On Mon, Jun 19, 2017 at 3:26 PM, Hugo Latapie (hlatapie) <[email protected] > wrote:
> Thanks Linus. The approach here does look extremely promising. > > > > Bridging the gap between these various camps is the holy grail that few > are even searching for much less attempting to implement. > Thanks, But its not that I have some brilliant insight; its just that I have a minor insight, and prefer to actually experiment, and see if it works, rather than theorize in some grand fashion. Theories and insights often seem grand when one is in the shower, but if they won't work, I prefer to find out ASAP, rather than to continue imagining how great they are. --linas > > > -Hugo > > > > > > *From:* Linas Vepstas [mailto:[email protected]] > *Sent:* Monday, June 19, 2017 1:10 PM > *To:* Hugo Latapie (hlatapie) <[email protected]> > *Cc:* link-grammar <[email protected]>; opencog < > [email protected]>; Ruiting Lian <[email protected]>; > Word Grammar <[email protected]>; Zarathustra Goertzel < > [email protected]>; Hugo deGaris <[email protected]>; Enzo > Fenoglio (efenogli) <[email protected]> > *Subject:* Re: [Link Grammar] Cosine similarity, PCA, sheaves (algebraic > topology) > > > > > > On Mon, Jun 19, 2017 at 2:11 PM, Hugo Latapie (hlatapie) < > [email protected]> wrote: > > *arXiv:1702.00764* > > > > I've just barely started reading that, and from the very beginning, its > eminently clear how even the latest, leading research on deep neural nets > is profoundly ignorant of grammar and semantics. Which I think is another > reason why the direction we're on is so promising: apparently, just about > exactly zero of the researchers in one area are aware of the theory and > results of the other. > > Which I guess is a good thing for me, But its really really hard to read > that paper, and not want to scream at the top of my lungs, "those > ding-a-lings, don't they know about result xyz? what's wrong with them? Are > they all ignorami?" and yet, it seems to be a giant pyramid of results > built on results demonstrating a lack of knowledge, education about form > and structure. So it's a bit hard to take seriously, and yet, everyone who > is interested in deep learning seems to be doing just that... its > remarkable. > > --linas > -- You received this message because you are subscribed to the Google Groups "opencog" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/opencog. To view this discussion on the web visit https://groups.google.com/d/msgid/opencog/CAHrUA36m3thLpQQnLus03Q7fZuXxeRvpCoUW%2B3FZ0NtqFh%3DB3Q%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
