I finally worked all the way through Antifragile: Things That Gain from Disorder by Nassim Taleb.
An amazing book. Amazing obvious ideas.
I’m now thinking about how to apply this in IT. How do we build processes and IT infrastructures that gain from disorder. Systems that don’t fail from stress, but instead get stronger.
I’m thinking that the redundancy we’ve put into our network structures might get stronger over time due to disruptions. Over time, we adapt, put in more capacity, more redundancy, work out triangle WAN links between sites, design in excess capacity, etc. as a direct result of events that hurt the network. The net impact of the negative events is that the whole gets stronger. Does that count as one way antifragility increases in the system? We do seem to have less negative impact from outages because the ‘whole’ is more capable of adapting to outages.
Do our information systems get more antifragile over time as we experience issues, problems, outages, etc. and then adapt and improve and improve the ‘whole’ to lessen the impact of those events? Are we learning from these events and getting stronger as a result? Are our management teams getting more antifragile as we learn, adapt and improve over time?
Think about storage and how it used to be the drive failures resulted in outages in the data center. Such outages are hardly an event at all any more due to much higher MTBF on drives and due to RAID and other technologies that have been honed and improved over the years. Fragility has been reduced in our storage systems.
Lots to think about here.