In my recent conversations with non-native English speakers, I discovered their enthusiasm for using GPT to significantly improve their written communication in English.

I can relate to this challenge! Before utilizing GPT, I often spent hours revising emails and reports...

Additionally, I sought an impartial perspective on my communication to ensure that my ideas were organized, clear, and effectively conveyed.

While I could hire someone to proofread my work or simply consult a native English speaker (who may not necessarily be a professional copywriter), this often led to further questioning, revising, or even rewriting the whole document.

This is where LLMs became a game-changer. They provide a way to enhance communication without the anxiety of being judged or wasting someone else's time (or my own money) when seeking a fresh viewpoint on my writing.

However, many users find interacting with ChatGPT to be unnatural.

To address this, I developed a toolbox that integrates seamlessly with Word and Google Docs.

This toolbox offers:

- A robust translated draft from the original language

- A refined version following multiple revisions

- A comprehensive review

All of this is accessible with just a click from Word or Google Doc—no logging into another platform required

By streamlining this process and integrating these essential yet fundamental use cases into commonly used tools, we can improve basic communication skills and explore new methods to utilize LLMs in everyday activities.

I was not satisfied with a simple return from GPT when I submitted a text for different actions from Google Docs or Word. I wanted to know, in some cases, the changes and also the reasoning. So, I improved the toolbox by now having three outputs in a specific area.

The first is the final result after executing the instruction (e.g., corrections, rewriting, etc.).

The second is a differential. You might ask, why not have the differential in the text as is already the case? The reason is that between the corrections in red and the pop-ups of “suggestions,” the text becomes totally unreadable over time. I need to see the difference clearly and quickly. 

The third output is GPT’s reasoning behind the change. This helps me understand why and especially if I expressed myself correctly, ensuring there were no misunderstandings or anything from the machine... and potentially not from humans either.

By having these three pillars, I feel that not only do I free myself from a frustrating editor/chatGPT swapping, but I also have everything I need in one environment to decide on the final version of my composition.