Hello everyone, I'm 陳鈺澔, my English name is Danny. I graduated from the Department of Electrical Engineering at the National Taiwan University of Science and Technology. Initially, I worked in Computer Vision after graduating from graduate school.

However, as we all know, AI began to explode around 2023. We all know that people in the AI ​​industry must keep up with the times. If you only focus on your own expertise, you will soon be swept away by the tide of time.

Therefore, I started CryptoMind in December 2025. At first, my purpose in creating this platform was simply because I felt empty after reading many theoretical books but lacking a practical application. I was also involved in cryptocurrency, and on a whim, I wanted to build a cryptocurrency recommendation agent for my own testing.

However, as everyone knows, around that time, Bitcoin exploded, and the entire field became increasingly narrow due to the introduction of cryptocurrency regulations. This made me want to shift my focus to other directions, combining my lifelong AI experience to do something.

For me, having capable people use my product is not just a matter of use, but more of a form of praise and approval. Whether this person is a PhD student or a kid on the street, their identity is unimportant. What matters is that I've achieved recognition for what I've done.

So, why did I want to create this platform? Because there's just too much information in the cryptocurrency market, scattered across different places. News websites tell one story, social media tells another. With the rise of AI, it's incredibly difficult for ordinary users to judge which information is trustworthy and which is a scam. Therefore, I want to provide a reliable information hub.

Furthermore, by incorporating everyone's habits and memories, users can customize the platform to create the most suitable agent for themselves, and switch between different companies' models to answer different questions.

But my partners and I have a vision that goes beyond this. If it's just about choosing models, tools, and customizing agents, then it's not much different from using GPT or Claude. Our goal is to build a trustworthy environment, but what does "trustworthy" mean? That's a very complex issue.

If someone posts a message today saying we can buy a certain stock, and that post gets tens of thousands of likes and hundreds of comments, is it true or false? For a market newbie, the probability of believing the message is far higher than doubting it.

But now AI can automatically comment and even operate computers, potentially creating a "Mandela Effect" in the virtual world: because everyone else believes it, I should believe it too, or it looks true. This same problem can occur, or even exist, on social media and major cryptocurrency exchanges.

Therefore, we can no longer rely solely on likes, comments, or followers to determine authenticity. Some teams choose to start with real-person verification, such as using biometrics to confirm if an account is backed by a single person. But even with privacy-preserving technologies, some users may still feel uncomfortable about verification processes involving iris scans, facial features, or other biometrics.

Therefore, we want to try another approach. Users are not required to disclose their real names; instead, they can establish a long-term, anonymous identity through blockchain technology. Users can prove they control an address without revealing their real name. The platform then gradually builds credibility based on the address's lifespan, long-term activity, past content, risk records, and endorsements from other trusted addresses.

This system does not directly equate credibility with the amount of cryptocurrency held or the number of transactions, as these numbers can be manipulated through funds and self-dealing. What truly matters is the source of the endorsement, the reliability of the endorser, the existence of financial relationships, and the accuracy of information provided in the same field in the past.

However, how to identify these aspects is something we are considering, and we aim to integrate relevant AI features to create a DApp that truly helps and benefits people, not just something created for the sake of creation.

Therefore, adhering to this ultimate goal, we are gradually improving our infrastructure, combining personalized memory/skills, financial agent assistants, AI accounting systems, forum and friend functions, fraud reporting systems, and a system for proposing new ideas that emerge along the way.

Our hope is to make the platform truly useful to users, expand and sustain its development, and eventually, when the user base is large enough (hopefully so), the existence of relevant records will become irreplaceable labels. Of course, these are grand goals, but at least we've started.

CryptoMind is now officially online. The website is as follows; anyone interested is welcome to use our platform:

https://cryptomind-ton.zeabur.app

While combining AI-Agent, we also remember to bring our own interests and knowledge into different fields, trying to create a platform that people can use and even retain.

Trust is a long-term accumulation process; it doesn't depend on who I am, how much money I have, or how great my status is, but simply on how much you prove your value. Just as if I didn't interact with the world today, the world wouldn't know I exist, let alone whether I'm trustworthy.

So even with ambitious goals, we must start from the ground up: building a truly usable and useful platform. This isn't just our professional mission, but also a way of giving back to our field. We hope that one day, a small group of people will genuinely use what we've created and proudly say, "I'm certified by this platform."

However, because our team of developers is small, and we have limited resources and time, thanks to the development of AI, we can at least continue working on this project after work.

We also hope that those with the willingness or ability to support us can contribute, not just financially, but also in terms of hardware. We hope to move the development to a dedicated server to manage platform information, reducing current cloud server rental costs.

Furthermore, we hope to expand the deployment of AI models and services like ASR and OCR models to provide members with more functionalities. This is the direction we will take if we raise funds.

Our developer and platform information page: https://aaaa47080.github.io/cryptomind-landing/

We have also open-sourced the current status of the relevant code development for investors' reference:

https://github.com/aaaa47080/cryptomind_public

中文版本：

大家好，我是陳鈺澔，英文名字是Danny，我畢業於臺灣科技大學電機所，一開始我是做CV從研究所畢業，但眾所皆知的是從2023年左右，AI逐漸開始爆發，我們都知道，做AI這行業的人總必須與時俱進，如果只關注自己所精通的知識，很快將會被掩沒在時間的洪流中，因此我從 2025 年 12 月開始做 CryptoMind。一開始，我做這個平台的目的，只是覺得在書本上看了許多理論，卻沒有一個實踐的地方覺得空虛，剛好自己也有在碰觸加密貨幣，突發奇想的我就想說搭建一個讓自己測試用的加密貨幣推薦 Agent，但大家也都知道的是，剛好從那個時候比特幣爆發，而整個領域隨著加密貨幣的監管推出變得愈發狹隘，讓也開始想轉往其他方向，搭配自己畢生所學的AI經驗來去做些什麼。對我來說有能人夠使用我的產品，不僅僅只是一種使用，更多的是一種表揚與贊同。不論這個人今天是個博士生，又或者路邊的小孩，身份不重要。重要的事，我做到了有人認同了我所做的東西。

那話說回來我為什麼想做這個平台？因為加密貨幣市場的資訊真的太多，而且散落在不同地方。新聞網站講一套，社群媒體又講另一套。加上AI的興起，普通使用者根本很難判斷哪些資訊能相信，哪些又是詐騙，所以提供一個可信的資訊匯集。加上配合每個人都會有的習慣與記憶，可以於平台中自定義進行客製化搭配出自己最適合的Agent，同時也可以在不同詢問場景中，切換成不同家公司的模型進行回答。但我與夥伴的理想不僅限於此，因為如果只有選模型、選工具和自訂 Agent，那這樣與使用 GPT 或 Claude 還是沒有太大的區別。我們的目標是建立一個可信的環境，但什麼叫做可信？這其實是一個很複雜的議題。

如果今天有人發了一則貼文，說我們可以買進某一支股票，然後該貼文被按了幾萬顆愛心，甚至有好幾百則留言，那這則貼文到底是真是假？對於一個剛步入市場的新手來說，他選擇相信這則訊息的機率，一定遠高於先去懷疑它。但如今 AI 已經可以自動留言，甚至可以自動操作電腦，很可能因此產生虛擬世界中的「曼德拉效應」因為大家都相信，所以我也應該相信，或者它看起來就是真的。不管是在社交媒體，還是在各大加密貨幣交易所平台，都可能發生甚至存在一樣的問題。

所以如今的我們，不能再只靠按讚數、留言數或追蹤人數來分辨真偽。有些團隊選擇從真人驗證開始，例如透過生物辨識確認一個帳號背後是不是一位獨立真人。但即使技術上使用了隱私保護方式，對部分使用者來說，只要驗證流程涉及虹膜、臉部或其他生物特徵，就仍然可能引起隱私上的反感。因此，我們想嘗試另外一種方式。使用者不一定要公開真實姓名，而是可以透過區塊鏈特性建立起一個長期存在的匿名身分。使用者可以證明自己控制某個地址，但不需要公開真實姓名。接著，平台再根據地址存活時間、長期活動、過去內容、風險紀錄，以及其他可信地址對他的認可，逐步建立信譽。

這套系統不會把持幣量或交易數直接當成可信程度，因為這些數字可以被資金和自我交易製造。真正重要的是認可來自誰、認可者本身是否可靠、彼此是否存在資金關係，以及這個人過去在相同領域提供的資訊是否準確。但又要怎麼樣去識別這件事情呢，我們也是有在想，並且又想融合相關AI特性打造出一個真正可以幫助人，對人有用，不是為了做而做的DAPP。

所以我們秉持著這個終極目標，逐步完善自己的基礎設施，結合個人化記憶／SKill、金融Agent助手、AI記帳系統、論壇與好友功能、詐欺檢舉系統等等，以及過程中一個個出現的新點子提案系統等等，希望是讓使用者真正覺得有用的同時，擴大並持續維持平台上的發展，而直到基數夠多（希望如此），變可以將相關紀錄的存在，而這個紀錄也將變成真人無法取代的標籤。當然這些都是宏大的目標，但起碼我們起步了。

而CryptoMind 目前已經正式上線，網址如下，歡迎有興趣的人有可以使用我們平台：

https://cryptomind-ton.zeabur.app

結合AI-Agent同時我們也不忘乎將自己的興趣所學帶入不同領域，嘗試做出可以讓人使用，甚至可以留住人的平台。信任是一個長期被累積的事件，他不取決於我是誰、我我有多少錢、我的身份是否偉大，而僅僅是你做了多少證明你自己存在的價值。就像如果今天我不去與世界互動，這個世界變不知道有我的存在，更不知道我是否可信，所以即便我們有著遠大目標，但還是要從根開始做起，那就是打造一個有任使用的，真正有用的平台。這不只是我們對我們專業的使命，更是我們想要對我們所學專業的回饋希望能真正有朝一日，世界會有那麼一小撮人平常就真的在用我們所做的東西，並且他可以驕傲地說，我有得到這個平台的認證。

但因為我們幾個開發者人數不多，同時不管是財力或者時間都有限，得力於AI的發展，讓我們下班至少還有精力可以持續做著這個項目，但同時我們也希望有意願或者有額外能力的人可以資助我們，他不緊緊只是資金，其餘硬體設備也都可以，我們也希望能將相關開發挪於落地伺服器中自行保管相關平台資訊，減輕目前的雲端租用伺服器費用，同時也希望在這之上，擴展部署落地AI的相關模型，以及更多服務像是ASR、OCR模型給與相關會員擁有更多功能使用。而這也是我們如果籌集到資金後會運用的方向。

我們開發者與平台相關介紹頁面：https://aaaa47080.github.io/cryptomind-landing/

同時，我們也開源了到目前為止的相關代碼開發狀態，提供與投資人檢核：

https://github.com/aaaa47080/cryptomind_public