This is the simple text version of a talk I gave at ETECH today
(download slides – 8Mb PDF)Amidst all the technofetishism of the ETECH alpha geeks this year, there has been a very encouraging theme of humanising technology and understanding the need for attenuation as a way of protecting the scarce resource of our attention from the onslaught of information overload
This theme was picked up by several speakers, such as the eloquent and thoughtful Linda Stone, who predicted a move away from Always-on behaviour to some form of protected time and space, plus Tim O’Reilly and Felipe Cabrera who both used the metaphor of the ‘Mechanical Turk’ to talk about the importance of augmented human intelligence rather than artificial machine intelligence. Eric Bonabeau reminded us of the way people make decisions using pattern matching skills, in an entertainingly humanist follow-up to his previous ETECH talks on computer-based evolutionary intelligence
Geeks and infovores have their own reasons for needing attenuation and the management of attention as a resource, but if we are in the business of saving souls, then it is the matrix of the modern enterprise where the need is greatest. This is where people suffer the worst form of information overload, because they have so little control over the flow of inputs but are required to process them as part of their job. This is where primitive enterprise systems regulate life, and people cannot manage their context, attention or even the language they use to express themselves
When Dinosaurs roamed the enterprise
The enterprise is a lost world full of large, lumbering dinosaurs who evolved during the early dot.com boom of the 1990’s and who survive by subjugating humans within their CRM/CMS/DMS/LMS/ERP systems, and also because they are too big and expensive to kill. Even though modern companies see themselves as organic, network-based entities, the systems that people are supposed to use for communication and collaboration are based on an IT-driven command and control model that is grounded in 1950’s management thinking
Meanwhile, in the outside world, where people don’t have to use crippled centrally managed IE6 browsers, the dot.com bust swept away a generation of software predators and ordinary people began making tools for themselves. When the 1990’s bubble burst, developers went back to basics to create tools like blogger, movable type and the early wiki platforms. They found that if these tools all connected together, they could communicate and share more easily, and they began building a new world based on what has been dubbed an ‘architecture of participation.’ Now you can download free tools that do 80% of what companies paid millions for in the 90’s, and you can make them do whatever you want.
For the past few years, transport mechanisms like RSS and the connected conversations of weblogs have been creating a rudimentary knowledge ecology that has achieved what so-called Knowledge Management failed to do within the enterprise. The spread of open source platforms provides a viable alternative to expensive enterprise systems, and these social tools are infinitely more adaptable. Yet somehow, inside the enterprise, managers continue to buy the arguments about process, workflow, security and control that software vendors use to keep them in the stone age. Some organisations are spending 80% or more of knowledge sharing budgets simply on email and document storage, and yet they know this is just a finger in the dyke. Vendors sell features, not value, and consultants sell fear
The wrong sort of information overload
Old-style enterprise IT has created a vast problem of information overload, where the wrong sort of information is forced on people in ways that inhibit initiative and decision making. Management by e-mail turns people into Pavlov’s dogs, and is unsustainable from both the individual and the enterprise point of view. Email began as a cool way for academics to share ideas, but it does not scale to the level of usage we see today. People are expected to receive, process and dispose of a sequential bombardment of hundreds of emails and assigned tasks, but the majority of internal emails are little more than corporate spam
This approach forces people to focus attention on their inbox in a psychosclerotic fashion at the expense of using their peripheral vision, sense making skills and intuition. It is not only a gross waste of time and money, it is actually damaging to an individual’s decision making ability, as well as being a factor in workplace stress. Information overload without control of information inputs has been shown to be worse for concentration than smoking dope and can be extremely frustrating
Many companies, even today, remain obsessed with codification and storage rather than socialisation. Some have millions of documents in primary EDMS storage, but they have literally no idea which ones are important. Thanks to centralised storage and centralised access, nobody can find what they need and eveybody fears they are missing something important. But contrary to the KM orthodoxy, you cannot actually codify knowledge – it downgrades to information when stored in databases, and crumbles to dust
The rise of social software
The rise of social software tools, services and ideas provides an historic opportunity to liberate people from the stultifying grasp of enterprise systems. Cheaper, better and faster are all possible, and that is an immense business opportunity for companies in this space
Blogging, wikis, social tagging, aggregation and syndication, IM, shared presence and lightweight group forming tools are increasingly being used within the enterprise, whether officially sanctioned or not, because they get the job done. But it is not just about different tools, it is also about a new relationship with information and people. It is about managing feeds and information inputs, not individual content items as in the old CMS model. It is about recognising and supporting the different interaction modes (synchronous, asynchronous) and models (read, edit, publish, comment, bookmark, etc.) that are required for different work contexts. It is about the importance of social markup of existing data and the creation of new layers of personal and group metadata
Social tools are reaching critical mass, and more and more IT/IS departments are realising they need to let go, stand aside and let people get on with their jobs rather than act as gatekeepers
How do people make decisions and come up with new ideas?
The main purpose of Knowledge Management within organisations is to support individual and collective decision making, but how do we actually make decisions? As Eric Bonabeau reminded us yesterday, humans are brilliant pattern matching engines. We don’t do a full analysis of all available data and then make the best possible decision. Instead, we sense a wide range of available inputs to determine the best first fit: when faced with a lion coming towards us at speed, we don’t traverse a hierarchical database of all yellow things with hair and try to find an exact match with what our eyes are seeing. We shout ‘LION’ and run like hell. As Dave Snowden says: “the only humans who analyse all the data and then make a rational choice are autistic, but economists insist this is the way we all work.
The pattern matching systems of our brains actually have lots of unused bandwidth that is not used by traditional sequential information processing. The brain takes in more than we know, but is very good at filtering out wha
t we don’t need and simplifying what we see using archetypal patterns and representations. In ‘Blink’, Malcolm Gladwell cites experiments where gamblers’ physical reactions actually seem to presage conscious awareness, which suggests we see more than we think we do. We are very good at sucking meaning out of thin-sliced data. The problem with existing enterprise software is that there is simply not enough diversity of inputs to stimulate intuitive decision making, which is why the most adept individuals in organisations rely on more basic, offline techniques to know what is going on
The information overload people face within organisations today is the worst kind: the wrong sort of information, with no individual control over inputs
The answer to information overload is …. more information!
Paradoxically, the answer to information overload is more information, but taken less seriously. We need a new relationship with information inside companies. Instead of trying to gather all data and then analyse it in real-time, we need to help people build a better radar, go with the flow and trust in their own ability to sense what matters and make decisions. We need more peripheral and contextual information and less sequential memo and email processing. We need to create multiple paths to things we need to improve findability, as Peter Morville says, rather than focus just on storage and retrieval. Classification within organisations is calcification – it kills context. Social navigation and filtering is cheaper and more effective
We need a new relationship between people in companies. We are hard-wired for socialisation, but enterprise tools are based on 1950’s management thinking that treats us like machines in a production line. The focus of information sharing in support of decision making should be on weaving the social web and allowing people to negotiate their own language and meaning by adding social markup to existing corporate data
This is not rocket science, and it is easy to achieve, but doing so requires a culture change in both IT and management thinking. Ambient social knowledge sharing requires some basic underpinning activities:
- We need to develop a social fabric for knowledge sharing: a critical mass of weak ties to support the exchange of knowledge within context. This is best achieved by stimulating connected conversations, in the way that blogging does, to encourage creativity rather than just read-only modes of interaction, and by developing social networks to support ownership of those conversations within specific domains of knowledge. This requires true personalisation where everybody has their own blog or channel and individual voices give personality to the system.
- We need to break information down into small pieces by giving everything a URI and a feed so that people can choose their information inputs and access them using RSS/ATOM syndication and subscription. Ideally, they should be able to construct feeds from search queries against existing corporate data. These feeds need not only be from human sources: corporate ‘blogjects’ may also feature in the information ecology, such as financial instruments, machines or other systems that broadcast their status.
- Social tagging, collaborative filtering and the “web of links” can bring existing information to life by adding contextual markup to stored data, documents and content. That way, social bookmarking, tagging and feed selection become primary mechanisms for finding things that are useful rather than the blunt instrument of enterprise search or the deadzone of a corporate taxonomy. It also means people can re-mix corporate data to create temporary collections that help them with a task or a project.
- Create simple group spaces that are local or specific to teams or disciplines, so that people can share what they are doing within a trusted context. These should include some combination of group blog for sense making, wiki for co-production and action, a tag space and shared links to useful data and external sources. They might be open or entirely closed to give people a safe context for collaboration. Perhaps also consider some form of real-time event support where people can ‘jam’ and discuss new ideas of ways of working.
- Consider some basic ambient spatial knowledge sharing support through shared screens and displays that teams can remote control, and possibly basic ambient devices. Architecture and office design are also worth considering, as the classic open plan layout gives management far too much latitude for interruption and individuals and groups too little room for private work.
- Create a culture in which sharing and openness are aspirational values to be rewarded, and in which an ecological approach to knowledge – rather than storage and accumulation – is encouraged.
- Finally, and most importantly, free people from e-mail enslavement. Allow them to choose their own information inputs and sources to support their work, but let go of the need to process and react to everything that comes their way in favour of immersion in a flow of information and social contact. Within a framework of agreed objectives and targets, let people find their own way to achieve them. One of the key lessons of social knowledge sharing using blogs and RSS is that other people act as your filter and information you need tends to find its way to you, somehow.
How do we get there from here?
Conventional enterprise systems are good at ‘heavy lifting’ and storage, but bad at the user interface and the ‘last mile’. It makes sense to keep the back end and give people smarter, social tools to discover, store, share and create what they need to do their job. Bring out feeds from legacy systems and create a mediating layer of services to manage them. Then wrap it all up in a simple, social interface that creates network effects from individual usage
We have been applying some of these lessons and pointers within organisations with a surprising degree of success. There are still live minds within the enterprise, despite years of software abuse, and if you can reach them with a more human alternative then they tend to respond by taking these tools and techniques and making them work
See this collection of links on Del.icio.us for some of the references I used in putting this together.