The world today is different in many ways from how it was 40 years ago. From considering these differences, this article suggests that the most hopeful way forward to support learning in an increasingly complex world is not institutional, nor purely individual, but involves people collaborating peer-to-peer, generating and managing knowledge commons. The MyLK project has made a promising start in addressing the related technology issues — but where do we go next?

The social and economic challenge for learning, education and training

The social, political and economic worlds are all increasing in complexity. At the same time, economic inequality has steadily increased — the forces of economics mean that wealth and political power are being further concentrated in the hands of an elite of the richest and most powerful individuals and corporations. The combined complexity of all this is well beyond the grasp of any single person. To make any sense of everyday living, every individual needs simpler models that they can understand, and one tempting solution is to accept living in an “epistemic bubble” or “echo chamber” — terms that mean that we only see or hear, or we only believe, the news and views of people like us — those who agree with our world view. In an “echo chamber”, any challenges to the accepted world view are dismissed as “fake news”, and the sources of challenge are actively discredited. Because of the proliferation of these “echo chambers”, there is no longer any one authority or information source that is trusted by the large majority of all citizens.

We have seen jobs being replaced first by mechanical automation, and now increasingly by AI systems. Many remaining jobs have less and less security. Not only poorer people, but now also the middle classes have to work harder and longer to maintain their standard of living. The inflated cost of land and housing contributes to this: wealthy people speculate on property value, are able to live a ‘rentier’ existence, extracting money from other people using their property, while the less well off have to pay more and more to rent whatever the rich own, with decreasing chances of actually owning the assets currently needed to secure a comparably stable and fulfilling life.

Why have people not rebelled against this? It’s a hard question to answer. Perhaps the answer lies both with individualistic world-view myths, and also lies hidden close to one of the values inherent in this project work: the ethic of education and personal development. The dominant cultural myths in the USA in particular, but also in other western cultures, still seems to be that anyone can attain any goal, providing they make enough effort, or are smart enough. So if you are poor, it must be your own personal failings. The continuation of the myth is that you can become smart enough through education. How do you get that education, though? Recently, it has been by paying more and more money to those in control of educational establishments, or at least paying for training materials that have IPR restrictions, and are not available freely. Sadly, this often ends up as a lottery. Many people put in their money, but only a few come out top, achieving what they were hoping for. The elite cream off the profit — they are the ones who actually benefit from these myths. As these myths have started to be questioned, it appears that those with power are trying to hang on to that power — perhaps most clearly suggested by the Facebook / Cambridge Analytica scandal, where voters were influenced by personally targeted, highly biased ‘news’ stories. Sowing the seeds of “fear, uncertainty and doubt” has long been a tactic to keep people subjugated.

The implications of this for learning and careers are profound, and even alarming. A few professions still have a relatively stable career structure, like law and medicine, and perhaps the emergency, security and civil services. But other professions, even including many sectors of teaching, seem to be increasingly overworked, underpaid and insecure. It becomes harder and harder to give people any meaningful careers advice, as traditional careers begin to fragment. It is less and less easy to know what courses or qualifications that young, displaced, or redundant people should take, in order to achieve the kind of worthwhile employment that will allow their needs to be satisfied and their aspirations and even modest ambitions achieved.

Alongside this difficulty and challenge, there are also positive developments. Resources for learning that are easily accessible on-line have been multiplying quickly. Wikipedia is probably the best known single site, and it has assembled an unrivalled body of second-hand, encyclopedic knowledge, with reasonable quality standards, maintained by ordinary people as a knowledge commons. As well as the purely written word of Wikipedia, video services like YouTube now include vast amounts of first-hand training materials that cost nothing more than accessing the internet. Free structured courses available to everyone online, often called MOOCs, may combine the written word with other media, and sometimes they add a social element, which enables learners to help with each others’ learning. And while some publishers are fighting rearguard campaigns, more and more academic publications are now openly available, as many have argued they should rightly be.

Because of all of this change, fragmentation, inequality and complexity in our society, any body of knowledge — including any labour market ‘intelligence’ as it has been called — has limited validity. Any centrally produced guides to learning resources or careers are likely to be incomplete even when they are published, and their reliability quickly drops. For institutions, and individuals, engaging in career-oriented learning, this may seem like an impossible situation. Is there any good way forward? Surely most people share the opinion that learning is very worthwhile in many ways. We can indeed improve ourselves. We can learn knowledge and skills relevant to particular occupations, and we can also learn the ‘soft’ skills that, among other things, enable us to share and collaborate. This means that we do not need to buy in to the myth that ‘education’ comes at a high price, nor that the most worthwhile resources necessarily cost a lot of money.

Being aware of these myths, and being aware of the ways that we can too easily be manipulated, is among the most important knowledge to have today, supporting life skills that are now becoming essential. While political discussion too often involves ideological rhetoric, there are other, calmer places to reflect on what is going on in the world, in our economy, in our society, in our culture. One helpful author, introduced to me by the e-portfolio expert Darren Cambridge, is Robert Kegan. Building on the work of William G Perry, Kegan has given us a fascinating way of understanding how what he calls “orders of consciousness” develop in people, and this supports a reflective perspective on cultural change. His insights need to be taken into account for any realistic way forward, so we need to learn about them as well, as and when we are ready. Less notably, my own book on electronic portfolios is now freely available on my web site, and the third and final part of the book is also devoted to understanding the wider perspectives on where we are and how we can move forward in learning and development.

Beyond the institution and the individual

If the individual himself or herself is overwhelmed by complexity and choice, and traditional institutions lag further behind what is relevant today, there still remains a third way, which is already in play in small ways in online life at present. We can help each other, as equals, or peers. ‘Peer’ in English implies equality of status, similar to ‘pair’ in French. A young person who works “au pair” (English uses the French term) is considered to have a status similar to other family members, unlike a servant. “Peer-to-peer” (sometimes translated as “pair à pair” in French) describes communications between people (or indeed machines) that are counted as equal in status. In the vision of a helpful peer-to-peer world, we share (rather than restrictively ‘own’) common, non-personal information, and that information is not limited, and its flow is not constrained, by structures or channels governed centrally.

How does interacting as peers help in practice? In our own small ways, we are all experts in our own particular life situations, and the sum of that expertise extends far beyond what can be gathered together by one person or in one book. I have my own opinions, from my experience, about what knowledge and skills have helped me in my life situation, and the resources that I have found useful, and the closer your life situation is to mine, the more likely those resources are to be helpful to you, too. For example, if I am exploring a vegan lifestyle, resources that show me how to bake delicious cakes without eggs might be very useful, in a way that would be of no use to someone who doesn’t bake, or someone who is happy enough using eggs. On the other hand, that information may be elementary to a chef in a vegetarian or vegan cafe, who might have been baking vegan cakes for years. They might be more interested in quite different things, for instance, where to source ethical bulk ingredients that fit with the values of their clientele.

So, everyone has potentially valuable recommendations about learning resources, relevant to situations that they know well; and the less closely related their experience is to my real life situation, the less valuable their recommendations are likely to be for me. So how do we find recommendations for learning resources that are relevant to us? Definitely not by trying to look through everyone’s recommendations — we simply don’t have the time for that, let along patience or ability to process all that information.

To make P2P resource recommendations work, we need a way of prioritising the recommendations that are most relevant to us. This is a little like what happens on many websites most of us use frequently. When you use Amazon, you are offered suggestions based on what other people bought, who either bought or looked at similar things to you. Who knows how deep Amazon profiling may go, in their attempts to tailor suggestions ever more closely? Recommendations appear on other services as well. How are these recommendations made? Are some of the ‘recommendations’ a cover for paid promotions? Perhaps one way of sensing this is to look at the ‘resources’ recommended. Resources, as we tend to think about them on the Internet, include Wikipedia pages; YouTube videos or playlists; open courses like MOOCs, that may contain various particular specific resources. We could, perhaps in the future, extend this to face-to-face traditional courses, to clubs, to professional associations, etc. If the resources are genuinely neutral, with no promotional bias, that is good. If they advertise something, they are more likely to have been recommended as promotion, rather than as what is most helpful.

Peer-to-peer recommendations for learning resources

How, then, do we address this challenge in the case of vocational learning resources, and in particular, using the MyLK system? I’ve written about this before — under the title of How can MyLK advise users about helpful learning resources and courses? — and here is a recapitulation with some different examples.

One way starts by keeping track of what various users’ interests are. A user’s interests in MyLK are simply the topics that he or she chooses when ‘tagging’ a particular resource that they have seen and taken an interest in, enough to record it, and indeed to rate it. But people may rate a resource as relevant to various interests, and the recommendations one person wants to see are from other people who share the same interest in that resource, not those who see it from a different point of view. For example, imagine an intentional, sharing eco-community that grows food with ‘permaculture’ in mind. If there were a video about that community, it may be a useful resource to those interested in permaculture, or climate change, or the new economy, or non-traditional lifestyles, or perhaps several other interests. Simply giving resource recommendations based on the fact that other people gave similar ratings to resources would miss the reality of those different orientations, and would likely therefore to be less helpful than recommending the resource based on the ratings of people interested in the same issues.

Are skills the same as interests? Not quite. A skill is also an interest if the person with that skill wants to develop it. But people may have many occupational skills that they use in their work, but feel no need to develop further. Equally, people have interests which they would prefer to keep separate from their work, or that are not associated with any employment that they might seek in the future.

Another of my articles in this series — MyLK needs career maps: how could these work with and for users? — addressed the challenge of peer-to-peer resource recommendations through career maps and career paths. Using career maps is another way of ensuring that the people giving and taking the recommendation are in similar life situations. If two people have similar occupations, at similar levels, then resources rated as helpful by one are likely to be helpful to the other, certainly more likely than resources rated helpful by random people in unrelated occupations. In that previous article, what I didn’t do was to mention the directional nature of recommendation value. By this, I mean that if someone else has in the past been roughly where I am now, their recommendations of resources relevant to that stage are likely to be much more useful than recommendations from someone following on behind me. So it is not simply a question of how close your occupation is to mine, but also whether you are more advanced or less advanced. In fact, the closer you are, the more is the difference in value between recommendations of people just ahead or just behind.

Peer-to-peer organisation of the relevant knowledge commons

Imagine, now, that we have some kind of P2P recommendation system, built after considering the matters I have outlined above. What I did not address in previous articles was how this whole system of recommendations could be based on commons-organised knowledge, and this seems to me one central issue in the long-term viability of the kind of service which MyLK is prototyping.

I’ve mentioned the word ‘commons’ earlier as well. It may need some explanation. For those who are unfamiliar with the term, I recommend two introductory resources: The Commons, Short and Sweet; and, longer and more detailed, the Commons Transition Primer. A commons is, essentially, some set of resources that are managed peer-to-peer, that is, managed and controlled by a peer group of people who could naturally be called ‘commoners’. It’s safer not to say that the resources are ‘owned’, as ownership is quite a tricky concept in this context. Best to start with an example.

Wikipedia is a classic knowledge commons. It is free to use for everyone, and because it is knowledge, rather than a physical resource, one person’s use in no way takes away from anyone else’s use. What makes Wikipedia work is the way it is managed. It is all managed by volunteers who have Internet access. The way in which changes are recorded and notified means that if there is ‘vandalism’ on a page, others who look after the page can very quickly step in and revert it. Automated services can detect clear cases of vandalism, but it still takes humans to correct subtle errors or well-intentioned mistakes. Just playing a part in keeping Wikipedia articles correct gives a definite sense of inclusion and authorship. Apart from Wikipedia, there are many other wikis or wiki-like resources that are effectively knowledge commons.

So, should the MyLK service really be a commons? Turning that question on its head, we can ask, could it be managed effectively without being a commons? Certainly, there are good examples in the world of projects not run purely as commons, such as the Linux operating system kernel, for which Linus Torvalds is known as the “benevolent dictator for life”. Software code is relatively simple to evaluate. Programs work, or they don’t; they take more or less time to execute; they look easier or harder to maintain. But what about learning? Learning resources are not so easily measured or standardised, so organising them centrally is increasingly hard, and individuals simply do not have the capacity. So the ideal solution is to distributed the organisation, and that points again towards a commons approach.

How can MyLK’s requirements be met peer-to-peer?

Looking at the practical side of these matters, there are two related questions that need answering here. First, what information is needed to allow MyLK or a similarly powerful resource recommendation service to operate effectively; and second, what underlying information is needed to support it.

MyLK’s direct information requirements

To provide the most helpful learning resource recommendations, there has to be a way of assessing which other users will give the most useful recommendations to the user looking for guidance. There are four potential sources of information that can help, in ways that have been hinted at above and elsewhere:

  1. the knowledge, skills and competences of users;
  2. the interests of users;
  3. users’ current job or role;
  4. users’ career history.

All of these need to be requested of users, and the issue of how to motivate users to enter these is challenging in itself, but is something that will take more study and trials to answer, and probably does not change much depending on whether the whole system is managed peer-to-peer or not.

The supporting structural information

If MyLK simply had a list of users’ interests and occupations, written using their own terms, the recommendations might sometimes be odd. For instance, if someone described themselves as interested in “design”, or with the occupation of “designer”, who knows what learning resources that might involve? In order to achieve even reasonable reliability, users need to be able to choose from a common vocabulary of terms, first, for interests, skills, knowledge, and second, for occupations, which are the steps on any career path.

What ESCO provides

ESCO, European Skills/Competences, qualifications and Occupations, is a classification that provides a vocabulary both for occupations and for skills and knowledge. The classification for occupations can be used, and is used by MyLK, to provide the terms chosen by users to register both their current occupation, and their career history, should they decide to share that. The classification of skills (etc.) provides, for MyLK, a good start on the users’ interests, as well as their skills and competences.

Each occupation classified by ESCO is associated with a set of skills and knowledge which are considered to be essential or desirable for that occupation. Putting this information together has been organised centrally by the ESCO staff, using knowledge from many domain experts. And MyLK currently uses ESCO to give the vocabulary for the topics of interest to the user. But while ESCO is an excellent start, it does not deal effectively with hobbies and recreations, and more general topics of interest. And as traditional careers break up, occupations and interests will overlap more, and an expert classification from an occupational perspective will become of less use. So in the case of ESCO, while we do have a central authority model at present, in the future that looks less plausible.

The other resources that has been widely discussed in the course of the MyLK project are career maps, as in my previously mentioned article. Not wanting to repeat myself at length here, I will just say that there are many very rough career maps, put together by people with vocational experience, but those career maps do not start to cover the range of occupations given by ESCO, let alone the envisaged extensions to that. I discussed there how career maps could greatly improve the accuracy of recommendations, as well as being the basis for P2P requesting and offering informal careers advice.

How could these resources be turned into commons?

One might well say that to be a fully P2P service, the information and other internal resources needed by that service really need to be managed as commons. There are no easy ways to achieve this, and the discussion here raises more questions than it gives answers. I do believe, though, that the way forward, for the medium- to long-term future, will be through addressing these issues.

Occupation classification and vocabulary

There is a long history of standardisation for descriptions of occupations, partly at least to enable labour market intelligence gathering. The Wikipedia articles on the International Standard Classification of Occupations (ISCO), sponsored by the ILO, and the Standard Occupational Classification System of the USA and other countries, are good resources to start learning about these. ESCO is the most recent European offering, and it is based on an elaboration of the framework from ISCO-08.

These existing occupational classifications in any case give a useful start to thinking about occupational skills. If they have free open licences, they can potentially form the basis of an open knowledge commons vocabulary. But what happens when new occupations emerge? And what about regional or local variations, where a similar name may be given to occupations where the associated knowledge and skills are different? The vision I am putting forward here is to have a classification that is both responsive to up-to-date input from people who actually have jobs in those areas, and manages to avoid duplication or overlap, which could be caused by people having different names for very similar jobs.

Though it may not be easy to imagine in detail, one hope might be in a kind of peer-managed knowledge commons, possibly taking some inspiration from Wikipedia. As Wikipedia has discussion pages where contributors can discuss matters before making changes to the pages themselves, so, proper discussion, along with a small community of people who actually have that occupation might ensure that the information stays up to date. Again, as with Wikipedia, new pages can be started when there is a need. Individuals would only need to look after one, or very few, pages, in collaboration perhaps with a few others, so the task would be feasible. The knowledge and skills needed for an occupation would be one thing that was kept constantly updated, and it would give a good basis for deciding when a new occupation was needed, which would be when a new occupation required a different set of knowledge or skills. Wikipedia has very many redirected page titles, where non-preferred names for a topic are redirected to preferred names, and the same mechanism could be used for jobs. In addition, the same could apply if an occupation became outdated — references to that job could be redirected to the closest up-to-date occupation.

A completely different approach would be to eliminate occupation titles as primary objects, and shift the primary basis for classifying occupations to being on the knowledge, skill and competence needed. Established occupation names could still be given for better known occupations, but new occupations need not have an explicit name until they became more common, and the name was needed for recruitment purposes. Another different possibility would be for people to define their occupation as a variant on a known occupation. To give a simple example, ESCO’s occupational term “software developer” could be further specified according to the programming languages used.

Career maps

I mentioned in the earlier article the idea of ‘crowdsourcing’ career maps — this could, given the user’s permission, use information from a recruitment site, from a service such as LinkedIn, or simply from users inputting their past occupations manually. Inputting jobs in a way that can be compared with others, needs a common vocabulary of occupations, as discussed above. But given that common vocabulary, there is still the matter of building up a repository, or database, of career paths, where the related career paths together constitute a career map. This seems more like a very natural commons, in that everyone may contribute, and everyone is able to use it. The people who do enter their own career path may get thanks from those who seek their advice, or perhaps those who use their recommendations, based on their career path.

There are some finer details of creating the overall career maps, which might need to be governed by peer-based structures, but perhaps these matters could also be dealt with by the same organisation as took on the harder role of maintaining the classification extended from ESCO.

Knowledge, skills, etc.

This is a mixed bag, in terms of creating a commons-based working system, so each part is addressed separately.

Knowledge

The knowledge topics could, very simply, be taken from Wikipedia itself. The main challenge here is that the English Wikipedia is by far the most extensive one, and many articles do not have versions in other languages. Where there are different language pages on similar topics, the pages do not always have exactly the same scope. Nevertheless, it would be feasible to allow people to refer to any language version of Wikipedia, and compare by taking the closest English equivalents.

Skills

The EQF documentation states that “In the context of EQF, skills are described as cognitive (involving the use of logical, intuitive and creative thinking) and practical (involving manual dexterity and the use of methods, materials, tools and instruments).” As acknowledged by the structure of ESCO, skills and occupations are mutually interlinked. Occupations need skills (as well as knowledge); relevant skills can be seen as the (context-free) abilities needed to carry out certain operations necessary to be able to perform a role, or to be employed in an occupation effectively. ESCO makes no specific distinction between skills and competences, though it does make the easier distinction between knowledge and skills.

To manage a skill vocabulary as a commons is a huge challenge. Any changes or additions to skill definitions would need to be approved by the people maintaining the definitions of occupations that use those skills. It would be even harder than maintaining a knowledge commons about occupations, but still, there is no reason to suppose that it would be impossible. Indeed, probably the only effective way to maintain such a classification or vocabulary would be to constitute it as a knowledge commons.

So-called ‘soft’ skills may be a particular challenge, as they are often assumed rather than explicitly specified as requirements for occupations. Even when they are specified, they are hard to define. Yet they are particularly important in the context of collaboration and co-operation, working in a peer-to-peer context. It will be especially important to work towards agreeing terms and vocabularies, and perhaps frameworks, of soft skills.

Other dimensions of ability

In 2008, the EQF had a third dimension called “Competence”. Since then, the EQF have changed the English name of this dimension to “responsibility and autonomy”. For the EQF, “responsibility and autonomy is described as the ability of the learner to apply knowledge and skills autonomously and with responsibility”. ESCO has no separate categories for this. It means, in effect, about being able to use the context-independent knowledge or skills in the particular contexts of work or study. This is something that can only be effective assessed in the context of an actual job, and therefore is something to do with experience. One way of dealing with this would simply be to accept the 8 levels of the EQF as applicable to relevant knowledge and skills in the context of an actual occupation.

One other factor that almost always has great significance in recruitment is experience. While the number of years experience of a job has no absolute relationship to the knowledge, skill, responsibility and autonomy, there is likely to be some connection between years in a job and the level of responsibility and autonomy reached. And while many occupations can be roughly matched to the 8 EQF levels, many established careers have their own internal system of progression through levels, often reflected either in job titles or in professional qualifications. The armed forces have ranks, and other public occupations may have similar official status titles.

Interests

We recognise that not all things we want to learn about are related to paid jobs or officially classified occupations. In MyLK, based on the ESCO division between occupations and skills, an interest is taken to be something that someone may want to learn, and the vocabulary for that is therefore the same as the combined knowledge/skills/competence vocabulary.

To move beyond ESCO into commons-based management of interests, it is perhaps enough to follow the leads above, putting together knowledge pages from Wikipedia, and the abilities as described above.


There is much more that I could say here, but this is enough to point to a way forward towards a world of more genuinely fair opportunity and peer support. This is, in my mind, one of the vital steps towards an economy, and a society, that is more equal, just, sustainable, collaborative, and supportive; where everyone is far more likely to be able to meet their needs creatively, without the insecurity and stress that has become more endemic in developed countries over the last 40 years.

Summing up

To conclude, I suggest that it is not too hard to imagine an effective commons-based system to support learning and development, in a peer-to-peer world; and that because of the increasing change, complexity and fragmentation of the world we are living in, this is in my view the only effective solution in the long term. MyLK is a highly promising first step. For MyLK and in the future, this is a start to listing what should happen:

  1. to support effective learning resource recommendations, have a process for knowing, about any two people,
    • whether they share similar learning interests
    • how close they are in their life situations about which they want to learn
    • who is further advanced along their path, and thus has the relevant experience to share
  2. to support this, have classifications or vocabularies of occupations, abilities and interests that are more extensive and fluid than ESCO
  3. record individual career paths, and construct career maps from these real data
  4. work towards managing these in the only effective long-term way, as knowledge commons
  5. do all this, in principle, through peer-to-peer organisation
  6. have good enough coordination between peer participants to enable this organisation
  7. develop earlier, and more strongly, the required ‘soft’ skills, to facilitate this coordination
  8. follow this promising path towards building a new, collaborative, solidarity economy and society.