Monad update

PROJECT HERE: Monad

MONAD with more brain

I have been tinkering and proposing new ideas, the inital theories and mathematical framework behind Monad took a year to develop, so rapid progress has been made in the last months. Mostly theorizing and little actually coding.

Now have had the health and opportunity to start again, most of the time went into testing and improving coding frameworks, optimization and wrestling and wrangling LLMs to do my bidding.

Finally gotten a somewhat self recursive loops begnnings to work, the theory is more solid than the code at this point, since most of it is too novel to ever have landed in an LLMs dataset.

What the MONAD system is supposed to be able to do now is to compare data on its own, data that is already in the system vs data fetched from the real world and stop hallucinations and trial real knowledge and facts. Testing shows the MONAD system is able to do that but work has just started and most of it is still in prototype stage to test theories. Actual long term learning is still wip, but there is a plan for that also (short term memory and long term memory can be toggled on or off)

What lead to this new theory and design was how babies actually learn, I considered many options for cold starts, when a system with such vast information, where was it to start it’s knowledge gap journey? But easiest was to just spin from the last user message, just like babies when they hear their first words are set on a trajectory.

More is coming sooner than later, hopefully, as theories and code mature.