(no subject)
Feb. 8th, 2008 12:09 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
I had one of those "So THAT is how it's all going to work!" moments today where a whole bunch of technologies slotted together in my mind and I could see the future.
It's a bunch of technologies, all of which are here now, at least in breadboard form in labs round the world. All I have done is put them together in a "new" pattern. I am sure
strangedave is going to say "Pshaw! That's old news!"
So my assumptions:
Now, say you went into a shop and got lousy service. When you got out you could tag the shop spatially with a tag of "rude".
Any of your friends wandering past the shop looking for a similar service would see that one of their friends tagged it negatively.
The next stage up is personal tagging. We all carry an increasing amount of RFID tags on our person, so something that queries tags can be used to build up an instant "fingerprint" of a person. Obviously some sets of RFIDs are going to be less than useful (Such as Walmart tags that they stick on everything) but many (Such as your bus pass) are going to stick with you.
When you meet someone for the first time and query them, your RFID reader will squirt out a generic query and scan the responses. If it can make a positive ID it'll flag that person with the "tags" other people have logged. If it can only make a partial match it'll show you screencaps of who it thinks you are looking at so you can confirm a matchup. If it doesn't know this person you'd be able snap a quick screencap and upload it for further tagging.
Now this simple tagging would be way too easy to abuse so the system needs to be able to understand your relationship to the people doing the tags. Firstly your own tags will show in black as they are assumed to be "true". (As you can edit your own tags if there is an error, you can always correct them). Secondly, if you have marked a fellow user as trustworthy, their tags will show in green. Otherwise unconnected peoples tags will show in grey. Finally if you have marked any fellow user as untrustworthy, theirs will show in red. (this last step is to minimise the effect of trolls tagging indiscriminately. If you see a tag that you know is wrong, you could tag the tagger as untrustworthy.
More importantly, you could set a threshold of how many people have tagged in any category and then see the group consensus, further building lists of trusted taggers and untrusted taggers.
I am only just touching on the possibilities here. This is a brute force Wuffie system right there.
It's a bunch of technologies, all of which are here now, at least in breadboard form in labs round the world. All I have done is put them together in a "new" pattern. I am sure
![[info]](https://stat.livejournal.com/img/userinfo.gif)
So my assumptions:
- Wearable computing delivers some serious portable grunt. Equivalent to a current era microPC. At the moment the stuff exists as wearable backpacks but you'd want it to be small enough to be comfortably wearable, possibly as a "personal cluster" over several pockets.
- Augmented reality displays are available, either as glasses or contact lenses.
- Everyone would have "webcam" grade cameras interfaced with the below
- Municipal broadband is ubiquitous via something like WiMax
- GPS + ping based positioning can generate accuracies of 1m or less in urban environments
- A small RFID query module is available.
Now, say you went into a shop and got lousy service. When you got out you could tag the shop spatially with a tag of "rude".
Any of your friends wandering past the shop looking for a similar service would see that one of their friends tagged it negatively.
The next stage up is personal tagging. We all carry an increasing amount of RFID tags on our person, so something that queries tags can be used to build up an instant "fingerprint" of a person. Obviously some sets of RFIDs are going to be less than useful (Such as Walmart tags that they stick on everything) but many (Such as your bus pass) are going to stick with you.
When you meet someone for the first time and query them, your RFID reader will squirt out a generic query and scan the responses. If it can make a positive ID it'll flag that person with the "tags" other people have logged. If it can only make a partial match it'll show you screencaps of who it thinks you are looking at so you can confirm a matchup. If it doesn't know this person you'd be able snap a quick screencap and upload it for further tagging.
Now this simple tagging would be way too easy to abuse so the system needs to be able to understand your relationship to the people doing the tags. Firstly your own tags will show in black as they are assumed to be "true". (As you can edit your own tags if there is an error, you can always correct them). Secondly, if you have marked a fellow user as trustworthy, their tags will show in green. Otherwise unconnected peoples tags will show in grey. Finally if you have marked any fellow user as untrustworthy, theirs will show in red. (this last step is to minimise the effect of trolls tagging indiscriminately. If you see a tag that you know is wrong, you could tag the tagger as untrustworthy.
More importantly, you could set a threshold of how many people have tagged in any category and then see the group consensus, further building lists of trusted taggers and untrusted taggers.
I am only just touching on the possibilities here. This is a brute force Wuffie system right there.
no subject
Date: 2008-02-08 05:58 am (UTC)