Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
Full (풀) Grass with Korea Homies
Recorded: Oct. 10, 2024 Duration: 0:56:06Space Recording
Full Transcription
I opened the space before 5 minutes for start AMA because this is my first space so I'm
very nervous.
So I want to confirm a specific punch function for Trio and waiting for chorus coming in advance.
Many global followers in there and I hope many Koreans are listening for my space because
I'm not fluent in English but I'll try my best.
For coming tonight event, there is some reward for participating in this space.
So if you want to join a Gigabod referral, please like and retweet for this space or Twitter
article and fill the Google form.
You can see the Google form link in my profile.
So if you want to join Gigabod NFT referral, please fill the form.
okay are you coming there
see your account okay
see your account
okay
okay
okay
okay
okay
okay
okay
okay
okay
okay
okay
okay
i'll buy some
CEO account
even like
responding
okay
i send the guest for speaking to founder of grass
Let me give a time to say hello to my Korean subscribers.
Because it is a space focusing on Korean community.
English is a little hard to say.
So, we need to be able to speak about it.
But if it's really hard,
you can be able to participate in the 졸업준비위원회
to help you.
So, if it's not possible,
you can ask for the question.
If you have any questions about quality,
you can be able to help you.
Okay, thank you for everyone.
Please wait a second.
Let me
under a share extract.
Okay, let's try.
Okay.
Greg, could you turn on your mic and speak to confirm for start AMA?
Annyeonghaseyo.
Okay.
Okay.
Okay.
Okay.
Let's start AMA from now.
Hi, everyone.
And thanks for coming.
Actually, it is my fourth space in X.
Today's space is also for the Korean community.
So, Korean translation will be provided for the Q&A.
Honestly, I'm not very fluent in English.
So, I'm very nervous now today and hope to make a successful AMA.
But, I'll do my best to make tonight successful.
Today's guest is GRASS.
GRASS is a project great to redefine the infrastructure of the internet.
By simple Chrome extension, they've attracted a lot of people to join GRASS community.
And GRASS sprouted a lot of people together as a community through on Gigabuzz NFT.
As you know, GRASS is also making excitement for TGE by releasing an airdrop checker ahead of GRASS token runs.
One of the most common question was about the GRASS token, which we are going to talk about today.
Before the AMA start, I'd like to introduce and greet the GRASS team who are here today.
Could you tell about yourself, GRASS?
Yeah, more than happy to.
So, I went to school for nuclear engineering about 10 years ago.
After that, I worked in finance for a few years, mainly because at that time, a lot of the skills carried over.
And I thought it was a cool thing to do.
But I quickly realized that was a big mistake and I wanted to be working on something a bit more innovative.
So, I went back to academia and worked on a PhD in computational physics.
During that time, I also started and sold a company.
And I met my two co-founders, Chris and Gordy.
The two of them are brilliant.
So, Chris, our CTO, he has a background in low latency market making.
His job was to make sure that billions of dollars made it to market in very real-time.
And Gordy was one of the first employees at one of the largest real-time AI companies.
The three of us met two years ago to actually work on solving a massive ethical problem in the web scraping industry.
And that problem was that major corporations were sneaking software onto regular people's devices.
And then using that software to scrape the web with millions of nodes.
And they were not even paying those people.
So, what we realized was this is a perfect application of crypto.
And we set out to build a distributed network that anyone can join.
All you need to do is install a grass node.
And then you're selling some of your unused internet bandwidth.
And then that bandwidth gets used to scrape the web.
Now, after we started working on this, the scope of what we've been building has grown exponentially.
To the point where we realized our mission isn't complete until we've achieved an internet-scale web crawl.
So, in other words, our mission isn't complete until we can crawl the entire internet the way that Google does.
And hopefully one day displace them.
So, anyway, that's a bit about us.
And once again, thank you so much for having me here.
I can't wait for this.
Okay, thanks.
Let me give a translation.
Hi, everyone.
Hello, everyone.
Well, thank you all of you, everyone.
Good afternoon.
We will all be able to help you out today.
Thank you very much.
NFT를 통해서 많은 사람들의 커뮤니티를 이끌었습니다.
아시다시피 그레스는 그레스 토큰 출시를 앞두고 에어드랍 체커를 출시하여
TGE에 대한 많은 기대감을 높이고 있고요.
가장 많이 받은 질문 중 하나는 오늘 말씀들이 그레스 토큰에 관한 것이었습니다.
AMA를 시작하기 전에 그레스 팀원에 대한 소개를 들었는데요.
혹시 덴초이 지금 자리에 계시면 바로 답변에 대한 번역 제공해 줄 수 있을지 부탁드리겠습니다.
아 네네. 이번 이제 나오는 영어 문장부터 제가 준비해서 발언하도록 하겠습니다.
네. OK. Thanks for the introduction of our 그레스 팀.
So, let's jump right into the Q&A.
Today's space will cover many topics including general info about 그레스,
and token, NFT, also product and business of 그레스, so community, future, and more.
So, first question is about what is 그레스?
Could you explain more about the vision and mission of the 그레스?
Absolutely.
So, the way we view the Internet is from Web 1 to Web 2, it's become more and more closed.
And from Web 2 to AI, AI is just making the Internet even more closed than it was before.
If you think about it, in Web 2, you join the Internet using a small number of applications
instead of actually accessing the websites directly, and if AI finds product market fit,
which it will eventually, then most of the population will access the Internet through one portal,
and that one portal is an LLM or an AI model.
Now, you can see how this can go wrong in many different ways.
And LLMs, or AI models, they only have one way of accessing the Internet,
and that is using what are called web indices.
Web indices are huge crawls, basically big maps of the entire Internet.
And in order to build these maps, you need tools like Gras.
In fact, there are only two companies in the world right now who are capable of building these maps.
They're Google and Bing.
And it's Gras' mission to not just build an alternative to them, but to displace them.
Because we believe that a world in which a user-owned map of the Internet doesn't exist
is not a world that we want to live in.
And this is something that should be out there, that anyone should be able to join.
And it's something that we just won't rest until it exists.
Yes.
대진은 혹시 번역 간단하게 제공해 줄 수 있으실까요?
Ah, 네.
번역이 생각보다 어렵네요.
네, 생각보다 어렵고 오히려 말하는 게 더 쉬운 것 같기는 한데
저보다 영어를 잘하시는 분들이 많을 거라고 저는 믿고 있지만
저의 이제 하찮은 영어 실력으로 좀 번역을 도와드리자면
저도 긴장을 해가지고 지금 땀이 많이 납니다.
그래서 이번에 조금 어색하게 번역을 했는데
다음 거부터는 제대로 하도록 할 거고요.
바로 번역을 해드리면
웹투와 웹쓰리가 이제 어느 정도 이제 점점 가까워지고 있다고 보고 있고
AI를 통해서 이제 점점 더 이제 많은 사람들이
인터넷에 더욱더 한 개의 포털을 통해서
액세스가 가능해지고 있는 시대가 다가오고 있다고 말을 했고요.
그리고 지금 가장 대표적인 주제로 구글하고 빙이 있는데
그레스는 뭐 같이 가는 프로젝트가 아니고요.
이를 웹쓰리적으로 이제 대체할 수 있는 프로젝트라고 말을 했습니다.
그리고 이제 그레스를 통해서 누구나 이제
참여 가능한 프로젝트를 만들고 싶다고 말씀을 하셨는데
다음 번역부터 더욱더 잘 하도록 하겠습니다. 감사합니다.
네 여기서 부가 설명을 좀 드리자면은
이제 그레스 같은 경우에는 기존의 웹쓰리 프로젝트들이
기존 산업과 그 AI의 접근성에 대해서
조금 더 이제 형식을 부수고 그리고 기존 산업의 대장들만
가져가는 기회들을 이제 여러 가지
웹쓰리 참여자들에게 나누는 거에 목표가 있다고 할 수가 있어요.
그래서 이제 뭐 데이터를 수집해서 그거에 대한 AI 에이전터를 만들거나
혹은 AI 에이전터 제작사의 그런 데이터를 판매하거나
그런 것에 대해서 여러 가지 기업들이 정보를 일방적으로 수집하고
우리에게는 어떠한 이익도 주지 않은 채
좀 큰 산업이나 그런 데 넘겨서 수익을 취하죠.
그레스 같은 경우에는 이런 것에 대해서 기여를 하는데
모두에게 좀 노드 운영을 하거나
혹은 크롬 익스텐션 설치를 통해서 기회를 주고
머니타이전 할 수 있는 그런 기회를 준다라고 이해하시면 좋겠습니다.
OK, next question is about the name of the grass.
Jandy, could you tell me story about why the name is grass?
Absolutely.
So, if you think about a blade of grass,
one piece of grass by itself is not very useful.
You can't do anything with it. It's very weak. It'll break.
And it kind of just gets lost in the wind.
However, a whole field of grass,
there's unlimited possibilities of things that you can do on top of that.
You can have a picnic. You can play sports.
You can just relax and enjoy.
But without that field, you have nothing.
It's just dirt and you can't do any of those things.
And it's very similar to how grass works,
the grass-to-crypto protocol,
where one node on the network
by itself cannot achieve an internet-scale web crawl.
But with 50 to 100 million nodes,
it is something we will be able to achieve together.
So that's the main inspiration for the name.
OK, thank you. It means the power of gathering.
This is a way I'll do it.
I'll do it.
Grass-to-crypto
In the eyes of grass,
I can't see the power of gathering.
So that's a full-crypto
I can't do it, but I can't do it.
If there is a pool, it will be a building of a building of a building,
a building of a building, and a building of a building.
So Grass Project's effect is similar to this.
It's a building of Grass.
One of the data and node offers is not much of a problem,
but many of the participants who have been able to build a building
and can contribute to the industry.
So that's why we're talking about the network power.
Okay, next question is about progress and achievement about Grass.
Could you tell me about progress and achievement about Grass?
And I'm curious about, do you feel satisfied about progress of Grass?
So Grass has achieved some really, really impressive things,
especially in the last few months.
A lot of people here might be aware that Grass open sourced
one of the largest repositories of Reddit data a few months ago.
Something that people might not be as familiar with is the fact that Grass
Grass actually scraped the largest multimodal data set of all time
that will be available to any company to use.
Now, I'd say what's probably the most impressive thing that Grass is doing
is actually some of the developments that have been happening with LCR
or live context retrieval for those who don't know.
And I'll quickly explain what LCR is for the audience here that might be new to Grass.
LCR is a technology that Grass has developed that uses the entire network.
And the way it behaves is basically you can simply plug it into an LLM
and give the LLM online capability.
So if you have a large language model that you're using,
usually if you ask it a question about something that's very recent
or if you give it a link to go and visit,
most models these days will tell you,
sorry, I was only trained up until data to this date.
But with LCR, these models, before answering your question,
can go and scrape any data from the internet using the Grass network
and then supplement the prompt and then answer your question with real time data.
And you can do this without having to retrain the model.
So you take a model, any model in the world,
and suddenly it has online capabilities.
And this is something that we find extremely important
because every Fortune 500 company that we speak to tells us they want to use AI.
However, the biggest drawback for them is the fact that their applications require access to real-time information.
And in many cases, real-time information from the perspective of millions of networks around the world.
So that's something that we're extremely excited for.
And to give you some context, during alpha testing,
the network is actually scraping enough data to train ChatGPT from scratch on a daily basis.
Now, you asked me if we feel satisfied, and the short answer is no.
The network hasn't even started yet.
The project is basically going to begin the day that Grass launches.
And we won't feel satisfied until Grass has achieved an internet-scale web crawl.
And not only done that, but actually taken down a few internet giants along the way.
Okay.
이번 질문은 이제 Grass가 어떤 업적을 달성했는지,
그리고 현재 Grass의 업적과 이때까지 진행한 성과에 대해서 만족하는지에 대한 답변인데요.
혹시 Denchoi님께서 간단하게 번역을 제공할 수 있으면 부탁드리도록 하겠습니다.
저 근데 진짜 웃음만 나오는 게,
코가투님, 저 먹이려고 지금 어려운 거 시키시는 거 같아서.
맞습니다.
아, 진짜 어렵네요.
일단은 제가 최대한 번역을 했는데,
중간에 놓쳐서 코가투님이 해주셨으면 좋겠다라고 속으로 기도를 하고 있었어요.
근데 역시나 저한테 토스하셔서 제가 한 내용을 말하고,
혹시나 코가투님께서 첨가가 가능하시면 해주시면 딱 맞을 것 같습니다.
제가 보충 설명도 하겠습니다.
좋습니다.
Grass가 이제 성취한 게 엄청난데,
최근 몇 달간 성취한 게 엄청나다고 이야기를 했고,
Grass가 이제 오픈소스가 대단하다고 합니다.
그리고 Grass가 제공해주는 멀티모듈 세트가 있는데,
어, 거기서 이제 최고로는 LCR이라고 합니다.
이게 뭔지는 저도 몰라가지고, 어, 들어봤는데 이제,
뭐, Grass가 독점적으로 제공해줄 수 있는 기술인 거 같아요.
그 차이점 또는 이제 비교할 대상군으로는 LMM이라고 있는데,
어, LMM은 이제 LCR하고 다른 게 이제 걔네는 뭐,
뭐, 연결을 할 때 안 된다라고 뜨지만,
이제 LCR은 Grass를 통해 뭐,
인터넷에 실시간으로 액세스해서 모든 게 제공이 가능하다고 합니다.
그리고 뭐,
포춘지 뭐, 유명한 뭐,
탑티어 회사들이 이제 AI를 사용하는 뭐,
모델을 쓰고 싶어 하는데,
그쪽에서 요구하는 모델이 현재 제공이 안 되고 있지만,
뭐, Grass는 가능하다.
그리고, 어, 네, 딱 요정도로 했는데,
코가2님이 추가해 주시면 감사할 것 같습니다.
네, 알겠습니다.
조금 보충 설명을 하자면은,
이제 아까 LCR에 대해서 얘기를 조금 더 하면 될 것 같아요.
이제 LCR 같은 경우에는 나중에 질문이 또 넘어갈 건데,
아까 말했듯이 이제 Grass가 채택한 방식입니다.
이 LCR에 대비되는 이 기술 이름이 RAG 모델이라고 해요.
근데 이제 어떤 이게, 이게 정확성 면에서는 조금 더 뛰어나다고 합니다.
그럼에도 불구하고 이제 LCR을 선택한 이유가 있다라고 얘기를 하고 싶은 것 같은데요.
LCR 같은 경우에는 조금 더 법령적으로 채택이 될 수 있고,
여러 가지, 여러 가지, 여러 가지, 여러 가지 어플에 쉽게 이렇게
이렇게 Integrated가 될 수가 있다고 합니다.
또한 실시간으로 이 AI 정보를 제공하는 데 훨씬 더 강점이 있기 때문에,
이 LCR 모델을 통해서 다른 산업에서 요구하는 그 스펙을 맞출 수 있다고 합니다.
그래서 지금 Grass가 이러한 면에서 여러 어플리케이션에 쉽게 적용될 수 있는
AI 정보 모델, AI 정보 모델을 개발하는 성취를 이루었다 라고 얘기를 하고 있고요.
만족을 했냐라고 했을 때는 No라고 합니다.
아직까지 할 게 많고, 그리고 또 우리는 기존 인터넷 산업에서
판도를 뒤바꿀 만큼의 성과를 내기까지
아직까지는 만족할 수 없다고 얘기를 했습니다.
OK, thank you.
The next topic is the main event, I think.
It's about the token, the utility of GRASS token.
Could you tell me about the token of GRASS?
Absolutely.
So, when LCR goes live,
any company in the world will be able to use GRASS tokens
to pay for this functionality.
And to give you some context on what type of customers would be paying for this,
it would be every major AI lab in the world,
any Fortune 500 company,
and basically any company that needs real-time access to data insights.
And in order to use the network, they will need to use GRASS tokens as gas.
Because the infrastructure that powers GRASS actually necessitates the token
in order to validate all the web transactions that are going through the network.
And in order to push web transactions, you need to pay for them with GRASS tokens as gas.
Now, as you can imagine, many companies, they probably don't want to pay with cryptocurrency
because a number of companies just haven't reached that point yet.
And that's fine.
They will be able to pay for GRASS services using fiat.
However, the foundation will do some conversions on the back end and turn those into funnels that would accrue value back to the token.
Or back to the network, I should say.
So, yeah, the best way to think about it is the GRASS token, one, secures the network.
Anyone will be able to stake it and delegate their stake to a router.
And then by doing that, they're helping secure the network and earning rewards for that.
And secondly, any major corporation will be able to interact with the network by buying the token and using it as gas.
Okay, okay.
Denchoi님, 이거 혹시 번역 제가 할까요? 아니면 해 주실 수 있을까요?
아, 이건 조금 쉬우셨나 보네요, Koga2님.
그렇죠.
네, 쉬워가지고.
아, 네, 열심히 준비해놨는데, Koga2님께서 하시죠.
아니요, Denchoi님이 해주세요.
아니요, Koga2님 진행하시는 거니까 저는 최소한으로 하겠습니다.
알겠습니다.
그럼 어려울 때만 얄밉게 부탁하도록 하겠습니다.
일단 세 개의 유지 케이스가 있다고 얘기를 합니다.
첫 번째는 이제 GAS로 쓰인다라고 얘기를 하고요.
GRASS 자체가 이제 자체적인 네트워크를 솔라나 기반으로 이제 구축할 거라고 얘기를 했었기 때문에
GRASS token 자체가 이 네트워크 안에서 트랜잭션을 발생시키는 데에 GAS로 사용될 수 있다고 합니다.
두 번째는 이제 프로덕트를 이제 다른 고객사들이 접근하고 이용하는데
이 GRASS가 구축한 데이터 센터에서 무언가를 구매하고 원하는 정보를 산출해낼 때
이 지불 수단을 GRASS token으로 할 수 있게 한다고 합니다.
그래서 그거에 대해서 조금 더 메리트를 느낄 수 있게 하는 방법도 있을 것 같고요.
그리고 이제 FIAS로도 물론 결제를 할 수 있게 할 예정인데
GRASS token은 이제 FIAS로 결제를 돼도
이제 백엔드에서는 그거를 이제 GRASS token으로 컨버전을 시켜서
GRASS token이 사용되는 구조로 사용을 하겠다고 합니다.
그리고 세 번째는 이제 GRASS 네트워크를 좀 더 안정적으로 유지하기 위한
스테이킹을 통해서 이제 네트워크를 유지하거나
뭐 딜리게이트를 하거나
그런 방식으로 쓰이는
전형적인 토큰의 유틸리티를 가질 거라고 얘기를 하고 있습니다.
Okay, next question is also about GRASS token.
I hope that the GRASS token can achieve sustainability.
So, next question about the...
So, I'm very curious about the...
What of the sustainability of GRASS token?
Could you tell me about this?
Yeah, more than happy to.
So, a lot of people might have noticed
that unlike many deep-in projects, GRASS does not just print
inflationary rewards to all of the nodes.
This is because GRASS has sustainability in mind.
And instead, the network is running points campaigns
in order to achieve very specific milestones.
Now, the first milestone is
that GRASS is able to scrape enough data per day
to train ChatGPT from scratch.
And because of that,
the network is looking to decentralize very soon.
After the network has been decentralized,
it's not going to be printing emissions to nodes
on a daily basis the way many deep-in projects do.
It will be running a second points campaign
with the mission of achieving another milestone.
So, during that time,
there are no inflationary emissions going to nodes.
But anyone that wants to use the network
still needs to do that using GRASS tokens.
So, I think the foundation did a very good job
of designing the tokenomics.
So, I'm pretty excited to see it in practice.
Yes, that's it.
Can you help me?
Did you go out?
Yes.
Can you tell us this
Jules?
How are you asking?
Are you going to talk about the token inflation?
So, I'm thinking the token inflation
who might not pay attention to L300
becomes larger than the moment.
Inflation to light the day to happen.
It feels it can happen to we present
to our innovation.
and other database
is going to look at patterns
which are the top in economics
our goal is not고
those special can go down
I think it is more important We provide
So, we will be able to do that.
As I mentioned earlier, the GRASS itself will be a utility.
So, we will be able to design it.
Okay, next question is about NFD, Gigabot.
Gigabots are also required for our today's space.
So, if we want to join Red Pro, please like and RT this space and feed the Google form, please.
So, could you tell me about the utility of Gigabots?
Yeah.
So, one thing I've been saying a lot recently is that there's no utility, just GRASS.
However, I will add something more on that note.
I know that saying no utility, just GRASS is a huge hint, but I'll give a bit of alpha on the spaces that hasn't been given yet.
And the alpha is that the final airdrop checker for GRASS, Stage 1, is going out on October 21st.
Uh-huh, uh-huh, I'm sorry, there's no utility here.
Oh, that's it.
Well, 일단은 utility I don't have to say about that, but we're just curious about the information.
But let me tell you a lot of more information.
I'm curious about this is that it has a new information that I've got to show you.
I think we can see that there is a lot of research data that we have discussed.
So, I don't think that's the right thing.
So, the reason why NFT is being able to maintain the purpose of that is
that it will not be revealed in January 21st.
Okay, next question is also very important.
It's about business of GRASS.
The monetizability of business.
What business will GRASS continue to do based on data correction?
What is the real monetization path of the project?
In other words, how does GRASS earn real revenue like dollars?
Yep, so I've mentioned this earlier, mainly in the context of LCR.
So, if companies want to scrape the web or use LCR to access the web for their AI models, they need to pay on a per gigabyte basis or on a per transaction basis.
We expect this, you know, we expect real-time data to accrue more value than anything else to the network long term.
Short term, however, there has been an explosion in video model training.
And in that world, raw training data is still incredibly valuable and scarce and very sought after.
And at the moment, you know, I mentioned earlier GRASS has the largest multimodal data set in the world.
There are many companies that want to access this data set.
And to give you an idea, I guess, of how in demand it is, GRASS has billions of minutes worth of multimodal content.
And you can, you know, you can probably look up what the range is for how these things are valued.
But it's quite a large data set.
And even looking at it conservatively, it's still very close to a billion minutes worth of content.
Okay, I guess.
Okay, okay, it's necessary.
I was talking about the short term,
about the
does 얘기를 하고 있는데 그게 무엇이냐면요 이제
multmodel database buat inglhazz 같은 경우는
지금 어떤 data center 보다 많은 아마 수십억의
multmodel database를 가지고 있고
multmodel contents를 가지고 있기 때문에
지금 당장 모여진 데이터 판매 외에도
현재 모아진 정보들을 종합해서
그것을 판매해서 또 다른 수익원을 만들 수 있다고 합니다
The next question is a difficult area to me, but we are talking about tech.
Accuracy is using live context retrieval, LCR model, but in these days, RAG models have emerged as an alternative approach for improving accuracy.
So, what are the competitive advantages of using the LCR model compared to RAG model? Could you tell me about this?
Absolutely. So, the issue with a lot of traditional RAG is that it has maybe 65 to 70% accuracy if it's implemented well.
And the reason that the accuracy is like that is because one, it's using static data for a very specific domain.
Traditional RAG doesn't work very well if you have lots of data. If you add more data, it gets less accurate.
So, if you want to use a big repository to help supplement a model and giving it knowledge, traditional RAG systems break down pretty quickly.
Secondly, most of them use something that's called cosine similarity to look up which things to fetch from the RAG database to give to the model.
Cosine similarity is a pretty good tool, but the big risk with doing that is that cosine similarity is not 100% accurate.
So, you end up potentially fetching the wrong row of data.
And if you do that, it's irreversible. The large language model has no way to self-correct.
Now, the way LCR operates is by first creating an index of the entire web.
And then, instead of calculating a bunch of distances using cosine similarity,
it uses a very similar technique to actually the large internet giants in the way that they build their searches
by creating knowledge graphs of all of this data.
And then, when LCR receives a prompt, it actually has much better accuracy.
And early tests are showing more than 12% better, which is huge in the AI world.
So, when it receives a prompt, it knows exactly which website to look up.
And then, it uses the entire GRASS network to go and scrape that website in less than one second.
So, to answer your question, accuracy is the biggest one, but beyond that speed as well,
and the ability to leverage a lot more data.
If you try using a traditional RAG system like a vector database with the entire internet,
it's basically impossible. But LCR enables this.
Okay. Okay.
Okay.
Okay.
일단 정말 어려운 답변이어가지고, 사실.
Okay.
그러면은, 일단은 조금 쳐보고,
그 가루 좀 위에 뿌려주시면 딱 맞을 것 같아요.
알겠습니다.
아, 네네네.
일단은 RH 모델하고 LCR 모델에 대해서 이제 질문이 들어갔었고요.
파운더께서 말씀하신 내용을 말씀드리면,
일단은 전통적인 모델은 정확도가 60에서 70% 정도로 이제 유지가 되고 있는데,
그 이유는 이제 스태틱 데이터를 사용해서이고,
또한 이제 더 많은 데이터를 입력하고 출력을 요청하면은 정확도가 점점 떨어지는 모델이라고 합니다.
그리고 이제 뭐, 전통적인 모델이 이제 코사인 시뮬러리티라는 뭐,
그런 걸 통해서 이제 데이터베이스를 불러오는데,
어, 요거를, 요건 되게 좋은 툴인데, 요거 또한 리스크가 있고,
100%의 정확도를 제공해주지 않기 때문에,
결국에는 틀린 정보 전달이 될 수밖에 없는 구조이다.
다만, LCR은 이제 전통적인 모델과는 다르게,
어, 정확도가 엄청나고, 속도 또한 정말 빠르고,
결과적으로는 그레스가 킹왕짱이다. 이상입니다.
알겠습니다. 일단은 계속 강조하는 게 이제 어크리스인 것 같아요.
그래서 정확도에 대해서 얘기를 하고 있는데,
이제 정확도 향상을 위해서 오히려 이제 LCR 접근 반식을 선택했다고 합니다.
그래서 기존 모델은 아까 말한 것처럼,
장기적으로 갈수록 이 데이터를 수집하는 데 있어서 정확도가 떨어질 수 있지만,
LCR은 계속해서 실시간으로 데이터를 학습하고,
많은 모델링을 함을 통해, 많은 데이터 수집을 통해서,
더 장기적인 비전으로 정확한 AI 기술을 제공할 수 있기 때문에,
LCR 기술을 택했다고 합니다.
Okay, next question is how easy the grass to use?
So I'm wondering if anyone can easily use and utility,
utilize a grass product without any barriers to entry.
Could you tell me about how easy to use the grass is?
Yeah, it's incredibly easy to use and this is intentional.
It was a very important design decision that anybody with no knowledge of crypto or blockchain,
no knowledge of AI, no technical knowledge, can easily join grass.
So right now you can install the desktop node,
which is the best way to join the network with only three clicks.
And at the moment, in order to participate in rewards,
people do need to link a wallet.
However, in the coming months,
grass will actually be putting out its own wallet solution
that'll enable anyone to just sign up and have a wallet already pre-built for them.
The reason this is so important for grass is because it requires mainstream adoption
in order to be successful, in order to achieve its goals.
If every single person in the crypto space joined grass tomorrow,
there still wouldn't be enough nodes for grass to win against the huge Web 2 giants.
So we need help from outside of crypto as well.
And that's the main reason why we put so much time and thinking into making sure that the process of joining is as smooth as possible.
And, you know, it's one of those things that you may as well do.
There are companies out there that are already using your resources without asking you for permission.
So you might as well just join grass and take back a little bit of control.
Yeah.
Okay.
Of course I'll do it since anything.
If it is ...
Are your nouveaux things, or for a...
Yes ...
Well, I'm going to take a close I think I'll do it least once I want it in my case.
Oh, yes?
Okay.
Once again.
I'll say I'm in subcaring time.
As you were偏�值 if there are sevenged-hit,
I kind of understood everything it takes me together.
이것도 얼마나 쉽게 했고 왜 쉽게 했는지에 대해서 얘기를 한 것 같습니다.
그리고 앞으로 제공될 프로덕트에 대해서도 계속해서 이렇게 쉽게 조인할 수 있는 방향성을 추구할 것이라고 해요.
지금도 아시다시피 여러분들도 단지 클릭 3번이면 그레스에 조인을 할 수가 있는데
So, this is a simple and easy way to use it.
Grass is web 3.0, but not mainstream.
So, that's the goal to achieve.
So, we can use Grass as a onboarding.
So, we can use Grass as well to use it.
So, we can use Grass as a Chrome extension.
The most important thing to think is easy to use.
Okay, let's jump into the more casual subject about community.
Many web 3.0 projects are entering the Korean market
and investing a lot of interest in Korean market.
So, I'm curious about what the Grass team think about the Korean market
and what are your plans for entering the Korean coin market in the future?
So, the Korean market is incredibly important for Grass.
Now, not only from a node perspective, but also from a community perspective.
But I'll speak to both.
From a node perspective, Korea is actually one of the most highly in demand geographies for Grass nodes.
And when OCR goes live, Korea will be a very important place for there to be a large saturation of nodes in order for it to work at 100% functionality.
So, Korea is, yeah, like I said, very important from a technology perspective.
From a community perspective, it's also very important.
Korea has one of the strongest, one of the most loyal Web3 audiences in the world.
And it's our belief that for any crypto project to be taken seriously, the Korean audience is very, very important.
So, yeah, I mean, it's one of those things where we're extremely grateful for everyone that's in our community, including the Korean community.
Okay.
In fact, the Daniel, the BD of Asia has to spread grass in Korean markets.
So, I want to say thank you for Daniel to have spread grass in Korea.
So, let's take a look at this
I think that's what we're doing.
Community, I think that's what we're doing.
We're doing web3 project,
and we're doing it.
We're doing it.
We're doing it.
We're doing it.
We're doing it.
We're doing it.
We're doing it.
Daniel is a great-assia-bd.
So we came up with our new world,
and we brought them to your community.
We want to spread the conversation,
as we wanted to make it more support.
We're doing it.
Unfortunately, the app repayment was a lot.
It was a lot to say.
The most of us in this video,
the most gums are a great-assiening
to make our community feel very happy.
So, I appreciate the help of that.
We can hear it.
Alright, we are now in almost the final step
of this journey.
past, is GRASS only going to work on Solana, or can it be cross-chain?
So GRASS works as a sovereign data roll-up. At the moment, the plan is for it to settle to Solana.
That being said, GRASS also has the flexibility to settle on any chain. But for the time being,
and for the foreseeable future, GRASS is building on Solana. The decision to do this
was pretty easy considering how great Solana technology is. It's the fastest chain and it's
the cheapest chain. And it's the only chain that enables GRASS to operate without creating unnecessary
costs for the Web2 AI customers that use GRASS.
Okay, I understand. In the future, we will continue to use Solana as well as Solana.
Because Solana is now the business of GRASS.
So, the next question is about the product.
about the product of GRASS.
The only product of GRASS is our Chrome extension
and desktop app.
So I want to hear about that.
Another product of GRASS will be announced in near future.
Yeah.
So as many people know, there are four ways
to join the GRASS network today.
You can download the desktop node.
That's the best way to join the network.
You can download the Saga app.
That's the Solana phone app.
You can add the community node, or you
can add the traditional Chrome extension.
GRASS is working on a few other exciting distribution
mechanisms.
One of them is a hardware device.
Another one is an Android app, which is coming out in Q4.
And there are also a few other fun ones
that GRASS has slated for next year, including things
like Roku TV applications, and potentially
Xbox and PlayStation apps as well.
OK, very interesting.
Well, first of all, there are four ways that you can add to it.
You can add the desktop node, community node, and Chrome extension, and Solana phone.
We still don't have the opportunity to get it.
But we still don't have the opportunity to get it.
The Solana phone is also GRASS app with the
has been used to be used to have a lot of
And many products that have been used to be added.
Android app, starting to start with, and
The next is hardware, so the hardware is a device.
Glasses with a device that has to be seen.
And the next year, in 2025, local TV has a device that has to be seen.
And Xbox, we can say that it's a word or a word,
or a word or a word.
Or a word.
But gaming has to be seen in a specific way.
data to collect data.
I think it's just a Chrome extension,
but it's just a different place.
Grass's product can be found in a different place.
All right.
Thanks for many detailed questions.
This is the final question of the AMA and end of the AMA.
So I want to ask about Alpa for people participating
at this pace, can you share something about coming up
exciting event?
Yes.
So I mentioned it earlier, but I'm
going to mention it one more time just
in case anyone tuned in a bit later or missed it.
The Grass final airdrop checker is going
live on October 21st, 2024.
And yeah, that has never been announced before.
So I'm happy to do it here.
계속 강조하는 거를 강조하고 싶다고 합니다.
지금부터 11일 뒤죠.
10월 21일에 final airdrop checker가 나온다고 합니다.
지금은 단순히 수양만 확인할 수 있는데
final를 안 했고 하니까
claim function이나 이런 게 포함될 수도 있다.
뭐 이런 얘기인 것 같아요.
그래서 다른 분들도 참고하실 거는
여러분들이 기다리는 그 소식은
10월 21일에 있을 것 같습니다.
I'm really looking forward to this event.
So I want to see the official announcement for final airdrop
checker in October 21st.
So this is the end of today's space.
And thank you for all joining our space.
Before the turn of the space, I want to notice about event again.
So if you would like to participate in today's Gigabot referral,
please like RT this space and submit the Google form in my Twitter post.
Thank you for joining this space.
And thank you, Chris.
Have a good day.
Thank you so much once again for having me.
This was a lot of fun.
And maybe we'll do it again sometime.
Thank you guys.
감사합니다.
