Toward a truly personal computer

Download Report

Transcript Toward a truly personal computer

Toward a truly
personal computer
Pattie Maes
MIT Media Lab
Firefly Network, Inc
[email protected]
ACM 97
ACM 97
THE NEXT 50 YEARS OF COMPUTING
ACM 97
ACM 97
THE NEXT 50 YEARS OF COMPUTING
Copyright  1997 ACM, Association for Computing
The files on this disk or server have been provided by ACM. Copyright and all rights therein are maintained by
ACM. It is understood that all persons copying this information will adhere to the terms and constraints
invoked by ACM’s copyright. These works may not be reposted without the explicit permission of ACM.
Reuse and/or reposting for noncommercial classroom use is permitted. Questions regarding usage rights and
permissions may be addressed to: [email protected]
ACM 97
PATTIE MAES
ACM 97
50 years of ACM
50 years of AI
Artificial Intelligence (AI):
goal: build intelligent machines
justification:
– understand intelligence
– practical applications
ACM 97
AI’s holy grail
ACM 97
Cog project (Brooks, MIT)
ACM 97
Cog project (Brooks, MIT)
ACM 97
Next 50 yrs: IA rather than AI?
Intelligence Augmentation:
human
+ machine
= super intelligence
ACM 97
Next 50 yrs: IA rather than AI?
Intelligence Augmentation:
human
+ machine
= super intelligence
ACM 97
History of prosthetics
Overcoming physical limitations:
–
glasses
– hearing aids
– cars
– bicycles
– voice synthesizers
– ...
ACM 97
Why do we need prosthetics
for the mind?
Overcoming cognitive limitations:
–
lousy memory
– only dealing with one thing at a
time
– probabilities, logic non-intuitive
– slow to process large amounts of
information
– ...
ACM
97
Why do we need prosthetics
for the mind?
mismatch complexity of our lives &
our cognitive abilities:
–
too many things to keep track of
– information overload
– learn & remember more
– ...
ACM 97
People are good at:

judgement
 understanding
 reasoning, problem solving
 creativity
ACM 97
Computers are good at:

remembering lots of facts
 searching lots of information
 being in many places at once
 multi-tasking
ACM 97
Some examples of
intelligence augmentation

memory augmentation
 “extra eyes, ears”
 automation behavior patterns
 information filtering
 matchmakers
 transactions
ACM 97
Remembrance agent (MIT
Media Lab)
ACM 97
Remembrance agent (MIT
Media Lab)
ACM 97
Memory augmentation

help remember people, places,
names, actions, ...
 provide "just-in-time"
information
ACM 97
Memory augmentation

help remember people, places,
names, actions, ...
 provide "just-in-time"
information
ACM 97
Remembrance agent
ACM 97
Extra eyes, ears, ...

monitors for bits as well as
atoms:
unusual price stocks
– has certain site changed?
– need more milk?
– is there fresh coffee?
– ...
ACM
–
97
Extra eyes, ears, ...

monitors for bits as well as
atoms:
unusual price stocks
– has certain site changed?
– need more milk?
– is there fresh coffee?
– ...
ACM
–
97
Automation behavior patterns
(Media Lab)
ACM 97
Automation behavior patterns
(Media Lab)
ACM 97
Information Filtering
INFO
personal
information
filter
FILTERED INFO
user interest profile
user
ACM 97
Information Filtering
INFO
personal
information
filter
FILTERED INFO
user interest profile
user
ACM 97
ACM 97
ACM 97
Yenta (MIT Media Lab)
agent
(user profile)
ACM 97
Yenta (MIT Media Lab)
agent
(user profile)
ACM 97
Kasbah (MIT Media Lab)
ACM 97
Kasbah example selling agent
Sell: Macintosh IIci
–
Deadline: March 10th,1997
– Start price: $900.00
– Min. price: $700.00
– Strategy: tough bargainer
– Location: local
– Level of Autonomy: check before
transaction
ACM
– Reporting Method: event
driven
97
Kasbah example selling agent
Sell: Macintosh IIci
–
Deadline: March 10th,1997
– Start price: $900.00
– Min. price: $700.00
– Strategy: tough bargainer
– Location: local
– Level of Autonomy: check before
transaction
ACM
– Reporting Method: event
driven
97
Putting it all together

Example scenarios:
–
monitoring agents & remembrance
agents
– shopping agents & matchmaking
agents
– eager assistants & filtering agents
– ...
ACM 97
Hardware: “wearable
computers”
ACM 97
Hardware: “things that think”
embedded
sensors
processors
communications
ACM 97
Software: “Agents”
Software that is:
 personalized
 proactive
 autonomous, long-lived
 adaptive
ACM 97
Software: “Digital Ecologies”
collections of people & machines:
 perform tasks in radically
distributed way
 very adaptive
 collaboration, competition,
natural selection & evolution
ACM 97
“The network is the
computer”
TM
small efforts by many, rather than
large efforts by few result in
increased:
–
efficiency
– adaptivity
– robustness
ACM 97
Design challenges
TRUST between human and
computer:
 understanding
 control
 privacy
ACM 97
Design challenges
TRUST between human and
computer:
 understanding
 control
 privacy
ACM 97
ACM 97