Dear FIS'ers, 

Thanks so much to all of you for your wonderfully thought provoking comments, 
questions, and feedback! 
In this note, I'll attempt to address some of the queries/comments that have 
been shared. 

John Collier says: 
[----As I understand it, John?s approach is specifically based on using 
Information Technology mediated groups of agents to derive the existence of a 
collective intelligence, but I would like it to be explained in what this 
intelligence consists . In other words, are we dealing with knowledge-as-such 
(stored and shared data) or capability for effecting change. John P. does say 
that crowd capability is directed at processing knowledge, but does this 
exhaust the content of the concept of intelligence as capability?----] 

This is a great question John, and one that has largely been ignored in the 
field, or is at least still contentious in my mind. 
In a sense, the avoidance of this issue signals that this domain is still 
immature, and further that there is opportunity to shape the domain in this 
respect, if one were so inclined. 


The Woolley & Malone et al. crew (in the Science 2010 paper) focus on the small 
group level (very small!), and define CI as a group ability to perform. This 
intimates a process model of CI. Much like Steven I'm not a big fan of this 
work, principally because there is no place for IT in this notion of CI, and it 
seems to me that communication is driving all outcomes (which I think is 
essentially Loet's point too). It does reveal, or rather reminds us, how very 
nebulous the cognitive concept of general intelligence is in it's own right, 
and so in this respect the work is useful. 


The other popular work attempting to define CI, is from Pierre Levy (1999), who 
defines it as [----...a form of universally distributed intelligence, 
constantly enhanced, coordinated in real time, and resulting in the effective 
mobilization of skills...----]. As opposed to the group-level, Levy is here 
focused on society as a level of analysis. Thus far, I've yet to see any work 
that attempts to reconcile these different levels of analysis, though off the 
cuff these two definitions of CI seem to share a process-orientation, with a 
focus on performance/mobilization outcomes. 


>From my view, broadly speaking, I think that we can certainly say with some 
>confidence that there should be a difference between IT-mediated CI and 
>non-IT-mediated CI (or at least this is my hypothesis). In the former, frozen, 
>yet adaptable artifacts (ie software/algorithms/AI) are involved, while in the 
>latter they are not. You could of course argue that 
>language/society/culture/communication is a technology/artifact in it's own 
>right, and philosophically I wouldn't necessarily disagree, but we can also 
>agree that software is demonstrably different than culture for example. 


It would appear that the explicit codification of knowledge is a bridge between 
the two categories that I draw. 
Further, both forms of CI would most certainly have some level of individual 
human intelligence in common too. 


For me the key difference is that in the IT-mediated case of CI, the existence 
of AI/algorithms denotes a demonstrably different system of interacting parts. 
If this is true, we should expect different dynamics (and probably different 
outcomes) for each system. 



I very much look forward to any further thoughts from the group! 


Best, 
John 
_______________________________________________
fis mailing list
[email protected]
https://webmail.unizar.es/cgi-bin/mailman/listinfo/fis

Reply via email to