Context: Shannon and my associations First category Grouping) Learning : Effectiveness : Redundancy : Low Entropy : Sparse : Many Patterns To Find : More relaxed : Large b) Production : Efficiency : Surprise : High Entropy : Compressed : Looks like noise : Tense : Small My gut says Database might be First category Grouping) and Algorithm might be b) but I’m not 100% if they’re opposite. My intuition says they are but I need confirmation which I’ll look for later as it’s unlikely to be an original thought overall, just original to me.

Context: Shannon and my
[read full article]
 

I’ve never been crazy about the idea of naming it. Seen way too much abuse of that. “God says”, “God wants” this or that when it was clearly the opinion of the person saying it. Hinting at’s good enough I think. There’s more wonders than we’ve yet discovered. There may be some things that are just a bit beyond humans. Stuff like that.

I’ve never been crazy
[read full article]
 

in that concept-similarity would be a more fundamental notion and will imply determine their values. But no satisfactory mathematical theory of concept-similarity has been developed yet. In SCOP, similarity, membership, typicality, and any other measure, is modeled as a measurement-operator that acts on the state of the concept, in the same way as context do.

But no satisfactory mathematical … [read full article]

 

I’m still amazed at my visceral reaction. I re-researched bits of exemplar theory and thought of … well, examples where I use examples for generalizations. But it might be that I generalize _too_ close to the specific. Drawing broad generalizations “hastily”. Not hasty to me as it’s just how my brain works — but apparently I should put a bunch side by side and draw lines. I think I generalize “differently” than I’m supposed to. Chronically. As in, that’s just how I am. So if I’m working from exemplars, I get results that don’t always match up to reality — or frequently don’t. But prototypes, which are general rules, don’t have to correspond to anything specifically but can be broad enough within their own ranges to encompass any exemplar sufficiently that one might throw at it that is accurate to the meaning of the prototype. Oh, gosh that means I’m endorsing gaussian, smeary vague mono … but wait, no it’s not because it’s hierarchical. It has different scales. Not a scale free system. Each scale can be functionally independent until you get to the interfaces, although upon zooming in they do interact. Hm.. ok, I need to start mapping these on paper visually.

I’m still amazed at
[read full article]
 

Minimal Human Instruction Set? So: I read about a proposed Universal Tactile Language for Visually Impaired with multiple disabilities. 55 proposed symbols from teachers from several schools that were spread out in the northeastern USA. Well, out of the 55, which ones worked? I found the follow up study. They didn’t test them all but of those they did, 14 got a 40-60% success rate but 7 were mastered by most: “Wheels on the Bus,” “Twinkle Twinkle Little Star,” “Itsy Bitsy Spider,” more, finished, juice, sensory. If this was an instruction set for a computer to bootstrap functionality, could it scaffold into a dictionary of functions with more and more meaning? Well, maybe. Wheels on the Bus could be the continuity of time, a crucial feature of computing that standard mathematics has as optional. It allows for step/time based functions. So, I’ll put Wheels on the Bus as expressing that. Twinkle Twinkle is “expanded workspace” / buffer / pointer system. It’s visible yet, out of direct reach. Pointers. Maybe Push/Pop stack or “memory pages” (like Spreadsheet Worksheets). So I’ll go with that. Itsy Bitsy Spider today I realized fits the monomyth. it’s in four distinct states: status quo, rain conditions, drying conditions, loop condition (“again” – a counter). This means it’s loop – example here is a game loop for comparison: while (user does not exit) check for user input run AI move enemies resolve collisions draw graphics play sounds end while It will presumably rain regardless of the spider’s position but if the user is in the water spout/gutter, the rain which gets the inside of the water spout slippery, any thing in the water spout will slide down, whether user or other kinds of enemies not introduced. User interactions with AI include: more, (stroke red velvet cloth) finished, (pull on spool at the end of a string) juice, (pick up juice box shape) sensory. (pick up lotion bottle shape) Sensory is akin to an averaging / smoothing / abstracting function as it smooths the skin, calms the nerves, smoothing. Juice is energy. More/finished sound like diffusion controls at the very least. More on that later. https://journals.sagepub.com/doi/abs/10.1177/0145482X1310700303

Minimal Human Instruction Set?
[read full article]