Maggie Appleton @maggieappleton.com 路 Jun 28

Gemini in Google Colab is a shining example of how *not* to integrate LLMs into a product. So much potential. Ruined by terrible design. Giant diffs where I can't accept/reject on a line-by-line basis. Unclear what context it has. Can't ref specific cells. Can't copy/paste code into chat! Wild

57 likes 4 replies

?

Replies

Maggie Appleton 路 Jun 28

I hope they shipped this under some bananas deadline because a VP forced them to. No one could have dogfooded this and said "yup, that is a useful tool that does what I need it to in most situations"

Loque 馃 路 Jun 28

wow that's interesting and funny how they copied the cursor approach... which to honest, is challenging to keep up with - it should be about small changes to help reduce cognitive load, with the 'why' spelled out, and an overall plan/goal, not complete re-writes that take a long time to parse 馃珷

John Allsopp 路 Jun 28

Similarly Gemini in Google docs. Terrible terrible UX and UI.

Archived fry69 路 Jun 28

Part of this problem may be the underlying JSON file structure for Jupyter Notebooks. Marimo solves this (notebooks are valid Python file) with interesting side effects/improvements ->