Playing with Claude Code and Claude Cowork

AI

Over the past couple of weeks, I’ve dabbled with Claude Code and Claude Cowork to see what it’s all about. I love tinkering with technology, after all.

Claude Cowork is essentially a less technical way to do the same thing Claude Code can do. It can be used to manipulate files in specified folders as well as connect to the many online services you may have. You can then automate tasks, do research, clean up files, create or manage notes, and the list goes on.

{{more}}

As usual, I love playing with this stuff. I like to know how it works. I’m interested in discovering the systems that can be automated by this technology.

Yet, at the end of the day, I have no particular use for it in my personal life. My life is intentionally simple. I don’t need automation. I don’t need an LLM to create notes or make connections. I don’t need it to tell me what I should be working on.

On the other hand, my work life is very complicated, especially recently. I don’t want to use Claude or other LLMs for work because of the nature of the content. My organization does have Microsoft Copilot.

I use Microsoft Copilot every day to take meeting notes, find information across the organization, do research on particular problems, and write simple scripts.

Recently, I’ve been using Copilot to troubleshoot issues, and when I’m done I have it create a Word document that summarizes the issue, the troubleshooting process, and the solution. I then add this document to our project files. Next time I have a similar issue, the solution pops up in my search, and it saves a ton of time. Copilot has replaced search, in a good way.

Despite its usefulness, there is no way these tools could take over my job, or anyone else’s job in my company. That would be a disaster. They require specific instructions and guidance to extract the usefulness from them.

Certainly, this will change over time, but I’m optimistic that we will change as well, and will learn and adapt as time goes on.