human (un)intelligence

[gusl]

Suppose you (B) and your friend (A) are expecting to meet a friend of theirs (D) who you've never met before. You're running a bit late, and just as you get there, you see someone (C) who is walking away, so must decide ASAP whether this is the person you're supposed to meet, so you can decide whether to yell. Are C and D the same person?



A can't see C.

If humans were digital creatures, B could instantly communicate C's image to A. Better yet, A could have instructed B on how to recognize D as effectively as A herself.

However, as a human, A and B must use English to tell each other these things. This is an awful solution! It doesn't bother me so much when machines are dumber than humans. It bothers me more when humans are dumber than machines.

Comments

[surrey-sucks.livejournal.com]

Why is A standing where he/she is? Is A running late too? (That was a little ambiguous.)

[gustavolacerda.livejournal.com]

just when I thought I had made a good cover story...

Just imagine that A is tired and taking advantage of B's wake.

[surrey-sucks.livejournal.com]

You math/physics/computer people think much differently than I do!

[gustavolacerda.livejournal.com]

what kind of person are you?

[surrey-sucks.livejournal.com]

The total opposite.

I don't really know, but I prefer chemistry and biology.

I'm also the kind of person who over thinks things. Such as: do person A and person C have cell phone? And if so, why doesn't person A call his/her friend to tell him/her that you guys are late?

[0436.livejournal.com]

Gosh, I wouldn't use "dumb".
Gosh, I can't ever relate with you.

[gustavolacerda.livejournal.com]

Interesting. This kind of overthinking may be a cause and/or consequence of you not being a math person. Mathematical thinking involves abstracting away from such details.

Ironically, the intended intepretation of cover stories / word-problems often involves some subjectivity, some (possibly culturally-relative) common-sense. To take an example from combinatorics: "There are 10 men, 5 white ties, 5 black ties. How many ways are there of matching men to ties?". Now, replace "ties" with "women", and the interpretation (and answer) will change: unlike ties, women are not exchangeable.

[gustavolacerda.livejournal.com]

I use "dumb" as a technical term. Do you have a better term?

[the-locster.livejournal.com]

Better yet, A could have instructed B on how to recognize D as effectively as A herself.

What you mean like showing B a picture of D beforehand? :)

[gustavolacerda.livejournal.com]

Note also: this is presupposing that everyone is heterosexual and all M-F matches are possible, etc, etc. I did warn about cultural relativity.

[surrey-sucks.livejournal.com]

this is presupposing that everyone is heterosexual and all M-F matches are possible, etc, etc.

See, these are the types of things I would start thinking about.

I know you meant the original problem to be answered mathematically, but I lack the ability to think that way; I am just not mathematical at all. Because I am deficient in thinking mathematically, I start trying to find flaws in the question asked, thus annoying everyone :) And that is the kind of person I am!

[http://users.livejournal.com/_tove/]

If humans were digital creatures that didn't have some of the dumb skills we have, even A might have a hard time recognizing C, especially since C is facing away, and probably wearing different clothing and lit differently than the last time A and C interacted.

[0436.livejournal.com]

High five.

[en-ki.livejournal.com]

If I were B, I would yell "hey, D!" and if C were not D, C would ignore me or would be somewhat bewildered. In the latter case, A would then show up and help appropriately.

As _tove seems to be saying, humans' tendency to abstract (to the point of not being able to communicate raw data efficiently) is a tradeoff more useful to us than not. Sure, I'd like to be able to freely store and copy and alter and communicate arbitrary subsets of my sensorium—sign me up as soon as the devices come out (cell phone cams anyone?)—and that's why I got interested in copyright law in the first place.

I wonder if there is a quickly-teachable Soundex for faces.

[gustavolacerda.livejournal.com]

ooh... you mean like this??

[parentheses.wordpress.com (from livejournal.com)]

B and D have seen each others' Facebook profiles.
B phones/shouts to A to call D.
A gave D's number to B in case B arrived first - B calls D.
B shouts, "Hey D!" Worst case, it's not D: "Sorry, wrong person!" Then there's a chance B and C have a chat.
...

[parentheses.wordpress.com (from livejournal.com)]

Do you know about verbal overshadowing? Could be a nice experimental setup to test you verbal code for facial features. Verbal overshadowing is moderated by individual differences in verbal ability - something to bear in mind...

[parentheses.wordpress.com (from livejournal.com)]

"Interesting. This kind of overthinking may be a cause and/or consequence of you not being a math person. Mathematical thinking involves abstracting away from such details."

Does mathematical THINKING involve this, or the intended end result?

[gustavolacerda.livejournal.com]

Cool. The process of giving description in English labels is essentially lossy compression: there's no going back. By communicating to B, A is impairing his own recognition.

Re: "moderated", is high verbal ability correlated with a stronger overshadowing effect?

[gustavolacerda.livejournal.com]

The point is that the information cannot be directly communicated from A's brain to B's brain. Not portable.

Currently, this is because of the channel: English is a terrible format for images.

But even if we solved that by adding some sort of direct brain-to-brain channel, there might be coding compatibility issues. Computers have standard formats (like GIF), but humans might not. However I believe they've shown that the neural codes for low-level vision (V1, maybe V2) are pretty much the same for everyone. [Add caveats/qualifiers/provisos here]

[parentheses.wordpress.com (from livejournal.com)]

"By communicating to B, A is impairing his own recognition."

Not quite :-) Unless A has just recently seen D's face...

Higher verbal ability => less of a detrimental effect of verbalising the facial description on facial recognition.

[gustavolacerda.livejournal.com]

hypothesis:

higher verbal ability => more precise language => more faithful (i.e. less lossy) compression => less of a detrimental effect of verbalising the facial description on facial recognition.

[gustavolacerda.livejournal.com]

btw, have we met? how did you find my blog?

[parentheses.wordpress.com (from livejournal.com)]

Yes, that was their idea.

[parentheses.wordpress.com (from livejournal.com)]

Yes we have (figuraleffect.googlepages.com)

[easwaran.livejournal.com]

I don't think it's because English is the channel. We certainly don't use verbal descriptions to recognize people we know, but I think we also don't use an image to recognize people we know. That is, there are lots of people I would be able to recognize instantly, but can't necessarily visualize what they look like well enough even to know if they have blond or brown hair, or how tall they are. The bigger problem might be that our recognition abilities are completely outside of our conscious abilities.

[gustavolacerda.livejournal.com]

dude, why so many blogs / identities?

[figuraleffect.wordpress.com (from livejournal.com)]

:-) I'm just down to two blogs and one website now!