From: Pope Salmon the Lesser Mungojelly (rainbow@beautywood.org)
Date: Tue Oct 25 2005 - 15:17:21 MDT
On Mon, 24 Oct 2005 15:04:57 -0500, Chris Capel <pdf23ds@gmail.com> wrote:
> Specifically, I have in mind things like wearables that are able to
> keep track of what and when your activities are; manage your nutrition,
> circadian rhythms, and some hormones, keep track of your social
> interactions, your task list, and the state of various information
> feeds you're subscribed to; do background data mining based on your
> interests and recent activity, and other low-level AI (or even
> simpler) tasks.
As well as considering what rigs (i.e. wearable computers) can do
autonomously, we should consider what they can do in response to quick
vocal commands. Even if they're not good enough at making decisions to be
trusted with much of the judgement work, they can still be so closely
integrated with their users that they transform their users' thoughts into
actions with a fluidity we've never seen before.
For instance consider the activity of sending a picture to a friend. At
the moment I have the technology to do that, of course, but it requires
rather a long series of actions: Take the picture, plug in the camera,
upload the picture, write an email, attach the file. A near-future
streamlining of that could look more like this: After I press the button
to snap the picture, I say (or type if I like) "Computer, send that to
John and Bob and Mary." It's not so much a big deal that this would save
me a few minutes of work; rather what's a big deal is that I probably
would never have bothered to send that picture to John and Bob and Mary if
it hadn't been so easy.
Raw computing power is only meaningful to the structure & future of
society to the extent that it's actually used to do particular
computations with particular results. We have neurons in our heads whose
"total computing power" by whatever metric is much greater than that
required to do, for instance, digital manipulation of photographs, but
there is absolutely no way of photoshopping in an unaided human brain and
therefore the consequences of an increase in ability to photoshop should
not be expected to be linear in respect to the total number of
mathematical operations being done or any such nonsense. It's useful to
look at these exponential trends in computing technologies, because
they're factual and easily graphed/understood and they do really point to
the magnitude of the transformation we're undergoing, but we shouldn't be
fooled by those graphs into thinking that the most dramatic
social/practical consequences are going to wait around for the later half
of the curve in order to spring.
> Ideally, it would be able to detect your mood to know whether you're
> in especial need of certain kinds of exercise or relaxation or
> entertainment in order to stay in peak condition.
Of course most people wouldn't like it if their rig TOLD them that they
ought to get some exercise, but maybe it could find a subtler way to make
it happen. I think that we'll see computers interacting a lot more with
the unconscious minds of their users. They'll be analyzing unintuitive
sets of data, reaching conclusions that you wouldn't understand even if
they explained it (e.g. evolving color schemes based on heart-rate
monitoring). Feedback cycles are closing here.
It's an entirely reasonable question to ask, "In what ways have computers
ALREADY begun to integrate with us on unconscious levels?" This question
should turn up more & more dirt as we go forward, but I believe it already
does turn up a little. User interface design for example has been
progressing on unconscious levels. In the near future the human
unconscious will be deeply probed by automatic creation & mining of
metadata.
> And a related question: Do the sort of intelligent agents that would
> drive these systems pose any takeoff risk?
Yes obviously. That's why we're taking off at the moment. Get rid of all
of the intelligent agents (i.e. make people design chips by hand) and you
wouldn't see a Singularity this century. We are on the curve, we are
inside the curve, and there are no realistic methods of controlling it at
hand. So we strap in, right?
<3,
mungojelly
This archive was generated by hypermail 2.1.5 : Tue Feb 21 2006 - 04:23:18 MST