Thread by @swardley

X : Your table on next gen practices?
Me : You mean this one? It's from 2010, published in 2011. Not really next gen today.
X : The chaos engines bit.
Me : Netflix Chaos Monkey, Amazon's "Master of Disaster" (early to mid 2000's), old manufacturing concept ...

... the constant drive towards ever greater levels of resilience through the constant introduction of random failures and chaos.
X : Is that the same as adaptation?
Me : Not quite. Adaptation is a much broader concept including themes like inheritance, mutation, exaptation ...

... but in general, nature (as a whole system) is the most resilient system we know. It has both engineering and ecological resilience. CS Hollings work is definitely worth a read.

X : Where does anti-fragile fit in then?
Me : I'm not the person to ask, I don't recognise that meme as being meaningful. As far as I am concerned it only exists because people redefined what resilience means. If you find it useful or it helps you then that's good for you.

X : You're not a fan?
Me : Of growing from disorder? That's part of resilience. Of the meme? No. If you find it useful then that's great for you, just don't go telling me that resilient systems don't learn from disorder or that resilient systems are confined to being robust.

X : Can you give me examples of resilient systems that learn from disorder?
Me : Sure. Amazon's "master of disaster" or Netflix's Chaos Monkey (I was a judge on), a host of older manufacturing examples and nature itself. That's why I pointed to chaos engines in 2010/11.

X : They're anti-fragile.
Me : Hmmm ...

X : Manufacturing?
Me : It probably started with the introduction of failure into manufacturing lines to ensure that Q&A function was working. Decades old.
X : What about growth?
Me : Evolution is a necessity in competitive systems to maintain a status quo - see Red Queen effect.

... for interest, this is why in highly resilient systems you don't design them as much as you grow them. You provide the conditions for a resilient system to emerge with all that adaptability and resilient goodness and what you do is more akin to gardening.

X : Do we do that in engineering?
Me : Biomimicry is a growing topic - engineers have been learning the concepts of evolution, exaptation, degeneracy, learning, competition, training etc for some time. Future starships will be grown not designed.

X : Examples?
Me : Lots in defense but take the Chaos Monkey example. Netflix's Open Source competition created new cases that hadn't been considered. The conditions were set for new modes of failure to be discovered and harvested.

X : Are you saying the next gen have gone beyond "Design for failure".
Me : Probably ... if not already then soon enough. That was next gen in 2010. Things evolve, we continue to learn from nature. It's still far more resilient than anything we've created.

X : Why probably?
Me : Well, there will be cases but maybe not enough to declare it as a next gen thing yet. There might be constraints to overcome. The path is clear enough though. Give it time.

X : Our organisation isn't even where you said the next gen was in 2010.
Me : That's common. Many organisations are 20+ years behind. That's ok, as long as your competitors are also 20+ years behind then no-one gains an advantage.

X : What happens if an advanced player moves into our space?
Me : You're toast.
X : Anything that can done?
Me : Regulatory barriers? If that fails - sweat and acquire i.e. sweat assets, return value to shareholders, acquire similar distressed competitors, sweat more ... repeat.

X : That doesn't sound like much of a survival plan.
Me : Depends upon how close you are to retirement? It's simply a way of buying more time. Maybe you'll strike it lucky?