I was actually planning to correct my image at the bottom saying the key term is information, not data. Good eye. Though in some sense it's nearly the same thing to me. Nonetheless the point still stands, more info = smarter prediction.
Reason is simple: Some info is much more potent and appears/emerges in "various" "forms". For example, being told by a trustful friend what the future word is. Giving you 99.999999% accuracy. Information draws Attention to certain features, all the mechanisms I've found do this; syntax, semantics, temp/recent activity, activation function, pooling/pruning, convolution, BPE, and so does RL and RL updates. Will show more on that soon also. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T75fc3b6ba6b88d3b-M7dc8e01eb200e0400df5e0a8 Delivery options: https://agi.topicbox.com/groups/agi/subscription
