泠雨 (@caelyreth) header
泠雨 (@caelyreth) profile picture

独立开发者|UI & UX|Agent|System 希望我的作品可以找到你直觉里的方向 𖤣𖥧𖥣。𖤣𖥧𖥣。 近期在开发写作软件 Neverland GitHub: https://teapawt.org/t.co/VcJfgJqgP8 Ko-Fi: https://teapawt.org/t.co/QeqsyyUR7m

Joined October 2022
Photo by 泠雨 (@caelyreth)Photo by 泠雨 (@caelyreth)Photo by 泠雨 (@caelyreth)Photo by 泠雨 (@caelyreth)Photo by 泠雨 (@caelyreth)Photo by 泠雨 (@caelyreth)Photo by 泠雨 (@caelyreth)Photo by 泠雨 (@caelyreth)Photo by 泠雨 (@caelyreth)Photo by 泠雨 (@caelyreth)
泠雨 retweeted
Day 28/ tibo's game has landed i spent $5.99 on this domain, but i had a blast making it with grok bot and cursor! i hope you enjoy it and find all the easter eggs 🤍 if i did it right you should be able to play it inside of X too tibo.dog
1,515
890
834
14,052
2,176,609
みんなが React と Vue の話しかしないので悲しい Svelte ユーザー
9
11
4
105
8,003
泠雨 retweeted
you must overcome ai-induced ego death: the sooner you accept that the models are smarter than you, the sooner you’ll see your true value was never just your intelligence and you'll be so much happier
36
76
10
1,130
35,179
I wont be updating any of my Mac apps for touch. Touch on a laptop is a dumb gimmick. Just another touch bar thing.
41
28
9
525
24,711
泠雨 retweeted
我一直觉得,随便给别人的 GitHub 账号打分,再分个三六九等,是一件挺不礼貌的事情。 现在是什么都要卷吗?连写开源项目都得像做题一样,最后拿个分数证明自己? ghfind 甚至把开发者分成「夯」「顶级」「人上人」「NPC」「拉完了」。 我知道它的算法是开源的,但公开算法就代表算法合理吗? 谁决定 PR、Stars、Issue、项目规模各自应该占多少权重?这些标准经过什么验证?凭什么用一套人为设计的评分规则,来定义一个开发者的实力和价值? 一个人可能长期维护私有项目,可能擅长架构设计、定位复杂问题,也可能只是根本不喜欢经营 GitHub。 这些东西一个分数能体现出来多少? 更让我难受的是,真的有人会拿着这种第三方评分去评价别人。 甚至还有 GitHub App 可以根据这个分数给别人的 Issue 和 PR 打等级标签。 我不反对做数据分析,但我很反感把一个活生生的开发者压缩成 0 到 100 的数字,然后拿着分数给别人排座次。 如果以后连公司招人都开始参考这种东西,那跟应试教育有什么区别? 我说这些不是因为自己分数低,也不是在乎自己得了多少分。 哪怕今天给我打 100 分,我也一样觉得这种风气很荒谬。 写代码已经够累了,能不能别什么都搞成排行榜?
60
32
25
290
33,737
Gemini 4 Carbon is our next frontier model. It’s state-of-the-art across cybersecurity and long-horizon capabilities.
741
1,388
1,096
14,505
1,675,068
泠雨 retweeted
Somewhere the world feels a little quieter.
16
231
9
1,614
48,610
泠雨 retweeted
Today we're introducing Genie 4 — our newest interactive world model from @GoogleDeepMind 🌐
299
639
259
6,482
1,084,271
泠雨 retweeted
Somewhere between midnight and forever.
88
1,798
97
8,747
284,470
Unpopular opinion There's a bunch of amazing graphic design resources and packs you can buy to support graphic designers, instead of using AI that just copies the same styles. None of the AI generated graphics are original, it's just stealing from the original creators. Go support graphic designers instead of ripping them off with AI. 🖤
32
223
8
3,308
64,248
泠雨 retweeted
tech desperately needs a new design aesthetic we literally have agi now and flying cars but every ui looks the same with this apple/linear minimalist rounded corner tailwind shadcn slop look we’re not thinking big enough on how things should look and feel chatting with an agent should look and make me feel like i’m living in the future it should make me feel hope and awe and optimism that humanity actually got to agi and it will lead to unimaginable prosperity it should make me tap into my 9 year old self, who wondered if we’d *ever* have talking humanoid robots, and make me feel totally amazed by the realization that this future is not even a possibility now, it is a CERTAINTY! the time is so ripe now the coding agents have gotten so good that we can be experimenting and playing with new ideas 100x faster than before now is the time for excessive experimentation
108
57
14
914
98,701
泠雨 retweeted
Post 2 exposing Kimi AI That's the creators campaign docs I got from Kimi AI after their Kimi K3.1 pro release and thats the narrative they want creators to push, of course I'm out and I'm never gonna say something I don't test myself and give my honest opinion about and I don't want to be involved in this China vs US political situation.
149
149
137
1,727
263,816
泠雨 retweeted
HTML/CSS is Turing complete, so why not compile programs to plain HTML/CSS and runs with zero bytes of JavaScript? zero-js.vercel.app
39
81
7
1,403
109,418
After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution
4,020
8,269
6,821
76,547
40,164,946
泠雨 retweeted
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics. The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra. The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
5,666
20,010
14,810
119,860
75,386,429
泠雨 retweeted
I am addicted to linux pornography – at Jaipur, India
78
152
35
3,248
200,921
gpt 5.6 最让我受不了的就是: 我让它做一盘番茄炒蛋,它往里还加了东坡肉。 我说有必要加东坡肉吗?它说你说得对,然后把东坡肉去掉。 我说好,你提 PR 吧。再一看,它 PR 写着「番茄炒蛋(无东坡肉)」并且注释里会写一大堆为什么本道菜不需要加东坡肉。
584
1,032
333
10,834
1,414,752
New tsgo feature to be considered: Content mappers to support Vue/Svelte/Angular It allows an external integration to transform a foreign file’s original text into valid TS syntax and to provide mappings between the original and transformed content. github.com/microsoft/typescr…
7
13
2
146
37,567
泠雨 retweeted
最近很真切的感受到,大部分人用了 LLM 之后,生产效率完全没提高,反而把一堆人齐刷刷推上了达克效应的“愚昧之巅”… 最要命的是他还不是自己爬上去的,所以再也下不了了…
94
27
11
550
58,632
mojo is now open source modular.com/blog/mojo-open-s…
23
164
26
1,548
112,400