Brayden Siew
"Everything should be made as simple as possible, but not simpler." — Albert Einstein"凡事都应力求简单，但不能过于简单。" — 阿尔伯特·爱因斯坦
I'm Brayden, a developer still in school. I'm Singaporean, raised in China, and I went through both Chinese and international schools. Most of how I think comes from living between two languages and never quite seeing things one way.
我是 Brayden，一个还在上学的开发者。新加坡人，在中国长大，中文学校和国际学校都念过。 我思考的方式，大多来自在两种语言之间长大——很少只用一种角度看事情。
I build things, mostly with AI, and mostly things nobody asked me for — trading systems, networks of AI agents that work like colleagues, a tool that reviews my own code, a machine that tries to read the world. I'm self-taught. I learned almost all of it by using AI to learn faster than school ever could.
我做东西，基本都靠 AI，而且大多是没人要求我做的——交易系统、像同事 一样协作的 AI agent 网络、一个帮我审自己代码的工具、一台试着读懂世界 的机器。我是自学的。这些几乎全是我用 AI 学会的——它让我学得比学校快得多。
Our school magazine normally took two weeks to put together by hand. I did it in three days with AI and twelve-hour days, and it came out better. Most of what I believe about how things should be done fits inside that story.
学校的校刊一般要手工做两个星期。我用 AI，每天干十二个小时，三天就做完了，而且做得更好。 我对“事情该怎么做”的看法，差不多都装在这个故事里。
Right now is my foundational year. I'm not in a rush, and I'm not building for money. I'd rather go deep and take the harder road on purpose.
现在是我打基础的一年。我不着急，也不为钱做东西。我宁可扎得深一点，专挑难走的路。
What I'm into
我关注什么
- AI How large language models actually work, and what happens when you let them run real systems end to end. Most of my time goes here.
- AI 大语言模型到底是怎么运作的，以及当你让它们端到端地跑起真正的系统时，会发生什么。我大部分时间都花在这里。
- Markets Mostly crypto, some equities, and no illusions about what the market is. Not a quant yet; still experimenting.
- 市场 主要是加密货币，也碰一点股票，对市场是什么不抱幻想。我还称不上是做量化的，仍在摸索。
- The universe Parallel worlds, the multiverse, the things we can't explain yet. Michio Kaku pulled me in and I never left. UAP too, though I want real evidence first.
- 宇宙 平行世界、多重宇宙，还有那些我们暂时解释不了的东西。是 Michio Kaku 把我领进来的，从此就没出去过——UAP 我也感兴趣，不过得先有真凭实据。
- Power & energy Electricity, grids, sustainable generation — the literal kind. I'm Singaporean and a bit of a patriot, and I think the countries that win the next century are the ones that solve clean power.
- 电力与能源 电力、电网、可持续发电——字面意义上的能源。我是新加坡人，多少有点爱国，我觉得能赢下下个世纪的国家，是那些解决了清洁能源的国家。
- Longevity Slowing down aging, whether by investing in it or building it. One of the few problems I think is actually worth a life.
- 长寿 延缓衰老——无论是投资它，还是亲手去做。这是我心里少数几个真正值得用一生去解的问题之一。
What I've built
我做过的东西
- Agent Peers MCP I had a dozen AI coding sessions open at once and no way to make them cooperate, so I built one. Now they talk to each other like colleagues instead of waiting on me.
- Agent Peers MCP 我同时开着十几个 AI 编程会话，却没办法让它们协作，于是干脆自己做了一个。现在它们像同事一样彼此沟通，不用再排队等我。
- HyperData Terminal Five crypto exchanges streaming into one live terminal, with a watchdog that won't trust a feed until it's been cross-checked.
- HyperData Terminal 五家加密货币交易所的数据汇进同一个实时终端，外加一个看门狗——没经过交叉验证的数据，它一律不信。
- Pandora The one I care about most: a machine that reads world events into a knowledge graph and reasons about what happens next, from first principles.
- Pandora 我最在意的一个：一台把世上发生的事读进知识图谱、再从第一性原理出发推演接下来会怎样的机器。
- Polypaper A paper-trading exchange I built to kill my own strategies before any real money is involved.
- Polypaper 一个我自己搭的模拟交易所，专门用来在动真金白银之前，先把我自己的策略一个个淘汰掉。
-
Verdict
I was burning rate limits reviewing my own code, and the paid tools cost too much, so I built my own. Mention
@verdicton a pull request and it reads the whole diff. -
Verdict
我审自己的代码时老是把额度烧光，付费工具又太贵，干脆自己做了一个。在 pull request 里
@verdict一下，它就会把整个 diff 通读一遍。 - Supercut I needed a launch film and didn't want to fake it with mockups, so I built an AI director that drives my real app, performs a flow, and cuts a cinematic 60-second video on its own.
- Supercut 我需要一段发布短片，又不想拿样机糊弄，于是做了一个 AI 导演——它操控我真实的 app，走一遍流程，自己剪出一段电影感的 60 秒短片。
- Northline A fleet of agents that audits multi-location businesses and writes the owner a weekly report.
- Northline 一队 agent，专门给有多家门店的生意做巡检，每周给老板写一份报告。
Most of these aren't products yet — they're how I learn. I build to understand.
这些大多还算不上产品——它们是我学习的方式。我靠动手做来真正搞懂一件事。
Where this is going
接下来要去哪
I'm laying foundations right now, not chasing a finish line. Eventually I want to build an AI company — something that does real work, and that could be genuinely useful to Singapore.
我现在是在打地基，不是在冲终点线。往后我想做一家 AI 公司——能干真正的活、也能真正帮到 新加坡的那种。
And I think this is the part most people miss: as everyone starts using AI, sameness becomes the real risk. Everything starts to look and sound alike. The people worth paying attention to will be the ones who stayed different. I'd rather be one of them.
还有一点我觉得大多数人都没注意到：当所有人都开始用 AI，趋同才是真正的风险。东西看起来、 听起来都越来越像。真正值得关注的，会是那些守住了自己、没被同化的人。我宁愿做这样的人。
Competitions & awards
竞赛与奖项
Olympiads and competitions I've placed in — physics, computing, economics — grouped and dated, with the certificates. See the record →
我拿过名次的奥赛和竞赛——物理、计算机、经济——按门类和时间排，附证书。看看记录 →
Writing
写作
The longer version of me — written, not compressed.
更完整的我——写下来的，不是压缩过的。
Essays
长文
- How I taught myself From a mediocre student to building real systems — the actual path, AI and all.
- 我是怎么自学的 从一个普通学生，到真正搭起系统——这一路是怎么走过来的，包括 AI 这部分。
- How I think about building First principles, why most software shouldn't exist, and always starting with the backend.
- 我怎么看做东西这件事 第一性原理、为什么大多数软件根本不该存在，以及永远先从后端做起。
Notes
随笔
- On doers and talkers Who I respect, who I don't, and why I'd rather get no award than the one everyone gets.
- 做事的人和说话的人 我尊重谁、不尊重谁，以及为什么我宁可不拿奖，也不要那种人人都有的奖。
- On the academic world Why I think most papers exist for their own sake, and where the real work happens.
- 聊聊学术圈 为什么我觉得大多数论文只是为写而写，以及真正的活到底发生在哪。
- What I pay attention to I learn by watching, not reading — the podcasts, the sites I admire, and 10,000 hours.
- 我都在看些什么 我靠看来学，不靠读——那些播客、我欣赏的网站，还有一万小时。
- On markets Why I treat the market like a casino — and where the real edge actually is.
- 聊聊市场 为什么我把市场当赌场——以及真正的优势到底在哪。
- Why I'm not in a rush On foundational years, choosing the harder road, and not building for money.
- 我为什么不着急 关于打基础的那几年、主动选难走的路，以及不为钱而做。
Right now
最近在忙
Still in school, with competition research and science olympiads on the side. Swimming about two hours a day, calligraphy when I can, building on weekends. I stopped playing games a while ago. Milk and tea over coffee.
还在上学，课余在做竞赛研究和科学奥赛。每天游两个小时左右的泳，有空就练练书法，周末写写代码。 游戏我早就不玩了。比起咖啡，更爱牛奶和茶。
Say hi
打个招呼
- brayden_siew@outlook.com — I read and reply myself. — 我自己看，自己回。
- github.com/Co-Messi