Clay Shirky’s statement — “It’s not information overload, it’s filter failure” — is an oft-quoted line when discussing online sensemaking. I was discussing filters last week during an interview on personal knowledge mastery which will be used to inform a program we are developing for a client organization, a large global corporation. The interview reminded me that it’s time to refine my work on knowledge filters because times have changed since I first wrote up the work of Tim Kastelle and his five forms of filtering in 2011. I slightly revised these knowledge filters in 2018 and recently discussed the importance of trusted filters.
One current challenge with machine filters (heuristic & algorithmic) is that in most cases the end-user does not know what logic or code is driving them. One machine filter that many of us use is Google Translate, which you could say is either the result of the wisdom of crowds, or the blind leading the blind — you choose.
“The main issue is the mechanism used by Google Translate itself. It does not actually translate anything, but it scours the web for similar or identical translations performed in the past, constantly learning and building upon what it has learned. This might sound great, but this also means that any time you plug your word, phrase or paragraph, or upload a document into Google Translate, it then becomes public domain.” —Robert Gebhardt
More and more I do not trust machine filters unless I know who created them and how. The other issue with public algorithmic filters is that they are open to being gamed. The search engine optimization field is basically designed to game Google search. The Amazon recommendation engine is constantly bombarded with fake reviews.
If we are to rely less on machines and more on fellow humans we will have to put more effort into our knowledge filtering. Inside large companies, human filters can be identified, promoted, and supported. The identification of knowledgeable people should be an important management function. The organization can also help people to codify some of their knowledge, especially through stories. I have noted before that stories connect knowledge. Stories can provide the contextual glue, holding information together in some semblance of order for our brains to process into knowledge. Stories also help to develop empathy and in the longer term, trust. Knowledge in trusted networks flows faster.
A good company knowledge management structure can address a lot of internal filtering needs. Developing external filters is more difficult. First, people need time to find and select knowledge filters. Again, the organization can support this internally, giving people time to seek external knowledge, and provide resources to curate it for internal needs. But in the final analysis, it is up to each knowledge worker to develop their own knowledge filters. Fifteen years ago Lilia Efimova identified this new work-learning contract. Today the need to take control of our own work and learning is only growing.
To a great extent PKM [personal knowledge management] is about shifting responsibility for learning and knowledge sharing from a company to individuals and this is the greatest challenge for both sides. Companies should recognise that their employees are not ‘human resources’, but investors who bring their expertise into a company. As any investors they want to participate in decision-making and can easily withdraw if their ‘return on investment’ is not compelling. Creativity, learning or desire to help others cannot be controlled, so knowledge workers need to be intrinsically motivated to deliver quality results. In this case ‘command and control’ management methods are not likely to work.
Taking responsibility for own work and learning is a challenge for knowledge workers as well. Taking these responsibilities requires attitude shift and initiative, as well as developing personal KM knowledge and skills. In a sense personal KM is very entrepreneurial, there are more rewards and more risks in taking responsibility for developing own expertise. —Lilia Efimova
External knowledge filters are available in our social networks and communities. These are not the same — a network is not a community. Social networks, especially online platforms, are excellent places to connect with people we do not know. We can follow experts in almost any professional field. We don’t need to have a personal relationship in order to learn from others. These social networks are also places where we can learn from others who are not like us. With an open mind, we can become more empathetic. In closer-knit communities with rules to govern our behaviour, we can have safer and deeper conversations. Communities can be seen as ‘knowledge-commons’.
Commoners must be willing to monitor how their resources are used (or abused) and must devise a system of sanctions to punish anyone who violates the rules, preferably through a gradation of increasingly serious sanctions. When disputes arise, commoners must have easy access to conflict-resolution mechanisms. —David Bollier in Evonomics
Effective sensemaking in our professions and our lives requires access to diverse and deep sources of knowledge. The key is to balance and complement our social networks, our communities, and if possible, with whom we work.