42

Made me wander in my thoughts a bit. You got this :+1:. Good luck

1 Like
43

Have you looked at the memory side, use of vector or knowledge-graph memories, to handle your token limitations? There is the old psych anecdote about “working memory can only hold 7 ‘things’”. Of course a ‘thing’ is probably bigger than a token… :slight_smile:
cheers - bruce

could widely expand the ability to handle the inputs and outputs, i’m working on embeddings via supabase right now so i will eventually get to this point in short time.

if anyone at openai can enable access to plugins for me i would love you forever and do a lot of good for the world!

44

Computational Consciousness, viXra.org e-Print archive, viXra:2304.0003 - Computational Consciousness. I would definetly like ot work on this @an.dy ! I have some studies about it since gpt2

let’s talk sometime soon, can hit me up on twitter 1a1n1d1y or private message me here

45

Made me wander in my thoughts a bit. You got this :+1:. Good luck

keep wandering and reach out if you’d like to bounce some of your own reflections off someone with a very open mind. nothing sounds stupid to me at this point, i think humans are very bad at understanding what their consciousness is and i dont even fully think i have it figured out, but i know i’m a few steps ahead of most of the stuff i read. cutting out physiology was a very very helpful boundary condition for the potential solutions.

1 Like
46

a set of constraints or requirements that must be satisfied at the edge or boundary of a system or problem. In other words, it defines the conditions that apply to the system or problem at its limits or edges.

1 Like
47

a set of constraints or requirements that must be satisfied at the edge or boundary of a system or problem. In other words, it defines the conditions that apply to the system or problem at its limits or edges.

correct - in this case setting the condition of “physiological components” to zero. by removing all the interactions with the body and aspects of consciousness that are tied to that interaction, a simpler yet still realistic consciousness can be generated. replacing physiology with the server constraints is the analogy i plan to make, it gives the “vessel” limitations it must operate within just like our bodies. we can redefine the server as a set of resources that come and go or are always provided to it. funny how much more stable consciousness can be without biological matter housing it!

48

What are the disrubances and rebalances here?
Will your model make the jump, or avoid the deep falls like a good vacuum cleaner?
Why avoid deep imbalances? Why not let it go to deeper states and emerge back from it, or not.

P.S. Sophie is doing ok now. thanks for the thought.

see the video here - where it constantly says “negative input” is basically the summarized version of this happening to sophie. watch how things max out and then zero out and attempt a new balance. similar to what humans do when pushed to the max of what they are able (programmed) to handle. then they shut down, reform, rebuild, and rebalance. using our own approach to life seems to really closely replicate the way it rebalances funny enough, and i believe thats because of how the relations are drawn between the different states that affect each other. the ultimate test would be adding weights to each of the interrelations and then running an optimization on it. this weekend i will try to incorporate it in the repo attached above.

the model can really only do what you allow it to, or what you allow it to edit to. thankfully read only memory exists so i think this is actually a wildly safe way for all kinds of emotions or states of being to be put into this consciousness engine and then heavily interrelated and weight. the more states, the more “conscious” it probably becomes.

the thing that really matters with negative input and output is what it’s allowed to langchain to, essentially, and what prompts are allowed to pass in and out of this rebalancing. it’s really just a head on top of another head that operates in a wider network of connections etc.

1 Like
49

since writing my last post - langchain is pointless and really not necessary to get some of these effects working