Com­put­ers as Metaphor, Minds as Com­put­ers: Notes To­wards a Dys­func­tional Ro­bot­ics




Symposium:


Session Title:

  • Signs of Life: Human-Robot Intersubjectivities

Presentation Title:

  • Com­put­ers as Metaphor, Minds as Com­put­ers: Notes To­wards a Dys­func­tional Ro­bot­ics

Presenter(s):



Venue(s):



Abstract:

  • Panel: Signs of Life: Human-Robot Intersubjectivities

    This paper will pre­sent a range of ideas un­der­pin­ning the de­vel­op­ment of John Tonkin’s new pro­ject, a se­ries of dys­func­tional ro­bots that ex­plore dif­fer­ent ap­proaches to think­ing about cog­ni­tion and per­cep­tion. Com­pu­ta­tional the­o­ries of mind have been used both by cog­ni­tive sci­en­tists as model of how to build an elec­tronic mind, and by cog­ni­tive psy­chol­o­gists as a means of un­der­stand­ing the human mind. They see the mind as an in­for­ma­tion-pro­cess­ing sys­tem and thought as a form of com­pu­ta­tion. These sym­bolic ap­proaches to think­ing about the mind have been chal­lenged by more em­bod­ied and em­bed­ded ap­proaches to cog­ni­tion and per­cep­tion. This has been re­flected through the de­vel­op­ment of a num­ber of bot­tom-up ap­proaches to AI and ro­bot­ics, such as neural net­works and be­hav­iour based ro­bots that are based on ideas of re­ac­tiv­ity and sit­u­at­ed­ness rather than higher level sym­bolic mod­el­ling. The ner­vous ro­bots that are being built for this pro­ject awk­wardly hy­bridise bot­tom-up AI ap­proaches with more clas­si­cal sym­bolic ap­proaches that use high level sym­bols drawn from a folk psy­chol­ogy con­cep­tion of the mind as being the home of in­ter­nal men­tal processes such as mo­tives, de­sires, pho­bias and neu­roses. They use a range of com­pu­ta­tional ap­proaches, for ex­am­ple Brooks’ sub­sump­tion ar­chi­tec­ture, to cre­ate lay­ered hi­er­ar­chies of stim­u­lus / re­sponse re­flexes. Ex­am­ples in­clude a claus­tro­phobot and an ago­ra­phobot, as well as needy/dis­mis­sive ro­bots based around at­tach­ment the­ory. One of the aims of this pro­ject is to ex­plore the lower bound­ary of com­pu­ta­tional com­plex­ity that still evokes some sort of self-iden­ti­fi­ca­tion and re­sponse in the au­di­ence.


PDF Document:



Category: