如何参与ChatGPT发展?

每一次技术进步,对一部分人来说,是灾难;对一部分人而言,是机遇。对新技术的钝感、无视、鄙视,可能带来灾难。而积极了解、深入学习、开放大脑,也许带来机遇。

这一轮ChatGPT热潮,给账户倒卖、API接口、知识整合后再兜售的急先锋们带来了一笔快钱。

对不会编程,或不熟悉计算机技术的人而言,除了利用ChatGPT提高生产力之外,还有什么可以参与到ChatGPT发展的事情可以做呢?

对此,我问了ChatGPT:

Me: How can one who know little coding take part in ChatGPT?

ChatGPT:

As an AI language model, I suggest you can take part in ChatGPT in several ways, even if you know little coding. Here are some ideas:

  1. Participate in discussions: You can read and participate in discussions on various topics that interest you. You can share your views, ask questions, and engage in healthy conversations with other users.

  2. Share your expertise: If you have expertise in a specific field, you can help other users by sharing your knowledge. You can provide valuable insights, tips, and advice to help other users solve their problems.

  3. Collaborate with others: You can collaborate with other users on various projects, such as writing articles or creating videos. You don’t need to know coding to contribute to such projects.

  4. Ask for help: If you need help with something, you can ask other users. You can search for answers or post questions on the relevant topics. Other users may help you with their expertise.

  5. Use non-coding tools: There are many tools and platforms that you can use to contribute to ChatGPT. For example, you can use graphic design tools to create images or videos, or you can use content creation platforms to write articles or blog posts without knowing how to code.

Overall, there are many ways you can participate in ChatGPT, even if you know little coding. You can leverage your skills, expertise, and interests to contribute in various ways.

这回答得,有些参考价值哈。

开玩效率工具——“幕布”“Typora”

由于学习记录的需要,开始玩一些效率工具。简单记录一下。

幕布

非常合适做阅读笔记。

最近读了两篇TCGA miRNA-seq的文献,想把文献的思路、两篇文献的差异、对自己工作的启发记录下来。发现用幕布来记录,可以方便得到有层次的思维导图。这样,基本就实现了与用笔和纸写写画画的效果,而且不乱,还可以扩展。再了解一下幕布的进阶功能——可以插入公式、小图标(可以让笔记更有层次,更易于分类)。

随手把明天开始上的单细胞分析课程整理了一篇。总算积极主动地为上课做一次准备。

现在正在学的卫生统计学,大概也可以把笔记电子化一下。

读文献,整理文献思路也很好用:

目前为止,我在用免费的幕布基础版,还没有升级的要求。

Typora

记录代码的好工具。我主要用它来记录linux代码和操作流程。R语言,则直接在R studio上记录。python,用Jupyter notebook记录方便。

Typora,很便宜。我直接买了个正版的。

Mendeley

文献管理软件,elsevier出品。

用mendeley,不用endnote,是因为endnote太贵了。用免费的mendeley,可以轻松管理文献,可以在word中标注文献出处,高效完成论文中的文献管理。

mendeley有一个非常赞的功能——给文献打多个标签。这样,可以方便地根据标签集中阅读某些内容的文献,如miRNA,Parkinson’s。

 

生信学习笔记

学习生信的步骤:

1. Start with a foundation in Python/R and bash

In Python/R: Just get to the point where you can read in data and run a statistical test.

In Bash: Be able to navigate around directories, open files, and run a program (like sort, uniq, or wc) and output the result to a file. Check out my intro video on bash for quick tutorial on getting started on the command-line.

2. Do a small project

3. Occasionally do tool safaris

4. Build tools to fill gaps as they come up in your research

为什么要学linux?

初学python,在windows下完成一些简单的编程,对python有了一些基础的了解。也遇到几个明显而且急迫需要解决的困难:

  1. 读懂别人写的代码。有相当多的代码,我仍然无法理解,也就无法学习或借用。
  2. 进入社区,参与代码讨论。不会把自己的代码打包,发布到社区上。不了解有哪些社区合适我,也不了解它们的玩法。
  3. 运算速度慢,却不会分析如何提高运算速度。
  4. 不会调用外部程序。

进一步学习Martin Jones的第三部python书,开始更多地在linux上举例。于是,我就跳进了另一个坑——linux。这个坑和刚学python时的体验很相似:大量的时间耗在软件安装和配置上。

但随着学习的展开,模糊地认识到,python可以方便地调用命令行(linux常用命令行),而很难直接操作windows的图形界面。这为python调用其他程序提供了便利。