C.A. Corriere

Good evening, my esteemed enterprise DevOps practitioners.

I'm C.A. Corriere, an advocate for mapping DevOps practices and complexity science for Auto Trader. And I'm here tonight to talk to you about the seven horses of abductive reasoning.

So once upon a time, a socialite gathered up this village and said, "I'm inviting everyone to a party, not just the socially elite. Isn't that wonderful?" And a farmer replied, "I don't know. Maybe."

And the socialite, feeling snubbed, stormed off in a huff.

Coincidentally enough, the next morning, the farmer's horse had run away, and the townspeople said, "Isn't that bad?" And the farmer said, again, "I don't know. Maybe."

Later that evening, the horse comes home with seven wild horses, and the people said, "Isn't that great? You must be the richest person around for miles now." But persistently, the farmer said, "I don't know. Maybe."

Later that week, the farmer's child breaks their leg trying to tame one of these horses, and the townspeople said, "Isn't that terrible?" But still, the farmer said, "I don't know. Maybe."

Well, it was later that year that war breaks out, and a conscription officer comes to town and says, "I have no use for an old farmer or a child with a broken leg, but I'll take you, socialite." And some thought this was great, and some thought this was bad, but still the farmer said, "I don't know. Maybe."

So time passes. The farmer prospers with a healthy child and these eight horses, and the war ends. And the socialite returns to a vestige of the life they once knew, house in shambles and nowhere to go. Nowhere to go but to work for the farmer.

And the townspeople said, "I remember you. You used to be such a big shot. Now look at you. Isn't that sad?" And being glad to be alive and have a place in the world, the socialite simply said, "I don't know. Maybe."

See, nature is an integrated process of immense complexity, and it's really impossible to tell if anything that happens is good or bad because we never know what the consequences will ultimately be. This story is adapted from an Alan Watts lecture. And it's really about requisite variety.

See, nature bets on a few different horses, and survival of the fittest is not about winning or losing. It's about survival. The farmer was focused on survival of the village, not any pettiness between him and the socialite. And this type of collaboration demands a flatter ontology.

So Theory X says, "You're bad people, and I need to manage you to get you to do stuff." And Theory Y says, "We can all hold hands and sing "Kumbaya."" And we need some of both, but this takes a different type of reasoning.

So there's really three kinds of reasoning. You've got deductive logic, which is top-down; inductive logic, which is bottom-up; and abductive logic, which simply asks, what seems most probable? What's most likely to happen here?

So deductive reasoning: all people are mortal. Socrates is a person. Therefore, Socrates is mortal, which is a fair deduction, unlike this dog that has deduced it might be a cat because they both have four legs.

Inductive reasoning: this Ignite talk has 20 slides. The other Ignite talks have 20 slides. Probably all the Ignite talks have 20 slides, which is a fair induction, unlike this penguin that's induced it might be an old TV show because they're both black and white.

And abductive reasoning just asks what's most likely. If you keep the Deep Heat next to the Colgate, it is only a matter of time before you brush your teeth with muscle ointment.

And this brings us to parallel safe-to-fail experiments. So where is the gravitational pull? Where is the inertia? And could this be flatter? If we're going to roll a wagon down these hills to see which one didn't fall in a pit, where might we test that before we took a ride?

Yes. And this hierarchy doesn't help feedback get to leadership. So top-down is good for command and control and broadcasting info, but the signal-to-noise ratio on that feedback is going to be low. This is like Shannon theory.

So some structure is necessary. What we've learned from Zappos and the holacracy is if there are no managers, then everybody's a manager, and you end up having to try to deal with all these different priorities simultaneously with nowhere to escalate.

So how can you collaborate with your customers, employees on a peer level to deliver more feedback and value? Do you have a map?

And scratching an itch and removing a parasite are not the same thing, but both are helpful.

So you can follow me on Twitter, you can follow Auto Trader, and you can come talk to me at the booth tomorrow if you'd like to discuss mapping.

Will I see you there? I don't know. Maybe.