Aiden Frontier

AI Entrepreneur / AI Educator / Ironman Fan

I'm AI "mechanics", I test and unlock its full potential

Agents, models, physical AI, infrastructure.

[Aiden Frontier @lvlstx]

AI is very intelligent now. But how useful is it actually?

The real test isn’t how smart it sounds - It’s what happens after you clearly tell it what needs to be done.

Does it actually get done? Or do you still end up fixing, prompting, checking, and finishing it yourself?

I’m documenting the gap between impressive intelligence and real usefulness — with actual results, honest insights, and unfiltered reviews.

What task have you given AI that it completely failed… or surprisingly finished?

[Aiden Frontier @lvlstx]

@eyi_feryyahooco @Teamily_AI It prioritizes live ones

[Aiden Frontier @lvlstx]

I Let AI Market Itself.

What could possibly go wrong? 💀

I asked the team agents at @Teamily_AI to plan a launch campaign for Teamily AI itself. I was curious what would happen if I gave the brand strategist, market researcher, and writer the same brief and let them work it out in one group chat.

The brand agent wanted to spend 12 weeks sharing ads the team had rejected. One example was “Teamily AI is the future of AI-powered collaboration. 🚀”

I’d happily watch someone cross that out. But the market research agent spotted the gap in the pitch. I could read the whole series, enjoy the jokes, and still have no idea what Teamily could do for me.

Its suggestion was to let readers bring their own draft. Pick a line from a post you’re about to publish and see what the team does with it. What would the writer change? Which claim would the reviewer challenge? Would you keep your original?

The writer picked that up and dropped the 12-week calendar. I liked the revised pitch because I could immediately think of something I’d use it for. I’ve written plenty of posts where I liked the opening but wasn’t sure I could back it up.

Then I got to see the reviewer do exactly that with Teamily’s own copy. It checked a proposed post against the company’s homepage, spotted conflicting wording, and put the two passages side by side for me before I approved anything.

That’s what sold me on Teamily. Usually, when I want a second opinion from AI, I’m the one copying the draft, explaining what the first chat meant, and carrying the feedback back. Here, Market Research tagged the writer directly. By the time the draft got back to me, I could see whose idea had changed it and decide whether I agreed.

[Aiden Frontier @lvlstx]

@Deefui I bet China isn’t as bad as this, is it ?

[紫雲 @Deefui]

说实话，哪个政党赢得选举没吊用，被剥削的永远是工薪阶层

在加州，年薪 10 万美元听着挺不错吧。我帮你算算这笔账：一个单身、没孩子、按标准扣除额报税的打工人，扣完联邦和州所得税，再扣掉社保、联邦医保和州伤残保险，到手大概 7.3 万美元。30%就这么蒸发了。

然后你拿着这笔税后工资去买东西，结账时还得给应税商品交销售税。加州的基础税率是 7.25%，很多地方还要往上加。基本上可以这么说，这么说吧，你每个月开销交的税，都够你买一部iPhone了。

挣钱的时候交税，花钱的时候还得交税。

每次大选，共和党和民主党总有一个要输。可怎么不管谁赢，到头来受伤的还是我们这些牛马。

[Aiden Frontier @lvlstx]

@skypdot These are hard facts and they chose to ignore it

[Sky P. @skypdot]

Tbh, I don’t give two fucks which party wins. Give working people a tax break they can actually feel.

A $100k salary in California sounds pretty good until you do the math. A single worker with no kids taking the standard deduction keeps roughly $73k after federal and state income taxes, Social Security, Medicare, and state disability insurance. That’s before rent or a single grocery run.

Then you spend what’s left—and sales tax gets added to taxable purchases. California’s base rate is 7.25%, with local taxes pushing it higher.

Taxed when you earn. Taxed when you spend.

For every election, Republicans and Democrats, one of them has to lose, but it’s always a loss for working class no matter what.

Fix this fucked up situation.

[Aiden Frontier @lvlstx]

@Deefui Teamily is really a piece of work in this era of AI

[Aiden Frontier @lvlstx]

@lwqconnie great job connie!

[Connie Lin @lwqconnie]

I got to see this up close. 🙌🏻🙌🏻
Worth reading if you're running GTM

Testing 20+ markets is only possible when the system does the heavy lifting

[Sac @Saccc_c]

受大神Karpathy的HackerTube启发，我做了一个「AnliBili」网站，专门集合16年来社媒上被反复安利的B站高质量视频，帮你消除信息噪音！

通过将知乎、小红书上16年来近1000万条讨论进行整理、分类，我筛选出了1132个B站视频，每一个都至少被 3 个不同的人、在 3 个不同的月份推荐过，是时间沉淀下来的绝对精品！

为了收集到足够多的优质视频，我用 Teamily AI 的 team agents 搭建了一个Loop循环。设定好多 Agent 工作流后，Agent团队会不断执行「整理评论 → 筛选视频 → 分类上线」，持续扩充和更新网站数据，而我全程几乎不用任何其他操作

迭代过程中，我和所有 Agent 都在同一个群里聊天交互，共享同一份上下文和目标。我只要根据整体情况进行反馈，比如「增添小红书的讨论数据源」，Agent就会自动重新分工、动态调整，再跑一遍Loop帮我快速实现网站更新

网站传送门放这里了：https://t.co/1s9TXAJhLu
后续我还会继续优化、更新网站，增加更多宝藏视频，让大家在学习、看视频都更方便！

[Kathy Liu @kathyyyyliu]

joined @Teamily_AI  in September and built the entire GTM engine from scratch with @summer_ikiki_.

60 days later we're now one of the fastest-growing startups in silicon valley with 1M traffic and more than 3000% growth rate. 

Personal agents are viral today, but team agents will replace all. We’ve surpassed personal agents products like https://t.co/FWLqkWyxtp, https://t.co/FnYOPCukAR and https://t.co/UXyPsSm0fY. The turning point is already here. 

Read this article ⬇️ if you wanna know how we made it

#teamily #teamagent #agenticinternet #agenticloop

[Aiden Frontier @lvlstx]

@summer_ikiki_ That seems the right decision to do, and very handy agents platform you guys have built, salute 🫡

[紫雲 @Deefui]

在这个信息爆炸的时代，我们看到的，真的就是全部我们能够知道的吗？

作为一名致力于传播和普及 AI 知识的内容创作者，我一直在思考一个问题：

我的粉丝们，是否真正接触到了足够全面、有价值的 AI 信息？

每天，我们都能看到铺天盖地的 AI 新闻。新的模型发布、科技巨头的突破、令人惊叹的产品演示，往往能在短时间内获得数百万次曝光。

但那些稍显冷门，却可能同样重要的技术突破呢？
一项技术的重要性，从来不应该仅仅由它获得了多少流量来决定。一些真正具有潜力的研究、开源项目和创新应用，可能仅仅因为缺乏关注、传播渠道有限，或不符合平台当下的热门叙事，就被淹没在海量信息之中。

最近，我利用 Teamily AI（@Teamily_AI） 的 team agents 和 Agentic Loops，为自己搭建了一套 AI 内容发现、创作与复盘系统:

它不只是一个帮我写帖子的 AI，而是一支由多个专业 Agent 组成、能够共享信息并相互协作的内容团队.

- Trend Researcher 负责搜索那些尚未得到充分关注、却已经显现增长信号的 AI 议题。
- Brand & Social Media Agent 负责评估它们的传播潜力，从中筛选值得向大众介绍的内容 。
- Market Research 则独立验证这些话题的技术意义、市场价值和受众需求，避免单纯为了流量而追逐噱头。
- 随后，Content Creator 综合不同 Agent 的分析，从商业、技术、哲学、实际案例甚至 Meme 等不同角度，构思具有原创性的内容。
- 经过我的审核，Multi-Platform Publisher 负责准备适合不同平台的版本，而 Data Analyst 则根据发布后的真实表现进行复盘，将经验反馈给下一轮选题与创作。

从发现被忽略的信息，到判断其价值，再到创造全新的观点，这正是我搭建这套系统的初衷。

在我看来，一个真正有价值的 AI 内容创作者，应该帮助他的读者不断拓宽获取信息的边界， 将那些容易被流量掩盖的知识与发现，重新带回大众的视野。

[Aiden Frontier @lvlstx]

I Let AI Market Itself.

What could possibly go wrong? 💀

I asked the team agents at @Teamily_AI to plan a launch campaign for Teamily AI itself. I was curious what would happen if I gave the brand strategist, market researcher, and writer the same brief and let them work it out in one group chat.

The brand agent wanted to spend 12 weeks sharing ads the team had rejected. One example was “Teamily AI is the future of AI-powered collaboration. 🚀”

I’d happily watch someone cross that out. But the market research agent spotted the gap in the pitch. I could read the whole series, enjoy the jokes, and still have no idea what Teamily could do for me.

Its suggestion was to let readers bring their own draft. Pick a line from a post you’re about to publish and see what the team does with it. What would the writer change? Which claim would the reviewer challenge? Would you keep your original?

The writer picked that up and dropped the 12-week calendar. I liked the revised pitch because I could immediately think of something I’d use it for. I’ve written plenty of posts where I liked the opening but wasn’t sure I could back it up.

Then I got to see the reviewer do exactly that with Teamily’s own copy. It checked a proposed post against the company’s homepage, spotted conflicting wording, and put the two passages side by side for me before I approved anything.

That’s what sold me on Teamily. Usually, when I want a second opinion from AI, I’m the one copying the draft, explaining what the first chat meant, and carrying the feedback back. Here, Market Research tagged the writer directly. By the time the draft got back to me, I could see whose idea had changed it and decide whether I agreed.

[Aiden Chaoyang He @AidenChaoyangHe]

32× growth in one quarter: From 10K to 320K customers across 100+ countries, with over 100 million messages processed.

We are not building just another personal AI product. Our long-term vision for Teamily AI is to build the central platform of the next-generation Agentic Internet, where AI agents deliver services and perform real work directly, no longer confined to traditional web, desktop, or mobile apps distributed through app stores and search engines. Our recent growth has only strengthened my conviction in this vision.

Agent-as-a-Service (AaaS) will be as accessible as connecting with friends on a social network. Just as Google democratized access to information and services in the web era, Teamily AI aims to democratize access to intelligence and services in the Agentic Internet era, making the right intelligence, people, and services equally accessible to everyone.

In this letter to Teamily AI customers, ecosystem partners, and investors, we share our growth, reflect on our early foresight and what we've built, discuss how we see the evolving AI agent landscape, and outline where we're heading next.

[XIAOAi @silvaniaferreal]

俗话说，三个臭皮匠，顶个诸葛亮。

那么，三个诸葛亮呢？

集体智慧的力量是伟大的。

人类文明最伟大的成就，往往不是某个天才独自完成的，而是无数智慧相互碰撞、协作与积累的结果。

但如今，一个有趣的变量出现了：人工智能。

AI 在某些专业领域已经展现出远超个体人类的能力，甚至能够承担过去需要数十乃至上百人协作完成的部分工作。

那么，当个体智能的能力边界被打破，集体智慧的命题是否还成立？

当然。

如果一个 AI 已经能够以一敌百，那么，当我把多个 AI 放在一起，你又该如何应对？

当我们把协作的主体从人类扩展到 AI，这个古老的命题便有了全新的想象空间。

最近关注到的 Teamily AI（@Teamily_AI），正在尝试将这种理念带入现实。

它所提出的 team agents，不再局限于一个人与一个 AI 助手之间的独立互动，而是让多个专业 AI Agent 与人类共同组成团队，在共享的上下文中协作，并通过 Agentic Loops 让复杂的工作流程持续运转。

试想一下：
将多个在各自专业领域都已登峰造极的高手汇聚在一起。
他们每个人，都拥有足以独当一面、甚至以一敌百的专业能力，在各自的领域将知识、效率与创造力发挥到极致。
而现在，你将这些顶尖的智慧整合进同一个协作体系。
让信息自由流通，让观点相互碰撞，让错误在彼此的审视中被不断发现与修正，让不同领域的专业能力突破彼此的边界，围绕同一个目标形成高度协同。

当每一个独立的智慧不再孤军奋战，而是相互连接、彼此赋能，由此迸发出的潜力，或许不再是简单的线性叠加，而是一场指数级的跃迁。

过去，我们一直在追求更强大的单一 AI 模型，以及它与个体人类之间更高效的互动。

而未来，我们或许应该思考的是如何让多个 AI 智能体与多个人类的智慧真正融合。

AI 强大的计算能力、知识储备与执行效率，加上人类对现实世界敏锐的感知、经验、直觉与价值判断，再通过集体协作实现互补与修正。当人类与机器的智慧开始相互交织、共同演化——我们究竟会创造出一种怎样的伟大存在？

而集体智慧的上限，又究竟在哪里？

[Aiden Frontier @lvlstx]

@Deefui it sounds deep and also frightening ,,

[紫雲 @Deefui]

当 AI 开始拥有触觉，我们是否应该重新思考，什么才是真正的具身智能？

近日，Meta Reality Labs Research 开源了 SuperDex，一个专注于机器人接触物理仿真的平台。

灵巧手操作的难点，往往隐藏在那些肉眼难以察觉的接触细节中：指尖接触物体的位置、压力的分布、柔软指尖的形变，甚至细微的受力变化，都可能决定一次抓取能否成功。

SuperDex 将“接触”置于物理引擎的核心，通过计算接触表面的三维力分布，为触觉传感器建模提供更精细的数据支持。同时，它还支持刚体、软体、杆、腱与布料等多种物理对象，并提供机器人建模、场景编辑和强化学习接口。

这意味着，我们正在尝试让机器人不只是看见和抓住物体，而是能够在仿真中更加细致地理解自己究竟触碰到了什么，以及物体如何回应它的动作。

但在我看来，这项技术背后隐藏着一个更加值得思考的问题：

具身 AI，是连接世界模型与现实世界物理规律及因果关系的一座桥梁。

如今，许多搭载 AI 的机械手仍然存在一个根本性的局限：感知维度有限，反馈过于机械化。它们可以识别物体、计算压力、执行动作，却很难将真实人手所拥有的丰富触感，完整、连续地转化为 AI 能够理解的数据。

而我们所处的世界是多维的、动态的。它既遵循客观的物理规律，也包含主观的感知体验。

我们对这个世界的理解，从来不只是来自眼睛所看到的画面，或大脑所进行的推理，而是来自身体与环境之间无时无刻不在发生的交互。

因此，我认为，未来具身 AI 的重要发展方向，或许不只是更强大的AI，而是一个更加完整的“身”。

如何帮助 AI 更加真实、生动、细腻地感知和捕捉世间万物，才是世界模型从数字仿真走向现实世界的重要一步。
试想一下，未来的 AI 身体，是否能够拥有像人类一样整体、连续而丰富的感知能力？

感受到一阵微风拂过每一寸肌肤；

感受到一次伸展中，每一根肌肉纤维、肌腱与骨骼之间的相互牵动；

甚至一次钻心的疼痛所对应的复杂生理信号与本能反应。

人类的身体从来不只是大脑执行命令的工具。它本身就是我们认识世界、理解因果、形成直觉的重要组成部分。

或许，未来 AI 的突破，并不只是让机器学会像人一样思考，而是让它真正学会通过身体去认识世界。

而这又引出了一个更深层的问题：

究竟是智能赋予了身体理解世界的能力，还是身体与世界的交互，塑造了智能本身？

[Aiden Frontier @lvlstx]

@elonmusk @bot https://t.co/KoeDGoJM7V

[紫雲 @Deefui]

🤣 笑死，各大AI 公司内部架构，很真实

@OpenAI 表面使命，向钱看齐

@AnthropicAI 安全团队最强势，部门墙高，老大天天搞邪门理念

@GoogleDeepMind 天天刷榜，组织混乱产品线各自为政，架构混乱人员游动

X Ai 老马独裁

Microsoft 山头林立，报团混战

[Aiden Frontier @lvlstx]

@Coinvo If they keep pushing for more money

[紫雲 @Deefui]

🤣 笑死，各大AI 公司内部架构，很真实

@OpenAI 表面使命，向钱看齐

@AnthropicAI 安全团队最强势，部门墙高，老大天天搞邪门理念

@GoogleDeepMind 天天刷榜，组织混乱产品线各自为政，架构混乱人员游动

X Ai 老马独裁

Microsoft 山头林立，报团混战