Build Intelligence on Land You Own
Your AI isn’t limited by intelligence. It’s limited by who owns your data
I used to know these old school computer people who would keep harping on about why local first and private data is so awesome. To be frank, I just thought they were nerds finding problems for solutions.
However, with the rise of AI, things have shifted materially.
You see, your data right now is owned by various corporate overlords. Google for emails, Apple for digital life, Meta for social stuff. Now that’s usually not a bad thing since they give you a service. However in the age of AI these services restrict your data in very specific ways that make it hard for computers to get that data.
This is an issue because when you want your AI to be able to do things on your behalf in another service, it will have limits around what it can and can’t do. Many of these companies mask the problem by giving you access to their integrated AI solutions that make you feel productive and powerful inside their walled garden.
However, when you physically have all the data you care about on your hardware (either as files or in a locally run database), your AI has full access and context around what it needs. Couple this with the fact that open source models are becoming more and more capable, you have the power to have continuous digital intelligence working for you 24x7 at no additional cost (except electricity).
In contrast, those who have their data owned by external services will be at the mercy of their digital overlords and pay a heavy rent for their heavily marked up AI services.
The calculus for owning your own data has never been higher.
As an example, I now:
Have systems running that scrape my own data from various services
Run local databases that let my own agents crawl through whenever they want however they want
Moved off Notion so my notes can be in Obsidian which can be used by Claude Code
Utilise old hardware running various services that can plug into the rich local ecosystem I’ve built up
Every passing quarter I reduce my reliance on external software and data traps to increase the amount of compute I can personally harness.
All this to say, start working towards owning your own data because you’re going to need to for maximal leverage in the coming future.