I really enjoyed the article but I have lots of questions about the accuracy with which it's possible to model and account for such a dynamic and complex system. In addition, wanting to be able to account for rare or unusual circumstances upon which there may be limited data for how things interact only makes this harder!
Back in the day when corporations were run by engineers with access to embryonic infotech, a dynamic approach to systemic resilience made some progress, although undermined by the rise of legal-financial agglomeration and its complementary opposite, market-oriented disaggregation.
Given that UK Government is dominated by legal and financial types in thrall to market mythology, you have your work cut out to get this into the frame, even though the reality principle keeps puncturing complacency.
In theory, a complex global interconnected network ought to be *more* resilient: the fact that sometimes they are not suggests we are not networked enough? But depends on the sector, and I think necessitates a degree of redundancy to create spare capacity.
In the general case, I think they are more resilient, but as we seek to operate in more complex and efficient ways, the net effect of composed and aggregated networks might be a decline in resilience.
I believe this would only be true if nodes of the network were equally weighted. Having critical central nodes critically dependent on smaller connecting nodes can lead to cascading failures across the network - see financial crashes due to complex and spread out liabilities.
Yes. In retrospect, I should have mentioned digital twins. I am more positive about ecosystem models because I do not think that we need to capture the full complexity. As below, I think gross system dynamics might be sufficient.
I run the podcast for the Technology Society at Cambridge University and we'd love to have you on to talk about some of these ideas - is there an email address I could please contact you on?
I really enjoyed the article but I have lots of questions about the accuracy with which it's possible to model and account for such a dynamic and complex system. In addition, wanting to be able to account for rare or unusual circumstances upon which there may be limited data for how things interact only makes this harder!
It is not necessary to be accurate, I think. Qualitative and approximate analysis is probably sufficient.
Back in the day when corporations were run by engineers with access to embryonic infotech, a dynamic approach to systemic resilience made some progress, although undermined by the rise of legal-financial agglomeration and its complementary opposite, market-oriented disaggregation.
Given that UK Government is dominated by legal and financial types in thrall to market mythology, you have your work cut out to get this into the frame, even though the reality principle keeps puncturing complacency.
Is the problem also that we tend to focus on 'what' without also asking 'how'?
Always!
In theory, a complex global interconnected network ought to be *more* resilient: the fact that sometimes they are not suggests we are not networked enough? But depends on the sector, and I think necessitates a degree of redundancy to create spare capacity.
In the general case, I think they are more resilient, but as we seek to operate in more complex and efficient ways, the net effect of composed and aggregated networks might be a decline in resilience.
I believe this would only be true if nodes of the network were equally weighted. Having critical central nodes critically dependent on smaller connecting nodes can lead to cascading failures across the network - see financial crashes due to complex and spread out liabilities.
Interesting it reminds me of:
1) a current vogue for “digital twins” in process engineering.
2) attempts to model ecosystems which have struggled to capture their complexity.
As ever I imagine it will all hinge on the availability of quality data.
Yes. In retrospect, I should have mentioned digital twins. I am more positive about ecosystem models because I do not think that we need to capture the full complexity. As below, I think gross system dynamics might be sufficient.
This was a fascinating read, thank you.
I run the podcast for the Technology Society at Cambridge University and we'd love to have you on to talk about some of these ideas - is there an email address I could please contact you on?
Glad you liked it. Of course, you can initially contact me on my website https://finkelstein.uk