Demos, Emotions, Rao

[Avery Andrews 961012]

Brief notes on serveral topics:

(Rupert Young)

The backslash/forward slash differences between DOS and Unix are
one possible problem (I use both on a daily basis, and am constantly
screwing things up because of that); another is that it can make
a difference whether the current default drive is `c' and you
type `a:install', or the current default is `a' and you type
`install'. Once upon a time I fantasized about porting these
things to X-Windows, but this idea crashed into the reality
of how hard it is to program GUIs. I have gotten to the point
of being able to write simple Windoze programs, so maybe some day...

(Emotion Posters)
There was a newspaper account down here not so long ago of
some researchers in England who claimed to have found that
people who had just watched a sad movie were much better at finding
solutions to problems that people who had watched happy ones;
the cheerful folk went for quick, easy & superficial solutions,
while the gloomy folk look deeper and found better and more
creative ones. Which suggests that sadness and gloom are what
it feels like to be reorganizing at some of the higher levels.

(Hans Blom & Bill Powers)
It's not clear to me that by `prediction' Rao means prediction
of future events; he may be talking about `predictive parsing'
methods whereby one uses guesses about the high-level categorization
of something to generate predictions about various lower-level
features, which are then checked. E.g. `this whole thing looks sort
of like a face, so this little bit here ought to be a nose'.
Such methods are used in natural language parsing. But Rao also works
on eye-movements, so he might also be about genuine temporal predictions
even for object-recognition (we do a lot of scanning when identifying
things). I wish I had tie to study his stuff more, I certainly don't
understand much of the paper at the URL I posted.

Avery.Andrews@anu.edu.au