The society is rapidly entering the information age. The technological evolution throughout the last few years has made it possible for the information revolution to take place. We have witnessed many new inventions and a lot of improvements in information technology. This has in it self been impressing, but they are all worthless if we can't find areas of use for these tools. Somehow we have to take advantage of improved storage capacity, more powerful processors, new communication services and improved bandwidth. We have a lot of state-of-the-art technology, and we have a lot of information. How can we use technology to improve use of and creation of new valuable information? And how can we stimulate, encourage and assist people to co-operate, communicate, learn and develop through human computer interaction?
The universities throughout the world have traditionally been a place where learning and research have been the important factors. These factors remain, but the environment changes and it always have. But it is not only the areas for research that change, it is also the methods and the tools. The way in which we learn, communicate and construct knowledge will change with the evolving information paradigm. This approach, "The Virtual University", is only one example of how we can improve this by using present and future technology. The product emphasises the traditional learning activities, such as conversation, tutoring, discussion and group-work, but it happens in a new, still familiar, setting and with some additional features made possible by technology.
In my view the future scientific- and learning activity will consist in constructing new material out of information that already exists. The future process of learning will consist in collecting information, restructuring it and presenting it in a new form. We can compare this with the philosopher Ludwig Wittgenstein's ideas in "Tractatus Logico Philosophicus" (1921). He claims that a fixed set of atomic facts constitutes all the propositions you can claim about the world, and that discovering the world consists in rearranging these facts. In a way, this is also true about information. By rearranging, integrating, discussing and collecting different sources of information, you can come up with brand new perspectives on whatever you are interested in. The grey professor sitting in his dark chambers producing books out of his lonely mind is no longer an ideal. The huge mass of information that we can access nowadays is a much greater resource than the thoughts of one man alone. The possibilities are enormous.
Another important aspect of the learning process is learning through conversation. You can not learn about the world without interacting with the other actors interpreting the world. That you can learn a lot from conversations is well known. Plato knew this, and so does any student or researcher today. Luckily technology has the power to bring the world together.
The fundamental role of dialogue for learning can not be underestimated. More specifically, I am interested in the benefits to learners of being able to observe others participating in discussion. Such opportunities are becoming fewer with the evolving traditional computer-based courses and distance learning, as well as growing student numbers. To make this possible one has to investigate what happens in educational dialogues. If we can apply the set of rules for this domain into a computer based system we can perhaps be able to observe discussions between virtual persons in a virtual space, and even contribute in the discussions ourselves.
Recent neurological studies indicate that the role of emotion in human cognition is essential; emotions are not a luxury. Instead, emotions play a critical role in rational decision-making, in perception, in human interaction, and in human intelligence. Current computer systems are incapable of either sensing or responding to users who are frustrated, intimidated or pleased with what's going on in the virtual space. State-of-the-art educational technology systems have no way of telling whether a hard math problem is motivating to the learner, or making him/her feel defeated. I envision a world where computer systems are sensitive and responsive to the user's emotional state, where a computerised math tutor can give the learner new material at his/her own pace, and sense when a concept may be beyond the learner's reach. Furthermore we can not underestimate the importance of emotions in discussions. It should be possible for computers to be empathetic to your stress level, your frustration, madness or happiness during computerised conversations, and respond according to this. It should even be possible for the computer itself to get frustrated, happy, irritated or sad during these conversations.
The idea behind the product is to create a virtual university, which, in many ways, is an improvement of real-world universities. Many of the main tasks that a university performs from a student's point of view will have to be covered. These include classes, courses, seminars, discussion groups, self-studying, production of assignments and expert guidance.
It is often difficult for a student to be able to talk to and ask questions to other students or scientists around the world, especially the famous ones, whenever they want. With this product this should be possible.
The systems main goal is to provide an environment that simulates a university and the actors in a much more efficient, timesaving and sometimes perhaps more interesting way than in the real world. This can be achieved by supplying all the features mentioned above in a virtual world. It can be achieved independently of time and place. The scientists don't even have to know if they are having a conversation with one of the students logged in at the moment. This means that it won't take up their time. They can even be dead.
All this can be achieved by creating artificially intelligent doubles of the actors. In this product you can create your own virtual "personalities". You can create yourself, famous scientists, an expert in Social Anthropology that doesn't even exist in the real world, a universal expert that knows everything about any concept, or you can download other personalities which others have made, just to mention a few examples. Even better you can arrange discussion groups among all these personalities. Imagine a discussion between Plato, Aristotle, Sartre, Wittgenstein and yourself. Imagine you can ask a general question about anything, and you will get all the viewpoints from all of these persons. You can step into the office of Nicholas Negroponte to ask some questions about his viewpoints on set-top boxes or have a chat with Konrad Morgan about human centred systems design.
The following will perhaps explain how this is possible. The virtual "personalities" you can create or download are verbally enhanced artificially intelligent entities or "verbots". After you have created or downloaded them you will be able to ask them questions, make them hold lectures and arrange discussions among them. When you are satisfied you can make an abstract or a summary of the session by asking a referent bot to print out your request. The referent bot is your helper in the virtual world. It can carry out commands like printing, rearranging and structuring information, be the moderator in discussions, your personal librarian and your study guide.
The verbots learn by reading and experiencing. You can for example make Plato by feeding the verbot with all of Plato's writings. You can expand Platos wisdom by making it read what others have written about him. By asking questions in natural language, he will answer based on what he has learned. By telling the verbots to explore information by them selves, they will search the web or other sources for information on that specific topic. The verbots will also learn through experiences of conversations and discussions. This means that they can update themselves on their own.
It is important to understand that this type of verbot is not a copy of the real person. The Plato verbot will not be Plato. It gets more interesting when people decide to make a double of them selves. These verbots will have a much better chance of being like their creators. They can be shared with others, but can also be used by one self. In this way you can arrange discussions with yourself. This can be quite interesting if you wish to test whether your thoughts are contradictory or if you simply want to get to know yourself better.
Artificial Intelligence (AI) problems are crucial concerns. Examples of AI problems are computer vision (building a system that can understand images as well as a human) and natural language processing (building a system that can understand and speak a human language as well as a human). I will explain more about these and other problems in the next chapter.
We will need a real-time system for tracking people and interpreting their behaviour. This will enable us to create an interface that is controlled by gesture rather than by a keyboard or mouse. Recognition of sign- or body language and the ability to read what it means will have to be interpreted. This skill is very useful in an interactive interface like this one. It is in fact an essential feature with this type of communication between the user and the computer system.
Many common gestures are universal, but there are certainly many individual differences, not to mention the variety among different nations and parts of the world. We have to be highly concerned about these difficulties while designing the system. In order to make the system understand and recognise facial expressions and other types of body language, it has to be taught how to do this by exploring millions of examples and still be able to learn from new experiences. The system should also be able to recognise people, just as we would recognise the guy we met at the HCI seminar last summer.
One interface possibility is to allow the user to be a part of the interface by creating an avatar of the user. The systems estimates of the user's head, hands, and feet position can make it possible to create convincing shared virtual spaces. The system places the user at a particular place in the virtual room populated by virtual occupants. To make a convincing 3-D world, the person must be placed correctly in the 3-D environment, that is, the representation of the person must be able to occlude, or be occluded by, the graphics. On the remote end information about the user's head, hand, and feet position can be used to drive the avatar that represents the user in the scene. It is important to note that the avatars need not be an accurate representation of the user, or be human at all.
In addition to this we need an application where the continually updating physiological signals provide the engine which drives the graphical display. For example, a full-face mask on the user provides detailed, fine muscle information, which drives representation in the avatar's computer-graphic face, and information about the user's physiology provides further visual feedback to conversational partners. A facial temperature sensor causes the avatar's face to "blush" at the same time that the user's face is blushing in the real world, thus providing an accurate representation of affective reality in the VR interaction. It is not necessary to represent the user as an avatar though. The important thing is that the system should be able to "sense" the user's affective state. This kind of enhancement greatly improves the possibilities for virtual reality environments to approximate physical and emotional presence. Still this does not solve the problem completely. The conversational partners may or may not be real or online at the time. They still have to express emotions. These features have to be learned by the virtual personalities through analogy with millions of examples, or by coding it into their "personality". An important concern is that people are different, and so are virtual ones.
The interaction design must also be able to respond to the users affective states. This means that the system must have the ability of empathy, not only to the users of the system, but also to other virtual entities and avatars. In order to achieve this, the system must be able to absorb information, remember the information, and model the current mood and the emotional life of the persons. Eventually it must be able to build and maintain a more complete model of the person's behaviour and the context they appear in. Then it has to provide a basis for the generation of synthetic system affect and last but not least ensure confidentiality and security.
The multi-modal interface must also combine computer vision, and machine learning with speech recognition. It must be able to learn words by interacting with a person, not only with natural audio input, but also with visual input. This means that the system must recognise objects visually and be able to understand what it is. In addition to this, the system has to learn relations between gestures and speech. Learning such relations can in fact help predict the occurrence of new words or help find salient words from a continuous speech stream. If I refer to an object by pointing to it, it will be easier for the computer system to interpret the meaning of my utterance. Likewise if I raise my eyebrows and shake my head while stating "you are really smart!" Another interesting feature is to extract speech events from background noise. This is needed in for example a very enthusiastic and heated up discussion. The system will of course also have to respond in natural language.
To sum up; we need a system that supports human-to-human communication, human-to-VR-personality communication, VR-personality-to-VR-personality communication and human-to-self (reflexive) communication.
Shape recognition is also a feature that has to be implemented. Not only recognition, but also interpretation of these shapes. In order to make it possible for the system to respond to models and visual representations of different types, it has to be able to understand them.
Many of these topics are already under research by a. o. The MIT Media laboratories (http://www.media.mit.edu/).
I have made some samples from the interface. You will find these illustrations on the web at http://www.morisbak.net/people/stein/papers/vr_univ/english.html
A: Office
This illustration shows an office. In the upper left corner you will find a blackboard available for showing illustrations, models, text etc. The person behind the desk is "the referent bot", but could have been anyone person, e. g. Plato. You can lead a normal conversation with the person using natural language. This situation could e. g. have been a guidance session.
A2: Office2
This illustration shows the office again, with one difference. The user is here represented as an avatar. The avatar performs the user's actions.
B: Class
This illustration shows a class. In front of the room is a blackboard. The tutor is in this case an alien. Right in front of you is a digital piece of paper and a virtual pencil for taking notes and writing questions. Other students are participating in this class. These are either verbots or other persons logged in to the same system as yourself. They can also ask questions and participate in the class in the same way you can.
B2: Class2
This illustration shows a further development of the same class. In the upper right corner the tutor is carrying on with explaining, while you float through space to have a closer look at the planets.
B3: Class3
This illustration shows the same as B2, but the user is here represented as an avatar.
C: Discussion group
This illustration shows a discussion group. In the upper left corner you will find a blackboard. The persons participating in the group are from the left Jean Paul Sartre, Ludwig Wittgenstein, Sylvie (a fictive expert in Social Anthropology), "the referent bot" and Aristotle. You can lead a normal conversation with these persons using natural language. Right in front of you will find a digital piece of paper and a virtual pencil.
The preceding thoughts allow for more than one possible design of the user interface. I will therefore present some alternative approaches. My choice of a final design will depend on evaluation of a questionnaire handed out to students. My first consideration is on where the border between the user and the system, if any, should go. Furthermore how this choice can be implemented to attain the specific interaction.
One possibility is to allow the user to be a part of the interface by creating an avatar of the user. Another is allowing the user to interact with the system from outside without being a part of the interface itself. A third possibility is to make use of Virtual Reality devices such as goggles, gloves, or full body suits, allowing the user to feel as if he/she was really a part of the virtual environment. As a fourth possibility it would be possible to use hologram technology to make the virtual system enter the real world instead of the opposite. Finally, it would also be possible to let the user switch between all the different features, or even combine them. An example of a combination would be using VR-suits to detect the movements of the user and controlling an avatar as a hologram, or on the wide-screen on your office wall.
Other interface considerations with the possibility of alternative designs are the way in which the verbots and other entities are represented. The verbots can be represented as "real people", as human-like cartoon characters or need not be human at all. An example of the last approach is the alien lecturing about the solar system. This also has influence on the way users are represented as avatars. One possibility is to let the users be represented as who or whatever they want.
The interface metaphor I am using in all the different designs is a real world university environment, with add-ons. By add-ons I mean e. g. the possibility to fly through space during a class. This means that this product will have many additional features, which is impossible to have in real world universities. I have not explained much about these virtual reality possibilities elsewhere in this paper, but they are enormous. It is possible to use all our modalities in developing experiences in the virtual world.
You will find a full version of the questionnaire @ http://www.morisbak.net/people/stein/papers/vr_univ/skjema.html (Norwegian). The questionnaire consists of 14 questions and was handed out to 6 people. All of the respondents have academic backgrounds. Their academic backgrounds are varying according to subject and level. The selection was based on my assumptions of a target group for this product.
The questions asked and the answers given were:
|
Respondents |
A |
B |
C |
D |
E |
F |
||
|
1. Gender |
M |
M |
M |
F |
M |
F |
||
|
2. Age |
25 |
25 |
22 |
24 |
24 |
21 |
||
|
3. Educational level |
||||||||
|
Less than three years: |
|
|
X |
X |
|
X |
||
|
More than three years: |
X |
X |
|
|
X |
|
||
|
4. Area of studies |
||||||||
|
Humanities: |
X |
|
|
|
X |
|
||
|
Social Sciences: |
|
X |
|
X |
|
|
||
|
Mathematics and Natural Sciences: |
|
|
X |
|
|
X |
||
|
Psychology |
|
|
|
|
|
|
||
|
Economy |
|
|
|
|
|
|
||
|
Law |
|
|
|
|
|
|
||
|
Other |
|
|
|
|
|
|
||
|
5. How often do you use a PC? |
||||||||
|
Never: |
|
|
|
|
|
X |
||
|
Less than once a week: |
|
|
|
X |
|
|
||
|
1-2 times a week: |
|
|
|
|
|
|
||
|
3-4 times a week: |
X |
|
X |
|
X |
|
||
|
More than 4 times a week: |
|
X |
|
|
|
|
||
|
6. What do you use a PC for when you study? (1 = most important area of use, 0 = "I don't use this feature", - = question unanswered) |
||||||||
|
Writing projects: |
1 |
1 |
1 |
2 |
1 |
- |
||
|
Spreadsheets: |
0 |
0 |
2 |
0 |
0 |
- |
||
|
Statistics: |
0 |
0 |
0 |
0 |
0 |
- |
||
|
E-mail: |
0 |
4 |
3 |
1 |
3 |
- |
||
|
Searching the Internet for information: |
2 |
3 |
0 |
3 |
2 |
- |
||
|
Other: |
||||||||
|
Programming |
2 |
|||||||
|
Obligatory projects |
4 |
|||||||
|
7. How valuable do you think these study activities are? (U = Uninteresting, LV = Little valuable, V= Valuable, VV = Very Valuable, E= Essential) |
||||||||
|
Classes: |
V |
V |
V |
E |
VV |
V |
||
|
Seminars: |
V |
V |
U |
V |
VV |
LV |
||
|
Discussion groups: |
VV |
V |
V |
V |
E |
VV |
||
|
Reading: |
E |
VV |
E |
E |
E |
E |
||
|
Writing: |
VV |
E |
LV |
VV |
E |
V |
||
|
Discussions with other students: |
VV |
E |
V |
VV |
E |
VV |
||
|
Guidance: |
E |
V |
LV |
V |
E |
V |
||
|
Others: |
||||||||
|
Task solving: |
|
E |
|
|
|
|
||
|
Laboratory projects: |
|
|
V |
|
|
|
||
|
Field courses: |
|
|
LV |
|
|
|
||
|
8. How would you appreciate these activities if you could choose your tutors, discussion partners, and/or guides? They can be from anywhere in the world, alive or dead, and you can use them as long as you want and at any time. (W = Worsening , LI = Little improvement, I = Improvement, GI = Great Improvement) |
||||||||
|
Classes: |
GI |
GI |
GI |
GI |
GI |
I |
||
|
Seminars: |
I |
GI |
LI |
I |
GI |
GI |
||
|
Discussion groups: |
I |
GI |
LI |
GI |
GI |
I |
||
|
Discussions with other students: |
LI |
LI |
LI |
I |
LI |
LI |
||
|
Guidance: |
GI |
GI |
I |
GI |
GI |
LI |
||
|
Others: |
||||||||
|
Task solving: |
|
LI |
|
|
|
|
||
|
Laboratory projects: |
|
|
I |
|
|
|
||
|
Field courses: |
|
|
I |
|
|
|
||
|
9. If you had such a possibility, whom would you choose first to be your tutors, discussion partners, and/or guides? |
||||||||
|
A. |
1 |
Fredrik Barth |
||||||
|
|
2 |
Anders Hagen |
||||||
|
|
3 |
Richard Bradley |
||||||
|
|
4 |
Andrew Sherratt |
||||||
|
B. |
1 |
Grady Booch |
||||||
|
|
2 |
Jesus from Nazareth |
||||||
|
|
3 |
Nicholas Negroponte |
||||||
|
|
4 |
Rosa Luxemburg |
||||||
|
C. |
1 |
Jim Morrison |
||||||
|
|
2 |
Christoph Lehninger |
||||||
|
|
3 |
Buddha |
||||||
|
|
4 |
My self |
||||||
|
D. |
1 |
Contemporary or dead experts within different areas of research within the same area of studies |
||||||
|
|
2 |
Contemporary or dead experts within different area of studies |
||||||
|
|
3 |
People from different cultures or societies telling about themselves. |
||||||
|
E. |
1 |
Imanuel Kant |
||||||
|
|
2 |
Wittgenstein |
||||||
|
|
3 |
Heidegger |
||||||
|
|
4 |
Hegel |
||||||
|
F. |
1 |
Newton |
||||||
|
|
2 |
Einstein |
||||||
|
|
3 |
Gallileo Gallilei |
||||||
|
|
4 |
Pythagoras |
||||||
The respondents were asked to read through one page containing background information about the system (see appendix A "Questionnaire").
Next they were asked to see through 6 samples from the interface (see Appendix B "Illustration"). The samples all have a textual description (see appendix A "Questionnaire" (Norwegian) or chapter 5 "Screen shots" (English)).
|
10. How would you like the verbots to be represented? |
||||||||||||||||||
|
As human-like cartoon characters |
|
|
|
|
|
|
||||||||||||
|
Real representations ("real people") |
|
X |
X |
X |
X |
|
||||||||||||
|
Optional (e. g. an alien as a tutor) |
X |
|
|
X |
|
X |
||||||||||||
|
Other |
|
|
|
|
|
|
||||||||||||
|
11. How would you like to interact with the system? |
||||||||||||||||||
|
Hologram technology: |
|
X |
X |
X |
|
X |
||||||||||||
|
Virtual Reality: |
|
X |
X |
|
X |
X |
||||||||||||
|
Traditional (screen): |
X |
|
|
X |
|
|
||||||||||||
|
The use of an avatar to represent the user: |
|
|
|
|
|
|
||||||||||||
|
Other proposals: |
||||||||||||||||||
|
B: |
The system enters the user's brain. A kind of hypnotic state. |
|||||||||||||||||
|
12. Is there anything else you would want such a system to support? |
||||||||||||||||||
|
B: |
Informal discussions with the verbots after the discussion groups. |
|||||||||||||||||
|
E: |
Choose setting yourself, e. g. go back in time to Greece and participate in one of the dialogues of Plato. |
|||||||||||||||||
|
13. Would you like to use such a system? Why/why not? |
||||||||||||||||||
|
A: |
Yes. The system will be helpful in gaining information, especially in the Natural Sciences. Is more insecure about how the system would work within the Human Sciences where almost everything is relative. |
|||||||||||||||||
|
B: |
Yes. It would be a relief to discuss with people that are guaranteed to be competent, and which I have chosen on my own. Will also be interesting to see if my own viewpoints are accepted by different intellectuals throughout history, and by contemporary authorities within my own area of studies. |
|||||||||||||||||
|
C: |
Yes. Thinks that the system will have great value within Natural Sciences. Especially 3D virtual models of chemical and biological processes in e. g. the human body or in a bacteria. You could for instance have a tour inside the human body. |
|||||||||||||||||
|
D: |
Yes. Classes on demand. Can adjust the content of the classes as you want. Can ask as many questions as you want without being shy. The persons in the system know "everything", and have the ability to remember everything in comparison with a human. |
|||||||||||||||||
|
E: |
Yes. A splendid opportunity to sort out your own thoughts. You learn much more and quicker by discussing and asking questions rather than reading. |
|||||||||||||||||
|
F: |
Yes. To develop a deeper understanding of the wisdom of those who lay down the foundation stone for contemporary science, e.g. modern physics. |
|||||||||||||||||
|
14. In what degree can you agree with these statements? (D = Disagree, DL = Disagree a little, DK = Don't know, AL = Agree a little, CA = Completely agree) |
||||||||||||||||||
|
"I want computers to be more like humans" |
CA |
CA |
AL |
DL |
AL |
D |
||||||||||||
|
"I want computers to be tools, not human like partners" |
CA |
DL |
CA |
AL |
DL |
CA |
||||||||||||
|
"If we give computers human abilities they will be harder to use" |
DK |
D |
DL |
D |
D |
AL |
||||||||||||
|
"Computers can never be like humans". |
CA |
D |
CA |
CA |
D |
CA |
||||||||||||
I have reached 4 males and 2 females between 21 and 25 years. 3 students have studied less than three years and 3 students have studied more. 2 have Humanities, 2 have Social Sciences and 2 have Mathematics and Natural Sciences as their area of studies. I feel I have reached my target group, but I lack respondents from a wide spectre of study areas. It is important to avoid the failure to not take into account the differences in educational objectives, presentational requirements, and expected learning outcomes in different subject areas. It appears to be clear differences between arts, science and applied subjects in their teaching requirements, and this should have a major impact on the design of the system. It is also of course too few respondents all together. I would have handed the questionnaires out to many more, but the lack of time prevented me from doing so.
How often the respondents use a PC varies, but only one of the respondents claims she never uses one. This means that they are familiar with computer technology and likely to be interested in improvements of today's technology and software.
I am not surprised that the most important computerised task is writing projects. The respondents also use e-mail and search the Internet for information. These results may indicate that the need for information and communication are present. It might even indicate that the students collect information, restructure it and present it in a new form by using information technology. I won't rely too much on these conclusions though. The questions asked are not sufficient for stating that this is the case.
If I put numbers on the results in question 7 I find that the ranking of study activities is this:
(Uninteresting=0 Little Valuable=1 Valuable=2 Very Valuable=3 Essential=4)
I don't think this ranking tells us much about which activity is more valuable than the other. All of the activities seem to be very valuable. The result is still very important. It shows that I will have to take all of these activities under concern while designing the system.
If we see the results from question 7 in relation to the results in question 8, we find very uplifting results. It shows that the respondents believe that most of the activities have great chance of being improved if they can choose freely among tutors, discussion partners and/or guides. This means that if the product is successfully implemented after its intentions, it will improve the study activities.
Question 9 gives me a fine list of verbots. If I had handed out these questionnaires to many more I would probably be able to sort out the most popular verbots to implement for the first edition of the product. It also tells me that the students really have dreams about interacting with dead or alive celebrities.
Four of the respondents want the verbots to be represented as "real people", but three of them would like to have the freedom of choice. This means that we should probably listen to the minority and implement something that gives the users freedom to choose how they would like the verbots to be represented.
A Virtual Reality environment is the most popular solution to interaction with the system, but hologram technology is also very popular. 2 of the respondents state that they would prefer a traditional solution using computer screens, but one of these two also states that she likes the hologram solution. Respondent B has an additional proposal. He would like the system to enter the user's brain and transform the experience to a kind of hypnotic state. It is a very interesting proposal, but it will have to wait for later versions of the Virtual University. None of the respondents answered that they would like to be represented as avatars.
As the results from question 13 shows, it is a great interest for this product. None of the respondents answered that they didn't want to use such a system. This does perhaps indicate that there is a need for such a product.
Few of them answered that they wanted to add additional features to the system. The reasons for this can be many, for instance laziness, lack of understanding of the system or inability to see possibilities. Two of the respondents did answer that they wanted the system to support something else, and I think the ideas are great. The ideas were informal discussions with the verbots after the discussion groups (12 B), and the possibility of choosing historical setting (12 E). Respondent C answered in question 13 that he would like 3D virtual models of chemical and biological processes in e. g. the human body or in a bacteria. He also proposed a tour inside the human body. I think that these ideas are very good and I would like to include such features into the system.
The last questions of the questionnaire was put there to find out whether people feel comfortable with computer systems being more like humans or not. This kind of work represents a controversial area of human-computer interaction, in part since attributing emotions and emotional understanding to machines has been identified as a philosophical problem; what does it mean for a machine to express emotions that it doesn't feel? What does it mean for humans to feel "empathised with" by machines that are simply unable to really "feel" what a person is going through? The answers are very varying, and it gives me no final answer except that it is a highly controversial issue.
On this background I recommend further development of such a system. The system should support classes, courses, seminars, discussion groups, self-studying, production of assignments and expert guidance, and should take into account the differences in educational objectives, presentational requirements, and expected learning outcomes in different subject areas. By concentrating on these activities it emphasises conversation, tutoring, discussion and group-work, which are the traditional learning activities at universities. Verbally enhanced artificially intelligent entities or verbots should support these features. The verbots can be created by the user or downloaded from whatever network that will exist in the future. You should be able to create yourself and others as well as non-existent experts. By integrating advanced multimedia features that exploits all of the human modalities, it will broaden and improve the different activities even more.
It should also be possible to select different settings. You should e. g. be able to have a conversation with Michelangelo while he paints the Sistine Chapel ceiling or have a meeting in the UN-building.
I recommend that the first development of the system should be based on virtual reality technology. Hologram technology and the possibilities for making the system enter the user's brain should be explored along the way. Later versions should perhaps have the ability to switch between different interaction styles.
The representation of the verbots should be optional, but it is important that they can be represented as "real people" (copies of the real world persons).
The systems main goal should be to help learners learn by discussion and self-reflection. The users should be able to actively interpret information and transform it into personal knowledge by using the system. If we can achieve this we have found an excellent area of use for information technology. We have improved use of and creation of new valuable information, and we have stimulated, encouraged and assisted people to co-operate, communicate, learn and develop through human computer interaction.