paper presented at Human Centred Design for Intelligent Environments (HCD4IE) Workshop, HCI2016, Bournemouth, UK, 12 July 2016.
http://alandix.com/academic/papers/HCD4IE-2016-human-like/
2. context
EPSRC workshop on Human–Like Computing
likely programme in the area
similar funding initiatives elsewhere
high profile discussions about AI
3. What is Human-Like Computing?
tentative EPSRC definition:
“offering the prospect of computation
which is akin to that of humans,
where learning and making sense of information
about the world around us
can match our human performance”
4. why human–like?
(i) because it is a good model to emulate
understanding people helps design algorithms
(ii) for interacting with people
computers that better understand humans
(iii)new interaction paradigms
how to deal with human–like agents
(iv)learning about human cognition
human-like algorithms are cognitive models
appropriate
intelligence
regret modelling
spreading
activation
over
ontologies
5. issues
human or superhuman?
emulating human frailty?
for over confidence in algorithms
REF!!!
legibility
understanding black-box algorithms
+ my 1992 paper, recent EU directives
6. smart environments?
embodied agency
uncanny valley issues
but at least you know it is there!
disembodied and distributed agency
ghosts in the walls
the more it understands you …
… the more frightening it is!
c.f. designing
incidental
interactions
& new book chap.