Vũ Thuận Phát (Edgar Vu)

Founder of Cyno. Building businesses through technology. Exploring AI agents, robotics, and real-world systems. Sharing lessons as Edgar Vũ.

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

@DopaOnX Nực cười là bài post này cũng do AI Render

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

Câu chuyện IPO của Anthropic khiến tôi nghĩ nhiều về hạ tầng AI: một sản phẩm xuất sắc đi cùng nhu cầu vốn khổng lồ cho chip, điện, trung tâm dữ liệu và nghiên cứu. Nhà đầu tư đang trả tiền hôm nay cho doanh thu kỳ vọng trong tương lai.

Tôi nghĩ dòng vốn có thể còn nuôi cuộc đua này thêm một thời gian. Nhưng vốn huy động được chỉ giúp mua thêm thời gian; khả năng trụ vững vẫn phụ thuộc vào hiệu quả kinh tế. Thị trường sẽ kiên nhẫn đến đâu?

IPO có thể kéo thêm FOMO, khiến giá chạy trước năng lực tạo lợi nhuận. Nếu tăng trưởng hụt kỳ vọng trong khi nghĩa vụ hạ tầng vẫn đến hạn, cú điều chỉnh sẽ rất khốc liệt. Đây là một kịch bản rủi ro, chưa phải kết cục chắc chắn.

Đôi khi, chúng ta đặt cược vào bước tiến của xã hội trước khi bài toán kinh tế kịp hoàn chỉnh. Tôi vẫn chúc mừng Anthropic vì đã tạo ra một sản phẩm tuyệt vời. Còn mua cổ phần ở mức giá nào lại cần một luận điểm riêng.

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

Having just completed a project with Astra & Sol 6.1 (OpenAI), it’s time for the @OpenAI team to address three practical bottlenecks encountered during AI-assisted pair programming:

1. Coherence & Adherence to the Plan (Only 70–90% effective)

Even with a detailed plan and a specific QA framework in place, the AI ​​tends to experience "context drift" during complex tasks. It prioritizes generating code that works in the immediate moment over maintaining the integrity of the overall architecture.

2. Discrepancies in Figma-to-Code Translation

The transition from UI/UX designs to code is not yet seamless. Issues persist regarding spacing, micro-interactions, and interface latency. It is essential to enforce the use of fixed Design System tokens; you cannot simply allow the AI ​​to handle styling autonomously.

3. A Complete Lack of Security (Security by Default = Zero)

Unless explicitly prompted, the AI ​​will overlook critical aspects like input sanitization, access control, and data exposure risks. A "Security Guardrail" or a final pre-flight check is always necessary to safeguard the source code.

While AI serves as an excellent co-pilot, we are still a long way from being able to "hand over the keys" and let it handle projects from end to end.

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

Nhưng đây có đang lừa dối khách hàng hay không? Có nên cho khách hàng biết người nào thật sự đang hỗ trợ họ… trò bịp này có thể gây tổn thương chính những người đặt niềm tin vào những người này.

Những người thật sự không có bất kỳ kiến thức về tài chính và không khác gì lừa đảo.

[rob. @robiartec]

Un grupo de estudiantes chinos compró 7 Mac Mini en eBay por un total de 1.600 $, los conectó mediante Ethernet para formar un solo sistema y luego abrió una firma financiera basada en IA directamente en su dormitorio universitario.

Su primer cliente le pagaba a un asesor financiero 8.400 $ al año. Ellos cobraron 240 $ al año por hacer lo mismo, pero mejor.

Claude lee los informes 10-K en unos segundos, construye modelos de asignación sin sesgos de comisiones, ejecuta escenarios de optimización fiscal y proyecta la jubilación al dólar exacto.

Un asesor financiero que gestiona un portafolio de 500.000 $ cobra unos 5.000 $ al año solo por existir, y el 92 % de ellos rinde peor que un simple fondo indexado en 15 años.

Warren Buffett apostó 1.000.000 $ a que un simple fondo indexado del S&P 500 superaría a cualquier fondo de cobertura en 10 años. Ganó con 854.000 $ de ventaja.

Siete Mac Mini, un archivo CLAUDE.md y 240 $ al año reemplazaron a un equipo de analistas. Primer mes: 8 clientes. Segundo mes: 20 gracias al boca a boca.

1.600 $ invertidos una sola vez. El resto son solo líneas en una hoja de cálculo del cliente.

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

Nếu có một liều thuốc dành cho hối tiếc không tồn tại thì nó là một điều tuyệt vời đấy bạn.

[Elon Musk @elonmusk]

I have this great idea for a drug called Regretamine. One dose and all your regrets are gone!

Including, obviously, having taken Regretamine in the first place.

[Elon Musk @elonmusk]

I have this great idea for a drug called Regretamine. One dose and all your regrets are gone!

Including, obviously, having taken Regretamine in the first place.

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

I don’t mean to sound negative, but I find it frustrating how often “humor” on social media feels like a tactic to farm engagement and sell something. I enjoy a good joke, but when it feels manufactured purely for attention, it puts me off.

Am I reacting to something genuinely troubling about online culture, taking it too seriously, or holding others to a standard I don’t always meet myself?

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

I share that optimism, but one risk deserves a deeper discussion: the gap between how quickly AI can reduce the need for human labor and how slowly new livelihoods can emerge.

If productivity rises while employment and incomes come under pressure, who sustains demand for what businesses produce? And if investment increasingly concentrates in AI, will other sectors still attract the capital they need to grow and create jobs?

New jobs may eventually emerge, but the timing matters. People cannot put their lives on hold while the economy adapts. Retraining alone may not be enough if new opportunities are not growing at the same pace.

How do we turn AI-driven productivity into broadly shared income, support businesses that create new opportunities, and protect people through the transition? Society’s ability to adapt deserves as much attention as the technology’s ability to advance.

[Bill Gates @BillGates]

AI is the most profound technology I’ve seen, both in its ability to help people and to disrupt their lives. I’m optimistic about the good it can do, but we need to take its risks seriously. https://t.co/MZuSaceOYr

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

@elonmusk @bot Moving forward without a clear plan for our resources and the people behind them risks wasting both. We need to know where each resource adds value and how each person can contribute most effectively.

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

At the end of the day, that’s just how it works: build useful things, serve people, and let the money follow. If you're not serving anyone, what's the point...? Elon was right all along—he saw through the noise. Live naturally, contribute naturally, and everything else simply falls into place.

[AYi @AYi_AInotes]

马斯克的最新访谈治愈了我的AI焦虑，很多人以为 AI 抢走工作带来的是毁灭，马斯克在访谈里却给出了一个完全相反的推演：AI 真正干掉的不是人类，而是几千年来我们绑在有用性上的虚妄尊严。

马斯克给了一个很有意思的比喻，
他说未来人类的工作，大概率会彻底变成园艺。

你根本不需要在后院自己种番茄和土豆，超市里买到的永远比你自己种的更快、更便宜、品相更稳定。

但今天依然有很多人在阳台折腾花草、在院子里拔草，甚至花一整个下午为朋友做一顿自己种出来的晚餐。

这件事在纯粹的工业效率上毫无意义，但人类做这件事，本来就不是为了效率。

很多人聊到 AI 替代人类，第一反应永远是失业、断供和阶层滑落。

但这件事最反直觉的地方在于，AI 真正击碎的不是饭碗，而是人类过去几千年赖以生存的价值坐标：有用性。

过去几千年，一个普通人的尊严几乎全靠有用性撑着。

你能搬砖、能算账、能写代码、能做报表，社会按月给你发薪水，你确认自己被这个世界需要。

可当 AI 在任何一项具体任务上都比人类做得更快、成本更低时，有用性瞬间就通货膨胀了。

机器负责解决生存的效率，人类被迫退回生命本身的体验。

就像那些退休后反而过得极度充实的老人，每天早起练字、下棋、养花、钓鱼，没有一件事是在创造 GDP，但每个人都活得自得其乐。

未来所有能被标准化的实用技能，都会被算法和机械臂彻底端走。

而那些效率最低、最耗时间、最不划算的事情，反而会变成精神上真正的奢侈品。

手写一封字迹歪扭的信，亲手烤一块火候不匀的面包，花整个周末陪孩子搭一个不讲力学的树屋。

这些事情的全部意义，恰恰在于它们是低效的、笨拙的，并且留着你自己活过的温度。

如果未来靠打工换来的社会身份全部贬值，我先说我自己，代码和排版我秒甩给 AI，但给朋友做一顿慢吞吞的手工饭，我肯定是都要自己做的😄

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

Is it just me, or has the ChatGPT experience taken a massive hit lately?

• Endless latency & random downtime ruining daily workflows.
• Noticeable model drift: wandering off-prompt, hallucinating, and throwing random errors.
• Zero meaningful UI/UX improvements.
• Recent integration attempts just make everything feel cluttered rather than seamless.

Stability should always come before new features. Needs a fix ASAP @OpenAI.

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

Có lẽ đúng điều đó AI đang dần khiến phần mềm chết đi và thay đổi quá trình kinh doanh mạnh mẽ.

Chúng ta vẫn làm công nghệ nhưng ở chiều tương tác vậy lý chứ không còn trên Internet nữa.

[Mo Elgaraihy @EngMoElgaraihy]

🚨 رئيس أنثروبيك يقلب الطاولة ويفجرها: "إذا كنت تتعلم البرمجة لحماية مستقبلك، فأنت تقفز في النار!"

​"البوس الكبير" خلف Claude خرج بتصريحات مرعبة تنسف أوهام سوق البرمجيات التقليدي وتكسر أوعية المبرمجين دون رحمة:

​🛑 كتابة الكود الخام انتهت: البرمجة البحتة هي أول وظيفة سيفرمها الذكاء الاصطناعي تماماً. الثروات الهائلة في العقد القادم ليست في السوفتوير، بل انتقلت بالكامل إلى العالم المادي والبنية التحتية الصلبة كأشباه الموصلات والسيليكون!

​⚡ المعادلة الجديدة بسوق العمل: الـ AI سيقوم بـ 95% من الشاق وتنجز أنت 5% من القرار.. ليتضاعف الإنتاج 20 ضعفاً! لكن احذر الفخ القاتل: بياناتهم الداخلية تثبت أن الاستخدام الخاطئ للذكاء الاصطناعي يُضمر المهارات البشرية ويجعلك أغبى بمرور الوقت.

​كيف تنجو إذاً؟
1️⃣ المهن ذات اللمسة الإنسانية الحقيقية: مثل الطب.. قراءة الصور الطبية وحدها لا تكفي، بل التواصل البشري العميق وفهم الطبيعة النفسية للمريض.

2️⃣ مهارة "كشف الكذب" (التفكير النقدي الحاد): عندما يصبح بمقدور السيستم توليد أي شيء بنقرة زر، فإن القدرة على تمييز الحقيقة وسط طقس كامل من الزيف الرقمي هي الأصول الأكثر ندرة وثمناً!

​من لا يزالون يتزاحمون كورسات Java وPython بكل وسعهم.. حان الوقت للاستيقاظ. عصر السوفتوير يركض بأقصى سرعة نحو خط النهاية، وقواعد اللعبة تغيرت للأبد!

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

Unemployed even after graduating from university…

I just finished reading an article that made me pause and reflect: What is it that we truly need to survive in this life—especially when even our utmost effort still fails to land us a job?

What do you think? Is a university degree still the only path forward?

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

It has absolutely nothing to do with it; they’re still just carrying sugar back to the nest. It’s the most idiotic experiment—completely irrelevant.

Honestly, it’s best to grasp the actual underlying principles; why bring up evidence that has absolutely no relevance?

[D Mk @wickedsmaahht]

アリの巣に2年連続で「無料の砂糖」を大量投下――社会主義がなぜ失敗するのか、アリのコロニーがその答えを即座に証明 @rich_anthill が公開した動画では、アリのコロニーに山のような砂糖が投下され、その後の30日間のタイムラプス映像が事の顛末を物語っています。アリたちは、まるでこの世の最高の恵みが訪れたかのように砂糖の山に殺到しますが、やがて「生きるために食料を探す」という本来の活動には一切関心を示さなくなります。投稿者の分析は辛辣です。「無料の砂糖がある以上、働く理由などなくなる」というわけです。 コメント欄は、たちまち経済学のケーススタディの場と化しました。「ユニバーサル・ベーシック・シュガー（砂糖の無条件支給）」、「アリの福祉」、そして「生産手段（今回で言えば砂糖）」を無償で与えると、なぜか生産性が消え失せてしまうという、お決まりの指摘が飛び交います。「各人から能力に応じて取り、各人に必要に応じて与える」という理想は、誰も貢献しようとしなくなる瞬間までしか機能しないようです。砂糖は記録的な速さで消え、巣は拡大し続ける一方で、アリたちは政治経済学における最も古典的な教訓の一つを、見事に実証してみせました。 自然界が代わりに実験を行ってくれたおかげで、私たちが自ら試す必要はなくなりました。アリのコロニーが、その答えを示したのです。 #アリの社会主義 #社会主義が失敗する理由 #アリのためのUBI #無料の砂糖 #アリの福祉 #もう働かないアリたち #コロニーの依存 #砂糖の大量投下 #タイムラプス #実験2年目 #アリの経済学

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

@elonmusk Haha, what’s going on, man? I never would’ve guessed that even you could be taken down by SI.

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

Writing code has become truly easy these days. Thanks to AI and various support tools, even a solo developer can create games, build apps, or bring all sorts of ideas to life.

But how do you get people to actually pay for what you’ve created?

That is an entirely different challenge—one that, truth be told, I am currently trying to solve myself.

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

It reminds me of the cruel policy of expelling immigrants from U.S. borders.

Those who have been granted entry and provided with decent living conditions ought to fulfill their responsibilities, rather than undermining the very administration that supplies them with jobs and various other benefits. It serves as a cautionary tale about how reckless actions can lead to negative societal outcomes.

Things will turn out fine; the lesson is clear, and now it is a matter of how to address the situation.

[Stitch @StitchKings]

If anyone knows how to donate to this man defending his LGS please let me know. Respect!🫡

Here is his website:
https://t.co/GmSDczVOQ6

#Pokemon ——-> #Lorcana
#France

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

Looks like some "dots" might be flying around—either way, it’s definitely intriguing!

[Tibo @thsottiaux]

@StefanoGPT @OpenAI

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

I’ve been wondering what the future holds for programmers. Job opportunities seem to be shrinking, yet the cost of hiring a developer continues to rise.

What do you think the future looks like for developers?

I’d love to hear a variety of different perspectives.

[Vũ Thuận Phát (Edgar Vu) @ThunPhtV4]

@codewith55 Most of the technology is released in the US.