2 ms·

"... I wrote the post in question and I actually registered with Hacker News for the first time just to reply to you. As far as your comment goes, I apologize i

"... I wrote the post in question and I actually registered with Hacker News for the first time just to reply to you. As far as your comment goes, I apologize in advance for being harsh, but get over yourself. ..." HN is attracting its share of "1 quip snipers", trolls that criticise but not really bother to expand why & generally make a nuisance of themselves ~ http://www.paulgraham.com/trolls.html http://www.paulgraham.com/trolls.html but we're not all like this. "... But if you look outside of that self imposed box you see there's so very much left to do in technology ... Now I've found that dating someone gives me a tremendous opportunity in bars because rather than be a part of the dating rat race I actually get to observe it as an outsider which is really enlightening ..." This was the standout part of the post that struck a chord with me. It's the reason I posted it. A hard problem at the moment I see at various levels is the "human" matching problem. I often wonder the reasons why this problem is tackled so badly in software? Is it because the same thing that makes nerds good at computing also degrade their ability to undertand "human matching"? It's not just dating where solutions to this problem can be applied. It could also be applied to anywhere you have a requirement for recruiting individuals to interact with other people. Part of the reason why the problem domain is not well understood by nerds could also be covered by the "homo-logicus", "homo-sapiens" divide Alan Cooper describes in "Inmates are running the Asylum". Funny enough this idea also partially explains Trolls. Nerds posting insults like jocks giving wedgies. The big problem with the matching humans from what I can gather talking to people who try pure software solutions, is the lack of ground-truth. What appears on a site just doesn't match real life. Nothing beats the opportunity to evaluate a human in real-life. Any augmentation to help filter this process may have pay-off in a lot of other areas.

benjamincanfly 18y ago
Thanks for pointing out the fact that my comment wasn't constructive. Didn't mean to troll, should have kept my thoughts to myself.
TomOfTTB 18y ago
Well first, thanks for posting this. Always nice to know someone enjoyed a post. On what you said... The big problem with the matching humans from what I can gather talking to people who try pure software solutions, is the lack of ground-truth. What appears on a site just doesn't match real life. Nothing beats the opportunity to evaluate a human in real-life. Any augmentation to help filter this process may have pay-off in a lot of other areas. I think you’re right that people lie but I think statistics can, to a certain extent, help with that. To give an example, let’s take a guy who loves Gilbert and Sullivan musicals but is ashamed of it so he says he hates theater productions. But then in response to other questions he reveals more truthful answers. Eventually, a statistical measurement will determine that Women who love musicals match best with guys who claim not to as long as they fit those other traits. So the computer never manages to decode the lie but it still, through the magic of statistics, manages to make the match just the same. I’m sure we’re years and years away from actually getting to the point of having an effective system like this but just the fact that we seem to have the power to do it is pretty amazing.

benjamincanfly 18y ago

Thanks for pointing out the fact that my comment wasn't constructive. Didn't mean to troll, should have kept my thoughts to myself.

TomOfTTB 18y ago

Well first, thanks for posting this. Always nice to know someone enjoyed a post. On what you said... The big problem with the matching humans from what I can gather talking to people who try pure software solutions, is the lack of ground-truth. What appears on a site just doesn't match real life. Nothing beats the opportunity to evaluate a human in real-life. Any augmentation to help filter this process may have pay-off in a lot of other areas. I think you’re right that people lie but I think statistics can, to a certain extent, help with that. To give an example, let’s take a guy who loves Gilbert and Sullivan musicals but is ashamed of it so he says he hates theater productions. But then in response to other questions he reveals more truthful answers. Eventually, a statistical measurement will determine that Women who love musicals match best with guys who claim not to as long as they fit those other traits. So the computer never manages to decode the lie but it still, through the magic of statistics, manages to make the match just the same. I’m sure we’re years and years away from actually getting to the point of having an effective system like this but just the fact that we seem to have the power to do it is pretty amazing.