5 Tips for using Logseq with ChatGPT
Вставка
- Опубліковано 9 лип 2024
- Using AI to help process your notes saves a ton of time, if done right. So here are 5 tips that I use on a frequent basis when refining my own notes.
==== 🧾 Table of Contents 🧾 ====
00:00 Intro
00:35 Tip 1 - Ask for Logseq formatting
01:10 Tip 2 - Be specific
04:49 Tip 3 - Optimize your notes in 3 levels
07:34 Tip 4 - Talk to make notes
17:50 Tip 5 - Easy export limited data for ChatGPT
20:17 Thinks to keep in mind with LLM and Privacy
26:40 Thanks for watching
==== 👋 Contact 👋 ====
Website - toolsontech.com/
Twitter - / toolsontech
Why some of the blocks are colored in orange, and what's the reason behind it? Do you not use headers or is it for another reason? I'm still figuring out what note style is the most useful.
I use a small css snipped to color bold text orange, that way it pops out and is easier to spot when reviewing my notes.
Thanks for the video. I'm surprised it is acceptable for to copy/paste the data between Logseq and the ChatGPT client. There are a few logseq plugins that connect directly to the ChatGPT API (or any other compatible API with alternative models) so you can have the LLMs interact directly with your notes as you like, and have the results tagged and orchestrated so that is always clear what notes are your own and what is generated by a LLM.
While I tried a few of the plugins, the tradeoff is usually a steep drop in quality on the interactive part. So either it's prebuild queries that get me a specific result and stop me from refining the get the exact result I need. Or chat is build into Logseq but a very poor implementation compared to the ChatGPT chat interface. I'll keep an eye on the plugins obviously as I expect them to get better.
Great video!! Any news on the database version?
Not yet, but you can be confident that as soon as their is I will be making a video about it. And posting on Mastodon/X