👉🏻👌🏻 (@pr0n) header
👉🏻👌🏻 (@pr0n) profile picture

Living in Truth 活得磊落真誠

Joined November 2007
Photo by 👉🏻👌🏻 (@pr0n)Photo by 👉🏻👌🏻 (@pr0n)Photo by 👉🏻👌🏻 (@pr0n)Photo by 👉🏻👌🏻 (@pr0n)Photo by 👉🏻👌🏻 (@pr0n)Photo by 👉🏻👌🏻 (@pr0n)Photo by 👉🏻👌🏻 (@pr0n)Photo by 👉🏻👌🏻 (@pr0n)Photo by 👉🏻👌🏻 (@pr0n)
全球機迷引頸以待的《GTA 6》今日正式發售,虛構的 Leonida 州街頭頓時擠滿急欲踩下油門的玩家。然而,比超跑引擎聲更早擊中香港玩家心臟的,竟是一把來自 1987 年的聲音 ── 有人將車載電台調至「萬花筒 FM」(Kaleidoscope FM),喇叭傳出的,是林子祥的《敢愛敢做》。那一刻,不少香港玩家忍不住笑了一聲:「不會吧,沒聽錯嗎?」 然後,心裏浮起一種難以言明的感覺。在一個那麼遠、那麼龐大的西方遊戲工業盡頭,竟然有人記得這首歌。 「萬花筒 FM」的設定本身已很有意思,這個電台專門搜羅世界各地歌手翻唱的作品,原曲《Nothing's Gonna Stop Us Now》是 Starship 筆下的荷里活式甜蜜童話,但潘偉源填詞時沒有照搬那層糖衣,反而把場景拉回香港街頭:交通燈、冷雨、餐廳的沙玻璃、路人的目光,全部成為「敢」的佈景。於是副歌那句「微雨中/身邊車輛飛過/街裏路人走過/交通燈催促過」,在 GTA 的飛車追逐中竟渾然天成,彷彿一早為某種速度與不顧一切寫好了註腳。潘偉源寫的從來不是單純的愛情,這是一座城市在高速運轉之中,個體選擇忠於自己的勇氣。 這份「敢」,恰好又是另一個香港人故事的核心。 當「林子祥 GTA」的討論在 Threads 上炸開,有網民翻出一則以「#香港人做遊戲」為題的舊帖。樓主「weestrangelife」任職於英國愛丁堡的 Rockstar North,他在帖中平靜地回溯了一條漫長的路:拍過短片,做過電視、電影與軟件,為投身遊戲業追夢二十五年,歷經一番努力,他終於在 2018 年如願加入 Rockstar;從最基礎的遊戲翻譯與在地化測試做起,深入熟悉遊戲架構後,逐步晉升至開放世界設計團隊。遊戲發售前八十三天,他寫下一句:「終於終於總算要完成這個童年夢想」。而他身處的Rockstar North,正是近年裁員風暴中最受衝擊的核心工作室之一,能夠留守至作品面世,本身已是一場無聲的「敢愛敢做」。 有網民追問《敢愛敢做》是否由他親手放入遊戲,他幽默否認,卻留下一句更耐人尋味的話:「公司不少香港人其實」。這句話比任何官方宣傳都更有力量,它說明在那座頂級遊戲工業的金字塔尖,香港人的參與不是孤例。他們未必個個負責選曲,但他們帶着屬於這座城市的耳朵、記憶與品味,在一個以虛構美國城市為舞台的遊戲裏,留下一道不易察覺卻真實存在的香港痕跡。 一首 1987 年的舊歌,因為一次「彩蛋」式的相遇,在 2026 年的全球玩家耳邊重新響起。對外國玩家而言,它或許只是一首節奏凌厲的八十年代搖滾;但對香港人來說,那是潘偉源筆下「敢於交出真情/那算可鄙」的執拗,是林子祥一人唱盡的硬淨,也是香港設計師們用二十五年光陰換來的一瞬間共鳴。 萬花筒轉動時,碎片看似偶然碰撞,但每一片都有自己的來歷。當一座城市習慣了說「不行」,總有人仍然選擇「敢」。而那份「敢」,終究會被聽見 ── 哪怕隔着一整個虛構的海洋。 冷雨撲向我 點點紛飛 千噸高溫波濤 由你湧起 個個說我太狂 笑我不羈 敢於交出真情 那算可鄙 狂抱擁 不需休息的吻 不需呼吸空氣 不須街邊觀眾遠離 微雨中 身邊車輛飛過 街裡路人走過 交通燈催促過 剩下獨是我跟你 #GTA6 #林子祥
153
👉🏻👌🏻 retweeted
Same performance. New world. OTOY Studio VFX Tools: drop in your footage, describe the world you want, and run it on the canvas. Every movement, face and camera move stays exactly as you shot it. Only the world around it changes. Up to 4K, HDR and 4:4:4 color space, built for the highest quality film work.
7
20
2
171
10,676
大晚上九點多十點收到 Nikon @RED_Cinema 的電郵,說看到我早幾天把 Komodo-X 放到購物車。這麼卷的嗎?
1
243
👉🏻👌🏻 retweeted
Every driver needs their crew.
1,005
1,494
414
21,120
6,028,013
有些傳媒就是特別「壞心眼」,logo 這樣子放啊?
224
今天留言了幾個關於標準、正宗和話語權的串文,然後就看到這帖 #newspeak
据爆点周刊报道, 习近平给自己的历史定位, 已经不仅仅是毛泽东, 也不仅仅是斯大林了, 他的新目标是秦始皇! 根据爆点周刊的爆料, 习近平所追求的政治目标, 不仅包括长期执掌最高权力, 实现其提出的中华民族伟大复兴的中国梦, 以及推动台湾统一, 还包括在中国历史上完成一次更为彻底的民族整合. 按照习近平的构想, 什么五十六个民族五十六朵花, 以后只有“中华民族”一朵花, 而他, 就是这朵花的大花骨朵. 这让我想起前两天大家讨论的, 中国社会科学大学成立“冷门绝学学院”的事情. 十年前, 习近平刚上位不久, 那时他的权力还不如现在稳固, 但是他的内心已经十分稳固了, 那时候他就提出要求, 要研究西夏文, 契丹文, 女真文, 突厥文, 回鹘 文, 古藏文等等文字和文化领域. 没有人会相信, 小学生文化水平的习近平, 真的会对这些“冷门绝学”感兴趣, 所有人都知道, 他的“指示”和“要求”, 每一条都暗藏着他的政治野心. 比对一下爆点周刊的报道, 我们现在就明白了: 习近平“治国理政”的核心内容之一, 就是“中华民族的伟大復兴”. 一切意识形态的包装, 都要以打造“中华民族的伟大復兴“为宗旨. 所谓的“冷门绝学”, 明眼人都看得出, 这些文字和文化, 通通涉及中国历史上不同民族, 文化与文明之间的关係. 而中国开始重视这些文字与文化的研究, 其目的就是强化“中华民族共同体”这个国家论述, 为这个国家论述寻找更多的文献资料, 古代档案和考古发现等方面的支撑. 换句话说, 为他实现全民族的大一统提供理论基础. 即将进入第三个任期的习近平, 越来越搂不住自己了. 在多年的“为人类指明方向”之后, 他最近提出要“引领”“亚大大家庭”, 开始想当家长. 而现在的民族统一大业, 显然是要做三皇五帝到如今的第一人, 中华民族共同体的“体长”了. 这听起来简直就是疯了. 但是如果我们回顾历史, 很多大独裁者, 最后的精神状态其实就跟疯了差不多, 从希特勒到毛泽东, 到达权力巅峰之后, 尤其是到了政权开始不稳之后, 都多少有点疯癫的状态. 看来, 谁都不会例外啊. @baodiantimes
1
494
The in-camera visual effects and practical effects in Mother Mary are stunning. Editing and music are impressive too.
465
The post-merger reality: SpaceX absorbed xAI and Cursor, and @cursor_ai won xitter.cf/theinformation/status/…
Replying to @mark_k
Since the Cursor acquisition, the Grok Build and Imagine teams have been largely laid off. Resources are are now focusing on Grok Bot, which is built by Cusor team, plus the core Grok model training. Grok subscription used to be the best all-in-one package, it includes model, coding, image/video/voice, and deep X integrations, with no real competitor after OpenAI dropped Sora. Game now looks handed off to Meta, which has the full stack: model, coding, image/video/audio generation, agent, plus deep integration of Facebook/Instagram/Threads/WhatsApp.
415
The Weeknd's "Hong Kong" topped @youtubemusic trending in the US and the UK on a Cantonese hook from Faye Wong's 1993 "Monsoon." Cantonese isn't a niche dialect. Ethnologue counts 85M speakers. #香港 #廣東話 #王菲 #季候風 xitter.cf/theweeknd/status/21209…
3
407
Other frontier AI labs: give #Cantonese real support, in Traditional Chinese text and native voice
54
Replying to @pr0n @youtubemusic
Hong Kong cinema shaped Hollywood: Wong Kar-wai, John Woo, Chow Yun-fat, Michelle Yeoh, Sammo Hung Kam-po, Yuen Woo-pin, not to mention Bruce Lee #kungfumovie
72
Chinese models' spoken Cantonese sounded better but often slipped into Putonghua pronunciation and Simplified Chinese wordings. @GoogleDeepMind Gemini's pronunciation is weaker but it didn't mix in Putonghua or Simplified. Gemini is strong at Cantonese spoken word understanding, especially Cantonese-English code-mixing.
58
👉🏻👌🏻 retweeted
“Hong Kong” Single Artwork by : Christopher Doyle
386
1,643
233
12,725
554,759
👉🏻👌🏻 retweeted
Built by Gemini 4 Hydrogen purely in single-file code...no external libraries, no skills, no image generators, and zero embedded assets.
131
121
43
2,701
133,123
I have just setup ComfyUI workflows to test Grok Imagine earlier today 🤯🔫
Grok Imagine is taking a massive hit. According to The Information, @SpaceXSI has laid off around 60 employees from its image and video generation team, leaving only a skeleton crew to maintain the product. What's remarkable is that Elon Musk said they were "doubling down" on Imagine just a few months ago. Now the company appears to be shifting its focus toward coding and autonomous AI agents like Grok Bot. A dramatic reversal that raises serious questions about the future of Grok Imagine.
1
268
P-H3-Image’s FL2VA editing with references is so good. I was not expected its quality is on par with some frontier image models.
Today, we open-source P-H3-Image, an image generation and editing model built from MiniMax-H3 by @MiniMax_AI. A SOTA video model, turned into an efficient open image model. - 𝗛𝗼𝘄 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝘁 𝗶𝘀 𝗶𝘁? Generate or edit images 5× faster, in 8 steps instead of 50. A 4 MP image takes 2.6 s instead of 12.4 s. The model is 7× smaller than H3, so high-resolution images are a light download and run on a much smaller GPU. - 𝗛𝗼𝘄 𝗴𝗼𝗼𝗱 𝗶𝘀 𝗶𝘁? Strongest at photorealism, documentary and editorial photography, film stills, and graphic styles with short quoted text. It renders 2 to 16 MP and edits with up to 8 input images. -𝗛𝗼𝘄 𝗱𝗼𝗲𝘀 𝗶𝘁 𝘄𝗼𝗿𝗸? We reworked every component of H3: the prompt upsampler, the text encoder, the diffusion schedule, the transformer, and the decoder. Use it with Diffusers or ComfyUI. 📄 Tech Report: pruna.ai/blog/oss-p-h3-image… 🤗 @huggingface Open Weights: huggingface.co/PrunaAI/P-H3-… 👉 @ComfyUI Native Support: huggingface.co/PrunaAI/P-H3-…
347
👉🏻👌🏻 retweeted
Meet Pika H3 Flex, debuting today as the #1 open weight model on DesignArena’s t2v board. It’s built for professional iteration, with exportable audio stems, resolutions up to 4k, and faster-than-realtime generation. It can produce a 15-second clip with soundtrack in as little as 7 seconds. So, yeah. It’s kind of a flex 💅
398
390
403
4,849
1,533,790
今晚壓軸登場係咪 Mac Studio M5 Ultra 512GB
Muse Spark Open -- weights releasing today -- runs great on new Mac Studios thanks to work with @johnternus and Apple.
1
595
There are two ways to use AI in video editing. One is cutting costs: replace editors to save money, or post that AI will take everyone's job. The other is building capacity. Let AI do the grunt work, like syncing, logging takes and rough assembly. You put out more videos and still have time to tell each story well. AI lets experienced editors produce more and better videos. It's an assistant, not a replacement.
🤖 Made with AI
My thoughts on the agentic video editing wave: 1. There won't be one winner in agentic video editing. Speed, use case and workflow will always matter. Just as Cursor moves upmarket while Lovable caters to non-technical users, specialised video tools will continue to serve different audiences, even as general-purpose agents improve. 2. Every major AI leap unlocks new product categories. Since GPT-3.5, we've seen coding, 3D and presentations. Opus 5.5 and Google's vision-language models are now unlocking agentic video editing. The edits still look more like websites than videos, but this is the worst they'll ever be. 3. Editing costs will collapse, and the number of creators will explode. Basic edits could go from dollars to cents within a year, with huge speed improvements. Frontier models will still command a premium. As editing becomes effortless, recording content might be the biggest bottleneck. 4. The best editors might be the biggest critics of agentic editing. Just as experienced engineers initially resisted AI-generated code over quality and maintainability, I think professional editors will eventually come around as the tools improve. 5. The real magic is in iteration. Getting 80–90% of an edit right on the first attempt is great, but making it excellent requires much more iteration. The way creators give feedback, refine edits and make videos feel distinctly theirs is where these tools will become truly magical 6. Anthropic's focus on coding enabled agents to build videos using HTML/CSS, SVG, browser rendering and OpenCV. Packages like Hyperframes, Rebooshed and VEED's OpenEdit gave LLMs editing capabilities they previously lacked. Opus 5.5 now makes many of those use cases redundant. 7. Chat alone won't be enough for video editing. Chat will be central, but video operates across time and X, Y and Z coordinates. More sophisticated UI will be needed to make great edits. 8. Brand consistency and repeatability remain unsolved. A Casey Neistat video feels distinctly his because of intentional creative decisions. Creators and brands need that same predictability across hundreds of edits. Even vibe coding hasn't solved this well (yet). 9. Basic editing won't be a long-term moat. Anthropic, Google and OpenAI will eventually offer it. Deeper workflows, specialised use cases and excellent mobile experiences will drive differentiation. 10. Don't reinvent professional editing tools. Building non-linear timelines is incredibly hard. Agentic video tools should focus on creating great first cuts and making it easy to hand them off to existing professional editing tools. Overall, the next six months will be noisy! But video editing has never been winner-take-all. And I do NOT think that will change. Hope this helps.
244
「今日割五城,明日割十城,然後得一夕安寢。起視四境,而秦兵又至矣。然則諸侯之地有限,暴秦之欲無厭,奉之彌繁,侵之愈急。」──《六國論》 Su Xun (蘇洵) wrote that 1000 years ago about appeasing Qin (秦). Land is limited. Appetite isn't. The more you give, the harder they push. Ukraine gave up its nukes in 1994 for security promises. Russia took Crimea in 2014 and kept it. The full invasion came in 2022. Ceding land doesn't end a war. It schedules the next one.
May the bloodthirsty warmongers, who push for a war that never ends, burn in hell!
297