![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
I've been working further on trying to create a prototype associative filing system and have been giving a lot of thought to working out ways to get the computer to anticipate the user's needs. In some sense Google already manages a version of this by keeping a record of the most common searches made by other people and offering those as you type your search terms. It also seems to speed the presentation of results by re-using those from common searches. And they definitely use other people's links to rank recommended search results.
Many programs on our computers are in the very early stages of doing similar things. They maintain lists of most recently or most commonly accessed items. On the net many of us use tags on our blogs to categorise what we write (I don't). Many sites use tags to organise their content. These are all undeniably convenient, but it is the beginning of a trend that poses a subtle long-term danger, and we need to be aware of it.
If we allow the computer to filter the kind of thing it presents to us based upon what is most popular then we end up with a self-amplifying, positive feedback system. Initially it can be very useful, but over time it risks stagnation. Such risks may take a long time to become apparent with Google's search engine, and might be sufficiently offset by the variety of humanity to ever be a real problem, but an associative filing system on a single machine for a single user might gradually narrow their focus so that they're almost never presented with items beyond a slowly shrinking range.
Individuals would be encouraged by their computer to never step outside their comfort zone and never see anything truly new. If emotional tags were used to guide this it becomes more treacherous still. People's irrational beliefs and world-views would never be challenged and they would be presented only with data that appeared to support their erroneous views. We already have a worrying tendency to do this; we don't need our computers amplifying it. Its effect on whole societies could become quite dangerous in time. The scariest aspect to me is that its progress would be almost invisible, at every step offering genuine improvements, but over the longer term becoming deleterious.
So, I need to be aware of this risk when making my associative filing system and find ways to circumvent it. My immediate thought is that novelty and humor could be used to add greater value to fresh things, but how? There must be other useful (and easier to program) strategies too.
Many programs on our computers are in the very early stages of doing similar things. They maintain lists of most recently or most commonly accessed items. On the net many of us use tags on our blogs to categorise what we write (I don't). Many sites use tags to organise their content. These are all undeniably convenient, but it is the beginning of a trend that poses a subtle long-term danger, and we need to be aware of it.
If we allow the computer to filter the kind of thing it presents to us based upon what is most popular then we end up with a self-amplifying, positive feedback system. Initially it can be very useful, but over time it risks stagnation. Such risks may take a long time to become apparent with Google's search engine, and might be sufficiently offset by the variety of humanity to ever be a real problem, but an associative filing system on a single machine for a single user might gradually narrow their focus so that they're almost never presented with items beyond a slowly shrinking range.
Individuals would be encouraged by their computer to never step outside their comfort zone and never see anything truly new. If emotional tags were used to guide this it becomes more treacherous still. People's irrational beliefs and world-views would never be challenged and they would be presented only with data that appeared to support their erroneous views. We already have a worrying tendency to do this; we don't need our computers amplifying it. Its effect on whole societies could become quite dangerous in time. The scariest aspect to me is that its progress would be almost invisible, at every step offering genuine improvements, but over the longer term becoming deleterious.
So, I need to be aware of this risk when making my associative filing system and find ways to circumvent it. My immediate thought is that novelty and humor could be used to add greater value to fresh things, but how? There must be other useful (and easier to program) strategies too.