On consensing…
Posted on February 7, 2010
Filed Under Collective Intelligence | Leave a Comment
We live in a much faster-paced world today than our ancestors did, and we have to make decisions very quickly (something that, as it turns out, I really, really suck at… but less of that, and more about what I intended to write about in the first place).
We like to reduce decision-making to the equivalent of answering a multiple-choice questionnaire. Take ordering fastfood, or the way our voting systems work, for example (pick one: candidate A, B or C, etc). This has, as intended, taken out lots of hassle from choices we must make everyday.
Unfortunately, the ease of generating a response (any response) from a survey recipient seems to vary inversely with the meaningfulness or usefulness of the response – especially for complex surveys. More to the point:
1) the fitness of solutions founded on rapidly-made, quantised responses cannot scale across time, and
2) the fitness of solutions founded on any responses from any kind of mass survey / mass measurement must necessarily vary with time.
We need to move away from “dumb”, closed, uni-directional surveys/polls/rating systems and toward something that I like to call consensing…
1) Maintaining consensus-convergent solutions across time
This basically comes down to, ‘now that you know what the people want, keep giving them what they want’. Unfortunately at it’s core, this quickly reduces to a negative feedback algorithm: ‘are we *sure* we *still* know what the people want?’. Because what the people want will change, and whatever our original response (to reduce confusion I’m just going to say ’solution’) to the mass response, it needs to adapt if it is to remain at least as optimal as we thought it was, back when we first thought we had consensus. Shoot. That was a mouthful. Let me try again:
take#2:
If we are going to poll a demographic about something and we are going to apply their black-and-white answers *forever*, as (for example) corporations and governments are wont to do, the intended services/offerings will diverge from the consensus, if consensus there was. This divergence is inevitable and will occur even if only much later, when people have had time to do a bit of thinking and have now changed their minds. Or their value systems.
2) A “deprived consensus” might be fix-able with an adaptive (consensing) model
Divergence is more rapid when the quantised responses were constrained or inadequate to begin with (perhaps because of lack of imagination or skill on the part of the party making the enquiry). I think of a ‘deprived consensus’ (that’s a totally made up pair of words) as one where it is only possible for respondents to choose from a much narrower range (or invalid set) of responses than should have been made available for the enquiry in question.
Any mass-measurement – especially one seeking to measure opinion – should be revisited, whether on a scale of minutes or decades.
If there was a method by which responses could reconfigure the basis of the original enquiry (e.g affect its goals or change its assumptions) then the enquiry and the response set both become a unified system that can evolve toward a more useful basis for decision-making. Essentially, ‘consensing’ is the grativation toward a useful ‘attractor’ in the enquiry-response system. The ideal (and perhaps only) way to support that is through openness and transparency. Too often (unless it is a largely-pointless web poll), the cumulative results of surveys, quizzes and questionnaires are not made availabe to, or shared with the respondents. Certainly not its raw form (a summarised, sanitised version of things MIGHT be circulated).
I think something that would set consensing apart is to allow respondents to observe, in real-time, the accretion of their own (and collective) choices, and any metadata (e.g. reasoning or supposed rationale – perhaps this latter also quantised in some way) that supports such responses, and to have at their disposal tools which bring to the fore emerging patterns and shifts in value systems, affiliations among respondents, gaming strategies being employed to manipulate the enquiry / its outcomes, and so on…
Consensing* could be very interesting, folks…
*If you’d read my post about jobs for futurists, you’d know what a ‘Consenser’ was. You might even be planning to shed your current career in order to pursue this decidedly vague occupation…
Tenuously Linked (blame my tagging):
Comments
Leave a Reply