Vesper Gu

1.1K posts

Vesper Gu banner
Vesper Gu

Vesper Gu

@0x_vesper

Cooking @ECUPL_BA|🟩 Ambassador @ZetaChain_CH |或许是不知梦的缘故,流离之人追逐幻影

亚空间航行中… Katılım Aralık 2023
1K Takip Edilen264 Takipçiler
Sabitlenmiş Tweet
Vesper Gu
Vesper Gu@0x_vesper·
《关于我在交易所干法务实习生》 交易所的部门有很多,但我想法务部算是人最少的部门了吧。现存的CMC排名前20的交易所大多是从ICO浪潮开始的前几年成立的。虽然我不曾经历过那段蛮荒岁月,但也没有几家交易所在野蛮生长的头几年会成立法务部。即使到了行业合规化的现在,大部分交易所的法务部也更像装点门面的饰品。 很少见到同龄人会选择法务/合规这个成长赛道。一方面是愿意接触加密货币的法学生少之又少,国内有很多大学链协,但根据我的观测,背景集中于商科/传媒/工科,鲜少见到与我背景相同的。一位本校的学长也是从律所跑去干VC。VC的title和成长性也高于法务/合规。认识一位web3律师团队的高伙也苦于寻找新血,我和他的共同观点是“都来web3了,当然不会选择做律师、法务这种成长周期长又起薪中庸的职位了。”(此处指应届生) 至于我为什么从交易所listing回到律所,再回到交易所法务呢? 其实listing是个不错的岗位,我也很喜欢。(哪怕在一个非常小的所工作)离职后也和带教保持着良好的关系,经常听到他收到某某所的面试邀请的消息,给的薪资也很nice。但放弃做listing的原因是我认为这是一个资源密集型岗位,能否锁定某个天王项目的listing、能否从部门内斗中存活很关键。我对自己能否成为一个优秀listing lead保持怀疑hh。 回到法务/合规,从招聘群的日常聊天我能发现,市场上需要的优秀法务专家/合规经理与应聘人员是存在不小的差距的。换句话说,人才稀缺。天然适合洗钱的加密货币不断冲击现有的aml制度,只要行业想要走合规的路线,相关的岗位只会越来越多。而与此同时法学精英们还在走律所的路子,没准等到无望升par才会谋转型。除去交易所外,互联网券商、家办、基金也都会有这方面的需求。目前来看我没有找到一个与我同生态位的竞争对手,错位发展也是弯道超车的好方式。 至于我的选择是否正确,我会用自己的职业发展路线去印证。希望三年后能看到一个让我满意的结果。 给我的人生加点杠杆。😝
Vesper Gu tweet media
中文
4
1
20
1.6K
Vesper Gu
Vesper Gu@0x_vesper·
大家对于minimax招聘rwa相关岗位有什么头绪呢
Vesper Gu tweet media
中文
0
0
0
47
Vesper Gu retweetledi
Claude
Claude@claudeai·
Introducing Claude Managed Agents: everything you need to build and deploy agents at scale. It pairs an agent harness tuned for performance with production infrastructure, so you can go from prototype to launch in days. Now in public beta on the Claude Platform.
English
2.1K
6K
57K
21.3M
Vesper Gu retweetledi
kepano
kepano@kepano·
File over app File over app is a philosophy: if you want to create digital artifacts that last, they must be files you can control, in formats that are easy to retrieve and read. Use tools that give you this freedom. File over app is an appeal to tool makers: accept that all software is ephemeral, and give people ownership over their data. In the fullness of time, the files you create are more important than the tools you use to create them. Apps are ephemeral, but your files have a chance to last. The pyramids of Egypt contain hieroglyphs that were chiseled in stone thousands of years ago. The ideas hieroglyphs convey are more important than the type of chisel that was used to carve them. The world is filled with ideas from generations past, transmitted through many mediums, from clay tablets to manuscripts, paintings, sculptures, and tapestries. These artifacts are objects that you can touch, hold, own, store, preserve, and look at. To read something written on paper all you need is eyeballs. Today, we are creating innumerable digital artifacts, but most of these artifacts are out of our control. They are stored on servers, in databases, gated behind an internet connection, and login to a cloud service. Even the files on your hard drive use proprietary formats that make them incompatible with older systems. Paraphrasing something I wrote recently: > If you want your writing to still be readable on a computer from the 2060s or 2160s, it’s important that your notes can be read on a computer from the 1960s. You should want the files you create to be durable, not only for posterity, but also for your future self. You never know when you might want to go back to something you created years or decades ago. Don’t lock your data into a format you can’t retrieve. These days I write using an app I help make called Obsidian (@obsdmd), but it’s a delusion to think it will last forever. The app will eventually become obsolete. It’s the plain text files I create that are designed to last. Who knows if anyone will want to read them besides me, but future me is enough of an audience to make it worthwhile.
kepano tweet mediakepano tweet media
English
188
1.4K
9.7K
3.2M
Vesper Gu retweetledi
Andrej Karpathy
Andrej Karpathy@karpathy·
LLM Knowledge Bases Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating knowledge (stored as markdown and images). The latest LLMs are quite good at it. So: Data ingest: I index source documents (articles, papers, repos, datasets, images, etc.) into a raw/ directory, then I use an LLM to incrementally "compile" a wiki, which is just a collection of .md files in a directory structure. The wiki includes summaries of all the data in raw/, backlinks, and then it categorizes data into concepts, writes articles for them, and links them all. To convert web articles into .md files I like to use the Obsidian Web Clipper extension, and then I also use a hotkey to download all the related images to local so that my LLM can easily reference them. IDE: I use Obsidian as the IDE "frontend" where I can view the raw data, the the compiled wiki, and the derived visualizations. Important to note that the LLM writes and maintains all of the data of the wiki, I rarely touch it directly. I've played with a few Obsidian plugins to render and view data in other ways (e.g. Marp for slides). Q&A: Where things get interesting is that once your wiki is big enough (e.g. mine on some recent research is ~100 articles and ~400K words), you can ask your LLM agent all kinds of complex questions against the wiki, and it will go off, research the answers, etc. I thought I had to reach for fancy RAG, but the LLM has been pretty good about auto-maintaining index files and brief summaries of all the documents and it reads all the important related data fairly easily at this ~small scale. Output: Instead of getting answers in text/terminal, I like to have it render markdown files for me, or slide shows (Marp format), or matplotlib images, all of which I then view again in Obsidian. You can imagine many other visual output formats depending on the query. Often, I end up "filing" the outputs back into the wiki to enhance it for further queries. So my own explorations and queries always "add up" in the knowledge base. Linting: I've run some LLM "health checks" over the wiki to e.g. find inconsistent data, impute missing data (with web searchers), find interesting connections for new article candidates, etc., to incrementally clean up the wiki and enhance its overall data integrity. The LLMs are quite good at suggesting further questions to ask and look into. Extra tools: I find myself developing additional tools to process the data, e.g. I vibe coded a small and naive search engine over the wiki, which I both use directly (in a web ui), but more often I want to hand it off to an LLM via CLI as a tool for larger queries. Further explorations: As the repo grows, the natural desire is to also think about synthetic data generation + finetuning to have your LLM "know" the data in its weights instead of just context windows. TLDR: raw data from a given number of sources is collected, then compiled by an LLM into a .md wiki, then operated on by various CLIs by the LLM to do Q&A and to incrementally enhance the wiki, and all of it viewable in Obsidian. You rarely ever write or edit the wiki manually, it's the domain of the LLM. I think there is room here for an incredible new product instead of a hacky collection of scripts.
English
2.8K
6.7K
56.3K
20M
Vesper Gu retweetledi
moreless
moreless@moreless·
你接到了一通电话。 去北京报到。 中国的国家发改委(NDRC)要见你。 你坐在他们对面。他们对一切了如指掌。 公司的重组。 迁往新加坡的举动。 来自 Benchmark 的 7500 万美元投资。 来自 Meta 的 20 亿美元收购金。 你在北京裁掉的那 80 名员工。 以及你决定不再向中国市场提供的产品。 他们提问。你作答。 会议结束了。 你被禁止出境,无法离开中国。 没有指控。没有逮捕。没有审判。也没有明确的期限。 你可以在国内任何地方旅行。 上海。深圳。成都。你想去哪儿都行。 唯独不能离开这个国家。 你的本名叫肖红。大家习惯叫你“Red”。你今年 32 岁。 这就是你走到这一步的始末。 你在中国中部地区长大。攻读了工程学专业。 你开发过各种应用程序(App)。而且做得非常出色。 随后,你打造了 Manus。 这是一个人工智能代理(AI agent),它不仅仅会“说话”,更能真正地“干活”。 它是一位数字员工,没有国界,无需签证,不分国籍。 短短 8 个月内,营收便突破了 1 亿美元。 然而,你是中国人。 而美国刚刚颁布禁令,禁止美国投资者向中国的 AI 领域注资。 一夜之间,你的公司便失去了投资价值。 于是,你做出了全球所有聪明创业者都会做出的选择。 你在新加坡注册成立了公司主体。 美国人通常会选择特拉华州。 欧洲人通常会选择爱尔兰。 印度人通常会选择迪拜。 这并非犯罪,而是一种商业策略。 你的律师建议你这么做。你的投资人也建议你这么做。所有人都建议你这么做。 你将整个公司进行了迁移。新加坡、东京、旧金山…… 你关闭了北京的办公室。 裁掉了在中国境内的 80 名员工。 并停止向中国市场提供你的产品。 彻底的切割。如今,这已是一家全球化的公司。 Benchmark 为你开出了一张 7500 万美元的支票。 随后,Meta 打来了电话。 20 亿美元。 全资收购。并授予你副总裁(VP)的职位。 你所打造的 AI 技术将被整合进 Facebook、Instagram 和 WhatsApp——触达数十亿用户。 你年仅 32 岁,却刚刚缔造了历史上规模最大的一笔“中国背景 AI 公司被美国巨头收购”的交易。 你的母亲为你感到骄傲。 你飞回了中国。为了探望家人,也为了彻底告别过去的生活。 你没有丝毫犹豫。 那是家。你这一生,一直在回家的路上。 随后,你接到了那通电话。 而此刻,你正坐在北京,与国家发改委(NDRC)的官员们隔桌相望;他们正告知你:你不能离开。 北京传达的信息简单明了: 你生于斯。 你在此创业立业。 你在此求学深造。 代码始于此。 知识产权亦源于此。 一个新加坡的住址,并不能让你成为新加坡人。 一家开曼群岛的控股公司,并不能让你变成无国籍者。 一张印着 Meta 标志的名片,并不能让你摇身变为美国人。 他们将此斥之为“新加坡洗白”(Singapore washing)。 而你,就这样沦为了那个“典型”。 美国曾要求你离开中国。 而中国却告诉你:你从未真正离开过。 两个超级大国。两套截然不同的规则。 这两套规则同时作用于你。而无论哪一方,都不曾征询过你的意愿。 此刻,你的 AI 产品已在全球 50 个国家上线运行。 无需护照,无需签证,更不受任何限制。 它正服务着数百万计的用户,而你却只能枯坐于此。 你亲手打造出了那个能够抵达地球上任何角落的造物。 然而,那个唯一无法自由通行的,却恰恰是你自己。
George Pu@TheGeorgePu

You get a phone call. Report to Beijing. China's NDRC wants to see you. You sit down across from them. They know everything. The restructuring. The Singapore move. The $75 million from Benchmark. The $2 billion from Meta. The 80 employees you laid off in Beijing. The product you made unavailable in China. They ask questions. You answer. The meeting ends. You are exit-banned from China. No charges. No arrest. No trial. No timeline. You can travel anywhere inside the country. Shanghai. Shenzhen. Chengdu. Wherever you want. You just can't leave. Your name is Xiao Hong. You go by Red. You're 32. This is how you got here. You grew up in central China. Studied engineering. Built apps. You were good at it. Then you built Manus. An AI agent that doesn't just talk - it works. A digital employee with no borders, no visa, no nationality. $100 million in revenue in 8 months. But you're Chinese. And America just banned Americans from investing in Chinese AI. Overnight, your company is un-investable. So you do what every smart founder on earth does. You incorporate in Singapore. Americans do this with Delaware. Europeans do this with Ireland. Indians do this with Dubai. It's not a crime. It's a strategy. Your lawyers told you to do it. Your investors told you to do it. Everyone told you to do it. You move the whole company. Singapore. Tokyo. San Francisco. You shut down the Beijing office. Lay off 80 people in China. Make your product unavailable in the Chinese market. Clean break. Global company now. Benchmark writes you a $75 million check. Then Meta calls. $2 billion. Full acquisition. VP title. Your AI goes into Facebook. Instagram. WhatsApp. Billions of users. You're 32 and you just built the biggest Chinese-to-American AI exit in history. Your mom is proud of you. You fly back to China. To see family. To close out the old life. You don't think twice. It's home. You've been going home your whole life. Then you get the call. And now you're sitting across from the NDRC in Beijing and they're telling you that you can't leave. Beijing's message is simple. You were born here. You built this here. You learned here. The code started here. The IP started here. A Singapore address doesn't make you Singaporean. A Cayman holding company doesn't make you stateless. A Meta business card doesn't make you American. They're calling it 'Singapore washing.' You just became the example. The US told you to leave China. China told you that you never left. Two superpowers. Two sets of rules. Both applied to you. Neither asked. Your AI is live in 50 countries right now. No passport. No visa. No restrictions. Serving millions of people while you sit here. You built the thing that goes anywhere on earth. You're the one who can't.

中文
177
314
2.1K
827.5K
Vesper Gu retweetledi
Anthropic
Anthropic@AnthropicAI·
New on the Engineering Blog: How we designed Claude Code auto mode. Many Claude Code users let Claude work without permission prompts. Auto mode is a safer middle ground: we built and tested classifiers that make approval decisions instead. Read more: anthropic.com/engineering/cl…
English
408
591
4.1K
1.6M
Vesper Gu
Vesper Gu@0x_vesper·
快逃快逃,ai才是真正的破坏式创新行业,crypto在亚洲已经没有大蛋糕分了😩 还好我没有全部all in web3,现在还能努力读研转行😅
Yuyue@yuyue_chris

逃离币圈的高素质年轻人们:高薪伪装下的 “次要劳动力市场” 今天跟一个 Crypto 卡牌项目的员工聊,发现几周前他居然已经离职加入 AI 行业了。身边有不少相似的现象,这也是我最近观察到的一个很明显的趋势:不仅仅是那个 Multicoin 联创离开了,众多高素质、顶尖背景的同龄人才正在快速从 Crypto 行业流失 我并不认为这能简单归结为 “币价跌了”、“熊市难熬”,比起周期,我从我的视角来看,真正逼退这些聪明的同龄人的,是这个行业在职业生涯规划上极其致命的结构性缺陷,上升通道的彻底锁死 哪怕他们熬过了币圈的裁员,也不会留在币圈的原因可以用一句话概括:炒币自由不了,那日子总还要过吧 两个问题引出第三个问题: 1. 币圈的流动性来源于哪里?二级市场 2. 二级市场发生在哪里?交易所 3. 行业需要优质的年轻的人才加入吗? 第三个问题,应该无需长篇大论论证。至少仅仅从 GitHub 数据来看,自 2025 年初以来,加密项目周度代码提交量约从 85 万降至 21 万,活跃开发者降至约 4600 人 所以可以很明显发现,在现阶段的币圈,除了头部交易所,几乎不存在真正意义上的职场阶梯。但即便是把青春奉献给交易所,能熬过 2-3 年的人也寥寥无几。在这个阶段入场,既拿不到早期员工那种具有爆发力的期权,也早就失去了靠行业红利直接 “买断青春”、实现财富自由的机会(当然,不考虑某些老鼠仓小编和某位现已离职的某链运营哈) 用经济学家 Doeringer 和 Piore 提出的二元劳动力市场理论来拆解,一切就变得无比清晰:Crypto 行业的就业基本盘,本质上就是一个被高薪粉饰的【次要劳动力市场】 在传统的主要劳动力市场(如顶尖金融机构、科技巨头或成熟的咨询公司)里,雇主和员工之间存在一份长期的隐含合同,提供的不只是体面的薪水和社会地位,最核心的是知识与技能的复利累积。在那种环境里,职业发展呈现出一种螺旋上升的态势:个人资本(能力、人脉、认知)随着时间不断增值。一个行业专家,可以凭借过去几年积淀的核心技能,横向跨界跳槽到咨询行业看特定赛道,或者进入大企业的战投部 但在 Crypto 这个次要劳动力市场里,哪怕工资未必低,你依然要面对工作极度不稳定、社会地位边缘化,以及晋升机制的极度匮乏 这就引出了那个最致命的问题:从币圈出去,你还能跨界到哪儿?答案是残酷的:无路可退。 在 Crypto 行业积累的所谓 “专业技能”,无论是找 KOL 进行 TGE 的宣发、写智能合约、还是推特运营,或者扫链炒 meme,在传统商业世界里几乎毫无复用价值。技能的不可迁移性让币圈的从业经历更像是一道职场履历上的断层 在传统就业市场尚未完全被 AI 吞噬的今天,顶级年轻人才必然会算一笔长远发展的账。当暴富预期被抹平,剩下的只有无法复用的屠龙技和随时归零的职业履历时,高素质人群的加速离场是一种极其理性的止损 在这个高波动的黑暗森林里,如果不能蜕变成一个自负盈亏、绝对理性的超级个体(独立交易员或投资人),仅仅想作为一个打工人来出卖时间,那这无疑是对自己青春最大的做空,这也是那些优秀的同龄人不愿继续在 Crypto 从业的核心原因 行业领导者更应该重视人才流失的问题,与其选择什么是对的,不如先排除错误答案

中文
2
0
1
98
Vesper Gu retweetledi
Claude
Claude@claudeai·
You can now enable Claude to use your computer to complete tasks. It opens your apps, navigates your browser, fills in spreadsheets—anything you'd do sitting at your desk. Research preview in Claude Cowork and Claude Code, macOS only.
English
5K
14.3K
138.9K
77.6M
Vesper Gu retweetledi
Anthropic
Anthropic@AnthropicAI·
We invited Claude users to share how they use AI, what they dream it could make possible, and what they fear it might do. Nearly 81,000 people responded in one week—the largest qualitative study of its kind. Read more: anthropic.com/features/81k-i…
English
818
979
6.8K
2.9M
Vesper Gu
Vesper Gu@0x_vesper·
solana在上海的活动怎么突然取消了呢,有没有知情人士讲讲
Vesper Gu tweet media
ME News@MetaEraCN

Solana携手阿里巴巴打造中国最大区块链Builder基地 Solana @Solana_zh 上海Builder Station将于3月20日开幕,这是中国首个Solana开发者孵化中心,由上海市闵行区政府、Solana基金会及阿里巴巴虹桥中心联合打造,提供技术交流与项目支持空间。 活动亮点包括上海市政府领导致辞、多家Solana生态项目参与现场分享,并设有SafePal联名硬件钱包等抽奖环节。

中文
1
0
0
163
Vesper Gu retweetledi
Kimi.ai
Kimi.ai@Kimi_Moonshot·
Introducing 𝑨𝒕𝒕𝒆𝒏𝒕𝒊𝒐𝒏 𝑹𝒆𝒔𝒊𝒅𝒖𝒂𝒍𝒔: Rethinking depth-wise aggregation. Residual connections have long relied on fixed, uniform accumulation. Inspired by the duality of time and depth, we introduce Attention Residuals, replacing standard depth-wise recurrence with learned, input-dependent attention over preceding layers. 🔹 Enables networks to selectively retrieve past representations, naturally mitigating dilution and hidden-state growth. 🔹 Introduces Block AttnRes, partitioning layers into compressed blocks to make cross-layer attention practical at scale. 🔹 Serves as an efficient drop-in replacement, demonstrating a 1.25x compute advantage with negligible (<2%) inference latency overhead. 🔹 Validated on the Kimi Linear architecture (48B total, 3B activated parameters), delivering consistent downstream performance gains. 🔗Full report: github.com/MoonshotAI/Att…
Kimi.ai tweet media
English
334
2K
13.5K
5M
Vesper Gu retweetledi
0xMarioNawfal
0xMarioNawfal@RoundtableSpace·
Life without crypto.
English
120
91
869
191.5K