[FRIAM] sometimes an onion is just an onion...

[Marcus Daniels]

Others behaviors come to mind, like the agent that requires or expects fully contextualized unambiguous linear arguments.   This could be due to long term memory limitations, due to a desire to teach (supervisory learning), or an agent that defaults to priors in absence of systematic communication.   Premature registration could be a sign of a simple hardware failure or inequitable / entitled expectations of the correspondent.   In organizations, some roles might well be encouraged even if they aren't adaptive in general, e.g. the boss vs. the nuanced facilitator.

-----Original Message-----
From: Friam [mailto:friam-bounces at redfish.com] On Behalf Of glen ?
Sent: Wednesday, June 14, 2017 4:32 PM
To: The Friday Morning Applied Complexity Coffee Group <friam at redfish.com>
Subject: Re: [FRIAM] sometimes an onion is just an onion...


Beautiful!  Surely we don't need much imagination ...  Surely (!) there exist modal pattern recognizers we could (almost) drop into a simulation to at least implement your 4 agent types.  All we need is the right combination of search keywords to find them.  I wonder if Edelman's "neural darwinism" simulation system, which supposedly allowed an objective function to select amongst various multi-neuron clusters, would work.  Of course, I tried a few years ago to find code for the simulation(s) he claimed to have, and failed.

On 06/14/2017 03:13 PM, Marcus Daniels wrote:
> One can imagine a neural net with similar inputs and outputs but different depths of hidden layers inhibiting & exciting internal neurons of the network.  These would represent relevant contrastable features tied to previous similar experiences.  Together they'd compete to activate one or several neurons that correspond to one or several registrations.   
> 
> A lack of experience with ambiguity in inputs would be one explanation why premature registration would occur.  [Naïve agents]   Another might be no particular pressure to distinguish similar categories -- no cost for bad predictions -- so no reinforcement of connections to other neurons.  [Unengaged agent]  Another might be that training had occurred on similar but distinct data and re-training wasn't believed to be needed -- the learner had been educated in a curriculum-based (programmed) way and believed that the features in the environment were easier to contrast than they really were.  [Smug agent] 
> Finally, there's the no neurons available possibility... [Disabled 
> agent]
> 
> -----Original Message-----
> From: Friam [mailto:friam-bounces at redfish.com] On Behalf Of glen ?
> Sent: Wednesday, June 14, 2017 2:43 PM
> To: The Friday Morning Applied Complexity Coffee Group 
> <friam at redfish.com>
> Subject: Re: [FRIAM] sometimes an onion is just an onion...
> 
> 
> On 06/14/2017 01:29 PM, Marcus Daniels wrote:
>> The meaning is clear, but is this a term that is used in particular communities?   The reason I ask is that I deal with people all that time that do this, and I'd like to be able to whack a book over their head, since  they like to do that to others.
> 
> On 06/12/2017 10:39 AM, glen ☣ wrote:
>> Cf Brian Cantwell Smith in: 
>> https://global.oup.com/academic/product/philosophy-of-mental-represen
>> t
>> ation-9780198250524?cc=us&lang=en&
> 
> It's not clear to me how common the usage is.  In B.C. Smith's "On the Origin of Objects", he calls it "inscription error" instead.  In the book cited above, he states that he prefers "premature registration", mainly because it applies not only to programming/inscribing, but to things like what happened in this discussion (where Nick and Steve prematurely clamped on an onion metaphor I didn't intend).  I still prefer inscription error when I use it in a simulation context because the meaning is more clear.  In logic or rhetorical contexts, the standard "petitio principii" still works.


--
☣ glen

============================================================
FRIAM Applied Complexity Group listserv
Meets Fridays 9a-11:30 at cafe at St. John's College to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com
FRIAM-COMIC http://friam-comic.blogspot.com/ by Dr. Strangelove