AI Summary
Etherhub 与 AI 代理产品 AI-on 的开发进展:预测市场数据平台预计 10 月中旬上线,AI-on 已获付费客户积极反馈并持续迭代。
- Etherhub 定位为全球首个面向金融机构的金融预测市场数据收集平台,通过预测市场汇聚金融专家及普通用户的智力数据。
- Etherhub 是 Oort 与纳斯达克上市公司 Etherholdings 的合资产品,目前处于内部测试阶段。
- 9 月底 Beta 与 VP 上线目标未如期达成,但团队称一切按计划推进,预计 10 月中旬(约两周后)正式发布。
- 下次社区会议将披露 Etherhub 更多细节,届时产品预计已上线供用户下载试用。
- AI-on 已上线并有付费客户,多数反馈积极,产品挑战更多在市场营销而非技术层面。
- 典型用户反馈已落地:如为社媒经理代理新增 X/Twitter 帖子定时发布功能,支持提前 1-2 周排期自动执行。
- 团队认为 AI 代理赛道竞争激烈(如 MITA 等产品表现惊艳),但 AI-on 专注 Web3 公司营销这一细分领域,形成差异化优势。
结论:建议社区关注 10 月中旬 Etherhub 正式上线,并在下次 town hall 获取产品最新细节与试用入口。
Transcript
Sure, so the question is about Etherhub.
On the last town hall,
the target was to launch beta and VP by end of September.
Now it's October 1st,
so essentially the date is right now.
So where does the project actually stand?
What is the honest timeline so community can stay in tune?
Yeah, yeah. Okay.
So yeah, before I share the progress of
Etherhub development,
so I want to just give a recall on what Etherhub is.
Basically we're trying to build the world first financial,
financial prediction market focused data collection platform
for the financial institutions around the world.
So basically, you know, that we are trying to use this platform
to collect intellectual data from financial experts
or even just regular people.
And so on the prediction market, financial prediction market.
And this data could be super valuable for many financial institutions
for them to do like a predict the market or for them to understand
their potential retail investors and trainers.
Yeah, so now if you look at our data hub,
new launch the products tasks,
basically you will get a little bit better idea about how this,
you know, predicts a market focused data collection task looks like.
So the Etherhub is a joint venture products between Oort
and the Nasdaq listed company Etherhodings.
So the product is still under development,
but we already have shown on the roadmap.
We already have the first person.
Now it's under internal testing right now.
So yeah, I mean, everything is on schedule.
So we are still stick on the schedules launch date.
Which is like mid October.
So probably two weeks from now.
And I think next time how you know in next time how meeting,
we will cover more exciting detail of the stuff about Etherhub.
Because we are we're expecting that you know,
we will launch the you know the product to be alive and everybody can download it
and try it out.
So yeah, I mean that's the latest updates about the product aside.
Yeah, so Kima back to you.
Thank you, Dr. Lee.
So the Etherhub is a very big infrastructure story and the data layer sitting right underneath it.
Exactly what AI on the runs as well.
So that would bring me actually to the product that's actually live,
generating real revenue and evolving real time.
So actually we have a second question about the product.
And since Mike is not there, I would direct it to you.
And if you need Dr. Yang on myself can also jump in and maybe add some kind of comments.
So the AI on question.
It has been life with paying customers.
And what has been the most valuable feedback we've gotten from actual users so far.
And how is it changing the way the product is built?
Yeah, so first of all, we are glad to hear that the most of feedback from our clients are very positive.
But we also see some feedback valuable feedback for us to improve the product.
So basically this is the you know building agents AI on is pretty challenged.
The Chinese part is not a technical but most likely on the product marketing.
And you know that's AI on you know some some users take a very specific feedback.
For example, like a so for the social media manager you should add this feature and that feature.
For example, one of the comments is like, you know, yeah, also social media manager.
Yeah, agents to support the you know the X or the Twitter posts a scheduler.
So basically the marketing team can schedule the post one or two weeks in advance.
And then the AI agent going to automatically execute it and post it on Twitter or acts.
Something like that, right.
So and then we developed these features and now it's alive.
So yeah, I mean, these are the good examples of positive and valuable feedback.
But now we see lots of competition competitors on the market and you know especially for the news from MITA.
You know, this is a very hot.
Yeah, agent and personally I use it for quite a while and the you know the performance is amazing.
I mean, I use it to be my personal assistant to you know use it to.
Post the stuff here and there and the shopping for me and making schedules and book the you know the travel all the travel stuff and you know and ask any questions ask them use to you know monitor my email and you know give me notification if something you took my reply and they they all the news will draft the thing and reply the email for me.
And you know lots of things almost everything.
Yeah, I mean, this is a.
This is a pretty so that's why I said it's a pretty challenge for in the field of AI agent because you know the agent could be become more and more powerful and towards the age.
Right, the age.
So basically you can use it one agent to do everything you need right so yeah but but I think I also have the opportunity because we are very dedicated and focused on the marketing side with the especially for the Web three companies right so.
And this is a pretty niche and the very unique avenue that you know we have very specific design for it so that's what we and the feedback from our clients are very encouraging and very specific you know I think we will keep moving along the line and we will still see the opportunity down on the road.
I really love your answer about the news i've seen that you've been testing it quite a lot in the internal communications.
I also tested it and yes, I have to admit that AI on has very strong competitors have been there and please correct me if I'm being delusional or just being wrong but I feel like news and AI on are solving different problems entirely such as.
Like me uses a more like a consumer personal assistant that you do like scheduling with one file feeling personal errands and the ion is more of multi agent system that build for the three operators so the community management obviously can be done by both but I feel like the ownership and data sovereignty is a bit more privacy focused on our end.
And you know like I know that meta is already exploring a lot of commercialization opportunities and I own around some or disincentralizing for structure so it's essentially kind of different it's already decentralized by itself.
So yeah, I really love your answer about that and honesty really appreciate that and yes, the feedback from clients about AI on has been really really helpful in shaping on the product.
All right, if you don't have anything to add I'll jump into the next question which was covering the genetic account lock up.
So the genetic account lock up proposal has passed in all girls hello, could you hear me?
Yes, yeah, we're not for you for you.
Yeah, I guess so sorry for that.
So genetic account lock up continuing about that proposal passed in all girls and it has been in effect for over a month by now and we have worked with OAPA community very closely about the KPIs and what they need to provide in order to fulfill the proposal deliverability.
So we have been in touch with their representatives and they're working hard to increase the liquidity on gate market and essentially to you know mark check marked all of the responsibilities stated in the proposal.
So the communications go very good and smooth so so far it's good in about maybe months maybe in two actually in the months we will be having another review discussion with this community leaders and the further updates will be shared to all the community by Michael Robinson, the chairman of or foundation in detail.
So that's about that and yes, the governance is a part of shaping the network by the community and I wanted to speak with Dr Young about the technical side that you know to see how network grows from the inside the architecture, the protocol, the engineering decisions that usually don't get a lot of you know attention from the community.
Dr Young, I have a question for you from the.
Ather hub technical architecture so it's been targeting to launch MVP very soon and how does the technical build actually look like right now and is or owning the data layer the AI layer or both and what was the hardest engineering problem that our team been solving right now.
It's a great question so as everybody know that if the hub is a financial prediction application right i'll say while we're building this application they've what provides the data layer.
The data hub injection and decentralized storage and the data processing and all the AI and intelligence layer this top of that on top of physical network so that's what's main responsibility for this collaboration and.
Ather they may bring the data from the financial domain and you know also approach to their their community and their users and all the application service so the UI UX how to interact with with their community members is.
It's their fellow they they design this.
Cation that's how we split this responsibility between us the hardest engineering problem maybe or say on the financial prediction market this prediction task.
It actually arrives in different shapes so this OHL CV means like if you look into this kind of sticks we have like a high price low price close price open price.
Serious we have to integrate our program and the social TED context and use and I'll say events I know.
People can use different losses to do the prediction so it's it's difficult to it's a matter of challenging design a prediction task based on so overwhelming kind of information.
So we have to design different schemas different granularity is different label definition on so I was in the hardest part is to define a canonical way like a canonical format each task normalized into our data hub for storage and processing of them blue and pipeline on yeah so far.
I think our collaboration went very well and we are it's going through very smoothly and there will be available in the middle of tober yeah we're very confident people like it yeah.
Does it mean that the community will get exclusive early access by me of October.
I think yeah so technically it will be right so I think our marketing team will make some announcement later and see how people will get like early access to it maybe we can some great testing like white listed people.
I think I will back to you when to some announcement night.
I see well that's very good news for the community we have quite a few people trying to you know asking about this and we be and trying to get another access I think we can provide that community as one of the exclusive campaigns so this is very good and very good to know about the other helps architecture from the inside out.
Okay and another thing I wanted to address directly because community had been asking about it so we have how is the exchange good.