1. Is
Your
Intelligent
Computer
Smart
Enough
To
Be
on
My
Team?
Barbara
J.
Grosz
School
of
Engineering
and
Applied
Sciences
Harvard
University
Acknowledgements:
Nuance
FoundaEon
2. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Background:
A
Par@cular
Turing
Predic@on
} “.
.
.
The
original
quesEon,
‘Can
machines
think?’
I
believe
to
be
too
meaningless
to
deserve
discussion.
Nevertheless
I
believe
that
at
the
end
of
the
century
the
use
of
words
and
general
educated
opinion
will
have
altered
so
much
that
one
will
be
able
to
speak
of
machines
thinking
without
expec5ng
to
be
contradicted.
.
.”
“Computing Machinery and Intelligence”,
Mind, 1950. p. 442
3. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
What
Ques@on
Might
Turing
Pose
Now?
} SeOng:
Agents
and
people
work
together
on
nontrivial
task,
extended
in
Eme,
in
an
uncertain,
dynamic
environment.
} Can
a
computer
(agent)
team-‐member
behave,
over
the
long-‐term,
in
such
a
way
that
people
on
the
team
will
not
noEce
it’s
not
human
or
think
it’s
stupid.
Grosz,
Turing
Research
Symposium,
University
of
Edinburgh
&
Royal
Society
of
Edinburgh,
11
May
2012;
AI
Magazine,
33:4,
2012;
The
Atlan2c
blog,
August
2012.
4. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Why
Teamwork?
Canonical
Compu@ng
Environment
Now
From “Computing Machinery and Intelligence” and
“Can MachinesThink?” toTeamwork
5. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Why
Teamwork?
Arguments
from
Cogni@ve
Science
} It’s
fundamental:
“Every
funcEon
in
the
child's
cultural
development
appears
twice:
first,
on
the
social
level,
and
later,
on
the
individual
level
.
.
.
All
the
higher
func5ons
originate
as
actual
rela5ons
between
human
individuals.”
(Vygotsky,
1978,
pp.
56-‐57)
} Language
is
interpersonal,
cooperaEve:
} “Infants
depend
in
both
their
comprehension
and
producEon
of
poinEng
on
a
joint
ahenEonal
frame
(common
ground)
with
their
communicaEve
partners
.
.
.
And
they
do
this
for
the
fundamentally
coopera5ve
mo5ves
of
helping
and
sharing
informaEon
and
aOtudes
.
.
.”
(Tomasello,
et
al.,
2007,
p.
720)
6. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Teamwork
Is
Not
Easy:
Essen@al
Match
of
Abili@es
to
Tasks
and
Roles
or
.
.
.
Source: http://msdn.microsoft.com
7. Interac@ve
Intelligence
and
Health
Care
Coordina@on
new
twists:
loosely
coupled
team
members
team
of
peers
(limited/no
hierarchy)
relaEvely
long
Eme
horizon
(not
“one-‐shot”)
My
current
research
focus:
8. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Research
Program
Framing:
Design/Build
Collabora@ve
Computer
Agents
Models
and
algorithms
for
efficient
decision-‐making
under
uncertainty
Formal
models
of
collaboraEon
Methods
for
supporEng
communicaEon
and
informaEon
sharing
Methods
for
analyzing
and
learning
people’s
percepEons
of
agent
behavior
9. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Theore@cal
Founda@ons:
SharedPlans
Theory
} Each
team
members
commits
to
team’s
performance
of
the
group
acEvity;
requires
model
of
inten2on.
} Team
members
reach
consensus
on
a
(high-‐level)
“recipe”,
recipes
may
be
par2al,
revised
over
2me.
} Team
reaches
consensus
on
allocaEon
of
(subtasks),
taking
into
account
agents’
capabiliEes.
} Team
members
commit
to
assigned
subtasks.
} Team
members
commit
to
each
others’
success.
Theory
implicitly
requires
group
decision
making
processes.
(Grosz&Kraus, 1996,1999; Hunsberger,
1999; Grosz&Hunsberger, 2006 )
10. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Theory
As
Inspira@on
and
Design
Guide:
Collabora@ve
Interfaces
I:
What’s
the
right
division
of
labor?
Writer’s
Aid
(Babaian,
Grosz, Shieber, 2002)
11. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Theory
As
Inspira@on
and
Design
Guide:
Collabora@ve
Interfaces
II:
Detailed
descripEon
of
students’
interacEons
Condensed
presentaEon
of
students’
interacEons
Direct
feedback
from
teacher
to
students
Complemen2ng
and
working
well
with
people.
S-‐CASTS
(Reddy et al., 2009; Gal, Reddy,
Shieber, Rubin, Grosz, 2012)
12. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Key
Problem:
Interrup@on
Management
drive
home
Route
A
Route
B
0.9
0.1
drive
home
Route
A
Agent
(Observe/
Help)
Recipient
Communicate?
13. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Key
Problem:
Interrup@on
Management
drive
home
Route
A
Route
B
0.9
0.1
drive
home
Route
A
Inform
if
Expected
U@lity>Cost
Agent
(Observe/
Help)
Recipient
Surprise!
14. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
A
New
Representa@on:
Probabilis@c
Recipe
Trees
(Kamar,
2010)
May 20, 2015 Slide 14
Expected
Duration:
19 min
min10 min5 min3 min10 min14 min2 min2
15. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Inform
Decisions
May 20, 2015
Ece Kamar - Harvard University
Slide 15
Expected
Duration:
47.8 min
min5 min3 min10 min14 min2 min242 min
Route A
16. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Inform
Decisions
May 20, 2015
Ece Kamar - Harvard University
Slide 16
Expected
Duration:
28 min
min10 min14 min2 min2
0.11.00.0 0.0
Route A
17. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Interrupt
to
Inform?
EU( )
Updated Plan
EU( )
Original Plan
>
Communicate
if
the
gain
is
bigger
than
the
cost
of
communicaEon
and
interrupEon
18. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
The
Vision
and
the
Challenges
} “Turing
Challenge”:
Can
a
computer
(agent)
team-‐
member
behave,
over
the
long-‐term,
in
such
a
way
that
people
on
the
team
will
not
no2ce
it’s
not
human
or
think
it’s
stupid.
} Test
Environment:
Health
Care
Coordina2on
} Vision:
Care
Augmen2ng
SoQware
Partners
(CASPERs)
} AI
scienEfic
challenges
(MAS,
AI)
arise
from
} demands
of
coordinaEng
care
} needs
for
improving
communicaEon
between
paEents
and
medical
providers
19. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Parents
Primary
Care
Provider
Physical
Therapist
Neurologist
School nurse
Camp
counselor
Health
aide
Teacher
Speech
Therapist
GI
The
Care
for
Children
with
Complex
Chronic
Condi@ons
(Amir,
Grosz,
Law
&
Stern,
AAMAS
2013)
Project
in
collabora2on
with
the
Complex
Primary
Care
Clinic,
Stanford
University
20. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Parents
Primary
Care
Provider
Physical
Therapist
Neurologist
School nurse
Camp
counselor
Health
aide
Teacher
Speech
Therapist
GI
The
Care
for
Children
with
Complex
Chronic
Condi@ons
(Amir,
Grosz,
Law
&
Stern,
AAMAS
2013)
The Problem: care for children with
complex conditions is poorly coordinated,
leading to unmet health needs and
preventable health care crises
Project
in
collabora2on
with
the
Complex
Primary
Care
Clinic,
Stanford
University
21. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Team-‐Based
Care
Plans
for
Improved
Coordina@on
(LPFCH,
2014)
Goals Actions Caregivers
Move to
oral feeds
• Improve mouth muscle tone
• Adjust formula for weight
gain
PCP, GI, OT,
nutritionist
Start
daycare
• Minimize need for tube
feeds
• Assess therapy needs
Parents, PCP,
nutritionist,
home nurse
Go on
family trip
• Arrange portable equipment
• Arrange funding and
transportation
Parents, PCP,
PT, social
worker
22. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Team-‐Based
Care
Plans
for
Improved
Coordina@on
(LPFCH,
2014)
Goals Actions Caregivers
Move to
oral feeds
• Improve mouth muscle tone
• Adjust formula for weight
gain
PCP, GI, OT,
nutritionist
Start
daycare
• Minimize need for tube
feeds
• Assess therapy needs
Parents, PCP,
nutritionist,
home nurse
Go on
family trip
• Arrange portable equipment
• Arrange funding and
transportation
Parents, PCP,
PT, social
worker
Rationale: everybody “on the same page”
In practice: rarely deployed or consulted
23. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Study
of
Complex
Care
Teams
(Amir,
Grosz,
Gajos,
Swenson,
and
Sanders,
2015)
} “From
Care
Plans
to
Care
Coordina@on:
Opportuni@es
for
Computer
Support
of
Teamwork
in
Complex
Healthcare”
} Interviews
and
observaEons
of
team
members:
} Parents
(13)
} Primary
care
providers
(4)
} Specialists
(4)
} Therapists
(8)
} Care
coordinator
(1)
} Program
directors
(2)
} Family
services
coordinator
(1)
} Social
worker
(1)
} Analyzed
using
affinity
diagramming
24. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Barriers
to
Effec@ve
Care
Plan
Use:
Complex
Teamwork
in
Complex
Care
“FLECS”
teamwork
characterisEcs:
} Flat-‐structure
of
team
} Loosely
coupled
plans
and
acEviEes
} Extended
duraEon
of
plans
} ConEnual
distributed
revision
of
plans
} Syncopated
Eme
scales
25. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Flat
Structure
No
single
person
in
charge:
“We
have
different
goals
for
different
specialists;
it
is
hard
to
keep
track.”
(parent)
Need
to
priori2ze
goals
because
“everyone
wants
to
work
on
everything.”
(parent)
26. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Loosely
Coupled
Ac@vi@es
Loose
coupling
makes
appropriate
informa5on
sharing
hard:
“There
isn’t
an
example
when
I
wasn’t
missing
informa2on”
(specialist)
“We
need
to
relay
informa2on
back
and
forth...”
(parent)
27. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Extended
Dura@on,
Con@nual
Distributed
Plan
Revision
No
mechanism
to
support
plan
revision:
Full-‐team
mee2ngs
“totally
not
scalable”
(specialist)
“All
the
status
chats
have
to
be
provider
ini2ated,
and
so
if
you
don’t
remember
to
do
it
or
there’s
no
one
coordina2ng
it,
it’s
like
where
is
it
going,
where
do
you
even
look
for
it?”
(specialist)
28. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Syncopated
Time
Scales
Different
frequencies
of
seeing
the
pa5ent
} Primary
care
providers:
3
to
4
2mes
a
year
} Specialists:
2
to
3
2me
a
year
} Therapists:
1
to
3
2mes
a
week
Different
informa5on
needs:
“A
doctor
asks
if
she
is
walking
and
expects
a
yes/no
answer;
a
physical
therapist
will
ask
how
she
is
walking
and
how
much
progress
she
has
made.”
(parent)
29. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Team-‐Based
Care
Plans:
Ideal
vs.
Reality
} Principles
for
successful
care
plan
use
(LPFCH,
2014):
} “The
plan
of
care
is
systema2zed
as
a
common,
shared
document;
it
is
used
consistently
by
every
provider…”
} “The
team
monitors
progress
against
goals,
provides
feedback
and
adjusts
the
plan
of
care
on
an
ongoing
basis…”
} “Family-‐centered
care
teams
can
access
the
informa5on
they
need
to
make
shared,
informed
decisions.”
} Principles
are
largely
not
realized
in
current
pracEce;
root
cause:
FLECS
nature
of
teamwork.
30. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Team-‐Based
Care
Plans:
Ideal
vs.
Reality
} Principles
for
successful
care
plan
use
(LPFCH,
2014):
} “The
plan
of
care
is
systema2zed
as
a
common,
shared
document;
it
is
used
consistently
by
every
provider…”
} “The
team
monitors
progress
against
goals,
provides
feedback
and
adjusts
the
plan
of
care
on
an
ongoing
basis…”
} “Family-‐centered
care
teams
can
access
the
informa5on
they
need
to
make
shared,
informed
decisions.”
} Principles
are
largely
not
realized
in
current
pracEce;
root
cause:
FLECS
nature
of
teamwork.
Research focus: Develop foundations and
technology to better support complex
FLECS teamwork.
31. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Founda@ons
for
Design
of
Systems
to
Support
Complex
Care
Teams
SharedPlans
(Grosz
&
Kraus
1996)
:
A
computaEonal
theory
of
collaboraEon
“..the
capabili2es
needed
for
collabora2on
cannot
be
patched
on
but
must
be
designed
in
from
the
start.
"
32. SharedPlans
Representa@on
adjust
formula
funding
&
transportaEon
follow
family
prioriEes
move
to
oral
feeds
go
on
family
trip
improve
mouth
muscle
tone
{parents,
primary
care
provider,
specialists,
therapists,
community
members}
{primary
care
provider,
physical
therapist,
social
worker}
{primary
care
provider,
gastroenterologist,
occupa2onal
therapist,
nutri2onist}
arrange
equipment
{…}
{…}
{…}
{…}
33. SharedPlans
Representa@on
adjust
formula
funding
&
transportaEon
follow
family
prioriEes
move
to
oral
feeds
go
on
family
trip
improve
mouth
muscle
tone
{parents,
primary
care
provider,
specialists,
therapists,
community
members}
{primary
care
provider,
physical
therapist,
social
worker}
{primary
care
provider,
gastroenterologist,
occupa2onal
therapist,
nutri2onist}
arrange
equipment
{…}
{…}
{…}
{…}
34. adjust
formula
arrange
equipment
go
on
family
trip
move
to
oral
feeds
SharedPlans Representation
funding
&
transportaEon
follow
family
prioriEes
{parents,
primary
care
provider,
specialists,
therapists,
community
members}
{primary
care
provider,
physical
therapist,
social
worker}
{primary
care
provider,
gastroenterologist,
occupa2onal
therapist,
nutri2onist}
{…}
{…}
{…}
{…}
improve
mouth
muscle
tone
35. improve
mouth
muscle
tone
SharedPlans Representation
adjust
formula
arrange
equipment
go
on
family
trip
move
to
oral
feeds
funding
&
transportaEon
follow
family
prioriEes
{parents,
primary
care
provider,
specialists,
therapists,
community
members}
{primary
care
provider,
physical
therapist,
social
worker}
{…}
{…}
{…}
{…}
{primary
care
provider,
gastroenterologist,
occupa2onal
therapist,
nutri2onist}
36. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Key
Elements
of
SharedPlans
Theory
} Consensus
on
recipe:
} Recipes
may
be
parEal
and
evolve
over
Eme:
} Team
members
commit
to
performance
of
group
acEvity
and
to
each
others’
success:
(Grosz&Kraus, 1996,1999; Hunsberger,
1999; Grosz&Hunsberger, 2006 )
37. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Key
Elements
of
SharedPlans
Theory
} Consensus
on
recipe:
} Support
for
providers
establishing
agreement
on
high-‐level
approach,
establishing
mutual
belief.
} Recipes
may
be
parEal
and
evolve
over
Eme:
} Support
dynamically
evolving
plans.
} Team
members
commit
to
performance
of
group
acEvity
and
to
each
others’
success:
} Support
communicaEon
and
coordinaEon
at
appropriate
levels
and
Emes.
(Grosz&Kraus, 1996,1999; Hunsberger,
1999; Grosz&Hunsberger, 2006 )
38. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Key
Elements
of
SharedPlans
Theory
} Consensus
on
recipe:
} Support
for
providers
establishing
agreement
on
high-‐level
approach,
establishing
mutual
belief.
} Recipes
may
be
parEal
and
evolve
over
Eme:
} Support
dynamically
evolving
plans.
} Team
members
commit
to
performance
of
group
acEvity
and
to
each
others’
success:
} Support
communicaEon
and
coordinaEon
at
appropriate
levels
and
Emes.
Key
capability:
informaEon
sharing
without
informaEon
overload.
(Grosz&Kraus, 1996,1999; Hunsberger,
1999; Grosz&Hunsberger, 2006 )
39. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Key
Roles
for
Technology
for
Suppor@ng
Complex
Care
Teams
} Make
the
care
plan
“ever
present”
} Support
plan
revision
and
expansion
} Support
efficient
informaEon
sharing
Challenges:
} EliciEng
plans
} Inferring
context
in
plan
} Reasoning
about
informaEon
sharing
40. adjust
formula
Agreement on High-Level Approach,
Mutual Beliefs
funding
&
transportaEon
follow
family
prioriEes
move
to
oral
feeds
go
on
family
trip
improve
mouth
muscle
tone
{parents,
primary
care
provider,
specialists,
therapists,
community
members}
{primary
care
provider,
physical
therapist,
social
worker}
{primary
care
provider,
gastroenterologist,
occupa2onal
therapist,
nutri2onist}
arrange
equipment
{…}
{…}
{…}
{…}
41. Dynamically Evolving Plans
…
…
…
…
funding
&
transportaEon
follow
family
prioriEes
move
to
oral
feeds
go
on
family
trip
improve
mouth
muscle
tone
adjust
formula
arrange
equipment
{parents,
primary
care
provider,
specialists,
therapists,
community
members}
{primary
care
provider,
physical
therapist,
social
worker}
{primary
care
provider,
gastroenterologist,
occupa2onal
therapist,
nutri2onist}
{…}
{…}
{…}
{…}
42. adjust
formula
…
…
…
…
funding
&
transportaEon
follow
family
prioriEes
move
to
oral
feeds
go
on
family
trip
improve
mouth
muscle
tone
{parents,
primary
care
provider,
specialists,
therapists,
community
members}
{primary
care
provider,
physical
therapist,
social
worker}
{primary
care
provider,
gastroenterologist,
occupa2onal
therapist,
nutri2onist}
arrange
equipment
Communication and Coordination
{…}
{…}
{…}
{…}
43. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Ongoing
Work:
GoalKeeper
(Ofra
Amir)
44. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Parents
PCP
Physical
Therapist
GI
School
nurse
Camp
counselor
Health
aide
Teacher
Speech
Therapist
Neurologist
Beyond
Health
Care:
Plan
Coordina@on
45. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Victims
Police
Fire
Fighter
National
Guard
Politician
Business man
ICU
NGOs Construction
Coordina@ng
Rescue
and
Rebuilding
Psychologist
46. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Ongoing
Work:
Suppor@ng
Collabora@ve
Wri@ng
“Deploying
AI
Methods
to
Support
Collabora2ve
Wri2ng:
a
Preliminary
Inves2ga2on”,
Gehrmann,
Urke,
Amir
and
Grosz,
CHI
2015
47. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
No
man
is
an
island,
en@re
of
itself...
(John
Donne,
1624)
computer
48. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Giving
Turing
the
Last
Word:
Fundamental
Knowledge
} “The
whole
thinking
process
is
sEll
rather
mysterious
to
us,
but
I
believe
the
ahempt
to
make
a
thinking
machine
will
help
us
greatly
in
finding
out
how
we
think
ourselves.”
Typescript,
BBC
Third
Programme
15
May
1951;
reprinted
in
Shieber,
The
Turing
Test:
Verbal
Behavior
as
the
Hallmark
of
Intelligence,
2004
49. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
References
SharedPlans:
Barbara
J.
Grosz
and
Luke
Hunsberger,
2006.
The
Dynamics
of
IntenEons
in
CollaboraEve
IntenEonality.
In
Cogni2ve
Systems
Research
(special
issue
on
Cogni2on,
Joint
Ac2on
and
Collec2ve
Inten2onality),
7,2-‐3.
Luke
Hunsberger.
1999.
“Making
SharedPlans
More
Concise
and
Easier
to
Reason
About.”
In
Intelligent
Agents
V:
Agents
Theories,
Architectures,
and
Languages,
pp.
81-‐98.
Springer
Berlin
Heidelberg.
Barbara
Grosz
and
Sarit
Kraus.
1999.
The
EvoluEon
of
SharedPlans.
Founda2ons
of
Ra2onal
Agencies,
A.
Rao
and
M.
Wooldridge,
eds.,
Kluwer
Academic
Press,
pp.
227-‐262.
Barbara
J.
Grosz
and
Sarit
Kraus.
1996.
CollaboraEve
Plans
for
Complex
Group
AcEon.
In
Ar2ficial
Intelligence
86(2),
pp.
269-‐357.
Awarded
IFAAMAS
Influen2al
Paper
Award,
2007.
SharedPlans
for
dialogue:
Karen
E.
Lochbaum.
1998.
A
collaboraEve
planning
model
of
intenEonal
structure.
Computa2onal
Linguis2cs.
24,
4
(December
1998),
525-‐572.
50. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
References
Collabora@ve
interfaces:
S-‐CASTS:
Ya'akov
Gal,
Swapna
Reddy,
Stuart
Shieber,
Andee
Rubin,
and
Barbara
Grosz.
2012.
Plan
RecogniEon
in
Exploratory
Domains.
Ar2ficial
Intelligence.
176(1):
pp.
2270—2290.
WAID:
Tamara
Babaian,
Barbara
J.
Grosz
and
Stuart
M.
Shieber.
2002
A
Writer’s
CollaboraEve
Aid.
Proceedings
of
the
Intelligent
User
Interfaces
Conference
(IUI-‐2002),
San
Francisco,
CA.
January
13-‐16.
ACM
Press.
pp.
7-‐14.
Colored
Trails
MAS
Testbed:
Ya’akov
Gal,
Barbara
Grosz,
Sarit
Kraus,
Avi
Pfeffer,
Stuart
Shieber.
2010.
Agent
Decision-‐Making
in
Open
Mixed
Networks.
Ar2ficial
Intelligence,
174(18):
pp.
1460-‐1480.
Interrup@on
Management:
Ece
Kamar,
Kobi
Gal,
and
Barbara
Grosz.
2013.
Modeling
InformaEon
Exchange
OpportuniEes
for
EffecEve
Human-‐computer
Teamwork.
Ar2ficial
Intelligence
195:528–550.
Ece
Kamar,
Ya'akov
Gal,
and
Barbara
J.
Grosz.
2009.
IncorporaEng
helpful
behavior
into
collaboraEve
planning.
In
Proceedings
of
The
8th
Interna2onal
Conference
on
Autonomous
Agents
and
Mul2agent
Systems
-‐
Volume
2
(AAMAS
'09),
Vol.
2.
InternaEonal
FoundaEon
for
Autonomous
Agents
and
MulEagent
Systems,
Richland,
SC,
pp.
875-‐882.
51. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
References
Informa@on
Sharing
for
Care
Coordina@on
and
Collabora@ve
Work:
Ofra
Amir,
Barbara
Grosz,
Krzysztof
Gajos,
Sonja
Swenson,
and
Lee
Sanders.
2015.
From
Care
Plans
to
Care
CoordinaEon:
OpportuniEes
for
Computer
Support
of
Teamwork
in
Complex
Healthcare.”
In
Proceedings
of
the
33rd
Annual
ACM
Conference
Extended
Abstracts
on
Human
Factors
in
Compu2ng
Systems.
Seoul,
South
Korea.
April
18-‐23,
2015.
pp.
1419-‐1428
Ofra
Amir,
Barbara
J.
Grosz,
Edith
Law,
and
Roni
Stern.
2013.
CollaboraEve
Health
Care
Plan
Support.
Proceedings
of
the
12th
Interna2onal
Conference
on
Autonomous
Agents
and
Mul2agent
Systems
(AAMAS
2013),
Ito,
Jonker,
Gini,
and
Shehory
(eds.).
pp.
793-‐796.
Second
Prize,
Compu2ng
Community
Consor2um/AAMAS2013
Challenges
and
Visions
Track
SebasEan
Gehrmann,
Lauren
Urke,
Ofra
Amir,
and
Barbara
J
Grosz.
2015.
Deploying
AI
Methods
to
Support
CollaboraEve
WriEng:
A
Preliminary
InvesEgaEon.
In
Proceedings
of
the
33rd
Annual
ACM
Conference
Extended
Abstracts
on
Human
Factors
in
Compu2ng
Systems.
Seoul,
South
Korea.
April
18-‐23,
2015.
pp.
917-‐922.
52. Barbara J. GroszSmart Enough To Be
On My Team
IBM CSIG
May 2015
Thanks!
Collaborators: Ofra Amir, Krzysztof Gajos
Harvard SEAS; Sonja Swenson, Lee Sanders
Stanford University Medical School.
Research supported in part by Nuance
Foundation.
QuesEons?