Parallel Experiments
1.71K subscribers
62 photos
1 video
3 files
811 links
Stay informed. Stay authentic.

Welcome to the public part of my brain. Here I share curations and thoughts.

Created with ❤️ by @linghao.
Download Telegram
Mary Meeker's first Trends report since 2019. 340 slides on the state of AI.
https://www.bondcap.com/reports/tai
https://store.steampowered.com/app/2008920/Lorelei_and_the_Laser_Eyes/
今年似乎一直在玩同一类的游戏:发生在一个庄园或类似密闭环境中的解密游戏 — Blue Prince、Botany Manor 等。
但今天想推荐的 Lorelei and the Laser Eyes 是最近几年我最喜欢的游戏之一。
谜题属于偏简单的,硬核解密玩家如果对不上电波可能会踩雷。不过逐渐将散乱的线索拼凑成答案的体验依然很棒。
艺术风格赞,叙事到一半还开始加入对艺术史的反思。
#Annapurna 出品必属精品
👍2
Naming is extremely important in Computer Science and, frankly, everything. Good naming is hard. Being able to pick a good name shows a lot of good taste.

Context engineering (a term promoted by Karpathy: https://vxtwitter.com/karpathy/status/1937902205765607626) is much better than:
- Prompt engineering: "Prompt" is just too overloaded.
- In-context learning: This is more of a research term and feels awkward to describe engineering required for building good LLM applications.
- RAG: Today, when done right, RAG is a very specific kind of context engineering. But too many people conflate it with "anything that puts stuff in the prompt".

Image credit: https://github.com/humanlayer/12-factor-agents/blob/main/content/factor-03-own-your-context-window.md
2👍2
Forwarded from Reorx’s Forge
最近读到最喜欢的一篇文章是 Writing Code Was Never The Bottleneck [^1],它讲到软件工程的真正瓶颈从来不是编写代码本身,而是代码审查 (code reviews)、知识传递 (knowledge transfer)、测试 (testing)、调试 (debugging) 以及团队协调沟通 (coordination and communication) 等需要思考、共同理解和判断的人工流程。

这些观点让我很有共鸣,许多都是这些年来经历和思考产生的真实感受。而且我还想到,这个话题,还可以面向一个我更关注的事情——做产品,那么大体观点仍然是一致的,但也会有一些不同的侧重,比如第二段可以变成:

The actual bottlenecks were, and still are:
- design UI and business logic workflow, reflect them clearly in a PRD (Product Requirement Document)
- project and time management, the right pace of the development-testing-shipping circle
- marketing
- mental health control, life routine, working momentum
- the courage to start doing it, and accept failure to rework again


于是我让 Gemini 阅读原文,然后以 "writing code was never the bottleneck in making product for indie developer" 为题重写一篇新文章。我没有把上面这些我自己的想法输入给它,因此得以在它写的新文章中发现一些新观点。我感觉这是种很好的 AI 用法,即当我阅读一篇文章意犹未尽时,让他基于原文写出更多内容。阅读这件事情最好不要让 AI 代劳,但生产给自己阅读的内容可以。

下面是 Gemini 写的新文章:
https://telegra.ph/Writing-Code-Was-Never-The-Bottleneck-07-05

---

[^1]: https://ordep.dev/posts/writing-code-was-never-the-bottleneck
7
https://sakana.ai/ab-mcts/
Sakana AI 发布 AB-MCTS (Adaptive Branching Monte Carlo Tree Search) - 让 o4-mini / Gemini 2.5 Pro / Deepseek R1 协作,有点 mixture of mixture of experts 那味了 😎
Please open Telegram to view this post
VIEW IN TELEGRAM
❤‍🔥1
https://huggingface.co/blog/smollm3
Hugging Face 发布 SmolLM3,一个 3B 小模型。他们非常慷慨地提供了完整的技术细节,从 model architecture 和 data mixture 到 pre/mid/post training 的多个 recipe 🫡
https://store.steampowered.com/app/3743220/A_Solitaire_Mystery/

Baba Is You 作者 Hempuli 新作,20 个脑洞大开的空当接龙魔改版 😆
FYI: 昨天玩了几个小时遇到一些很明显的 game breaking bug,可以考虑过一阵子再入。
Please open Telegram to view this post
VIEW IN TELEGRAM
这篇是我读过最醍醐灌顶的对现代理论物理的全面“科普”。打引号是因为大部分内容其实并没有解释到门外汉也能看懂的程度…… 虽说如此,我读完以后还是对最小作用量原理和各种对称性等有了稍微不那么模糊的理解,突然感觉以后读科幻又能带着新的视角去看了!

现代数学和理论物理已经发展到怎样一个令人震惊的水平了? - 酱紫君的回答 - 知乎
https://www.zhihu.com/question/304611853/answer/1928827087810192602
4🤔1
https://store.steampowered.com/app/2475490/Mouthwashing/
随手打开一个游戏没想到是非线性叙事神作;延续制作组之前作品的诡异美术风格,让人感叹 indie 的无数种可能:只要有出色的点,不需要画面精良或是海量内容也可以做出好游戏。
流程不长,我甚至在 steam 退款时限内就打完了。
通关以后发现,故事其实非常简单,没有任何多余的要素,但你会开始深思故事的主旨…
https://william-rous.itch.io/type-help
最近风很大的变格推理游戏,纯文字+暴风雪山庄模式,真的很好玩!
而且它正在由 The Roottrees are Dead 的团队重制成 https://store.steampowered.com/app/3641000/The_Incident_at_Galley_House/ 预计明年发布 🤩
🤩3