designing future environmentsnunez/cogs1_f07/kirsh_slds.pdf · –how material science guides...
TRANSCRIPT
DesigningFutureEnvironments
David Kirsh
Dept of Cognitive Science
UCSD
Agenda
• Background: Technology is outpacing design
• Science: How are we coupled to our environments?
• Comparing Environments Scientifically
• Design considerations
Background of Inquiry:Technology
– Walls are data walls
– Internet everywhere
– Wireless everything
– Near field haptics
– Easy telepresence
– Effective digitization of paper
– Sensors make it easy to cross from
physical to digital
– Rooms are context aware
Background of Inquiry: Theory
• A new theory of how people areembedded in their environments– We are closely coupled to our
environments
• The environment we inhabit is partiallythe product of our own structuringactions
• dynamics of this coupling must beanalyzed at many time scales
• we learn to exploit the dynamics ofinteraction
World acts back
Part of a system
Basic Question
So what?
How does new cognitive theory explainhow to make new tech useful?
Answer
– How material scienceguides architects
• Provides constraints ondesign – allows us to saythat design will not work
– Physics guides engineers
• It is beginning to provide principles toguide designers. Similar to:
Gehry’s Disney Auditorium
The environment we inhabit is partially theproduct of our own structuring actions
One powerful idea
Create vs. Project Structure
Project structure mentally
ecesrrruutt
ee cs rrr uu tt
ece s rrr uu tt
vs.
ecesrrruutt
Create structure physically
The environment we inhabit is partially theproduct of our own structuring actions
One powerful idea
The environment we inhabit is partially theproduct of our own structuring actions
One powerful idea
Dynamic coupling
ecesrrruutt
ee cs rrr uu tt
ece s rrr uu tt crust strut e
restructure
a. dynamics exist at many time scales
b. we learn to exploit the dynamics of interaction
Another powerful idea
Time Scaleslog time
(secs)
complementary actions
maintenance actions
-
preparatory actions-
-
-
1 sec
10 sec
100 sec
1000 sec
Complementary Actions
Close Spatial & TemporalCoupling
• external action must occur at the rightmoment wrt internal processes
How many dots?
which is longer?
which is longer?
which is longer?
Fingers can be used tocompensate for perceptual
limitations• exploit vernier perception
Fingers can be used tocompensate for perceptual
limitations• exploit vernier perception
We manipulate localresources to compensate forlimitations in other judgments
Tetris
Rotations help Judgment
most rotationsphysical rotationfaster than mental rotation
Even Experts Rotate Extra
To disambiguate mirror pieces
rotate
How to CompareEnvironments
Comparing Environments
Env 2Env 1
What makes one environment better thananother for accomplishing a task?
Why?
Behavioral measure of better
time errors errors on hardest stress interruption problem
solvable
E1 > E2
performance
problem hardness
Env 2
Env 1Env 2Env 1 >
0
Cognitive measures of better
• The same problem or task may be solvedor performed with less cognitive load.Reduced:
• computation
• memory
• stress
• sustained concentration
• Harder problems or tasks may be solved
Why is E1> E2? Simple Example
On which surface is it easier to solve the problem below?
A B
Problem:
Arrange the blocks so that• green touches red• red touches yellow• yellow touches yellow
A is better
green touches red red touches yellow yellow touches yellow
Larger surface allows more solutions
Simple Causes
• If abundant space→ no need to schedule, optimize,swap
• change problem constraints→ change solution possibilities
Would this be better for …
• Organize the location ofresources
• But:• Simple organizational
system
• Reduced opportunism
• Less conducive tomultitasking
• How does this personactually work?
More interesting Causes of E1> E2
• Change the cognitive congeniality of theenvironment
– embed hints, intelligent agents, make constraintsmore explicit (magnets repel or attract)
– Improve cognitive affordances – scope forepistemic actions
– Context aware computing –track activity andproject personal metadata
• change the cognitive workflow of a task– add paper– solve it collaboratively
More interesting Causes of E1> E2
• Exploit possibilities for complementary actions -i.e. internal-ext coordination
– use our hands to help us think
– manipulate physical objectsto save mental manipulation
• Tetris examples:
– physical rotation saves mental
– physical translation improves certainty
Externalization
Representation Shifting
Visual Thinking - Representation
Recognition vs. Recall
Epistemic Actions
Pragmatic vs EpistemicActions
Jigging the Environment
Bad Interface
Better Interface
Better still
Interactivity
Interactivity & Externalization
Change Env Change problem
Coordinating Mechanisms
Starbuck’s Cup
Speed Accuracy & Design
Learnability & Design
Complexity & Design
Error Recovery & Design
Variance & Design
Bean Counters & Design