On Wednesday, May 26, 2021, at 11:23 AM, James Bowery wrote: > In AIXI terms, the difference between lossless and lossy compression is the > difference between AIT's and SDT's notion of "utility": The former being > concerned with what "is" and the latter being concerned with what "ought" to > be via a "subjective evaluation".
I don't look at compression as either lossy or lossless crisp dual. There is the vast region between. For example, I would consider OCR (Optical Character Recognition) as a compression. The characters extracted are lossless and the noise and background is lossily compressed out. Extend this example to any character OCR'ed in any known language, to characters in unknown languages but a “language” required in a context of environment (like learning an object recognition set) or unknown characters in a known language context in an environment, to unknown characters in unknown languages in unknown environments. Relegating compression to either lossy or lossless is similar to the above example of OCR’ing out two known characters in a language. Which has a narrow purpose. But there are unknown compression scenarios in unknown languages in unknown environments, which is more of a General Compression but GC is still not AGI IMO... do you guys think GC is AGI? GC may be able to be reformed into AGI and vice versa… in fact I may test my AGI model by reforming it into a “GC” but the goal there is not a maximum narrow compression but rather an efficient general compression… ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T95f11a183fb9b6e1-M52f3134237396d5b41832f23 Delivery options: https://agi.topicbox.com/groups/agi/subscription
