Dr Shuangyi Yan

Academic in HPN group, University of Bristol. Personal opinions on dynamic optical networks, ML/AI applications, 5G and Beyond. 中文推@yesterdayno20

[Dr Shuangyi Yan @necho]

How long can we still claim we are doing research is counting down. The one without involvement of real hardware will be replaced easily now.

[Dr Shuangyi Yan @necho]

Asking same question and you can see a clear difference between Chinese AI models and word top class AI models. It will be hard to realise if you are just using one model, especially the underperformances model.

[Dr Shuangyi Yan @necho]

I never thought one day I would stress out my MacBook Pro.

[Dr Shuangyi Yan @necho]

Unfortunately, the creativity, innovation and aspiration are gone as written in the text.

[Dr Shuangyi Yan @necho]

@Agent47vsDiana There will be no big differences with this top models. The GPT5.4 even makes the Microsoft copilot usable.

[Dr Shuangyi Yan @necho]

With ChatGPT 5.4, most simulation papers can be reproduced and validated in a short time. It also means you can validate your ideas easily.

[Dr Shuangyi Yan @necho]

Woo, for the first time hit the rate limit of ChatGPT.

[Dr Shuangyi Yan @necho]

There's still a long way to go before we have effective on-device AI. The cloud computing model is also under pressure as it continues to scale up. There will likely be opportunities for edge AI with efficient interconnects.
https://t.co/oEzrcPWGT8

[Dr Shuangyi Yan @necho]

就这么决定了，这次出差回去就把OpenClaw安装起来。主要原因是发现电脑上的自动流程需要电脑开着，时开时不开，自动化的流程经常出现错误。

[Dr Shuangyi Yan @necho]

今年在路上，经常能看到BYD的电车。今天早上打车，更是打到了一辆BYD的海豹。电车出海，BYD应该是做的比较好的。

[Dr Shuangyi Yan @necho]

确实充分发挥了AI的优势。

[AI奶爸 @zstmfhy]

最近看到 @vigorX777
做的 ai-daily-digest 超实用！直接从 Karpathy 推荐的 90 个顶级技术博客抓最新文章，用 Gemini AI 评分筛选 + 中英摘要 + 分类 + 趋势图，一条命令出份高质量 AI/技术日报。懒人必备～

https://t.co/8tpX6PsZ2K

[Dr Shuangyi Yan @necho]

Are you serious X?

[Dr Shuangyi Yan @necho]

Really have to take LLM seriously. Now my codex 5.3 is working hard for research.

[Kevin Weil 🇺🇸 @kevinweil]

💥 Today we’re introducing Prism—a free, AI-native workspace for scientists to write and collaborate on research, powered by GPT-5.2.

Accelerating science requires progress on two fronts:
1. Frontier AI models that use scientific tools and can tackle the hardest problems
2. Integrating that AI into the products scientists use every day

Prism is free to anyone with a ChatGPT account, with unlimited projects and collaborators.

Try it today at https://t.co/cmEsJYRtW9—would love to hear your feedback.

[Dr Shuangyi Yan @necho]

When I’m not vibe coding, I’m vibe writing.

[凡人小北 @frxiaobei]

Anthropic 在 Skilljar 上线了一门大概两小时的课程《Claude Code in Action》。从标题就很直白，整门课就是围绕 Claude Code 来的，偏系统性学习这一挂，比单纯看安装或上手教程更有整体感，值得花点时间认真刷一遍。

国内用户这边有点小门槛，需要🪜开全局，或者把 https://t.co/xrVEhXixjF 和 https://t.co/0E3BxiUjAW 加进节点里，不然视频大概率打不开。

内容本身讲得挺扎实，节奏稍微快了一点，不过配合沉浸式翻译，或者直接看视频下方的文字版，理解起来完全没压力。

链接在这里：
https://t.co/KngF2bFJAY

[Dr Shuangyi Yan @necho]

X is extremely flexible.
If your timeline feels messy, ask any AI to generate a list of keywords to mute or block, then load it into your x app.
Do that, and your timeline instantly becomes a clean, high-signal stream of the information you actually want.

[Dr Shuangyi Yan @necho]

1.Use ChatGPT for figure descriptions and discussions.
2.Utilize Nano Banana to generate figures.
3.Employ Chrome to remove watermarks https://t.co/VKIreXqWbg

[Dr Shuangyi Yan @necho]

空荡荡的火车，窗外飞驰而过的村落，一切正无比快速的冲进2026这个巨变的时代。

[Amir Salihefendić @amix3k]

https://t.co/lS3qFn7fQA