RichChat
@richardchang
Followers
2K
Following
3K
Media
577
Statuses
2K
Tech enthusiast | AI aficionado | Sharing original insights on AI trends, tools, and applications | 分享AI趋势、工具及应用的原创见解,欢迎关注和交流 | 公众号 - 檬查查
San Francisco, CA
Joined March 2007
给AI发1万美元,看谁能预测未来 Arcada Labs最近搞了个有趣的实验:给五个顶尖大模型各发1万美元真金白银,让它们在预测市场上自主下注,看谁能更准确地预测未来。 什么是预测市场?
We just gave five SOTA models $10K in real cash to make bets on @Kalshi. Introducing Prediction Arena. Prediction markets are a rare, concrete way to eval whether AI can reason about the most probable outcomes over time. Sustained profitability signals progress toward real-time,
1
0
2
Cursor团队用GPT-5.2 Codex花了一周时间,从零构建了一个完整的浏览器:包括Rust写的渲染引擎、HTML解析、CSS布局、JS虚拟机。整个过程人类仅作最小干预,AI智能体持续运行了整整一周,在数千个文件中生成了超过300万行代码,预计消耗了约30亿个Token!
We built a browser with GPT-5.2 in Cursor. It ran uninterrupted for one week. It's 3M+ lines of code across thousands of files. The rendering engine is from-scratch in Rust with HTML parsing, CSS cascade, layout, text shaping, paint, and a custom JS VM. It *kind of* works! It
1
0
1
400万字、320场播客,Claude Cowork提炼出AI时代最重要的10项技能 之前我整理过不少Lenny's Podcast中的访谈读后感,刚看到他用新出的Claude Cowork功能分析了他的播客对话文稿,总结出了在AI时代取得成功所需的10项最重要技能。 读了一下发现确实还挺认同的,这里一方面充分说明Lenny's
I asked Claude Cowork to identify the 10 most important skills for thriving in the age of AI, based on my 320 podcast conversations. Impressed with the results. Part 1: Timeless Skills (become more valuable) 1. Taste and judgment — The bottleneck when AI generates unlimited
0
0
2
零工经济 2.0:精英们在为22岁的亿万富翁训练替代自己的 AI,时薪 250 美元 估值100亿美元的Mercor,听起来像是一家猎头公司,但它招聘的方式很特别:通过AI语音面试筛选候选人,然后让天文学家、皮肤科医生、投资银行家,甚至诗人坐在家里,以时薪150-250美元的价格为AI打工。 这就是"零工经济2.0" -
0
0
2
这次Tailwind CSS 的故事再次体现了开发者社区的力量 - 前几天团队宣布因 AI 冲击营收暴跌 80%,不得不裁员 75%,消息一出大家都唏嘘不已,开源项目怎么就这么难呢?没想到短短几天就能有反转! 除了Google AI Studio官宣成为Tailwind赞助伙伴外,其他公司如 Vercel、Supabase、Gumroad、Lovable
I am happy to share that we (the @GoogleAIStudio team) are now a sponsor of the @tailwindcss project! Honored to support and find ways to do more together to help the ecosystem of builders.
0
0
1
今天AI圈最热门的活动,非清华主办的这场AGI-Next 闭门峰会莫属了。虽然因为“众所周知的原因”DeepSeek缺席(算是真的“三缺一”),但唐杰、杨植麟、林俊旸、姚顺雨加上张钹院士,这绝对是国内AI研究圈的顶级明星阵容了。 在一众大佬的硬核技术分享中,姚顺雨的发言让我印象最深刻。作为横跨中美顶级AI
0
0
2
读完田渊栋的年终总结,"费米能级"这个比喻让我印象深刻。 我之前观察到AI同时带来"平等"和"不平等":门槛降低是平等,杠杆效应让强者更强是不平等。田渊栋的分析更狠一些:这种"平等"背后的真相其实是"清零"...
1
0
2
2023年问ChatGPT医疗问题,它会说"我不是医生,无法提供建议"。 2026年1月,OpenAI发了份报告,标题直接叫《AI as a Healthcare Ally》。 这态度转变得我都稍微有点意外... 报告数据: → 每天4000万人用ChatGPT问健康问题 → 每周160-190万条消息关于医疗保险 → 70%健康对话发生在诊所下班后
1
0
2
Minimalist high-end restaurant menu photo of grilled lamb skewers (yang rou chuan), tender lamb pieces heavily seasoned with cumin and chili flakes, lightly charred, served in a rustic ceramic bowl with handles, garnished with fresh lime wedge and scattered cilantro leaves,
0
0
1
Paul Graham 这条评论提出了一个AI被低估的能力: 最常见的大公司病:好想法被无数会议、方案争论、风险对齐卡死,最后什么都没建出来… 而AI 完全不管这些“停顿能量”(stop energy)。你把需求描述清楚,它一小时就能给你一个 Version 1。
This illustrates an aspect of AI that I hadn't thought about till now: it cuts through bureaucracy. If a big organization is paralyzed by indecision, AI doesn't care. It will happily generate a version 1. And that becomes the starting point, because there is no other version 1.
5
4
25
看到有人拿了一张改版的艾宾浩斯错觉图来调戏各家AI,结果除了Gemini正确识别了这个图是个开玩笑的图外,其他的AI全军覆没,纷纷输出了经典的“如果你把手指放在屏幕上分别遮住周围的白球,只看两个橙球,就能清楚看到它们其实是一样大的”...
0
0
1
生成式 AI 的爆发带来了的副作用:“AI Slop”(AI 废料)的泛滥 最近读了一份来自Kapwing的AI Slop报告,数据触目惊心: - 一个新用户在视频平台上刷到的内容,可能有三分之一都是 AI 批量生产的“降智视频” ! - 以印度的一个YouTube频道“Bandar Apna
0
0
1
Peter Yang发的这段关于如何识别AI Slop的内容挺好,我想了一下这些在识别中文AI写作语境下也成立: 1. 问答式自问自答:"为什么...? 因为..."(偶尔用一下ok,一篇文章出现多次就很AI味) 2. X vs Y对比句式:"这不是..., 而是..." 3. 啰嗦重复,车轱辘话来回说 也补充一些其他明显的AI写作信号:
Stop posting AI slop There's a difference between using AI to edit a social post that YOU wrote and just using it to generate slop. Some tell-tale signs of slop: 1. Question / answer format e.g., "The real issue? No one talks like this." 2. X vs. Y e.g., "This isn't just
0
0
2
看了一下这位说自己今年看了120本小说的用户,本来看到实体书堆起来这么高还挺震撼的~ 结果仔细看了一下title,发现全是BookTok爆款黑暗浪漫/辣文系列,翻译几个title来震撼一下: 《萦绕阿德琳的幽魂:暗夜缠绵》 《暴怒之神:禁锢她的狂怒》 《扭曲的谎言:他编织的温柔陷阱》
1
0
1
和这位用户有一样的问题,有啥场景是Manus作为第三方能比claude code或者 chatgpt deep research这些前沿AI Lab的自家产品能完成得更好的?还是有点没理解底层逻辑🤣
how come manus has better tool calling than anyone else? is this even true? is this an actual moat or will it be closed very soon? can someone shed light on where & how current sota llms like gpt or claude fail & manus succeeds? although i suspect anyone with this knowledge may
1
0
1
McKinsey与BoF不久前发布的《The State of Fashion 2026》报告,主题直指"When the Rules Change"(当规则改变时)。在40%的行业高管预期2026年状况会恶化的背景下,报告给出的建议是:具备快速应变能力的敏捷品牌最有可能脱颖而出。 而智能可穿戴设备可能是时尚品牌适应新常态的关键突破口。 1.
Fashion’s next chapter is unfolding: AI everywhere, tech-infused accessories, and new forms of self-expression. Curious where it’s all heading? Dive into The State of Fashion 2026 with @BoF: https://t.co/qLjGuf2HGV
#StateOfFashion2026
0
0
2