Dear friends,
This discussion is another instance of something that was already illustrated by Monty Phython many years ago:
https://www.youtube.com/watch?v=4MwJOnleriM
Some people just want to follow the gourde or the sandal.
These scientists just came up with a new and very interesting sandal.
They found it interesting for its technological values.
Other people find some other meanings in there, just like someone might see the face of JC on a french toast or on a cappuccino.
In this particular instance, seeing these patterns is not exactly apophenia but it is close: what you are looking at is the expected patterns that the neural network (NN) was trained for. This is not an unexpected behavior from the NN itself (if there wasn't this kind of connection between the training data fed into the NN and its output... now THAT would be unexpected). What is unexpected is the striking looks of these eerie images and the eagerness (paranoia?) of the NN in finding the requested patterns inside random noise.
What is happening here is something that partially models the process of conceptual categorization (cf. Clancey, 1994:173). It doesn't surface the highest level of abstraction yet,— the process of categorization of conceptuaization itself: this NN doesn't "know" that the pattern it produces has a "resemblance" to _whatever_; it doesn't know that it knows anything, so the point about "art" is moot (see also Csikszentmihalyi, 2006):
Best regards,
==================================
Carlos Pires
[log in to unmask]
[log in to unmask]
-------------------------------------------------------------
Design & New Media MFA // Communication Design PhD Student @ FBA-UL
Check the project blog:
http://thegolemproject.com
On 20/06/2015, at 00:50, Don Norman wrote:
> I guess I'm simply dull-witted, but to me (and to the scientists involved)
> the Google study has nothing to do with art: it is an attempt to
> understand what and how features were extracted from images. If people wish
> to see art in the resulting data visualizations, wonderful, But it is
> non-intentional on the part of the creators.
>
> Don
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]>
Discussion of PhD studies and related research in Design
Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
|