A 2020 Vision:
Education in the next two decades


James Levin

University of Illinois





Appeared in 2002 in the Quarterly Review of Distance Education, 3(1), 105-114.



It is difficult to predict the future. In our everyday lives, we implicitly depend on a "future will be like the present" prediction. More sophisticated forecasters often predict that potential changes will occur quickly in the short term, ignoring the institutional barriers to change. Similarly, they often are too limited in predicting long-term changes, since it is very difficult to think through the full range of unexpected side-effects that changes in one part of society cause in other parts.

With electronic technologies today, change is the only constant. Following Moore’s law (Moore, 1965), microelectronics has produced doubling of performance every year and half since the mid-1960’s, leading to an order-of-magnitude improvement in performance every 5 years. Projecting ahead over the next twenty years, this rate of change would lead to the development of electronic technologies 10,000 times the power of today’s devices. Just to see what this looks like projecting backwards to the 1960’s, everyone wearing an digital watch has the computing power on their wrist comparable to the mainframe computers of the mid 1960’s; anyone with a laptop computer today has the computing power equivalent to the supercomputers of the mid 1980’s.

So what will education be like in the year 2020? What should education be like in the year 2020? Technologies enable possibilities but they don’t determine future development. This paper will explore some possibilities enabled by technologies that may have positive implications for education and society more generally.


Current contexts for learning and teaching

Schools: Past, present and future?

The dominant form of formal education today is schooling. It is so much a part of our concept of education that we sometime forget that it is not the only framework for learning, and that the current form of schools and schooling has evolved fairly recently. Formal education existed before there were schools. Schooling is education that takes place in building that are mostly isolated from the rest of society, in which most of the learning activities consist of exercises. There is a separation between learning and doing, a separation between the location of learning and the location in which that learning is eventually to be put into practice.

Before schools were the dominant form of education, a few privileged learners worked with tutors. The majority of advanced learning, however, took place in apprenticeship settings, formal learning frameworks in which novices acquired knowledge and skills in the context of practice. Most people today associate the term "apprenticeship" with craft apprenticeships, but in fact apprenticeships are the most common form of learning in most professions. Medical internships and residencies, law internships, and other advanced graduate study are all apprenticeships.

If apprenticeships are used to teach our doctors, lawyers, and scientists, why are they not used more widely? One of the most important reasons is that apprenticeships are expensive. They take the time of the experts; they take the time of the novices and others involved. It would be impossible to support mass education with conventional apprenticeships.



However, new electronic media enable new forms of education. As more and more of the work in a society occurs online, it becomes possible to engage more and more learners in "teleapprenticeships." These are formal educational frameworks that engage people in learning through their remote participation in ongoing work settings.

In face to face apprenticeships, novices start on the periphery of the activity, observing and being given simple tasks that contribute to the going work (Lave & Wenger, 1991). As they acquire expertise, the novices move toward the center of the activity (Hutchins, 1996). One of the reasons it is difficult to involve large numbers of apprentices in a face-to-face setting is that they have to physically be in the work setting, which can be disruptive all by itself.

In teleapprenticeships, novices start by observing the ongoing activity on the periphery. To the extent that the activity is being conducted via email or web-based conferencing, the novices can be added as additional recipients of the email or as people reading the conference postings. There’s a naturally occurring phenomena called "lurking", in which many people involved in an online conference, listserv or newsgroup read the interaction without contributing to it. These "lurkers" are serving as peripheral participants of the interaction — in a teleapprentice interaction, this peripheral participation would become legitimate (Lave & Wenger, 1991).

Novices, while they start out as relatively passive observers, do not remain passive for long. They are given small tasks to do, not as exercises but as real contributions to the ongoing activity. In teleapprenticeships, novices could be given information seeking or information synthesis tasks. The creation of pages of annotated web links is a common activity on the web today — this kind of information "hunting and gathering" could be formally included in teleapprenticeships. The passive peripheral participation serves to give the novices a global understanding of the overall context of the activity, so that these small tasks can be understood as part of the overall activity, rather than isolated "islands" of activity.

One advantage of teleapprenticeships is that learners can participate in more than one teleapprenticeship at a time, so they do not need to commit prematurely to a very specific area. It is also easier to move from one teleapprenticeship to another as learning goals and needs change. It is also easier to involve larger numbers of learners without disrupting the work setting.

Apprenticeships are typically used for specialized learning. In fact, we can divide things to be learned into two categories, "common knowledge" and "specialized knowledge". Specialized knowledge is knowledge that somebody in a society needs to know but not everyone. In our society, someone needs to know how to do brain surgery, but everyone does not need to be a brain surgeon. On the other hand, there is knowledge and there are skills that everyone in a society needs to know. This varies from society to society, and varies from time to time as a society changes, but at any one place and time, there are some things that everyone needs to know.

We have explored a wide variety of teleapprenticeships, especially teaching teleapprenticeships (Levin, 1990; Levin & Waugh, 1998). These are frameworks that allow teacher education to take place within the context of remote classroom settings. For example, we have developed and evaluated a number of different ways to allow undergraduate teacher education students to participate in the activities of K-12 classrooms, even before they participate in the conventional face-to-face apprenticeships of early field experience and student teaching. Some of these teleapprenticeship frameworks worked well; others worked poorly. A new interactional framework works best when it satisfies important goals for the participants within available resources available to the participants. It especially works well when it allows goals to be achieved that could not be achieved previously, or when it allows goals to be achieved with fewer resources.

Teleapprenticeships may good educational frameworks for acquiring specialized knowledge. But how would learners acquire the required "common knowledge"? Can this be best acquired through the kind of "work" oriented interactions that characterize teleapprenticeships? How is it acquired in other educational systems?

Some of this "common knowledge" is formally taught in our elementary and secondary schools. The "three R’s" of reading, writing and arithmetic consist of knowledge and skills that have been judged to be things that every adult in our society needs to know. However much more is learned in informal educational settings, such as play groups and other peer groups, family settings, after school activities, and in the gaps in formal schooling (recess, lunch, on the way to school and on the way home, etc.).

One way that children learn common knowledge in our society are by participating in joint activities with other children, often team activities in competition with other teams. In these settings they learn the skills, benefits, and costs of cooperation and competition. In formal education, these same skills are sometimes acquired in the context of school-based projects, which often have students working in small groups on a project, sometimes in competition with other students or teams of students.

Tele-task forces.

With new technologies, it is possible to create collaborative network-based projects, with diverse participants from widely distributed locations. There are many different frameworks that these collaborative activities can take (Harris, 1995; Levin, 1997; Levin & Waugh, 1998; Riel, 1998). Let us focus here on one called "tele-task forces". With this interactional framework, learners and mediators jointly tackle some task, and interact for a time in efforts to accomplish the task.

Like the teleapprenticeship framework described earlier, this tele-task force framework has parallels to the world outside of education. In some areas of society, unique problems arise that require bringing together a group of people to address the task, and so a "task force" is assembled. The group works together for a while, and then disbands after its work is done. The some of the same people may join together again on other task forces, but the particular task force terminates after a relatively short time period.

We have also been exploring tele-task force frameworks for learning and teaching. One good example was "the Zero-g Design Project", a year-long design activity, focusing on how to design artifacts and processes for people living in a free-fall environment (Cervantes, 1993; Levin & Cervantes, 2001 in press). The overall Zero-g Project was composed of a set of smaller design challenges, each of which was tackled by a "tele-task force" of students, teachers, and aerospace experts. If the teachers and students had participated as peripheral participants of an ongoing aerospace design effort, then this would have been more of a teleapprenticeship, but instead these design challenges were ones that were of interest to the aerospace participants but not ones that they were working on.

The first design challenge was to think about the mechanism of personal navigation in a zero-g environment. Since things don't fall, you can navigate down a hallway just by exerting a force at one end (for example, by pushing off from a wall), and then sail down the hallway. When you get to a place where you want to stop or turn, you have to exert another force. What if you were going down a hallway and, in mid-course, there was a person coming in the opposite direction on a collision course. How could you avoid the collision with minimal effort and disruption to your own travel and that of the other person?

This "navigation design challenge" was addressed jointly by widely distributed and diverse participants for about a month. Then two new design challenges were raised, the design of better ways to prepare, consume, and recycle food in zero-g and the design of recreational activities in zero-g. These task forces continued for approximately two months. Then the challenge of designing a school in zero-g was tackled, which took approximately three months. In all, the Zero-g World Design Project extended over a full academic year.

Why involve learners in teleapprenticeships and tele-task forces?

Why would we want to involve learners in these kinds of interactions with the world outside of classrooms and schools? One advantage of teleapprenticeships and tele-task forces is that learners are acquiring knowledge and skills within the context of use. One of the big problems facing our current educational system is the "transfer problem" - the things learned in the classroom often aren't used when the person is in the context in which those things should be used. Transfer is not as big of a problem in apprenticeship learning because the context of learning is the same as the context of practice. The context of learning in teleapprenticeships and tele-task forces is also much more like the context of practice than conventional schooling.

Another advantage is the motivation provided by having a diverse "real" audience for learning. In most of today’s classrooms, the only audience for learning activities is the teacher. Providing a broader set of people who care about the learning activities can be a strong positive motivation for learning (Cohen & Riel, 1989).

A third advantage is that these kinds of interactions allow for a more diverse set of people to interact in a way that can draw upon the diversity as a strength rather than a weakness. In most face-to-face synchronous interactions, communication among a diverse set of people is difficult. Interaction with network technologies in largely asynchronous settings allows for more diversity, and in fact, can be organized to turn that diversity into a major strength.

For example, normally novices are very limited in what they can contribute to a apprenticeship setting, because they lack the skills for real-time performance that can integrate with the ongoing activity. However, when much of the interaction takes place in asynchronous media such email or web-based conferencing systems, then the slower performance of novices is not as much a barrier.

In fact, in times of rapid change, expertise is a double-edged sword, and can serve as a barrier to problem solving. Expertise has a half-life, losing value as the world changes, and must be periodically re-examined to check for its continued validity. Novices who lack expertise in a domain sometime propose different approaches. In a relatively stable domain, these alternatives are most often not useful. However in times of rapid change, alternatives which previously were impractical or otherwise not useful can suddenly become a viable solution. For example, in the Zero-g Design Project described above, most of the ideas that the novice participants came up with were not very helpful to the aerospace expert participants, but some of the suggestions were found to be of interest because they applied to situations that none of the experts had much experience with.


The future of precollege education

What will the physical environment of learning be in the future? One extreme is the "video game arcade" classroom nightmare, which imagines that each student will sit in a classroom in front of a computer, and the teacher will sit in the front with an even bigger computer, and most of the interaction is through networks (even though they're all seated in the same room).

Once you have most of your interaction over computer networks, questions immediately arise: Why bother traveling to the same room to do this? Why not just stay at home? So the other extreme is a totally distributed model, in which learners stay at home and learn through interaction over computer networks.

My vision is that there will be "neighborhood learning centers", smaller than current schools, closer to learners' homes, and with adults there to organize the activity. What kind of activities? The reason why I like to call them "neighborhood learning centers" and "adults who organize the activities" is so that we can look more easily at what roles, learning frameworks, and social organization are most useful for such a setting. If we called them "schools" and called the adults "teachers", that carries along a lot of immediate assumptions (recess, hall passes, principals, etc.) that I'd like to at least address before including them in this vision.

How would learning be organized in these NLCs? I imagine more like the organization in a small day care center, rather than what is typical of a typical classroom. That is, there would be a number of different kinds of activities going on at once (in "centers", perhaps), with adults to provide overall supervision. There would be a lot of peer and cross-age interaction (something relatively rare in the prototypical classroom of today). The tracking of progress would be done to a large extent on-line, based both on the learners’ work on-line and on notes taken by the adults on hand-held wireless-networked devices.

There would be a mix of face-to-face interaction in learning and online interaction. Much of the "fact" learning would be done in computer-based learning environments. Much of the higher level learning would be done through tele-task force activities. Much of the advanced learning would be done through tele-apprenticeships, in which the learning would take place in the context of doing.

What are the chances that such neighborhood learning centers could come to play a significant role in learning and teaching? Any vision of changes in education that depends entirely on changes within the educational system should be taken with a very large grain of salt, given the many institutional barriers to change. On the other hand, if there are changes in the overall society, then parallel changes in the educational system are more likely.

The same line of reasoning that leads us to envision NLCs apply to much of the current office work in our society. The main reason for having the office workers of a company all gather at once place is that traditionally the major information flow was paper-based. Pieces of paper are physically carried from one desk to another, appearing in a employee’s "in-box", processed in some way and either physically transferred to a file cabinet or put in an out-box, to be carried to yet another desk. For this reason, people spend hours commuting to and from work in large urban areas.

Much of this information flow has been transferred to electronic media. Physical memos are being replaced by email. Physical forms are being replaced by web-based databases. Physical meetings are being replaced to some extent by increasing use of telephone, audio conferencing, and video conferencing.

Once a substantial amount of work interactions are taking place electronically instead of on paper, questions arise: Why bother traveling to the same office space to do this? Why not just stay at home? Thus there is an increasing interest in telecommuting. Right now, the time spent commuting is a cost paid by each employee themselves, both in terms of monetary cost and also in terms of time spent. Many employees would be very happy not to have to spend 30 minutes to two hours or more a day commuting, instead using that time for family or personal activities. The usual image of telecommuting is that of a person sitting at their computer in their informal clothes, doing their work while surrounded by the comfortable setting of their own home.

However, there are still economies of scale gained by gathering multiple people in one work place. Additionally there are social benefits of working with other people rather than working in isolation at home.

An attractive alternative to either of these extremes is that of a "neighborhood office center", a place near where people live that they can travel to in order to do their office work. The physical setup of a NOC might be similar to that of a standard office — the big difference would be that not everyone working at a particular NOC would necessarily be working for the same employer. The portion of the office space, furniture, communications media, coffee machine, and other resources used by a specific employee would be paid for by the employer. Much of the interaction among employees of a specific organization would be conducted electronically, with face-to-face meetings limited to those situations in which that interactional format is best suited. So a particular person might travel to a central location once a week or so to meet with other co-workers face-to-face.

A neighborhood office center might be co-located with a neighborhood learning center, so in many cases, family members would travel to the same location each day. In this way, some of current physical barriers between the location of learning and the location of doing would be lowered. Additionally, these neighborhood centers might help re-establish the importance of geographic neighborhoods, even while global neighborhoods become increasingly important.


The future of higher education

What does this vision of education in 2020 imply for higher education? Most of the examples given above were shaped for pre-college education, but the vision has equally important implications for post-secondary education as well. Many universities and colleges, especially the most prestigious ones, do not have a natural geographic pool for students, and so are at the most risk when new technologies make available distributed learning centers. The word "university" is derived from the Latin word "universitas", which means "all together". So geographic proximity is built into the root of the concept of a university. What might a distributed university look like? At one extreme, it could mean that each learner is located in a different location, without any face-to-face interaction. Many current online programs in higher education work on this model, with each learner interacting with teachers and other students through computers from their own homes or places of work.

In some cases, this kind of isolated individual learning will work well. However, in many cases, people will prefer to gather together with other learners, both for formally-structured learning and for informal interaction. New technologies make possible local community learning centers, where face-to-face learning takes place, mediated by expert adults, but where much of the learning also takes place in interaction with remote other people.

Perhaps more important is the question of what sort of activity would constitute advanced learning in the year 2020. Again, parallel to the description of learning in NLCs, much of the learning could take place within the context of tele-task forces and teleapprenticeships and other interactional frameworks that reconnect learning and doing. The low level skills and fact learning that makes up much of the "introductory course" learning will increasingly be acquired through interaction with computer-based simulations, especially multi-person simulations, with automatic tracking of learning and automatic adjustment of the simulations to maintain an optimal learning environment for each learner.

In fact, we can imagine a fairly seamless integration between multi-person simulated environments and computer-based collaborative interaction dealing with real problems. Again, new technologies lead to new ways to productively integrate learning and doing.

Learning and doing

It is easy to see the benefits for learners of interacting with the world of doing. But are there benefits to the doers of interacting with the world of learning? There are several advantages of this integration.


Network-based problem solving: Learners as mediators between doers

Learners can serve as valuable mediators between remote experts, helping to develop novel and potentially valuable solutions to local problems. New technologies provide new ways for remote people to interact. People that live in distant places are similar to us in many ways, and they are different from us in other ways, at least partly because of the differences between their environment and ours. Often these differences can serve as barriers to productive interaction. However, this diversity can also serve as a strength, if interactional frameworks are designed to benefit from the diversity.

Diversity can serve as a source of creativity. We have explored a framework for network-based problem solving (Waugh, Miyake, Levin, & Cohen, 1988) that draws upon this source. In this interactional framework, a group of distributed learners join a task force to address some shared problem they all face in their own locations. The first step is for each set of learners to describe and share with the others how the target problem is addressed in their own location. This can often involve interaction with local experts who deal with the problem. The next step is for each set of learners to examine the descriptions from the other locations, determining, for each action, whether a similar action is taken in their own location. This can also involve interaction with local experts. For the actions taken remotely that are not taken locally, the learners examine each to see whether the action could be taken locally. Finally, the learners propose to local experts those actions that are not taken locally but that could be taken as possible positive steps toward addressing the problem. If questions arise about the details of the proposed actions, the learners can serve as mediators between local experts and the remote experts, with the local learners passing on requests to the remote learners, who communicate with the remote experts.

In many situations, the learners who are novices in the domain of the challenge being addressed are more expert in the uses of new communication technologies than the local or remote domain experts. So this framework allows for and in fact depends on a diverse set of participants, when the strengths on each are of benefit to the others. It also builds upon diversity as a source of innovation, rather than a barrier.

Harnessing the power of naiveté

In a domain in which things are relatively stable, expertise is a major strength. In a domain in which there is rapid change, expertise is a double edged sword. Experts must frequently reexamine their expertise, to be sure that it is still valid given the changes that has occurred. Often the specific facts and skills have a relative short "half-life." In these domains of rapid change, there is also power in naiveté — novices will suggest ways of thinking about the domain that may not have been useful previously, but may be useful given changes. Because those approaches were not useful previously, they may be automatically blocked out by experts. Again, with frameworks made available by new technologies, there are ways to involve learners in activity in the world outside of education in a way that is useful to the other participants in the activity.

Long shot problems. An expert in a domain often has more challenges facing him/her than they have time or other resources to deal with. So they select the ones that are most likely to lead to productive results, given the resources needed to tackle them. There are many challenges that are not tackled because the probability of solving them is too low to be worth the effort. The successful solution of some of these challenges would be of significant benefit. Let’s suppose that the chance of any given expert successful tackling the challenge is 1%. Let’s suppose that a group of learners, in communication periodically with experts tackles this same challenge. Because they are novices, perhaps the change of making progress is 0.1%. Now suppose that 1000 groups of learners tackling the challenge. Suddenly the probability of making progress is much larger. Even if the learners do not make progress, they benefit from the learning that takes place during their efforts — the primary goal is learning and making progress on the challenge is a side benefit. And even if a particular group of learners does not make progress in tackling a challenge, they may well feel good about being part of the larger set of groups tackling the program, one or more of which were in fact able to make progress.

Grand & Everyday Challenges for Education. While tackling long-shot problems is valuable, it might be discouraging to tackling only such "grand challenges." Learners might incorrect conclude that they do not want to take on real challenges because they are really hard to make progress on. But the world is full of challenges, ranging from the relatively simple to the "grand" level. The "Grand & Everyday Challenges for Education" framework would provide a way for a wide range of people to pose challenges (through the Internet), contributing to a database of challenges. Then any given learner or group of learners could then search that database, to find a challenge that meets their learning goals and resources. Furthermore, "challenge workspaces" could be created, within which multiple groups tackling a particular challenge could post progress, debate alternative approaches, and electronically publish their results. The people posting the challenges would be available to clarify and refine their challenges, based on queries and feedback from the groups tackling the challenge. The motivation for the challenge posers to continue to interact with the learners is not some general motivation to volunteer, but instead a very specific motivation to get help with their own challenge.


The year 2020 is far enough away that we can envision substantial changes in the educational system, yet not so far away that we have no basis of making reasonable predictions. In this paper, we have presented some new interactional framework for learning made possible by new technologies, and explored some of the implications of these new frameworks for the physical settings of learning. There are an increasingly diverse set of media for learning — the challenge for educators is to determine the goals that each medium can best accomplish and the resources required. And this challenge will remain a continuing challenge, as technologies continue to change, impacting both learning and doing, and providing new ways for productively integrating learning and doing.



Cervantes, R. G. (1993). Every message tells a story: A situated evaluation of the instructional use of computer networking. Unpublished doctoral dissertation, University of Illinois, Urbana-Champaign.

Cohen, M., & Riel, M. (1989). The effect of distant audiences on students’ writing. American Educational Research Journal, 26(2), 143-159.

Harris, J. (1995). Organizing and facilitating telecollaborative projects. The Computing Teacher, 22(5), 66-69.

Hutchins, E. (1996). Cognition in the wild. Cambridge, MA: The MIT Press.

Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press.

Levin, J. (1990). Teleapprenticeships in globally distributed electronic networks. Paper presented at the Paper presented at the Annual Meetings of the American Educational Research Association, Boston MA.

Levin, J. (1997). Teaching Teleapprenticeships: Frameworks for building distributed communities of teachers and learners. Paper presented at the Paper presented at the International Symposium "Higher education in the 21st century — Bringing today’s education into the hi-tech world of tomorrow". National Institute of Multimedia Education, Chiba, Japan.

Levin, J. A., & Cervantes, R. (2001 in press). Understanding the lifecycles of network-based learning communities. In K. A. Renninger & W. Shumar (Eds.), Understanding change in the virtual community: Theory, method, and practice. New York: Cambridge University Press.

Levin, J. A., & Waugh, M. L. (1998). Teaching Teleapprenticeships: Frameworks for integrating technology into teacher education. Interactive Learning Environments, 6(1-2), 39-58.

Moore, G. E. (1965). Cramming more components onto integrated circuits. Electronics, 38(8).

Riel, M. (1998). Learning communities through computer networking. In J. Greeno & S. Goldman (Eds.), Thinking practices: Math and science learning. Hillsdale, NJ: Erlbaum.

Waugh, M., Miyake, N., Levin, J. A., & Cohen, M. (1988). Problem solving interactions on electronic networks. Paper presented at the Annual Meetings of the American Educational Research Association, New Orleans.