3 ms·

I don't know if our behavior can really be predicted on such a fine grained level. Sure, you could generate a dashboard based on most viewed information. This

by jiggy2011 13y ago

I don't know if our behavior can really be predicted on such a fine grained level. Sure, you could generate a dashboard based on most viewed information. This wouldn't rocket science and could be done today. OTOH, pulling up google maps a few days before fathers day? That could be for any number of purposes unless it knows that you always go shopping for gifts precisely 3 days beforehand.

dpcan 13y ago
Because of my Calendar and my todo lists, and my emails with friends and family, and my searches on Google, I do think that Google could know precisely what I may be trying to map as soon as I open my maps app on my phone :) I just emailed Mom asking what Dad might want for Fathers day, and I told my wife I'm headed out sometime this week in a text on my phone, and I searched Google for products that took me to Amazon.... So, when I pull up my map one afternoon soon after these communications, I think it will know what I'm up to - or could at least show me a popup saying "Headed Father's day shopping? Want some locations based on your recent web activity?" I say Yes, and all kinds of fun spots and deals popup.
- fumar 13y ago
  Google Now does some of these things. Now provides me for directions almost everyday. It also lets me know what what movie times are relevant for me. I was recently in NYC, Now gave me relevant things to do while I was there. They were great suggestions. It blew me away.
- jiggy2011 13y ago
  But how does it know which trips will be father's day shopping and which will be for some other purpose? Assuming you use your car more than once that week. The result it likely to just be constant annoying spam.
mayank 13y ago
> I don't know if our behavior can really be predicted on such a fine grained level. There was a project in the MIT Media Lab something like a decade ago where they stuck about 100 graduate students with Bluetooth-enabled cellphones (this predates smartphones) and tracked many of their behaviors over the course of a semester. Things like phone calls, personal interactions, movements, etc. A very interesting result from that study was that they built a model for predicting daily behaviors for their subjects with up to 79% accuracy [1]. The whole project was groundbreaking at the time, and they have some very interesting papers. [1] http://realitycommons.media.mit.edu/pdfs/eigenbehaviors.pdf http://realitycommons.media.mit.edu/pdfs/eigenbehaviors.pdf

dpcan 13y ago

Because of my Calendar and my todo lists, and my emails with friends and family, and my searches on Google, I do think that Google could know precisely what I may be trying to map as soon as I open my maps app on my phone :) I just emailed Mom asking what Dad might want for Fathers day, and I told my wife I'm headed out sometime this week in a text on my phone, and I searched Google for products that took me to Amazon.... So, when I pull up my map one afternoon soon after these communications, I think it will know what I'm up to - or could at least show me a popup saying "Headed Father's day shopping? Want some locations based on your recent web activity?" I say Yes, and all kinds of fun spots and deals popup.

fumar 13y ago
Google Now does some of these things. Now provides me for directions almost everyday. It also lets me know what what movie times are relevant for me. I was recently in NYC, Now gave me relevant things to do while I was there. They were great suggestions. It blew me away.
jiggy2011 13y ago
But how does it know which trips will be father's day shopping and which will be for some other purpose? Assuming you use your car more than once that week. The result it likely to just be constant annoying spam.

mayank 13y ago

> I don't know if our behavior can really be predicted on such a fine grained level. There was a project in the MIT Media Lab something like a decade ago where they stuck about 100 graduate students with Bluetooth-enabled cellphones (this predates smartphones) and tracked many of their behaviors over the course of a semester. Things like phone calls, personal interactions, movements, etc. A very interesting result from that study was that they built a model for predicting daily behaviors for their subjects with up to 79% accuracy [1]. The whole project was groundbreaking at the time, and they have some very interesting papers. [1] http://realitycommons.media.mit.edu/pdfs/eigenbehaviors.pdf http://realitycommons.media.mit.edu/pdfs/eigenbehaviors.pdf