I found it annoying that I couldn't directly address individual segments in long AI responses and that I had to resort to copy+paste. So I build myself a solution and realized a lot of you must have the same problem.
ChatNote pretty much kills this now. I'm surprised myself how well it works and using it all the time.
Here's ChatNote in the Chrome Store and here's for Firefox. I even build one for Edge for those among you who are forced to use it.
ChatNote works more or less across the board. Currently this would be ChatGPT, Claude, Gemini, Grok, Kimi, DeepSeek and LibreChat. Perplexity failed so far, but haven't given up, yet. Let me know if you need other platforms.
Would appreciate some feedback for the extension. It still has a few bugs here and there, but overall works fine. Just uploaded version 1.2 which is for now the final development stage and should be green lighted by tomorrow.
Not sure now how to further develop it. I thought about a freemium version with no character limit, free color choice and more prompt variety for how and in which language to explain AI what the comments are about. The latter can make quite a difference in how AI interprets the comments. Please let me know how you would proceed.
Side note: I'm a complete beginner in programming. This was my first project and I had to rely 100% on AI. Cost me overall around $20 in Kimi tokens and 100 hours to build and it's rather astonishing to me how easy this was.
If this is the future...wow!
The interesting part is turning a personal workflow fix into something that becomes a daily habit for other AI users.
Curious what you think will drive adoption more — saving time on long responses, or the ability to build a better feedback loop with AI?
Good question. Personally, it helps me more with the workflow. But I think it depends on the particular user and AI use case. Some have long responses, for whom the time aspect is more important, others prefer short replies, perhaps with sequences of questions. It's easier to click the editor open and reply directly than having to number the replies in the main chat window.
That makes sense. I’m curious whether that distinction is coming mostly from how you personally use it, or whether you’ve started seeing different users consistently value one side more than the other.