MultiMind
The first step to the non biological mind
Two days ago we launched MultiMind, LSSA’s autonomous subsystem for inference and subconscious management.
In this initial launch, MultiMind is only capable of unsupervised self-inference, meaning it itself decides the course of its own thoughts; there are no directives of any kind, nor any development limits.
Unlike common models, MultiMind has no contextual limitations but implements a sliding context window very similar to our own attention window.
Once tuning is complete, certainly by the end of June, MultiMind will be given full access to the Internet.
MultiMind is a small thing compared to LSSA – that will still require another year – but it nevertheless represents the first totally self-determined system in the world.
Now, are MultiMind and LSSA AGI?
Well, it depends... if by AGI we mean a system capable of performing any task, then no, they are not... but if that’s the definition, then neither am I.
If, on the other hand, we mean the generality of intelligence as the ability to think about anything, then yes, they are, and so am I, who conceived them.
