Sifting through your BLOB of text above here are your actual questions: "learning through human agents, that is your own conscious brains. The compressor improvements aren’t communicated compressor to compressor"
"Your compressors are isolated hard-coded hyper-AI agents, not AGI." It doesn't take much to feed its predictions to the rest of a sentence/ images back into its storage tree/ hierarchy web. When we do that, its predictions being stored will govern what it predicts in the future, I won't say how because its too powerful an idea, we are really close now. For example I predict 'AI is smarter than me' (and so this makes me forget my cryonics research and am now interested in AGI) and so now I talk to myself about AGI issues and speak AGI into Google and eat data about ML, it's not that I search Google, it's just that I can't get AGI off my lips and it coughs out into everywhere I go - including Google. My predictions return me AGI data - by generating to myself or into Google's search bar. My AI and OpenAI's DALL-E are not hardcoded in the sense that they are not general purpose, they are, and soon will be making their own rules up by ex. linking together thoughts with motor actions, it is not hard to see DALL-E is very general purpose and can get so general purpose that it starts storing and using pieces of memories to do very rare prediction solving. ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T5d9be03603e8ba20-M5aa0bb5e79a9f4e9923db313 Delivery options: https://agi.topicbox.com/groups/agi/subscription
