Though our professor gave the documented and oral instructions, it is still unbelievably clear that students felt unclear about HOW and WHY to solve it. The confusion arises in every aspect possible.

Since I am one of them(the class is mixed with CS students and half layman like me), I decided to observe myself regarding the confusion. I decided to interact with GPT, to get a better understanding and finally develop answers to the questions, instead of relying on GPT's response only.

So what I did:

The instruction for the assignment is like this:

Before diving into the details, the starter is a simple refer-to-what human language task based on the text itself. For example,

• These traces -> power trace

• The first two -> current, voltage - used to determine the power consumption

• The digital signal -> used to determine when the device is performing a specific task

Next is to match the text (instruction)-to-code task, where my colleague's feedback is like this:

For me, it's even worse(at the beginning), I cannot understand the code easily. I don't know where my attention should be. There are 2 outputs in [16] and [19], so i randomly take the [16] raw data into GPT,

Oh it reminds me of the things I did in the Preliminary stage.

Then I was confused with time params, so I started asking about delta T.

It turns out I was confused with the relationship between data-sample-rate and delta T, then I can easily calculate the energy consumption.

I knew what I want from GPT, so I prompted it to fill in the blank via (?)

I also quoted GPT response and asked

But interestingly, if we compared the same question given by GPT response and my (another) colleague, the latter is very straightforward.

Takeaway:

• In this practice, I have a "goal"(with many sub-goals) to understand and solve the assignment task.

  • Satisfied🕺🏻: conversational interactions.

  • Back-off🤯: A LOT of responses which add to my attention load.

Observation:

• Prompts are Goal-directed, every prompt contains a signal towards the goal, and the REAL prompt is likely formed after some active interactions. There is unlikely a one-shot problem-solving prompt.

• The subgoal(next prompt) is dynamically defined through interaction, I stared at the things(GPT response) I understood and explored things I didn't AROUND the understandable context. - Here is where the subgoal is defined step-by-step.

• Every interaction, there is a tiny component being picked up, towards the goal.

• It reminds me of a paper*[1]*, the insight there is adding noise to get better signal, where they leverage RL in their model.

• Maybe there is less need to verify the content of LLM FOR the users because we underestimate people's ability to verify the LLM "hallucinations" and overestimate our ability to do this for people. Instead, we might record the interaction and provide the user with a reflection of their learning process - e.g. in TLCA how they draft the input, how they prompt, and what profile they choose.

The paper is interesting in how it uses suboptimal human conversations for training the AI, specifically it employs “offline reinforcement learning to train an interactive conversational agent that can optimize goal-directed objectives over multiple turns.” The insight is a bit like adding noise to get better signal — a method used in audio processing and mastering, and in fact diffusion image AI models.[2]

This also reminds me of the huge gap in the eye of a domain expert and a learner, the linking path to the knowledge is not straightforward and quite personalized.

It immediately raises the question: Whom should we believe, and based on what? How can we establish ownership of an original idea? Through talking and showing pictures on social media?

How could the viewers verify? Through the trust of …(I don’t know what to write here). If I don’t know you(him) personally, how do I know what you(he) said, posted, or claimed what you(he) did is trustworthy?

This is intriguing when we consider trust in societal affairs. It also reminds me that this could be an interesting topic to explore within our Crypto Kitchen👩‍🍳.

I would prefer another approach on this matter coz I've never been a staunch believer in the notion of true originality (please excuse my rigid academic mindset, as you delve deeper into more books you will discover a multitude of minds with similarities), I’m more inclined towards the concept of “knowledge referencing”.

I’m a fan of acknowledging the source of the inspirations that fuel my work, that’s why I’m researching and building tools (word block 1, 2) on leveraging the semantic web, attempting to implement the ever-evolving idea of knowledge graph (here I referenced thoughts giant leaders Sir Tim, Ted Nelson, Alan Kay, Marvin Minsky and more), the idea is also about addressing the connections of the ideas and finding&proving the unique contributions of individuals on the web. Perhaps if we took this approach, our interaction could be more vibrant🧐 and less confrontational😠. Instead of defending the ownership of the idea, we could gracefully acknowledge the people who inspire us.

I can see your desire to declare the idea is yours, but this time WE DON’T CARE. The reason is beyond simple.

Unfortunately, I could not even reference you, since the fact is we didn’t even see you and your idea/project ever before while we were discussing and preparing the event,let alone copying. Otherwise, we will be happy to.

We have no intention of engaging in debates about originality; we have a multitude of tasks at hand. We kindly request that you refrain from tagging us and contacting our guest friends. If you're interested in a collaborative and humble conversation, we are open to that.

Additionally, I'd suggest exploring methods like inscribing your ideas on the blockchain or publishing papers whenever you come up with something new. This might be a viable way to gain the "confirmation" you seek for what you refer to as intellectual rights, though I can't guarantee justice in this regard.

Lastly, let's shift the question from "HOW" to "WHY."

Why do you feel the need to prove originality? As Ted Nelson wisely pointed out, ideas are variations. Perhaps it's more productive to focus on your version and your own purpose.

I wrote <what is a Crypto Kitchen?> here, we want to implement it for a good purpose and energy, to collaborate with like-minds, to engage interesting and knowledgeable souls (our friends we knew for some time) to share their knowledge, to bridge the eastern and western culture. That’s the purpose.

Fear is a common human experience. I've come to understand that some fears stem from a deep-seated desire, a longing, a sense of expectation. Sometimes fear is a magical signal, indicating what matters to you, even if you are not consciously aware of it. People often say not to expect too much from things, but I think that's a post-hoc perspective. You should expect things, or at least you will.

At times, this fear can even blank out your mind. I like an attitude where you push through despite the fear. Many situations call for it. I don't naturally possess the kind of confidence that's common in the West. I question a lot of things, I'm curious, and my reactions aren't quick. I enjoy slowly and meticulously pondering things before responding.

(yeah … My interface's response time is as slow as Ethereum without Layer 2 🤦‍♀️.

Recently, I prepared a presentation for a beloved Hypertext Research group at SG(we share the “narrow“ focus amazingly), talking about my Wordblock idea. I had prepared my presentation slides for a long time, and on the day of the presentation, I was desperately referencing Marvin Minsky's 'The Society of Mind,' attempting to integrate some concepts into my presentation. Even if to others, they seemed unrelated and illogical, I was passionately committed to it.

I was very nervous, and I didn't even know the source of this nervousness. It's like trying to open up your inner world, thinking you've opened a well-manicured garden, but others see an unordered jungle perhaps. I found that I genuinely couldn't answer everyone's questions rooted in their specific knowledge domains, but the questions themselves were more important.

For example, Prof. Simon asked why it's like Xanadu: "Since you are proposing a system that resembles Xanadu in many aspects, what reason do you have to believe that this is today feasible at scale? Technological advancement? Societal advancement? Greater societal need? Why should it work now when it failed then?"

Others raised questions about citations, links, and reference loops, as well as how to record and verify data ownership and reference relationships.

Although I'm better at expressing my own motivations and the imaginations that are mixed into the design of this little thing, I don't know how to answer these questions. I can only honestly say I don't know, but allow me to understand, read, and learn, and then provide my understanding and comments. I say I hope everyone can help me, whether it's about the application scenarios of this little thing or if it sparks any interest in any aspect.

Of course, this jungle leaves most people at a loss. But I also received a voice from Jérémy:

"I think the benefit of micropublications goes beyond 'succinct communication of findings.' It is more about directly getting to the point (whether this is succinct or not) with potential additional information (e.g., related work, proof) being present as linked (micro)publications. Since the reading of articles is not linear, this new way might exploit better hypermedia to create scientific publications. Of course, this new way of writing may create some additional challenges, but articles are not written linearly anyways, so it might make more sense for writers (and readers who read non-linearly)."

In my continuing conversation with Jérémy, I began to see a clearer and tighter connection, a shared expectation and effort between the Web 2.0 and Web 3.0 communities for a 'better internet' itself. In the iterations of the World Wide Web led by Sir Tim Berners-Lee, many concepts, such as Solid and GGG, are being explored now in the blockchain (or rather, we are continuing the profound thoughts that already existed in the World Wide Web), but few discuss the continuity and commonality of these ideas, what is being repeatedly re-used, but only with a more embellished narrative. An essential question is: can these two webs merge?

For example, in a concept of Web 3.0 proposed by Sir Tim in 2007, the social graph is defined as composed of each web space. After the transition from the publisher's view to the viewer's view in Web 2.0 (a transition period from the reader's view to the viewer's view), Web 3.0 will become the viewer's view, meaning:

users will be able to modify existing original content on the Internet, forming a network (Web 3.0 has a similar concept called a 'curation network').

A web space is a personified composition of web resources.

Based on my interactions with Jérémy and Prof. Simon exchanging resources, I drew a matrix in the last ten days. You can easily see that there are already a large number of papers/articles (carried in the form of “blocks” below). Each text block with context contributes a part that gets straight to the point of this issue.

https://whimsical.com/with-jeremy-LrDAc9dCHrwXYMECNUQCAz

You might sense that this is different from the one-way Internet jump link. Instead, it is built upon a decentralized web (abstraction layer) and referred to as micro-publication, I mean for now. Its mission is to establish bidirectional links on the internet. On one side, it connects to the original publications on the web, and on the other, it forms the fundamental units of the viewer's view.

micro-publication is the basic unit of the viewer’s view.

In other words, you can publish content as you do now (on personal blogs, academic platforms, writing platforms, etc.). When someone wants to take any action based on your content (quote, rewrite with an alias), they can create this web space to establish a link with your original content.

My article's theme is: How to turn the one-way links of the World Wide Web into two-way links? Publications are stored on the web, viewers make a modified version and store it on the abstraction layer. What's the abstraction layer made of? A decentralized web stack that integrates Web 2.0 and Web 3.0 components. Every word block abstracts the 'get-to-the-point' from the publication (articles, academic writings, etc.) and is linked through hashtags.

One question: how to confirm these bidirectional relationships?

(to be continued.

Wordblock is a block based note-taking tool with ownership control.

Wordblock extension

You can clip any text from any website, add tags, comments, and save the location (URL) as metadata.

You can write block-based memos, add tags and extra information(e.g. the real location: 📍🚶‍♀️38W,4S). For example, when you pick up a book in an antique bookstore, and you want to "steal" some words from it before you take it home.

The text is generated under the user's own DID signature, while data is stored on a distributed content storage network--IPFS to enable public search.

Wordblock.xyz - a global text block search engine

All the word blocks created will be pooled in one place for public search, or later public reference.

For example, you can process public block searches via tags, keywords, or user DID

Compared with local storage note-taking tools like obsidian or central storage Roamresearch, the difference is in data storage in order to achieve public block search.

Use cases

A social graph

After collecting those text blocks, it is easy to form P2P social relations*(e.g. walletChat)* via tag matching.

A decentralized multi-lingual literature library

The inspiration is when I visited an ancient Malatestiana library (the 1st public library in human history) in Cesena, Italy.

Each codex is tied to the desk with a root ironed chain, so that precise objects could not be stolen, it is now a symbol of the arts of literature, copies include 20 years italian, french and german s copy by using humane scripture called "literature antiqua", text includes classical latin authors, churchenergy , greek…

It's like a warm salute from afar history that tells how humans keep words and knowledge. In the context of today, we stand a chance to "chain text" differently and interestingly with technology.

Today, we care less about "preventing books from being stolen" but about "HOW things are connected", so the desire to build a distributed literature dictionary(library) where people from anywhere in the world could contribute their input even in different languages.

For example, after I read the Chinese version of《Mr.Gywn》, I para-read the original Italian version, then build the connections in between. It is one of the very fun techniques for acquiring a second language through intertextuality, while also a way to let the machine know how people from different linguistic computing systems tell the SAME story.

Read more: https://mirror.xyz/0x65a0Af703047dfDd270361659d02f4f0E8547202/YmbixVzwGJ0l46X6PDtVBbYY64Jd-SfRSD8usBZUzno

A data graph

we are in the early stage of defining data as an asset with maturing infrastructure like IPFS, Ceramic network, dataverseOS etc, the excitement of web3 is the data as streams that can be visualized and later used by individuals. We could first collect those fragmented attentions 👀 with ease, and then browse/ the data, or re-organize(reference) with personal signature.

Tech

Ceramic Streams

Wordblock uses Ceramic Stream to store the JSON metadata of a block. Metadata includes BlockID, Creation Date, Source URL, Author, Tags, Type, and Content. The DID of the user submitting the block to the Ceramic network is also bound to the metadata. The Wordblock data model can be shared between anplications on the Ceramic network.

web3.storage

Wordblock uses web3.storage for data backup and permanent storage. Our backend API can get all the blocks created by the given user DID to implement a global search on the wordblock.xyz page.

Reference

[0] The background story of word block (one of ;)

https://mirror.xyz/0x65a0Af703047dfDd270361659d02f4f0E8547202/YmbixVzwGJ0l46X6PDtVBbYY64Jd-SfRSD8usBZUzno

[1] wordblock github

https://github.com/fat-garage/wordblock

[2]wordblock gitcoin

[3]wordblock on IPFS grant

https://github.com/ipfs/devgrants/issues/278

[4] wordblock idea initiation

https://mirror.xyz/conaw.eth/hBj9GSkYzLpQM524VBVhjO8C5-KQFw9UmfrYkerlvZE

Wordblock 插件

你可以从任何网站上剪辑任何文本，便捷的添加标签、评论，连同位置（URL）自动保存为metadata。

你可以写基于块的便签，为其添加标签和额外的信息（比如你的真实的位置：📍38W,4S）。例如，当你在一家古董书店拿起一本书，你想在带回家之前从里面 "偷"走一些文字。

这些文字块是用户在自己的DID签名下生成的，而数据则存储在分布式内容存储网络--IPFS上，以实现公共块搜索。

Wordblock.xyz ：一个全局文本块搜索引擎

所有创建的文字块将被汇集到一个地方，用于公开搜索，或以后的公开引用（reference）。

例如，你可以通过搜 #标签、关键词/ 或用户DID 来进行全局搜索。

与本地存储的笔记工具如obsidian或中心化存储的Roamresearch相比，区别在于数据存储和身份的创建。Word block要实现的是公共块搜索。

使用案例

一个社交图谱

通过 #标签 匹配容易地启动P2P社交关系，例如未来可以结合 wallet_Chat 进行点对点聊天。

一个去中心化的多语言文学“图书馆”

这个灵感来自于我参观了意大利切塞纳（Cesena）的一个古老图书馆 *Malatestiana，*据说是世界上建立的第一个公共图书馆。

每个手抄本都用根铁链绑在桌子上，这样精确的物件就不会被偷走，它现在是文学艺术的象征，副本包括20年的意大利文、法文和德文副本，使用*人道的经文，称为 "Literature antiqua"，文本包括古典拉丁文作者、教会、希腊文。*

这可能是来自远方的历史的神奇互动，告诉我们人类是如何保存文字和知识的。在今天的背景下，我们有机会以不同的方式和有趣的技术来 "连锁文本"。

今天，我们关心的不是 "防止书籍被盗"，而是*”事物如何被连接 ”*。

因此，我们希望建立一个分布式的文学词典（图书馆），让世界上任何地方的人都可以贡献他们的输入，即使是不同的语言。

例如，在我读完《Mr Gwyn》的中文版后，我又去读了意大利文的原版，这种互文性天然的形成。如何通过互文性习得第二语言，是工具结合技术非常有趣的尝试，同时也是让机器知道来自不同语言计算系统的人是如何讲述相同故事的一种方式。

阅读更多: https://mirror.xyz/0x65a0Af703047dfDd270361659d02f4f0E8547202/YmbixVzwGJ0l46X6PDtVBbYY64Jd-SfRSD8usBZUzno

一个数据图（graph）

我们正处于将数据定义为资产的早期阶段，随着IPFS、Ceramic network、dataverseOS等基础设施的逐渐成熟，web3的兴奋点在于将数据作为流 这个叙事，即数据能被可视化，随后被个人支配并使用。Word block要做的是，我们可以轻松地收集那些零散在各个平台上的注意力，同时拥有这些数据/线索，以方便任意时刻重新组织（编辑、引用等等）。

建立在DID 下的行为将不仅是一种用户操作，而且是一种数据行为，可以被区块链反向引用，进而建立经济性活动或链接。

技术

Ceramic 数据流

Wordblock使用Ceramic Stream来存储一个区块的JSON元数据。元数据包括BlockID、创建日期、源URL、作者、标签、类型和内容。向Ceramic网络提交块的用户的DID也被绑定到元数据上。Wordblock数据模型可以在Ceramic网络上的任何应用程序之间共享。

web3.存储

Wordblock使用web3.storage进行数据备份和永久存储。我们的后台API可以获得所有由给定用户DID创建的区块，以实现在wordblock.xyz页面上的全局搜索。

英文版

https://mirror.xyz/conaw.eth/K0PeM2YtbEYtmnKIS76hML-YVfC_MzX59FxyY6IuuHo

Reference

[1] wordblock github

https://github.com/fat-garage/wordblock

[2]wordblock gitcoin

[3]wordblock on IPFS grant

https://github.com/ipfs/devgrants/issues/278

螺杆的每个裤子口袋都有硬币，不知道是有心还是无意，但总能至少找到一枚

有的时候他出去跑步的途中 会遇到一些被扔掉的刮得乱七八糟的刮刮卡 他每次注意到都会捡起来 值得一提的是 这东西并不显眼 鬼知道他是怎么注意到的 就像他妈妈从小讲的那样 饭菜里的头发丝儿别人都看不见 就他眼尖 多细小一根儿都能挑出来

10分的小硬币是最好用的 螺杆喜欢把数字刮干干净 然后检查好几遍 期待着哪个幸运的倒霉蛋当时走神 没注意到那出现了三次的 10mille（1000万）或者50块✖️3也不赖

这个游戏刮刮卡有两种赢法 一个看下面三排大数字和最上面那排的重合度 另一个看每个大数字下面附带的额外数字是不是出现了三次（比如像10Mille出现三次就是中奖了） 而人们总是专注于第一种

奇怪的是 他只热衷于捡 却从来没买过这玩意儿 虽然只是3、5、10块的事 但他每回经过小杂货店的时候 都好像有很多考虑似的

“我并不是舍不得花钱，只是...要是我真中奖了把运气一下子用完了怎么办，这可得慢慢分配”

这样好像能解释他为什么要捡别人刮剩下的了

因为那是别人的运气

他唯一买过一次的圣诞彩票，还因为这种“恐惧”立马转手送给了车站土耳其肉卷店门口的流浪汉。

后来流浪汉还在那里。

2

可能每个作家都幻想自己是某个作家构造出来的人物 以缓解那种浪费时间的罪恶感和终将一事无成的恐慌

“这不是一种选择 而是一种命中注定 因为他们除了遣词造句... 噢得了吧，只是浪费时间的高手”

“或许有的人的命运就是讲故事” 这种跳跃的心理是日常routine。

这时 螺杆跑步停了下来 因为他脑子里冒出了两个东西

这世界上的第一本书和

第一个作者 是谁？

这个疑问源于一种一直以来的感觉 这世界的作品 可能都是一个人写的 其他人都是这人用来伪装混淆的笔名 又或是 那个人超级富有雇了好多人把他写的东西到处搬来搬去 因为他感觉自己好像几辈子都在看书 也深谙几种语言的翻译转换逻辑 越看越发现

好多东西像是来回抄的 用看似变花样的词语和语言 说着一样的事情 于是

他下决心找到这个人

3

“你才读了多少书 就这样说” 螺杆把这个想法告诉好友凉鞋的时候，凉鞋嗤之以鼻。

“但没有人读过这世界上所有的书，万事总有个开始吧”

“是的，那你要怎么查？”

“我谷歌一下”

Pools

There are two pools of blocks:

1. the private pool of blocks, meaning you can only see what you have created.——the grey stripe

2. the public pool of blocks(the blockverse), meaning you can see blocks generated by anyone who use the extension and choose “public“. ——the orange stripe

Buttons

In the public pool, there are 3 buttons:

1. Archive

2. Irrigation（means “Like”)

3. Go to the context——URL web page

Search

You can search blocks by wallet address/ENS names, or by #tags

@vanishk.eth

https://twitter.com/Kaspar_Chen/status/1539616616526528513

@mask network

https://twitter.com/JESSCATE93/status/1539498726573895681

From Media🍱

@Foresightnews @iambabywhale.eth

https://foresightnews.pro/article/detail/6527#:~:text=%E6%8E%A8%E8%8D%90%E9%A1%B9%E7%9B%AE%EF%BC%9AThe,the%2Dword%2Dblock

@WeChat channel

From Friends👾

@MSP

Web3 could bring a new experience.

Block ID: Each piece of content carries a relationship

Bi-directional linking capability

The introduction of bi-directional links is for bridging web2 context with web 3 id-ized content. The concern is that the original and majority of the context is still on web2 pages(URLs), and the content we abstract from that context could have a parallel “identity(ID)“ enabled by web3 infrastructure.

Later, we could do things like block mapping and text computation on IPFS(probably not on IPFS, but some middle layer for semantic computing), while the context are addressable on web 2.

Treat every piece of content as an object, we could have a fresher and more interactive relationship with the content. Imagine you are building “the block website“!

This 👆thinking stems from a missing experience. For a long time I suffered from repeatedly OCR (optical character recognition), PDF highlights, web pages (URLs) clips,…etc from separate medium. And then to make it “in one place“, i become a tester & expert of TOOLs, i tried so many note-taking tools, plug-ins, and their corresponding plug-ins…and

You’d better not change your software, lose the PDF or anything, otherwise things are too tough to think about🤯. Mediums never exist as "objects" with “us(objects)”, we are trying to organize them until realizing they are dandelion.

Simply put, can I do it once (cOpy, highlight, clip, cut,generate) and these things can follow "me" ?

So the question is, how do I operate on the content itself, and establish a relationship with it?

A product idea_ ((W block))

Proof of text via wallet signature.

From the beginning, whatever text you generate is under your signature.

Text generation with ease

Whenever you generate text, either note-taking or clipping content with its context, you can easily keep that in word blocks.

Easy to reference on any editor

Take mirror for example,

1. You can pull up the plugin at any time to see the previously saved blocks

2. Copy the block ID (with bidirectional links)

3. For readers, jump directly to the specific part of the web page (like scroll-to-text)

Search

As you see, as we accumulate a block network of text, a semantic network needs to be constructed for block search.

Thoughts for the Block Network 🕸️ :

• Create a set of intermediaries (e.g., a group of people taking notes)

• Store the semantic relations of blocks on a network of these intermediaries

• spend up the retrieval of blocks (either yours or others' can be queried) keywords/categories/time/factors

Last but not least

Like i mentioned in the post, word block aims to explore a new experience between content and individuals, but in essence, it is still the relationship between people, between like-minded people.

1st, where we stand is the signature economy instead of an Information silos, we connect via open creation under MY signature.

2nd, "social" has a vivid landscape in web3, where people could be matched to each other for various purposes. if word block is a literature system, there will be “drivers“ helping match "words" or thoughts.

when people want to find similar "blocks", refer to each other's block, and be united by block curation, that's the kind of "social" via block-based knowledge graph!

3rd, I want to connect the web2 context(the existing literature history) with the web3 capability of composability, computability, and interoperability (make a parallel new world of things). I summarize as:

Dismantle web page, present in blocks, match via blocks.

🐳 Thoughts from Friends

@eth-worm

Recommendation:

@takenstheorem

@sidecode

More feedbacks…🏃‍♀️

https://mirror.xyz/conaw.eth/FaFnmmgWY52lJcXL1f_a-nQXxuT3Y46q45ASBoIsyPU