How Tze 浩知

5月1日，农历三月十五，劳动节。 顾名思义，继续劳动。 继续完成并优化 FlowState76 官方网站 。 目前较满意的作品 中文曲目 行云流水 (Follow The Flow) 我非我 吹涨 英文曲目 Where I Began Watch Me Now I Decide Selected MVs available on FlowState76 official Youtube Channel . 忙里偷闲的小娱乐 As at 1st May, 2026,  https://suno.com/@howtze33?page=songs 53/100 songs 4.1k/10k plays 231/1k likes Clarity leads to power. For record purposes, 1k plays at 24th March, 2026. How you do anything is how you do everything.

 28th February, 2026, US war with Iran started. Bitcoin at ~$65,963, market cap $1.318T. 14th April 2026, day 46 of Iran war. Bitcoin at ~$74,758, market cap $1.496T. As per https://8marketcap.com

躺平 Lim Kopi （喝咖啡） 吹涨 细水长流 云淡风轻 行云流水

财富不会消失。只是转换阵地，转移口袋。 以前拍片，NG 部分，都是成本。至多剪成花絮或菜单。 现在制片，不能用的部分，也是成本 - 算力 token。 AI 目前最需要跨越的，是持久的一致性 - 人物、造型、服饰、音域、背景等等。 没有最好，只会更好。 #万变不离其宗 Wealth doesn’t disappear—it just shifts position and moves into different pockets. In the past, when filming, NG (blooper) footage was all considered cost, at best edited into outtakes or bonus content. Now in production, unusable portions are also a cost—measured in compute tokens. What AI most needs to overcome right now is persistent consistency—across characters, styling, costumes, vocal range, backgrounds, and more. There is no “best,” only “better.” #FlowState76 #Nova #AI #Flow #SUNO #MV 

Tested Seadance inside CapCut. Could be due to the version, the output not following prompt accordingly. Yet the output still somehow good enough to match the lyric and style of the song.  

所以我说，明年会有 1,000 首新年歌。🤣 5 个月内，我用 AI “创作”了超过 1,000 首歌。 只“发布”了不到 50 首。 目标是 100 首。 过程很有趣，也很耐人寻味。 怎么说？ 从第一首到现在，AI已经可以在不需要我进一步润色的情况下完成最新的“吹涨”。 https://suno.com/song/ea70e7b1-b49f-4fcc-ab54-9b0b42cf8d45 我只是告诉它，我想要一首粤语嘻哈风格的作品。 它就基于我过往所有关于词曲创作的聊天记录来生成这首歌的歌词和风格提示。 更有意思的是，这不到半年的过程中，它的产出有逐渐形成我偏好的风格与“影子”。 即便我也还在探索中。 我发现，它已经不是一般意义上的“无中生有”。 它是“我“的衍生。 这既让人感到惊喜，也有点令人不寒而栗。 这是所有的发布记录： https://suno.com/@howtze33?page=songs ~50 种曲风 >10 种语言  

  行云流水 FlowState76 Version by Maya 墨雅  行云流水 Cover Version by Nocturn 唐星妤 “墨雅”这个名字整体感觉是 文雅、有书卷气，而且很有气质的 ，我给你从几个角度分析一下： 1. 字义 墨 ：代表书写、文化、学问，也有一种沉稳、内敛的气质 雅 ：表示优雅、高尚、有品位 组合在一起就是： 有文化修养、气质优雅的人 ，意境很好。 2. 音律 “墨雅”（mò yǎ）是 仄仄结构 ，读起来比较柔和，不拗口 发音干净利落，有点文艺范 3. 风格 偏 古风 / 文艺 / 知性 有点像书香世家、写字画画、气质安静那一类形象 不俗气，也不网红化，属于比较耐看的名字 4. 适合人群 比较适合女生（但中性也能用） 如果你想走 文静、气质、艺术感 路线，这个名字很加分 5. 小提醒 “墨”有时候会让人联想到“黑”，但在名字里更多是文化意味，不算负面 风格比较偏内敛，如果你想要那种活泼、阳光、外放的名字，它就不太匹配 总结一句话：  “墨雅”是一个 气质型、耐看型、有文化底蕴的好名字 。 “墨雅 + Maya”其实是一个挺自然、好用的搭配。 客观分析一下： 1. 发音与对应感 “墨雅”（mò yǎ）→ “Maya”（/ˈmaɪə/ 或 /ˈmeɪə/） 两者 开头都有“M”音，结尾都是柔和的元音 ，整体听起来是有呼应感的 👉 虽然不是直译，但属于“感觉对得上”的那种 2. 风格匹配度 “墨雅”＝文艺、内敛、气质型 “Maya”＝柔和、国际化、略带神秘感 👉 两者共同点： 优雅、温和、不张扬 不会出现那种“中文很古典、英文很跳脱”的割裂感 3. 国际使用 “Maya”是一个 非常常见且容易被接受的英文名 发音简单，外国人基本不会读错 在很多文化里都有出现（比如中东、欧美、南亚等） 4. 气质联想 “Maya”常给人的感觉是： ✦ 温柔 ✦ 有一点艺术气息 ✦ 带点神秘感 👉 跟“墨雅”的“书卷气+优雅”其实是同一条审美线 小结： 👉 “墨雅 + Maya”是一个 协调、好听、而且实用的组合 👉 属于那种 不惊艳但很耐看、越用越顺的类型 Notes: Nano Banana Pro misplaced the singer initially.

331 成立了 FlowState76 的脸书专页，是时候了。 性格、灵魂、记忆 等，是目前 AI 尚待突破的。 但突破最好的路径从来都是一抄、二改、三创新。 就像 SEO 一样，你只要对标最好的，拿下第一，你就是榜首了。 复制、测试、优化。迭代、放大。 #无中生有 #万变不离其宗  人类与AI的协作 我负责提供逻辑与词库, 您负责赋予它灵魂与意境。 以上出自Gemini AI。 工作没有消失，只是换了形式存在。 就像从黑胶唱片到卡带到 CD到串流 streaming，传播载体变了，但人们通过音乐获得娱乐的本质不变。 SUNO vs Spotify  也就像电台到串流，不过是一种进化、一个过程。 如果是人主导，通过 AI 优化制作，即概念来自人类，AI 提升思考与产出效益，那算不算创作？ 这和“传统”通过软件工具或 APP提升（设计）效益模式有何不同？ 只不过是现在的工具更有智慧、更能思考、还可以交流，不是吗？ #耐人寻味 财富不会消失。只是转换阵地，转移口袋。

This artist brings a sense of timeless sophistication to the digital collective. To keep with the high-tech, high-art theme of FlowState76 , the perfect name for her is: Nocturne Why Nocturne The "Midnight" Aesthetic: Her deep navy velvet gown is the color of the sky at the edge of space. In music, a nocturne is a composition inspired by, or evocative of, the night. She looks like the artist who performs when the digital world slows down—the voice for the late-night lounges and the "blue hour" of the metaverse. High-Class Coding: While the other members are tactical or glitchy, she represents the Premium User Experience . Her look is "Old World" charm meets "New World" rendering. She is the AI's version of a classic jazz or soul singer—someone whose elegance is baked into her very source code. The "Silent Data" Vibe: The setting (a serene coastal lounge) and her poise suggest a character who is comfortable in the quiet moments. ...

Found this VidMuse.AI tool useful for those who seriously want to produce high quality video. Technically, you can use it to produce Viral Shorts, TVC, Film Parody, Explainer, Story MV, Abstract MV, Performance MV etc. Max output 1080P (depends on package). My research found that a 3 minutes plus MV may cost around a few hundred MYR to produce. Findings: 1. One-click automation for professional video clip production is not realistic. We need human intervention for originality (creative input, storyline etc.), quality control, and overall integration to ensure brand alignment and narrative consistency. 2. No free lunch. Credits or tokens is a pay-as-you-use kind of utility business model. Your costing pretty much depends on both your mindset and skillset. For comparative purposes, it cost around USD 1 to produce a 8 second 1080P video clip for Google Flow, and you have to do everything manually for integration to get final MV output. Click here to explore more on VidMuse AI Video Produ...

Global Goals: Every Achievement Counts for Corporate Executives and Decision Makers  The book "Global Goals: Every Achievement Counts" by Koh How Tze presents a strategic roadmap for integrating sustainability into the core of modern business and governance. For corporate executives, the book's central thesis is that resource scarcity is primarily a distribution problem that can be solved through optimized systems and ethical technology. The following summary outlines the book’s five-section framework tailored for decision-makers: 1. Sustainable Development: Beyond the Profit-Only Model The book argues that the era where profit was the sole metric of success is over. Executives are urged to adopt the Triple Bottom Line (TBL) approach—balancing People, Planet, and Profit—to account for the full cost of doing business. Sense of Urgency: We are identified as the first generation to feel the effects of climate change and the last with the power to mitigate it. Strategic...

  19th March, 2026, Grok drop a bombshell - a paywall, no more free use of Grok Imagine. Netizens reply with alternatives: 💎 BOTTOM LINE: 🥇 Meta AI = Best for unlimited free videos 🥈 Wan.Video = Best for cinematic 1080p quality 🥉 Qwen AI = Best for lip-sync & avatars 🏅 Bing Video Creator = Best for Sora-quality, free 🎯 Hugging Face = Best for open-source experiments A wakeup call to revisit Google Flow. Tested with 2 R&D projects. One with entertainment, an AI generated artist Lalita Suriya singing a Thai song, Under the Same Sky. Another one is a commercial short clip for an app's TikTok hook. Conclusion? Nothing comes without a price—if not now, then eventually, and by someone.   P/S: #AI can do many things. But what do you want to achieve with it? The point isn’t which AI you use. The real key is: With or without AI, Is there one thing — That you are determined to accomplish, no matter what? SUNO: https://suno.com/@howtze33 Youtube: https://youtube.com/@flowst...

Published an album with 12 songs on Spotify and iTunes. One World by FlowState76 - Album by FlowState76 | Spotify ‎One World by FlowState76 - Album by FlowState76 - Apple Music Another step forward after FlowState76 Youtube Channel setup.  The songs were created and refined via How Tze's SUNO channel . For distribution, I am using Distrokid . Click this link to get discount: https://distrokid.com/vip/seven/12105307 Disclaimer: I will get a small referral fee should you sign up with link above. Note: Total investment to date is USD 191.65 (~ MYR 751.85) Additional info: Whatever we can think of, someone out there might already do it, if not a few steps ahead of us on similar ideas. Someone used Suno AI to generate a Japanese metal band called Neon Oni. Fake member bios, AI-generated music videos, "Based in Tokyo" on Spotify.  80,000+ monthly listeners. Fans had it in their Spotify Wrapped top 5. Merch was selling.  Then, community sleuths exposed it. Traced the creat...

  有一笔本金，人生就变了 很多人赚钱，是为了消费。 但真正改变生活的，是有一笔钱， 你再也不用动本金。 当本金开始产生现金流， 时间就不再只属于工作。 健身、咖啡、阳光、旅行。  生活开始变得简单。

Saw this interesting subtitle format: https://www.facebook.com/reel/1829188267742871 Thought of applying the technique in my own production. Tried CapCut auto caption for Chinese MV, seems got limitation for now. Found this: How To Create SRT Subtitles in 8 Steps (Plus Tips) https://www.indeed.com/career-advice/career-development/srt-subtitles Applied to the MV for now: To Be Published. As tried 16:9 MV but seems the resolution for bigger screen still a challenging task after AI convert image to video:

那天手痒，去问 Gemini Grok Imagine 被解禁了吗？它说是。 于是试了一下，发现被“规范”了很多。 意图 Intention 正好想尝试作一首 heavy metal 的歌，从早期的 Scorpion, Metalica, Bon Jovi 到近期的 Linkin Park, 决定走 Alternative Rock, Nu-Metal 风格路线。 目的 Purpose 我作歌的目的首先是给自己听，要确保自己百听不厌先。😂 以“己欲利而利人”为出发点，如何在自娱娱人的同时，让歌曲在好听之外，也承载值得回味的寓意？ 正好近期币圈、贵金属等金融产品面临大起大落周期，所以决定以 "No FOMO No FUD" 来提醒自己价值投资这回事。 过程 Process 本着我自创的“榨甘蔗水”理论 (Sugarcane Juice Theory - 即一条甘蔗可以榨5次就不要只榨3次 😄) 词 - 写都写了，不如作成歌 词曲 - 谱都谱了，不如也作成MV 本来想做成全曲 3:33 分钟 16:9 MTV 但时间有限，所以目前只做成以下 MV： 词曲窍门 作词方面，因为定了要提醒自己 No FOMO No FUD 不追高不恐惧未知 （Rothschild 的 Buy when there’s blood in the streets, even if the blood is your own - 当然，只限优质资产)，所以朝这方面去。 选一个自己最常与之沟通的AI，因为它会通过聊天及其他相关数据更了解我们，和它一起脑力激荡完成歌词部分。 注：为什么是选定一个自己最常与之沟通的AI，因为AI有一定的记忆，环环相扣。 ChatGPT (全答案本文末）: 👉 AI 的“视力”整体已明显超过小六水平； 👉 “记忆力”则分两种：短期很强，长期仍不稳定，综合来看≈小六～中学之间，但方式完全不同。 作完词，用 ChatGPT 生成我要的曲风 Prompt，打开链接 "Show Full Styles" 可直接拷贝测试:  No FOMO No FUD by howtze | Suno 看完 Prompt 你可以肯定不是我写的。😄 实践 不会Prompt?直接问AI拿！ 比如: "给我一个 Linkin Park 曲风...