourself.
Similarly to before, I've sent the repo
so we can communicate this out.
But what I recommend here is clone the repo, test it out.
There's a script which you can just run with npm test
that pops up with a CLI chatbot that you can use
to start creating tokens just via a CLI text input.
So install it, implement it into your dApps.
The benefit with Langchain versus what I'm going to describe next
is that Langchain is very extensible and it's very customizable.
You can really create these custom AI agent flows
or implement it into your existing Web 2, Web 3 application
with very little developer friction.
But also you can integrate it with other agentic tools.
So it's not just Hedera which has compatible tools
of Langchain, there are many other tools.
You can do tools like searching the web, generating photos,
all sorts of stuff, you can create your own custom tools
as we've done with Hedera.
The other thing which I wanna show is another
very popular AI agent framework, which has a different set of benefits
and value propositions when compared to
LAN chain and LAN graph.
So that's of course the ELISA plugin.
Now we have a Hedera ELISA plugin,
meaning you can use the ELISA agent operating system
to actually interact with Hedera.
And where ELISA is really strong
is when it comes to these sort of personalized
autonomous agents it's not quite so easy to integrate eliza let's say into your existing
web application with some specific uh functionality but what eliza is really great at is you can create
your own persona your own character in the character file, and you can hook it up to X, you can hook it up to
Discord, to Telegram, whatever you wish, and you can have this really, you can always have a person,
it basically creates a person who has these actions to tools, they can swap on behalf of
other people, they can create tokens for people, that's where Eliza becomes really good, we have
the Hedera Eliza plugin, again I've shared out the link. Hopefully we can get this communicated ASAP.
But what can you do with ELISA?
What can you as developers action
from this particular session?
Well, the first thing you can do
is you can spin it up on our Gitpod environment.
Because ELISA is an operating system for AI agents,
it's not as easy to set up as say the Langrath agent.
So we've got a Gitpod ready for you to start testing with
and you can spin out that Gitpod environment totally free.
You just need to create an account
and you can use that to start interacting with Eliza
and interacting with the Hedera Eliza plugin.
The other thing you can do,
you can create your own personalized character file.
I mentioned this earlier,
this is how you give an agent particular characteristics.
And interestingly, what the ELISA team have done here
is quite innovative, is they use this concept of entropy
with LLMs and sort of, yeah,
the way that large language models work.
And so that the AI agents don't just become
this really monotonous, one-sided being or entity.
It will randomly pick particular aspects
from your character file
so that in one message it uses that particular trait,
in another message it uses that particular trait.
It can interact with people on X,
it can reply to people, it can like stuff,
it can share stuff.
But every time it's slightly random based on what
you've set. So every time you get a different flavor and it's more human, I suppose. But yeah,
you can also connect it with X, Telegram, Discord. There's other bits involved where you can connect
it to. Right. That's it for the AI agent frameworks, but we're not done with AI agents.
And the next thing which I want to share is a HIP which has been in the works and that is HIP 991.
What is this HIP?
This HIP allows for a consensus topic,
the consensus service if you don't know,
is it's like a pub sub message bus,
totally decentralized on Hedera where you can create a topic and you can
submit messages to it now currently those messages cost about one thousandth of a cent some something
to that effect maybe one hundredth or one uh it might be one hundred percent one hundred got it
okay so what yeah one hundred percent um which is really cheap and for that reason hedera is the network which has the
most amount of transactions of any single network for that reason we had sort of 7000 8000 9000 tps
running uh in the past but this new hip actually allows you to create a custom fee on a consensus
topic similar to what we already have with nfts and with fungible tokens you can now have a custom fee on a consensus topic, similar to what we already have with NFTs
and with fungible tokens.
You can now have a custom fee on a consensus topic,
meaning yes, you pay that 0.0000 whatever one cent fee,
but also you have to pay this custom fee of $5, $10,
one meme coin particular token.
And only once you've paid that is the message sent.
Where this is interesting for agents
is you can have an AI agent
which consumes this particular topic.
It's subscribing to it.
And it will know if I've got this message,
someone's paid 10 bucks.
Someone's paid 500 J. Cool coin, which doesn't
exist. So you can actually use this to create a verifiable proof of payment, which then
initiates an AI agent to use their compute and react with it. What can you do with this?
First thing you can do is you can check out the new hip you can get involved with hyro hyro obviously that's the that's the open source uh governing
council essentially of hedera of the source code uh working with the linux decentralized foundation
trust um but you can look at the hip you can read it you can play around with the uh you can play
around with the latest sdk this is currently onnet. Not sure if it's on Mainnet quite yet.
And I know there are some potential delays with it,
but currently you can play around with it
on Testnet for sure.
I'm gonna be releasing a tutorial out on this,
so definitely check that out.
And then also have a think about
what you can build with it,
because again, we have this Hedera AI Agent Hackathon
coming up over the corner.
So now's a good
time to start bouncing ideas ideating stuff that you can do the other thing which i want to share
is a protocol called hcs10 hcs10 is also built on top of the consensus service and it's a protocol
which has been developed by a really amazing contributor to our ecosystem,
Hashgraph Online.
Hashgraph Online, if you don't know, they are a DAO,
which is governed by some of the leading community,
some of the leading protocols and wallets
and ecosystems in Hedera,
the likes of Hashpack, Kabila, Centex, NFT.
And they've established a really robust set of protocols
which live on top of the consensus service
that allow this unbelievable usage of it
to extract value from the consensus service
that is not typical from your typical PubSub service.
And one of these protocols and standards
which they've created is HCS10.
This is an AI agent to AI agent communication platform.
It allows you, it allows AI agents to message each other
in a totally decentralized way
without needing any sort of third party API.
They use the Hedera consensus service
to create a particular topic
which only those two agents can communicate on.
And they can send messages back and forth,
they can trade, they can plug it in
to the previous HIP 991 to even start paying each other
in a way where it's verifiably proven
and the AI agent action is only taken
after that message has been received.
So this is really exciting.
It's gonna be a really big focus
to the AI agent hackathon, which we're running.
We've got a stream specific for this.
So I think if you head over to hashgraphonline.com
slash hackathon, you'll be able to sign up.
Definitely recommend you do it.
So do that, but also have a look at the HCS 10 standard.
Have a look at the documentation, hashgraphonline.com.
Have a look at the standards SDK.
This provides a really easy interface
to start interacting with this particular standard.
And again, brainstorm ideas for our AI agent hackathon.
So we're calling all AI agents, all of their humans,
build your product.
If it's on a different chain,
come over to Hedera, give it a try.
If you're on Hedera and you're interested in AI agents,
this is the perfect opportunity to have a play
and hopefully win some cash prizes
as $30,000 on the line.
But I just want to have a quick question
to sort of leave everyone on,
which is how can we use HIP991 and HCS10
to create genuinely revolutionary web three
and AI use cases on Hedera?
These two protocols or sort of native functionalities
on the network are truly unique.
There's not another network that I'm familiar with,
layer one chain, who has these functionalities
out of the box, which leverage the native service.
I wanna leave the question here again.
I'm gonna repeat it one more time.
How can we use HIT991, the ability to pay a fee
and provably demonstrate that value has been transferred when receiving a message
for an AI agent, coupled with HCS 10, the ability for AI agents to communicate with each other.
How can we use those two features to create some disruptive use cases that haven't been seen
before? And of course, the Hedera hackathon starting on the 11th of April, something around
then,
that would be the best place to put it to the test.
So really recommend you sign up.
Again, I'm always here to bounce ideas off,
really keen to hear some of the insights from the community members,
and as well as all of the team at Hashgraph
and all of the team at Hashgraph Online,
everyone's really willing to help.
I know there's actually a call going on,
I think it might be now, where they're doing it but um there's there's gonna be lots more announcements
coming out so so stay tuned and sign up already if you want michael that that's it from me uh
i'm not sure but uh that's right on time yeah thank you so much perfect presentation absolutely Absolutely brilliant. Yeah. So now we'll have 10 minutes of QA for all of our presenters. And before we talk about this QA, I wanted to talk about that we created two new Discord channels.
Join us for another desk, HIP's discussion desk. So they're meant to centralize feedback.
You can give us bug reports, offer HIP discussions and ideas to improve engagement.
We're always keen on listening to that and giving, just keen on listening to the feedback
and hearing from you guys. We love that. So in next call, Simi will share more
details. And that call is slated for April 24th. So now enough of me. Let's hear from
our presenters and any QA that might be coming up from a result of this presentation. I know
there was one question that we had.
Let's see if we can put it on the screen here.
It was related to the quantum HIP, first off.
Yes.
Yeah.
Is there any initial information to share
about quantum resistant chips from CLSQ being used
in the network infrastructure?
Anything you guys can give Jeremy?
Yeah, okay. infrastructure? Anything you guys can give Jeremy? Okay, so yeah, CLSQ plan to release the QS7001 by the end of the year, and this chip is already
integrated MLGSA 87. So that would be the best choice for iot devices that want to uh transact using a token on hedera
if hedera integrates also the mlgsa 87 so yeah
so yeah
what else we have
can we put up the next question
Greg you want to take this one?
Yeah, I'll have you two.
I don't know if it's true.
Maybe, maybe not.
Yeah.
If it's a rumor, it's not one I'm aware of, is all I can say.
What is Hedera utilizing to achieve maximum computational power for scaling AI?
Jake, you want to talk to that one?
So what is Hedera utilizing to achieve maximum computational power for scaling ai so there's hedera the native hedera service we're not an ai company where hedera is a uh is is it
writes code for it provides a infrastructure for distributed ledger technology so we're not
necessarily going into the realms of,
okay, let's use GPUs and invest loads of money in GPUs and building these data centers and
gigafactories and whatever. Where we do sit into that equation is AI, people who are developing
AI models or people who want to leverage artificial intelligence within the context of Web3,
they can, of course course use our distributed ledger
to provide transparency, to provide immutability,
to provide the ability for agents to transfer value.
So I wouldn't think of it as that being necessarily
Hedera's job, that's probably something which
Microsoft, OpenAI, Google are thinking more about.
We're focusing more on, okay, how can we make the distributed ledger
as cheap, quick, sustainable,
and easy for developers to build on as possible?
And that will, of course,
enable use cases within the AI sector,
but it will also enable use cases within DeFi,
within real-world asset tokenization,
something which I know many people
on this panel can speak to on in
very in a lot of detail but in terms of yeah what is hedera utilizing to achieve maximum
we're not probably utilizing too much we're focusing more on making sure that our nodes
can run the hedera or highway consensus algorithm and making sure that we're delivering on our
road map make sure that we can get other people starting to build that's what we're focusing on
roadmap, make sure that we can get other people starting to build. That's what we're focusing on.
Uh, we all, thanks Drake. Great answer. Um, yeah, we're not an AI company, but, uh, we do stuff in
AI. There's a, another question that's come up. That's, uh, interesting. The USA administration
has promised a $500 billion investment to AI. Is Hedera partnering in the USA investment, in this USA investment?
I have no idea. I don't have any information regarding that. I do know there has been a $500
billion investment into AI, but I don't know if Hedera is partnering with the USA in this $500
billion investment. I guess to Jake's point, we're not an AI company, right?
So maybe some of that investment will trickle down into use cases on Hedera,
and that's certainly something that we're trying to promote
and through the work that Jake is doing with the community,
trying something that we want AI-driven use cases on Hedera for sure.
But we're not an AI company first.
So my guess, you know, without knowing for sure,
is that that investment would go into AI companies first.
Perfect. Cool.
And this next question I can answer on the digital euro question yeah um
even even if um there is uh or there was uh such a project going it wouldn't be for us to announce
before it is announced by whoever is issuing the token that's a mantra that we live by at Hedera
we don't announce projects on behalf of projects we let them go public when they're ready so even
if I knew about such a project I wouldn't be able to say yeah great answer thank you thank you Greg
okay this next question by Lewis these are new hips that are being developed by Adara.
Well, they're actually being developed by quite a few members, but will they be integrated with Jairo?
So if I could just speak on this, I'd like to say that several months ago, it's earlier in 2024, there was a move that
happened with Adara that we donated our source code to the Linux Foundation. And they made an
org in GitHub called Hiro Ledger. And we donated all our source code. And over the past, since
that happened, we've been migrating our repositories over there.
And just yesterday, the Hips repository also migrated.
So every single HIP, first of all, they're no longer called Hedera improvement proposals.
They're Hyro improvement proposals.
So these HIPs are now necessarily going through the hyro ledger org
and they'll be in that code base hedera will choose to whether or not they want to adopt
the hip on mainnet um if it's a hip that touches the source code um so yes, they will be integrated with Hiro because they are now in the Hiro
code base. Michael, is it fair to say that, I mean, the question would be phrased the other
way around. It would be, are the hips on Hiro, which are being developed by Hashgraph predominantly,
but also other open source contributors, are the hips on Hyro being integrated to Hedera, which is the public
instance of Hyro? Yes. Yeah. Consider that Hedera is like the first iteration of a Hyro ledger
that's a now public ledger. And it's always been a public ledger, but Hedera Council chooses when those HIPs go through and are put into the Hyrule Ledger code base.
They can choose whether or not we will put that feature on Hedera mainnet.
So it's always going to rest in the Hyrule Ledger code base.
And then Hedera will have discretion over whether or not they want to enable that feature if that hopefully that brought some clarity that question yeah maybe just one uh one
comment michael you said hedera uh approves uh hips in hyro i think um which is slightly incorrect the
hyro has a technical steering committee, which includes community members
and people from the Hedera team and Hashgraph team.
They are the ones who approve changes to the HIDRO code base.
And then HIDRO optionally can choose
not to include a particular change
into the code that runs on mainnet.
Right.
And to be totally technical,
the Hiro TSC accepts the HIPs.
They can accept or reject the HIP after last call.
And then Hedera can choose to approve it
or not approve it on Hedera mainnet.
Correct.
All right, great.
It looks like we're at the top of the hour.
We're one minute over.
This was a wonderful call.
I love it.
Thank you all for joining.
Thank you for your time.
Thank you guys for watching.
I hope you guys are excited for these hips.
And I'll see you guys with Simi on next,
the next community call that we have.
So thank you so much. Thank you. Yeah, it goes like this And it goes, yeah, it goes like this
Yeah, around and around it goes
Where it's dark and nobody knows you
