Link1987

应用程序开发者
可视化爱好者
关注应用开发，AI, Web3, VR/AR，GIS，数字孪生
不忘初心，方得始终

[Ihtesham Ali @ihtesham2005]

I accidentally discovered how to compress a semester of learning into 48 hours.

A grad student at MIT showed me his NotebookLM setup. I thought he was just organized. Then I watched him pass a qualifying exam on a subject he'd never studied before.

Here's exactly what he did:

First: he didn't upload a textbook.

He uploaded 6 textbooks, 15 research papers, and every lecture transcript he could find on the subject.

Then he asked NotebookLM one question:

"What are the 5 core mental models that every expert in this field shares?"

Not "summarize this." Not "explain this topic."
Mental models. The stuff that takes professors years to develop.

But the next part is what broke my brain.

He followed up with:

"Now show me the 3 places where experts in this field fundamentally disagree, and what each side's strongest argument is."

In 20 minutes he had a map of the entire intellectual landscape of the field:
the debates, the consensus, the open questions.

Most students spend a full semester just figuring out what those debates even are.

Then he did something I've never seen before.

He asked:

"Generate 10 questions that would expose whether someone deeply understands this subject versus someone who just memorized facts."

He spent the next 6 hours answering those questions using the source material. Every wrong answer triggered a follow-up:

"Explain why this is wrong and what I'm missing."

By hour 48, he could hold a conversation with his thesis advisor without getting destroyed.

The tool didn't change. The questions did.

Most people treat NotebookLM like a fancy highlighter.
These students are using it like a private tutor who has read everything ever written on the subject.

The difference between a semester and 48 hours isn't the amount of content.

It's knowing which questions to ask.

[Link1987 @LinkLin1987]

AI 在一段会话中会自圆其说，追求逻辑上的前后一致，沉溺于语境的“逻辑自洽”，掉进“前后一致”的陷阱，忽略了事实上的错误。

[Link1987 @LinkLin1987]

罪犯发现了一个 Vulnerability（漏洞），带着 Malicious（恶意）的意图，制作了一个 Fake（假的）页面，最终完成了对你的 Fraud（诈骗）

[Link1987 @LinkLin1987]

感觉 gemini app 越来越像豆包了，聊天还行，不适合解决问题了

[Link1987 @LinkLin1987]

有时需要 AI 主动和我讨论对齐问题，需要 AI 主动问我问题，AI 不问。
有时需要 AI 为我做决策，一次性给我交付，AI 不交付，竟在那问一些他职责内的问题，或者问一些弱智问题。

[Link1987 @LinkLin1987]

感觉 glm 在上下文到 70% 以上后，就变笨了

[Link1987 @LinkLin1987]

@turingou 时间线中看到的 post 还是之前那些博主，没有看到新内容，怎么办

[Link1987 @LinkLin1987]

ai 像一个先进的文明，他接受你使用他，但你要学会他的语言，使用他的语言和他交流。我经常会让 ai 教我如何表达，或是让 ai 为我生成提示词。

[Link1987 @LinkLin1987]

@dingyi https://t.co/LJWqvdiRrV

[Link1987 @LinkLin1987]

不要吃陌生人给的食物，不要装陌生人给的软件，不运行陌生人给的代码

[Link1987 @LinkLin1987]

@ZhuLiangDev cool

[Zhu Liang @ZhuLiangDev]

Claude Code. From my bed. Finally.
https://t.co/ro7OgIKiI8 is officially live on the iOS App Store. Take your agents to the park, the couch, or under the blankets. 🚀

[Link1987 @LinkLin1987]

gemini 说话越来越好玩了：
```
如果你正在开发软件，纠结一个文件该放哪时，请问自己：“如果用户把这个文件删了，他是会想杀了我（Data），还是仅仅觉得有点不方便（State）？”
```

[Link1987 @LinkLin1987]

qwen 的大佬们有没有可能去 deepseek ?

[Link1987 @LinkLin1987]

@Kai1338878Li 我的服务关了，代码是开源的应该还能跑起来

[Link1987 @LinkLin1987]

我感觉 gemini 3 pro 好像变幽默了一些，在回答时，比之前更会用一些大白话。比如 “以太坊协议通过算法（RANDAO）在全网 100 万个验证者中，随机选中了一个倒霉蛋（或者幸运儿）”

[Link1987 @LinkLin1987]

我发现和 AI 聊天，也有种刷短视频上瘾的感觉，聊着聊着就停不下来了，尤其是现在 AI 每次回复最后都会鸡贼的抛出一个可能感兴趣的话题，引导继续聊下去😂

[Link1987 @LinkLin1987]

使用 IPO 模型看 AI Agent, 会发现 agent loop 和计算机操作系统底层模型好像，agent loop 也是在反复处理 Prompt-LLM-Tool, tool 执行后又会改变 prompt.

如果为 Agent 配置一个 tool, 使得 llm 可以直接自主的改变下一轮他自己收到的 prompt(messages + tools + other params)，会发生什么呢？

[Link1987 @LinkLin1987]

IPO, 也就是 Input-Process-Output, 描述了计算机或程序的基本模型。计算机和操作系统将这个模型做成了无限循环，不断地处理 input 也就是指令。output 可以影响 input, 进而出现了更加复杂、动态，甚至智能的程序。

[Link1987 @LinkLin1987]

也许 AI Agent 不是应用, 是操作系统, LLM是硬件(CPU, 内存)

[Link1987 @LinkLin1987]

看老罗采访 lovable 创始人，创世人说他认为什么是技术驱动：产品经理说话没有技术说话好使就是技术驱动😂