Managing the Participant Viewpoint for greater Immersion
This is a Printer Friendly Article on the Virtual Worldlets Network.
Return to the web-view version.
Author Information
Article by Virtual Worldlets Network
Copyright 09/09/2007
Additional Information

If you run a virtual world, whose premise centres around a form of living or lifestyle significantly different from mainstream society; whether roleplay, persona play, or full blown alternate life, the last thing you desire is for people to enter it, completely ignoring the in-world culture and behaving any way they like.

To a certain extent, this issue is unavoidable. Free will frequently gets in the way, and is the reason tools to remove a person from the world - however temporarily - exist.

However, humans are social animals, and most, albeit not all, just go with the herd, following perceived behavioural norms. What if it was possible to modify behaviour of participants in your world, almost subconsciously, whilst they are within your world?

Normal behaviour takes it's cues from many sources. Several are internal to the individual and cannot/should not be tampered with. Others are taken from the environment, and mood of the place they find themselves in, at the time. Do you behave the same way at your local swimming baths, that you do at a cocktail party? Subtle, almost subconscious social cues at both locations modify your behaviour without really thinking about it. Now compare a battlefield to the cocktail party. Very different, practically incompatible life, yet again survivable through adaptation to cues.

Now, such gentle manipulation would not be 100% effective, and could not factor in happenings at individual's homes or business lives at that particular moment, and would not aid those too immature to grasp what was going on, but for the majority, it takes the raw edge off, and, over time, they may find themselves conforming much more than they intended to; whilst inside your world.

How to Accomplish This

Static backdrops do not work. This technique, of a lavishly built world that suggests certain modes of behaviour has been tried and tested for decades. It does not work. There is not sufficient activation of the social, group nature of the individual, when faced with a static environment.

Dynamic environments, changing to the presence of the participant, responding to their actions, actually do not help either. They are an improvement on the static, but they still lack that social aspect.

The next step up, is to include puppet avatars, or beings. These lack any fraction of artificial intelligence, but just drift along performing set actions, or saying set phrases. They mill around in the background, out of the corner of participant's eyes. More than just shopkeepers or farm hands, they actually seem to go about their day to day lives in the background. This gives the illusionary feeling that the world is populated, at least until examined closely. This does start to have the desired effect.

To increase it further, enough to be useful, the individual participant must feel immersed in a world where everyone around them is moving with a purpose in life. They must have requirements and desires, and be seen to meet both, whilst behaving within acceptable parameters for the type of environment the world is. Thus, the participants will start to identify with this behaviour, and ape it.

It does not take strong AI to achieve much of this: simple needs based AI as demonstrated in "The Sims" suffices. This is good, as to create convincing interpersonal AI for hundredsof autonomous agents; perhaps thousands - is beyond the capabilities of any world we have created to date.

Whichever way we look at it, modern artificial intelligence systems are not even remotely near passing a Turing test, let alone passing themselves off as authority figures, and social control leaders.

Artificial life and A-life technology is actually quite advanced. Whilst AI agent community members could not make speeches, and speak like sentient beings all the time, unless dealing exclusively with one another, they can move and act like such, at least, enough to be doing so in the background, at the edges of user perception.

If participants are observing many bot agents on a significant scale, interacting with one another throughout the fabric of a world, they are extremely likely to embellish their memories of the environment enough that they remember more complete and human behaviour than was actually present.

This then has the knock-on effect of reinforcing social stereotypes, particularly within those who do not so much follow their own path, as they do go with the flow.

Studies

Recent studies into the therapeutic and mental manipulation aspects of virtual environments have come up with some interesting data that, when extrapolated, may well pave the way for use of autonomous AI agents to reinforce behaviour patterns within virtual environments.

One study, by Ann E. Schlosser of the University of Washington concerned memory creation within virtual spaces. Testing out a group of subject's abilities to learn how to use a new camera system, Ms Schlosser discovered that the more immersive a VR system is, the more inclined participants are to create "vivid internally-generated recollections that pose as memories".

This means that, as in physical existence, within a suitably high fidelity VR environment, queues delivered by elements within the VR system, will trigger the formation of extrapolated memories within the user - they will remember things that did not occur, but which they are convinced did in fact, occur.

This presents the owners of alternative lifestyle and intensive role-playing VR environments with a possible psychological tool that could be used to reinforce desired behaviour patterns or lessen undesirable behaviour by means of subtle manipulation of participant subconscious - brainwashing if you will.

Such effects may be more pronounced that this study would suggest, if the effects of a second study are factored in.

Researchers at the University College of London have recently recreated intense psychological experiments which were halted in the mid 1960s due to severe ethical concerns.

Stanley Milgram's 'obedience to authority' experiment is infamous. It attempted to answer the question of whether Nazi war criminals could really 'just be following orders'. His experiments suggested people will do horrendous things if someone in a position of authority tells them to. But further work in this area was hampered by ethical concerns.

The University College experiments used VR re-conduct the same experiment, bypassing moral qualms for now. The moral debate is a bit beyond the scope of this report. What are interesting to us, are the results.

The UCL researchers found that participants in the virtual environment acted as though the situation was actually occurring. The original 1960s experiment created a situation where a random sampling of general public individuals watched another individual take a test. Every time the individual answered a correction incorrectly, the member of the public was supposed to shock them with a powerful electric jolt. In the scenario, the individual taking the test was an actor, faking being zapped, but the member of the public did not know this. An authority figure was telling them to shock this person, and other people were complying, so, most of the public administered what they thought were shocks.

In the UCL version, the same results were found. Despite both the actor AND the authority figure being replaced with a VR-generated agent and avatar, the test group - first year university students in this case, obediently did as instructed

"It has been argued before that [the] immersive virtual environment can provide a useful tool for social psychological studies in general," said Professor Mel Slater, the lead researcher at the UCL department of Computer Science, in his findings.

"Our results show that this applies even in the extreme social situation investigated by Stanley Milgram."

In Closing

In order for such a system to work, a great deal of the underlying world mechanics must also be implemented.

Therefore, such a world, would require a functioning economy, and labour system, and would have the side effect of creating meaningful work and activities for the human participants of the world.