Agile means Inefficient
A timeless truth
Modern industry loves to be ‘Agile’. The word doesn’t mean anything like what they think it means. Here’s a proof of that, using elementary logic.
Step 1: “A straight line is the shortest path between two points.”
Right?
Right.
Fix two pins to a board. Hang a piece of string between them. You’ll need a short string to make a straight line, and a longer string for any other curve or shape. Ob-vi-ous! *
Step 2: “Agility means, literally, choosing not to go in a straight line”
If somebody says ‘I am agile’, what do they mean?
Well, they might be implying that they are quick-witted, dynamic and fast — as opposed to daft and lethargic. But being quick is not enough. Agility also requires:
- Being able to change direction quickly, and often
- And, doing it by choice.
Don’t believe me? Not sure? Consider these examples.
A luge contestant and a ski-jumper are athletic and fast. They get to where they are going quicker than a freestyle snowboarder. But everybody must (surely?) agree that the snowboarder is more agile, since she changes direction and orientation at will.
A billiard ball is not agile, though it bounces around quickly — because it has no choice in the matter.
Somebody who slips and falls on their head might change direction very fast. But are they agile like a gymnast? Nope.
Being agile always implies deliberately choosing a longer route to where you are going than you could have taken (because it’s not a straight line, and you’ve changed direction… a lot).
Step 3: “Choosing agility on purpose means, deliberately wasting time, energy (and money)”
Conservation of Momentum is a law of nature. It means you must slow down to go round a curve. To change direction while driving, you burn more fuel, wear down your brakes, and lose more speed than if you took a straighter road. Stopping, starting, changing direction are all types of acceleration — which burn more energy than going directly from A to B.
The word agile comes from the Latin root agilis, meaning ‘having a tendency to act’. (It’s formed in the same way as fragile, from fragilis — ‘tending to break.’)
In our universe, every action requires a use of force and a loss of energy. Of course, we need to use force and energy — but there’s a limit to how much energy we can or should use.
Step 4: “People admire agility precisely because it is costly, inefficient and dangerous”
The first law of biology is homoeostasis: stay the same or die. During the vast majority of their lifetime, wild animals take it easy — resting and digesting on rocks, preening themselves, ambling to the waterhole, etc. Conserving energy. It’s all slow, planned and predictable. It is the stuff of normal, healthy life. Then, if a crocodile snaps out of the water, there’s a sudden flurry of panicked, agile activity — and then, as soon after as possible, all the animals all go back to what they were lazily doing before.
Agility is a trait that evolved for hunting, and to escape predators. It is life’s exception, not life’s rule.
No organism (or organisation) can be agile, more often than necessary, without burning up its vital energy reserves (or cash/credit reserves), which it stored up to enable it to be agile on the rare, brief occasions when agility matters.**
Because the ability to be agile, rarely, is a matter of survival, displays of agility grab our brains’ attention. (What do you notice most in the image below?)
Agile things are dangerous and exciting (and training our agility can be fun). But, constantly agile things are mad and moribund.
Conclusion: The Agility Spectrum
Let’s recap:
- the direct route to any objective is the most efficient route;
- choosing to be agile means taking an indirect and inefficient route on purpose;
- and so, Agile Development is a deliberate waste of time, life and money (to the extent that it is really agile)
The least agile things are like waterfalls — completely determined, hard to change, utterly inflexible. The most agile things are like dogs chasing their tails or mad, caged foxes — using a lot of energy to achieve utterly nothing.
The most effective way to achieve anything must clearly lie somewhere in between those extreme ends of the agile/unagile spectrum. And, the more you know your what your goal is, the more you have a clear vision of what success looks like, the less agile you will need to be.
Those who love (big ‘a’) Agile Methodologies will raise some objections to this. They’ll say that their methods are needed because we don’t know at the start of a project exactly where we want to get to, or which product features our customers will love. So we need to change path quickly based on trial, error and experimentation. Sadly, that is bullshit. The best algorithm for finding something is never a process of elimination by random guesses. When people need to find something quickly and effectively (search and rescue missions, say) they use Bayesian Search or Gradient Descent algorithms, or similar. Not random guesses.
Even if the exact location of our goal is unknown, we should search in the way most likely to find it soon — that is, the way that takes us there by the closest approximation of a straight line.
What tech projects need is MVA: the Minimum Viable Agility.
Notes:
- * [And, no, you don’t defeat this point by talking about manifolds and geodesics and spherical freakin’ trig. A straight line, or its equivalent in whatever space you inhabit, is still the shortest path between points.]
- ** [Too frequent agility burns through sanity reserves as well as energy stores]
