Why We’re Building Things in This Order

When it comes to AI (or tech in general), it’s beyond tempting to run towards the latest shiny object, especially after seeing the promise that AI, from Open AI to stunning open source projects running locally, show. The problem is, when you’re focused on what is right in front of you, you can miss what’s on the horizon, or where the AI gold rush will actually take us.

That’s why Fulcra is an AI facing company that isn’t building an AI (yet?) Instead, we’re focusing on collecting, de-siloing, securing and centralizing the data your life creates, before making those bets.

Here are some of the reasons:

- While the current state of AI has undeniably progressed leaps and bounds, it is not yet advanced enough to make the most of the wide array of personal data that is currently available. You can get good results for personal applications, but you seldom get -wow- results, or true insight that is specific to your life, even if you spend hours sitting in front of the computer training your version of a Chat-GPT like program.

- Those wow results, results that are unexpected, surprising and delightful are more likely to come sooner if there is a broad degree of individual experimentation.

- By creating a secure, accessible, centralized store of your personal data (that you own, and that is in your control), we can abstract away many of the annoying parts of finding out what is possible with your personal data, while creating a secure environment where people can easily share their -wow- results, as well as the ability for other people to have similar -wow- results without rebuilding the experiment from the ground up.

- Perhaps most importantly, if we don't start -now- by collecting and protecting the data we want to provide to the better models we expect to get in the future, we run the risk of losing that data, and the value that further advancements in AI can provide us.

- A bunch of other stuff we’re not ready to talk about yet.

We believe the first step in building an intelligent software agent that acts as you would, and that can act for you without supervision is not just carrying out the countless experiments that will show us how to use AI to get those types of results; it’s also building the store of context that agent will need about you, your life, your patterns, your preferences, your relationships to enable that result. By starting with data collection (and enrichment, more about that later) we’re building a strong foundation, setting the stage for a future where AI can truly know us well enough to actually help us, at a level far beyond simple tax level automation and text generation. It’s not enough to speak with our voices; the best help is anticipatory and preparatory. For that, our systems need to know us, and we need to trust that what those systems know about us is secure and completely within our control.

That’s what we’re building first.