Joho the Blog
An Entry from the Archives

I'm interested in indexes / lists like this too, and, though I like the idea of being able to play back certain things I did in the past, I really like the idea of allowing different people to "play" public versions of each others' indexes / lists.

(I think indexing collaborative / group behavior is way more interesting than indexing one's own.)

For my iCite net experiments, I am working with a small set of "semantic behaviors" that I can index (and publish) with a single-click, e.g., read, not-read, filed/archived, and (especially) cited. Automatic indexing would be nice for some things, but I think there are some interesting things to explore specifically with user interaction with the list / index.

In the last instance you make reference to invading privicy.

This idea running as a OS would be worry-some. Running as a client application with good controls and a experienced user would be ideal. IMO something like the zonealarm popup box system would be useful so you could teach the client app what content can or behaviors can be logged? I know it would be very ignoying at first but soon things would settle down.

The other thing I thought about, if the client app watched logs and caches. So for example my jabber client creates a log of all my communication which I choose to delete or keep. My browsers create cookies, histories and cached content based on what I do. I wouldnt object to a client application which read this information and made a semantic behavior log.

In the whole its an interesting idea and just like Jay, I would love to see a public play.

Ethan Zuckerman sent me email from Ghana that this is something like Gelertner's LifeStreams work. Ethan points to http://www.wired.com/wired/archive/5.02/fflifestreams.html?person=ted_kaczynski&topic_set=wiredpeople

We, i.e. ActiveWords, don't keep a history. We have always been cognizant of the privacy issue. We recognized that this was such a sensitive issue that any breach of it would have been fatal to our evangelizing our product.

With ActiveWords, every time you trigger an ActiveWord, we clear the buffer, and provide you with a clean slate so to speak.

I am familiar with Gelertner's work on a limited basis, and I personally like a lot of what he has done and written about.

I see a huge value to the idea that would allow me to point or search easily data in a continuous stream. An example might be that let's say I treated someone to lunch at a Chinese restaurant. I would probably post the expense to Quicken, and would make notes about any follow on items. At a later date if my friend E-Mailed me as to the name of the restaurant, I shouldn't have to puzzle through various applications to figure out what the name of the restaurant was, but rather just to search or point to the character string to find it, and in so doing see the context. It might prompt me to follow up on other things.

It seems to me that real issue is encryption. Hard drive space is getting real cheap, processor power more than adequate to accomplish these tasks. We have noted that the faster computers run, the better ActiveWords works. The lag in productivity comes from human factors, the “I know it’s here somewhere, just don’t know where syndrome!”

Notwithstanding the heat that it takes, I see people using Outlook as a file cabinet, and now with applications like Lookout, you can find stuff pretty quickly. It is not whether people do this kind of stuff but how.

Along what I call the ActiveWords Odyssey, I met a very bright woman who suggested that we had done a good job of implementing the basic communications metaphor. I didn't know what she meant, and she explained that we as humans have impulses, followed by utterances seeking predictable results. She argued this idea is fundamental, that it doesn't change, and that it won't change.

Hence if Jon's ideas are implemented, and if implemented by Google, they will merely manifest what we all want.

Again, in my mind it goes back to encryption, and how do we protect this stream of text, not whether we want or will use it.

Stop giving DARPA more ideas! Or worse, a possible architecture for CAPPS III! :)

Hello,
I fully appreciate all the comments. According to me Semantic behaviour index attract the people. I familir with little bit. I really like the idea of allowing different people to play public version of each other's index/list.