On Saturday, November 13, 2021, at 4:52 PM, Rob Freeman wrote: > Vector Parser - Cognition a compression or expansion of the world? - AGI-21 > Contributed Talks > https://youtu.be/0FmOblTl26Q
Rob, I enjoyed the talk you gave there is a lot of good information covering a range of topics. I agree with expansion, I view it really as both. Take human memory, some rare people can actually recall every day of their lives. I believe that we all could though we don’t actively recall due to various reasons. The brain’s visual memory though could be simply modelled as lossy compressions streamed onto an internal expansionist structure. This model would then be (many to less-many without inverse?) IOW a multi-source, single target lossy compression into expansion. By taking the various components of the compressionist systems model and essentially expanding them we can present extended operational models with various utility. An example would be a compressor that learns across compressive actions instead of only within one, or compressors that share knowledge by operating in parallel, or say a compressor that is single source multi-target like compressing into an assemblage of partitioned targets. Some of these then blend into other popular AI structures like neural nets. There is no fine line it seems. The expansionary model I’ve focused on is a single target duality where there are compressionary and expansionary forces occurring simultaneously since human general intelligence is a multi-agent intelligence and the single target duality is related to the consciousness in the agent where consciousness is communication protocol. Agents perform compression detection with other agents in order to communicate lossylosslessly with language... ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T5ff6237e11d945fb-M925c6de5097cea65e91de96f Delivery options: https://agi.topicbox.com/groups/agi/subscription
