|
p.505 knowledge management needs to develop processes and tools - associated with scenarios and real options - that will allow it to operate credibly in possible and plausible worlds, so as to extract value from them.
p.505 we are concerned with how knowledge-bearing agents... develop and deploy their knowledge to create value for themselves: or in other words how they manage their knowledge. Much of what passes for knowledge management today has its origins in practice, and qua practice, knowledge management has not much bothered with epistemological issues - that is, those related to the nature of knowledge and its justification.
p.506 Knowledge-bearing agents (hereinafter 'agents') use their knowledge in these worlds to take actions that secure future resources. The challenge for knowledge agents is to deploy their scarce resources in each of these knowledge worlds in such a way as to secure sustaining resources that exceed the resources committed.
p.506 we conceive of knowledge as comprising a set of beliefs which informs decisions by agents to take actions that consume the agent's (scarce) resources. With this conception of knowledge, each of the above cases involves deploying knowledge: that is, it involves taking action that consumes the resources of agents on the basis of sets of beliefs that they hold individually or collectively.
p.507 In a world characterized by complexity, variety and uncertainty, the challenge is to take actions that are appropriate to the nature of the belief that is held, and to match the latter to the level of uncertainty inherent in a given situation. Such a matching illustrates Ross Ashby's Law of Requisite Variety. This states that the variety of stimuli impinging upon a system must be countered by the variety of responses that the system can muster. As Ashby put it, only variety can destroy variety... A system that is incapable of filtering out noise from the set of stimuli that it responds to is condemned to dissipating its scarce resources unproductively as it overreacts by attempting to respond to every opportunity or threat, real or imagined... in intelligent systems - that is, systems capable of forging meaningful representations of the states of the world that they respond to - what constitutes noise and what constitutes information for them will itself be a function of their models of the world... The law of requisite variety is a call to action, and knowledge is an essential ingredient of effective action. But what is likely to constitute sufficient knowledge to take action? ...intelligent action is action that can handle variety adaptively within a given time frame - requisite variety has a time dimension. How, then, to zero in on the relevant models and appropriate beliefs in a timely fashion?
p.507 how might we economically manage our scarce knowledge resources under conditions of uncertainty? ...to be able to answer such questions, knowledge management will need to develop its epistemology - for our purpose, the question of what constitutes valid knowledge for who and under what circumstances. Epistemology provides the basis for action and thus serves as a foundation for the institutionalisation of practice... There is a need to identify the different circumstances... in which knowledge is considered valid and actionable.
p.508 Epistemology... is the study of the nature of knowledge and justification.... when it comes to resource allocation decisions... we are interested in knowledge that informs agent action.
p.508-509 The naturalistic perspective on knowledge... concerns itself with possibilities for action in which knowledge, as well as being representational, can be tacit or embodied in skills and know-how - what the Greeks termed techne. From this perspective... one... settles for justified belief as a ground for action... an agent may settle for true belief alone... Finally, an agent may... act on a 'hunch' where neither truth or justification are involved... the challenge for agents is to deploy scarce resources in each of the knowledge worlds in such a way that the agent secures sustaining resources that exceed the resources committed... Clearly, the most demanding constraints on our beliefs are that they should be both justified and true. As we go about our daily business, few of our beliefs are ever actually called upon to pass this demanding test... We can greatly expand what will count as valid knowledge, therefore, by dropping the truth condition and settling for a justification condition on its own. That is the strategy adopted by science. It is constrained by the requirements of justification... Alternatively, we can drop the justification condition, acting solely on the basis of true belief - a strategy that gives more scope to personal intuition, conviction and experience... we may be unable to get others to accept it... Finally, we can have a large number of beliefs that are unconstrained by either a truth or a justification requirement but that we are still willing to act upon, albeit in a more cautious and tentative fashion
p.511-512 Possible Worlds... Resource commitments to action... must be highly tentative and can only be based on the option value of a selected set of possibilities... Scenario thinking in organizations exemplifies the kind of thinking required in possible worlds. Over time, some subsets of possible scenarios may acquire plausibility on account of their coherence... knowledge in this space comprises beliefs in possibilities - hints and hunches about what linkages among beliefs might be... there is no history or evidence by which to corroborate and justify these hunches. The challenge for the agent is to recognize the potential for forging value from some subset of these possibilities by taking out options to exploit the value of whatever hunches can be navigated - either through plausible worlds or probable worlds (where profit potentials are created) - into actual worlds, where value is captured.
p.512 Plausible Worlds... The challenge for the agent is to recognize the potential for creating value from some subset of plausibilities by making 'speculative' investments in these opportunities unseen by others.
p.512 Probable Worlds... Through empirical testing and replication the outcomes can be replicated and a probability distribution assigned, thus creating socially justifiable probabilities.
p.512 Actual Worlds; In actual worlds knowledge comprises true, justified belief... the agent will certainly get what it pays for.
p.513 Plausibility and probability constitute alternative yet complementary bases for action. Whereas the first bases action on the coherence of an experience - does it make sense? [JLJ - plausible move sequences in a game, for example] - the second bases action on the robustness of the experience's replication - i.e. on its reliable correspondence to some recurring state of the world. [JLJ - evaluation function of typical continuation, for example] Both help to underpin action, but each may kick in at a different moment on the journey from belief to action.
p.513 If we think of a 'mindset' as the embodied epistemology that an agent employs to navigate a path from possible worlds to actual worlds, there are two possible navigating mindsets. We will call the navigating strategy using the path through plausible worlds the entrepreneurial mindset. An entrepreneurial epistemology seeks to extract real world value by enacting plausibilities. On the other hand, the navigating strategy using the path through probable worlds employs what we will call a managerial mindset. An embedded managerial epistemology seeks to extract real world value by enacting probabilities.
Entrepreneurial and managerial mindsets will each draw on their respective epistemologies as a basis for action. Whereas the managerial mindset seeks evidence and uses probabilities to justify its actions to others... the entrepreneurial mindset has to have the courage of its conviction in its beliefs
p.514 much of what today passes for knowledge management has its origins in practice... much of the knowledge that has been of interest to knowledge management... consists of rules of thumb, anecdotes, and best practices assembled and deployed within one organization... How is a claim of having knowledge to be justified? ...the creation of knowledge by the scientific community does not follow the 'logic' of knowledge creation as practiced in a corporate environment... whereas the scientific community's primary concern has been with the epistemological validity of the newly created knowledge, that of the corporate community has been with the economic utility of such knowledge.
p.515 because its primary concern has been with the utility rather than with the validity of knowledge beyond the boundaries of the firm, knowledge management as practiced by corporations has not so far felt any pressing need to secure its epistemological foundations.
p.515 An agent needs a good reason for selecting one hypothesis rather than another for further development and testing, since such activities consume scarce resources... In the world of the possible, if no prior distribution is available, no expectation can be built on the basis of recurrence... it will be the states of the world that suggest the appropriate trajectory to follow in moving from a possible to an actual world.
Given the lack of replicable precedents under conditions of innovation, the entrepreneurial mindset will tend to favor an initial move from possible to plausible worlds. It will then enact for itself... by constantly testing prior assumptions against accumulating empirical evidence in a process known as discovery-driven planning. The managerial mindset... is generally more disposed... to build its theories on the basis of available empirical evidence
p.516-517 Living systems endowed with cognitive capacities, however, have successfully evolved responses to representations triggered by the stimuli rather than to the stimuli themselves... this more cognitive strategy... does not attempt to match the variety of a given set of stimuli on a one-to-one basis with a given set of responses. Rather, through a filtering and interpretive process, it reduces the variety of the response called for by reducing the number of stimuli that it actually needs to respond to. [JLJ - we develop responses only to the stimuli that we need to respond to - i.e., our opponent's promising moves. The others are likely handled, so we do not waste time calculating responses to them.]
p.517 a process of exploration... a process of exploitation
p.517 When viewed from an evolutionary perspective, one of epistemology's key insights was that 'hypotheses could die in our stead'.
p.518 correspondence with the facts is typically not on offer in the early phase of a genuine innovation. The facts do not yet exist.
p.518 in a world of pure possibilities, for example, we can invest in low cost real options based on hunches derived from vaguely linked beliefs, which if successful, create real option potential that has option value. In this world, knowledge management can build on and improve brainstorming and scenario generation processes with a view to helping the firm to surface more and better possibilities and assess them as option candidates... Evaluation of such opportunities calls for knowledge management tools that identify those options opportunities with superior properties: upside potential, potential control of downside and potential sustainability.
p.519 The entrepreneurial mindset operates under conditions of novelty and uncertainty, where prior probability distributions, being non-existent, can offer little guidance, and in which coherence is the epistemological underpinning. The managerial mindset, by contrast, is constrained to seek justification from probability distributions and thus cannot operate in their absence, so that correspondence is its epistemological underpinning.
p.519 The entrepreneurial mindset attempts to enact bold yet plausible hypotheses that create their own reality; the managerial mindset acts on the basis of objectively verifiable facts and constraints from which it derives hypotheses with a high degree of probability. The first epistemic strategy will be appropriate under conditions of novelty and uncertainty. The second will be appropriate where accumulated and objectively verifiable experience is available.
p.519 there is a need for knowledge management tools that allow agents to develop and explore possibilities... and also to develop and explore plausibilities... Seed tools already exist... These are brainstorming and scenario planning for possibilities, and pattern recognition tools like text processing and analysis for plausibilities.
p.519-520 As the level of uncertainty goes up in many walks of life, the entrepreneurial approach to the management of knowledge will surely grow in importance.
p.520 When dealing with entrepreneurial proposals, move always from the analytically oriented business plan questions of 'how do you know that this will happen?' ...toward the more action-oriented entrepreneurial question of 'what actions will you undertake to make this happen?' ...insist on the construction of scenarios for the possible outcomes of such actions together with the identification of the options available should something like the scenario come to pass
p.522 Naturalistic epistemology... has pragmatic aims... it is happy to take as knowledge whatever subsets of an organism's beliefs it is willing to act upon providing that these contribute to its survival and... to its prosperity.
|
