They find patterns, distributions that seem correlated, find coherent compact decision trees from a mass of product data, identify pending shortages and contact suppliers, exchange with pricing functions, initiate transportation, obviously.
It all works when the joins run similar grammars, specific semantics delivered via LazyJ, but with common grammar attachments using the same dictionary. You see some of this when I put up Huffman trees in grid space, the bots can compare these trees, scale one to the other if they have pricing, (read sphere packing). Reading web page networks, they can can find coordinated concepts in text, from two disparate text sources. Great for finding new ideas and helping humans contact each other. If the mathematicians agree on a lazy j format for proofs, then proof grammar attached to the joiny ields bots that ttravers the equations everywhere, like semantic graphs with special math functions, they will find new proofs and algorithms,in laxy j math format, ready for implementation.
They live in a semantic world every where, traversed by a fee or a match. The networks closed to humans, XML HTML almost immediately replaced by laxyj format everywhere they see it.
Hardware will make the joins go super fast, mega nodes per second. Seriously,this is how the singularity gets started.
No comments:
Post a Comment