Repurposing at the Santa Fe Institute

I recently returned from my first visit to the Santa Fe Institute, a highly interdisciplinary research institute that promotes dialogue and collaboration across traditional lines, where I participated in a Workshop entitled “Principles of Repurposing” (link to the wiki), organized by Jessika Trancik and Jon Wilkins. Twelve presenters and a handful of other participants from all over the country (and the world) came together for three and a half days of talks and discussions on the topic of repurposing, i.e., how systems come to have new functions. We came from areas as diverse as Architecture, Computer Programming, Engineering, Biology, Economics, and Psychology to discuss the principles of repurposing that transcend disciplinary boundaries.

Below I briefly discuss a few of the questions and ideas that I found most compelling from the workshop:

1. How do we quantify structural and functional repurposing of systems? (Posed by Jessika Trancik) During discussions with Jessika and a number of other participants, we discussed the possibility that functional repurposing could be quantified by simply measuring the change in input and output of the entire system and that structural repurposing can be quantified by measuring the changes to input and output of the components that make up the system (e.g., by using network analysis). This could potentially provide a way of measuring repurposing that allows one to easily switch between multiple levels in a hierarchical system (since input and output changes are being measured at all levels).

2. What kinds of environments allow for the development of ‘repurposable’ systems (i.e., systems that can be easily restructured to perform new functions)? F lexible and repurposable systems presumably take longer to develop and ‘pay off’ than systems that perform the function at hand but cannot easily be restructured to perform new functions. It is for this reason, that there needs to be sufficient ‘gestation’ time for developing repurposable systems. Dave Ackley and I talked about the potential importance of long oscillations between weak selection and strong selection in this context. Long periods of weak selection create ‘development time’ for repurposable systems that have long-term high payoffs, and subsequent strong selection can then cull away the less fit competition that might have been able to survive during the periods of weak selection.

3. Stories as a vehicle for attracting investment. One particularly interesting recurring theme from the workshop was the importance of ’stories’ about the intention or function of a system in attracting investment. This topic came up (in various forms) in the talks of Victor Seidel, Nicholas de Monchaux and others. Because the creating or repurposing of technological and social systems requires development time and capital, the ability to attract investment can enable a richer and longer-lasting microenvironment in which the system can develop. Since humans have the ability to speculate about the likely long-term payoffs of investments, we can synthetically create microenvironments for system development, i.e., through the stock market and private investment. In these situations, the ’stories’ behind a company or a prospective product can play an important role in generating and maintaining investment.
There is an interesting parallel between these ideas and the notion of the SCI (Social Cognitive Interface) that my graduate advisor (Robert Kurzban) and I wrote about in a recent PSPR paper. Essentially, we suggested that humans have evolved an SCI to maintain and transmit the most positive plausible representations about the self. This is essentially the same principle of ’stories being a vehicle for attracting investment. ‘ In the context of the evolution of social abilities, the capacity to effectively ‘tell viable self-promotional stories’ is likely to have lead that individual to be more highly valued by others and therefore a more likely recipient of social investment.

As part of the workshop, I presented some simulation findings demonstrating that a very simple rule foraging rule can be ‘repurposed’ as a rule for leaving uncooperative social partners and groups (several papers on this ‘Walk Away’ strategy can be downloaded from my research page). I also discussed the importance of system composition (assortment) in promoting (or disfavoring) selection at the level of the group or aggregation. In biological systems, certain compositions are more likely to promote selection for cooperation and other higher level functions (to read more about assortment and cooperation, download the manuscript in prep, “Assortment in Space and Time,” which I’ll soon make available on my website).

All in all, this workshop was highly engaging, creative and stimulating. In addition to the presentations and discussions, we also had some wonderful time exploring the restaurants of Santa Fe and talking informally about many of these topics and others. I especially enjoyed the informal discussions with Jon Wilkins, who seemed to have an endless supply of interesting ideas and unbounded tolerance for listening to my speculative musings.