WR 20251205

Webto Wang -

题图：朋友给了我两只湖北泥藕，久违的煲了个排骨汤

2025 年最后一个月了，草稿箱里还躺着只有标题的文章 😂

【苹果】Bad Dye Job

After I published that post, I got a note from a designer friend who left Apple, in frustration, a few years ago. After watching Jobs’s Aqua introduction for the first time in years, he told me, “I’m really struck by Steve directly speaking to ‘radio buttons’ and ‘the key window’.” He had the feeling that Dye and his team looked down on interface designers who used terms like Jobs himself once used — in a public keynote, no less. That to Dye’s circle, such terms felt too much like “programmer talk”. But the history of Apple (and NeXT) user interface design is the opposite. Designers and programmers used to — and still should — speak the exact same language about such concepts. Steve Jobs certainly did, and something feels profoundly broken about that disconnect under Alan Dye’s leadership.

苹果的设计主管 Alan Dye 被 Meta 挖去搞 AR 硬件了，全球最大的果粉头子 John Gruber 可以说是直接破口大骂了。Alan Dye 掌管苹果设计团队十年，确实是可用性没有什么进展甚至有倒退的十年。iPhone X 的全面屏交互方案可能是为数不多的高光。按照引文的说法，这十年苹果的设计团队被一帮纯视觉设计师把持，只能说。。。似曾相识，以及，Cook 确实不太懂？

【AI】I Know We’re in an AI Bubble Because Nobody Wants Me 😭

The first thrill comes when you see an opening, a possibility that nobody else has spotted. There’s the satisfaction of working hard to chase down the opportunity, and then too often the despair when it turns out not to work. Even then, that means I’ve learned something, and being good at optimization means learning everything you can about the hardware, operating system, the requirements themselves, and studying others’ code in depth. I can never guarantee that I’ll find a solution, but my consolation is always that I have a better understanding of the world than when I started.

标题是作者的玩笑，但是目前云上的大模型和端侧用的迷你大模型从商业角度讲肯定是完全不同的策略，云上用的动辄千亿级参数的模型在找到真正的商业模式之前都是靠烧融资活下去的，那么能继续拿融资就事关生死。

但是因为里这段话很有意思，我觉得我们这些喜欢解谜的人是比较容易从工作中找到这种乐趣的，找到可见或隐藏的问题，试图用不同的方法改善，如果改好了就获得成就感，如果没改好也能获得经验值，下次做得更好，就和玩某些游戏差不多。

【AI】Wikipedia content guideline

Large language models (LLMs) can be useful tools, but they are not good at creating entirely new Wikipedia articles. Large language models should not be used to generate new Wikipedia articles from scratch.

现在在互联网上想要找未经 AI 处理的文字内容已经开始变得不那么容易了，至少从搜索引擎去找是这样的，其实给未来模型的训练带来了更大的麻烦。维基百科的指南说不要完全用 AI 写词条，与其说是建议不用 AI，不如反过来看，允许适量使用 AI 生成内容。

其他乱七八糟的