Show HN: Kuse – A visual, non-linear infinite canvas for ChatGPT (app.kuse.ai)
Why we built this:
- We found the typical linear-based chatbot is not enough for handling complex conversations. - Just like you'd use a whiteboard for detailed discussions with your teammates, Kuse provides a visual interface for dynamic, non-linear conversations.
How it works:
- At the core of KUSE is one magic block. Behind the scenes, it’s multi-modal LLM with many integrated tools: image generation, image editing, search, etc. - Within the infinite canvas, you can always divide conversations to bulletpoints or group multiple branches together. (It’s not just tree-based, it’s actually graph-based! Super flexible.)
Our beta testers are using Kuse to do brainstorming, travel planning, project management, research, content drafting, etc
It’s completely free now. Lots of features to build going forward: Collaboration, more whiteboard (text, drawing, etc) capabilities, semantic search, code execution, etc… Any feedback is appreciated. Thanks!
More use cases are inside Twitter/X: https://x.com/kuseHQ
28 comments
[ 2.9 ms ] story [ 69.3 ms ] threadWorkflow builder is an interesting idea, could you elaborate more? Did you see any similar workflow builder?
It's just a little uncomfortable to use the mouse for this, I would prefer it to be as smooth as some whiteboard Apps.
I'm curious about how well it handles real-time collaboration with multiple users, especially when AI-generated content is being added to the canvas. Has anyone here tried it out? I'd be particularly interested in hearing about the learning curve and how it compares to other tools like Miro or Figma in terms of performance and feature set.
Also, for those who've used it: how does the AI integration actually work in practice? Does it feel seamless, or is it more of a bolt-on feature? And how customizable are the AI models - can you fine-tune them for specific use cases?