[NetBehaviour] Quick thought about Neural Networks

Micheál O'Connell mocksim at gmail.com
Thu Mar 29 10:45:49 CEST 2018


I recently read Caroline Bassett's latest piece of writing (available here
http://sro.sussex.ac.uk/73253/ ) which revisits ELIZA and posits a more
complicated interpretation of  the controversy  surrounding its receipt.
ELIZA was not NN based but as the question of AI generally is being raised,
it occurs to me that whether something is interesting has little to do with
its apparent sophistication.  All kinds of human tendencies, bodies of
knowledge and overarching systems (including the 'art system') impact the
reading of a new entity, artefact or algorithm. Even Grey Walter’s
tortoises behaved surprisingly, the went into crisis on occasion,
‘flickering, twittering and jigging’. Grey Walter's rudimentary robots were
more obviously physically embodied dynamic mechanisms than any neural
network, but, as the nickname suggests, open to being at least
anthropomorphised, or misunderstood, under and over-understood. There's
surely a tendency, and need, to treat as black boxes, not only simple, but
complex technologies and processes. In my view an extended definition of AI
would include all of the human-made procedures, processes, instructions and
codes which surround us. Algorithmic processes and NNs are high speed forms
of bureaucracy, it could be argued. And machineries, and procedures may be
embedded withing others in a kind of object-oriented fashion. A NN's
meaning will change depending on what overarching arenas or systems are
brought to bear on it and what framework it is associated with. It may
indicate that itself for sure but not always. If you ran a NN designed for
business goals, one which dealt with large amounts of data in the
Serpentine and presented it operation and outputs in the way that you would
to an industrial client then I wonder what would happen


Micheál O'Connell
@mocksim <https://twitter.com/Mocksim_Latest>

On 25 March 2018 at 16:16, ruth catlow <ruth.catlow at furtherfield.org> wrote:

> Always very appreciative of ABC explainers of mystifying tech trends - so
> bravo!!!! and thank you.
>
> Currently Ian Cheng's BOB is on display at the Serpentine, about which I
> have mixed feelings.
> It is very very clever and technically accomplished.
>
> It also presents a view of "life" as something to be best studied and
> understood in a petri dish. The degree to which this is a comment on the
> sterility (or threadbare engagement with sociality) of AI development
> cultures at this time, is not clear to me. Perhaps we will know the answer
> to this question in the upcoming phase two of Cheng's exhibition called The
> Emissaries.
>
> One of the difficulties in parsing BOB as an artwork arises from the fact
> that (like a lot of products of digital culture) it is a black-box, or
> perhaps a better analogy would be an iceberg. We know that there is a huge
> and complex machine under the surface but we have no idea, and no way to
> gain understanding of what is going on. It's body is unlike ours, so it
> doesn't feel like we do. We can't learn together with it. All we can do is
> project our own idea of what an evolving system looks like onto its
> surface. And I think BOB is doing very clever pattern mirroring.
>
> So for me - with a creative Neural Net artwork I want to know about the
> "body" of the artwork and to be able to observe the glorious gap between
> what I might project onto it and what is actually going on.
>
> cheers
> Ruth
>
> On 25/03/18 05:20, BishopZ via NetBehaviour wrote:
>
> Been building some Neural Networks.
>
>
> There are two ways you can build them...
> Analytical or Creative.
>
>
> An Analytical NN takes a lot of data
> and boils it down to an answer.
> This is what the business world is so on about.
> It's a replacment for anything that was formerly
> random decisions, like which banner ad to display.
>
> You can also make a Creative NN,
> which taks a small amount of data
> and makes up a whole bunch.
>
> Examples of Creative NNs:
>
> 1. You press a button, and the NN creates an image. You tell it whether
> you liked the image or not.
>
> 2. You select 1 of 4 colors, the NN writes a poem for you. You tell it
> whether you liked the poem or not.
>
> Techniques for Creative NNs:
>
> (1) NN's only do what you train them to do. If you include the previous
> answer as part of the input data, then you can train it to not give the
> same answer twice.
>
> (2) The higher you set the learning rate, the more "in the moment" it
> becomes.
>
>
> Dear Netbehaviour,
>
> Any ideas, thoughts, comments? Is there room for creative NN artworks?
> Seems like this kind of thing could take over the interactive installation
> space. Maybe art-tech can become surprising again?
>
> If they make manipulation machines that drag on your worst fears, maybe we
> can make happiness bots that make you fearless?
>
> Any ideas, thoughts, comments? What if we made the entire museum into
> a rube goldberg of intelligent automation? Anything else we could do?
>
> Skynet's the limit.
>
> Bz
>
> --
> ((º Ω º))
>
> http://bishopZ.com
> _______________________________________________________________________
>
>
> _______________________________________________
> NetBehaviour mailing listNetBehaviour at lists.netbehaviour.orghttps://lists.netbehaviour.org/mailman/listinfo/netbehaviour
>
>
> --
> Co-founder Co-director
> Furtherfield
>
> www.furtherfield.org
>
> +44 (0) 77370 02879
>
> Bitcoin Address 1G7SPFpvHhVEqn5trpNEcyNWbDcyZXuAnh
>
> Furtherfield is the UK's leading organisation for art shows, labs, &
> debates
> around critical questions in art and technology, since 1997
>
> Furtherfield is a Not-for-Profit Company limited by Guarantee
> registered in England and Wales under the Company No.7005205.
> Registered business address: Ballard Newman, Apex House, Grand Arcade,
> Tally Ho Corner, London N12 0EH.
>
> _______________________________________________
> NetBehaviour mailing list
> NetBehaviour at lists.netbehaviour.org
> https://lists.netbehaviour.org/mailman/listinfo/netbehaviour
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.netbehaviour.org/pipermail/netbehaviour/attachments/20180329/0cdf3113/attachment.htm>


More information about the NetBehaviour mailing list