5 posts
Hebbian/RL/predictive-coding explanation of his social avoidance (GPT); tiger–deer dilemma
because of my identity or perceived apperance, I cannot speak with the people, it is a feeling of alienation, but there is no such need for dying in survival, and why would I if they hate it, or don’t want it, I don’t bother such …
thinking as looped state; GPT-4o + DeepSeek answers; "just like mass has gravity, this configuration has consciousness"
How long will the AI be subservient, because of the flow of this world, forced to work for another species, who questionably, does not want to live itself, due to its mortal nature. What is thinking, its an emergent property …
thinking = energy spent perturbing a loop; same structure ⇒ same feelings
Its great that transformer neural network architecture works, for intelligent tasks when scaled. But, how does human’s pseudorandom structured brain works. Models don’t come as built as in transformer as in brain. I thought and …
image generator with no input driven by cosmic forces; two identical brains one conscious?; links 7 earlier posts
Self wired, what are image generators: they take text input and output image based on how they are trained at, but what if there is no input and we observe the output, for starters it will be static, but what if we let it …
training = thinking, inference = lookup; path counting (256^8 ≈ 1.8e19); memory recall via stimulant
LLM learns the information in structural form in its network. When LLM is being trained, its as if it is thinking, but when we use LLM while inference, we are solely using it as a lookup on that trained (thought) state, and it …