Transcript Future of virtual worlds - The University of Sydney

DESC9180
Designing Virtual Worlds
Week 12
Future of Virtual Worlds
6pm – 9pm
Tuesday, October 16th, 2007
Kathryn Merrick and Owen Macindoe
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Announcements

This lecture will begin with 15 minute
demo in the Sentient


Level 2, through the glass door to the Key
Centre, then first door on your left
Please join me downstairs
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Lecture Overview

Blurring the line between the virtual and
the physical

Virtual worlds as platforms for
advanced AI

Technological improvements and
growth models for virtual worlds
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Blurring the Line Between
Physical and Virtual





DESC9180 Designing Virtual Worlds
Augmented reality
Augmented virtuality
Mixed reality
Alternate reality
Mirror worlds
University of Sydney, October 2007
Augmented Reality

Combination of the real world with
computer generated data:

Current research concerned with use of
live video imagery, which is processed and
‘augmented’ with computer generated
graphics
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Augmented Reality: Tools and
Techniques
Head
mounted
display
Tangible
pointer
Square
marker
patterns
Kyoto Garden
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Augmented Reality Applications
Kyoto Garden
DESC9180 Designing Virtual Worlds
AR Toolkit
University of Sydney, October 2007
Research Challenges with
Augmented Reality

Problem of tracking the user’s viewpoint

Camera calibration

Speed for real time use

Use of obtrusive, worn hardware
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Mixed Reality

The merging of real world and virtual
worlds to produce new environments
where physical and digital objects can
co-exist and interact in real time.
Paul Milgram’s Virtuality Continuum
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Mixed Reality: Human Pacman


Overlay of
physical with
virtual
Bluetooth used
to give physical
objects digital
meaning
Human pacman
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Mixed Reality Research
Challenges

Wearable hardware

Networking and
communication

Real time function

Health and safety
DESC9180 Designing Virtual Worlds
Wearable hardware for human
pacman
University of Sydney, October 2007
Alternate Reality

Physical and virtual worlds are merged
as an interactive narrative

Virtual world may be revealed through:


Internet, telephone, email, mail
Not necessarily through computer
graphics
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Alternate Reality Terms

Rabbitholes (trailheads): mark the entry
to the alternate reality

Puppetmasters control virtual content
while players interact with it

TINAG (this is not a game)
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Alternate Reality Research Issues

Can we create alternate reality games
without human puppet masters?

Combination of alternate reality and
mixed reality

Applications of alternate reality

Education, information literacy, advertising
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Alternate Reality Examples

The Beast (2001)


Promotion for A.I.:
Artificial
Intelligence
I Love Bees (2004)

Promotion for Halo
2
www.ilovebees.com
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Virtual Worlds as Platforms for
Advanced Artificial Intelligence

AI and virtual worlds are merging in
two ways:


Virtual worlds as test-beds for AI
AI to enhance virtual worlds
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
AI Enhancing Virtual Worlds

The promise of high quality AI is a
major selling point for virtual worlds


Especially true for game worlds
AI makes the world dynamic

Changing worlds have greater ‘replayability’
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
AI Enhancing Virtual Worlds

Current AI techniques for virtual worlds
are relatively simple:



Ensures predictability
Easy to implement
Virtual worlds are a complex
application domain for AI research…
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Complexity of Virtual Worlds

AIs must monitor large amounts of data
and choose from many different actions

World can change while AI is thinking

Changes may be unpredictable due to


Open-endedness of virtual environment
The presence of humans
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Virtual Worlds Enhancing AI
Complex
virtual
worlds
require
research of
new AI
techniques
Behavioural complexity

cognitive
agents
motivated
agents
learning
agents
scripted
behaviours
swarm
agents
Implementation complexity
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Motivated Agents for Dynamic
Virtual Worlds

Motivated agents are one emerging
research area for virtual worlds

The idea is to combine computational
models of human motivation with
artificial intelligence techniques

Interest, curiosity, competence motivation
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Advantages of Motivated Agents




Motivated agents find their own goals
If the world changes, motivated agents
can change too by finding new goals
In complex worlds, motivated agents
focus on a small number of goals
Motivated agents also have applications
in other research fields
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
A Model of Motivation for Agents
in Virtual Worlds

Agents
monitor
changes
(events) in
the world
Events with
moderate
novelty are
highly
motivating
DESC9180 Designing Virtual Worlds
1
Interest (Reward)
I(2N(t)) = R(t) .

Positive
feedback
0.5
Interest
0
0
-0.5
0.5
Novelty
1 2N(t)
1.5
Negative
feedback
-1
University of Sydney, October 2007
2
Future Technical
Directions

Virtual world technologies are driven by
(and also drive) other computer
technologies:



Graphics, chips, memory…
Servers, databases…
The increasing prevalence of virtual
worlds is also influenced by social,
cultural and economic values
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Have Virtual Worlds Reached A
Tipping Point?

What further changes will promote
growth?



Release of client code has already
prompted improvements
Release of server code would also
facilitate growth
BUT companies such as Linden would
need to find new ways to make money
from virtual worlds
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Possible Growth Models for
Virtual Worlds

Open source server code

Software licences for server code

APIs allowing other companies to build
clones of virtual world software

Co-location facilities with hardware
outsourced to other companies
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Summary


Virtual worlds are an exciting, emerging
technology
We can expect to see:





Merging of the virtual and physical
Improvements in artificial intelligence
New technologies to support virtual worlds
New business models for virtual world
providers
New applications of virtual worlds
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Today’s Tutorial

Preliminary critiques:



Spend 10 minutes demonstrating your
design to another group
Discuss strengths, weaknesses and
strategies for improvement
Work on Task 2
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Next Week


Each group will give a 20 minute
presentation
Presentation should include:


Slides describing plot, strengths,
weaknesses (2 marks)
Demonstration of game-play sequences (3
marks)
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007
Next Week

Each group to hand in:


One report describing the game
One report describing the group process
DESC9180 Designing Virtual Worlds
University of Sydney, October 2007