Our life's work, and something which could have reshaped the world 30 years ago, had we been more aware of the skills we'd need to ride and counter the internet.
A global brain, a technology that the world still has not replicated some 30 years and trillions of dollars later.
Not AI, but something that can now partner with AI.
Aurora is very much alive. We hope that this time, we can push it into the light of day.
Sat at my desk, long before the Internet, before we even had Microsoft Windows on our PC's I was manually typing data into a Symphony spreadsheet and asked:
"Why can't information flow naturally to me? I'd look smarter and be paid more."
Trillions of dollars invested in the Internet, and what we were about to invent still hasn't been replicated. Not only that, it's impossible for the world as you know it, for a reason we'll cover shortly.
I began to ponder what would be needed. I considered the technologies then available, which still dominate today.
Spreadsheets are invaluable but data is trapped inside a file. How many spreadsheets are there in the world?
Databases are incredibly powerful but data is still trapped, requiring skilled engineers to access and make sense of the data. And again, how many of them are there out there?
Documents are the meat and drink of knowledge sharing, but how many of them are there out there?
Those are three dominant data sources, but how many different applications also contain relevant data? And again, how many instances of those applications are out there?
My exploration was crystallising into a core understanding, which could be expressed as follows.
The solution had to be:
A single source of information.
Accepting structured and unstructured data.
Supporting spontaneous immediate contribution.
Allowing further refinement (eg: creation of relationships) in situ.
I looked at the then available technologies, most of which are still relevant and mainstream today:
RDBMS
Neural nets
Expert (rule based) systems
Blackboard systems
Spreadsheets and documents were clearly non-starters, RDBMS had to be taken seriously, but wouldn't serve.
Neural nets were in their infancy, but I wanted the real answer, not an approximation (today understood as AI hallucination).
There simply wasn't a technology that had the qualities I needed to create a single coherent store of all the world's information.
And then I made the biggest mistake of my life, technically three mistakes. Read about them here.