Steph Ango writes -
File over app is a philosophy: if you want to create digital artifacts that last, they must be files you can control, in formats that are easy to retrieve and read. Use tools that give you this freedom.File over app is an appeal to tool makers: accept that all software is ephemeral, and give people ownership over their data.
One of the reasons for the Obsidian renaissance has been the rise of the AI era, which thrives on parsing and interpreting open standards such as .txt, .csv, and Markdown (.md). Steph has already touched on this idea on social media, but Obsidian was uniquely positioned for this moment by embracing the “file over app” philosophy from the beginning and building around portable, accessible formats.
Notion has deservedly received a lot of positive attention from AI enthusiasts, and there are real strengths to its approach. However, exporting your data from Notion still often involves:
- Using their relatively limited export tools
- Relying on MCP connections, APIs, AI tooling, tokens, or subscriptions to transform and migrate that data into another format
I completely understand the appeal of the Notion model for many users — simplicity, collaboration, polished UX, managed infrastructure, and secure cloud storage are all meaningful advantages. At the same time, I’ve found myself increasingly drawn toward open formats that are not locked behind proprietary systems or dependent on a single application to interpret them.
Apple has taken some encouraging steps here as well. Last year’s addition of Markdown import/export support in Apple Notes was a notable improvement. But even today, directly reading from or interacting with the Apple Notes database remains cumbersome for AI workflows because of its proprietary structure.
What makes Obsidian compelling is that by embracing Markdown as a foundational standard, it gives users confidence that their data remains theirs. If Obsidian disappeared tomorrow, those notes would still be readable, searchable, accessible, and usable across countless other applications and AI systems.
That increasingly feels less like a niche preference and more like an important long-term design principle for the AI era.