I can access ChatGPT 3.5 and 4.0, Bard, and Claude 2. It’s interesting to compare the content they generate for the same prompt. What I particularly like, though, is asking one chatbot to critique the text generated by another or even critique its own generated text.
One of my typical writing sessions unfolds like this:
First, I draft a post myself, often not paying much attention to grammar or spelling.
Next, I ask, say, ChatGPT 3.5 to clean it up and possibly expand on it as it sees fit. Sometimes, I will ask two or more chatbots to clean up the post, then choose the version I like most. Alternatively, I may select specific sections generated by different bots and combine them to create my desired result.
Afterward, I may request a chatbot to add or remove items and adjust the emphasis on certain aspects of the post.
Then, I ask one or more bots to critique what’s been written.
Based on the received criticism, I update the text and put it through a bot for an almost-final pass.
Finally, I meticulously review the text, making further tweaks and changes I deem appropriate.
Oh, and one last thing — the consolidation takes place in Evernote before I transfer the final text to a post on my WordPress-based website. In WordPress, I can access Grammarly, which I use for one last edit.
Though slightly elaborate, it’s an exceptionally productive approach.
Tweet: Writing with GenAI chatbots
Knowledge Letter: Issue: 278 (Subscribe)
Tags: Bard (3) | chatbot (18) | ChatGPT (17) | Claude (6) | Evernote (5) | genAI (5) | Grammarly (2) | WordPress (2)
RSS: Blog Feed
Photo Credits: Midjourney (Public Domain)