> > > > so some of the criticism energy around/inside me atm releates to > > > > things that are technologically soudn and real, vs those that are not > > > > > > > > i spend a lot of time these years engaging technological fantasy, > > > > because part of me has worked very very hard to prevent me having > > > > skills and effectiveness. the fantasy is nicer and more helpful -- it > > > > lets me defend part of my skillset so long as i don't use it > > > > > > but i do try to do real things to, i just have to kind of approach it > > > sneakily and keep it small and rare and uncertain > > > > ... thought not completed > > > > _in other news_ maybe lets have _karl_ write a maze cutter and solver! > > > > this is an introductory task that is not intuitive to everyone, but i > > grew up with algorithms so it used to be intuitive to me. > > > > i challenged a language model to do it in 4 lines. this is because of > > how language models are trained. me, i could do it in four lines > > _only_ if i prepared by imagining or doing it in many more lines, and > > then shrunk it down. we could combine that by putting the imagination > > into notes, and keep the notes in 4 lines, but it's more relevant that > > how people work and what they present are different, as well as people > > working differently from machines > > or maybe we could move AI forward ! > > for some time i had two favorite contributions to opencog i would do > if energy built some day: > 1. it didn't have subcontexts which i think are highly important, they > are a simple addition > 2. it uses esoteric terms which reduces its developer base. given it's > AI, all its terms are analogous to everyday concepts. i wanted to make > a sugar interface that used everyday concepts so it would be intuitive > to use it. > > of course opencog is now getting supplanted by a new language that is > less flexible, the old approach of delineating the meaning of thought > still needs a home
or maybe something nitty-gritty i dunno but have you seen my recent hobby work with ndarrays? recently i implemented big integer multiplication --
