On Evolution, Inevitability and Inertia

or: How I learned to stop worrying and love the cloud

“Those who fail to learn from the mistakes of their predecessors are destined to repeat them.”

— George Santayana

The evolution of electricity

In 1600, William Gilbert discovered electricity. You and I can name hundreds or thousands of things that need electricity, but in 1600 it was merely a scientific curiosity. A source of wonder with literally no practical applications.

In 1800, Allesandro Volta built his famous voltaic pile, the first reliable source of electricity. In 1831, Michael Faraday invented the Faraday disc (an early type of generator). Hyppolite Pixii (a French instrument maker) built the first dynamo in 1832, and by 1844, electricity was used in an industrial process for the first time when the Woolrich Electrical Generator was used for electroplating at Elkington Silver Electroplating Works. Other companies soon saw how using electricity could give them an advantage and they started custom building their own.

Eventually, someone started producing and selling generators. The earliest instance I could find is the Ferranti-Thompson Alternator, launched in 1882, but I’m not sure if that’s the earliest.

Also in 1882, the first central power plant was established in lower Manhattan. Electricity was now available as a utility.

All of this obviously happened many years ago. We all grew up with electricity provided as a utility. If a device needed electricity, we’d plug it into a power socket. So, blessed with the power of hindsight, free from bias towards running our own generators, it’s obvious to us that if you have a reliable power utility, running your own generator is not only a silly waste of time, it will put you at a disadvantage compared to those of your competitors who focus on their business problems and leave electricity production to someone else.

It’s even more clear that if you’re spending time building your own generator, you’re at a disadvantage compared to those of your competitors who buy an off-the-shelf generator.

It’s hard to imagine someone who’s still wondering how electricity could even be useful, but they would obviously be at a disadvantage compared to pretty much everyone else.

Inertia

When ready made generators appeared, everyone stopped building their own. When the power utility appeared, everyone discarded their generators and relied on the utility.

Right? No. Not even close.

People had spent considerable amounts of time and energy building their own generators. Objections to using someone else’s productized generator can range from the relatively reasonable to the completely nonsensical:

• Our generators produce 160V at 125Hz. That’s what all the rest of our equipment has been designed to consume. There is no off-the-shelf generator producing the same voltage and frequency. We can’t switch.
• We’ve already built these generators.
• We know how to fix our generators when they break. We know nothing about fixing these other ones.
• The coils in our generators are wound counter-clockwise, not clockwise like everyone else’s. Clearly, this is a unique advantage for us.

Similar objections towards the utility from people operating their own generators:

• Some of our equipment uses 150V at 60Hz. The rest uses 260V at 100Hz. The utility only provides 220V at 50Hz. It’s too expensive to switch.
• We have a lot of experience operating generators. We’re confident we can produce electricity more efficiently than any utility.
• We have a lot of capital tied up in these generators. We’ll reconsider the utility in 5–7 years when the generators get closer to their end-of-life.

For completeness, the objections from someone still just wondering what electricity is:

• I don’t see the point. My steam engine solves all my problems.

We call these objections to change inertia. Even though more efficient, evolved forms of electricity exist, there’s resistance to change.

Taking a step back

As we saw with electricity, things evolve over time.

1. It starts out as a true innovation, a source of wonder. We’re not sure what to make of it or what it will evolve into.
2. If it turns out to appear valuable, people will custom build their own.
3. At some point, someone will turn it into a product and start selling it. Similar products will also appear, competing on e.g. features or price.
4. Over time, these competing products will reach a point where new features add very little value. Any of the products will be considered good enough for almost all customers. When this happens, a utility form will appear.

The stages of evolution

We also learned from the electricity example that if a utility is available, but you are stuck at an earlier stage of evolution, you’re at a disadvantage. A competitor using a more evolved form of a particular activity can focus more narrowly on problems unique to their business. This gives them an advantage over you.

The history of computing

Charles Babbage saw the potential benefit of computers already in the 1830’s, but for more than a hundred years after, the mere thought was mysterious to most people. The potential applications of a computer were hard to imagine.

In 1941, the first programmable, digital computer, the Zuse Z3 was built. Its applications were more clear at this point, and over the next few decades, many others saw the potential in computers, so they custom built their own.

In 1953, IBM launched the IBM 650, the world’s first mass-produced computer. Over the next 50–60 years many products came and went. It was chaotic at first with many different platforms, but eventually most people converged on an Intel based platform.

Around 2005–2006 we had come to a point where the exact make and model didn’t matter very much, because servers were more alike than they were different. Shortly thereafter, Amazon launched EC2.

Amazon, with no history in the hardware business, drove the change of computing power from product to utility. Not IBM. Not Dell. Not HP.

IBM, Dell, and HP were all faced with immense inertia.

They had a very successful business of selling servers. If they were to start selling compute resources on a utility basis, they would cannibalize their server business. Also blinded by their past success, they didn’t think the compute utility business would take off, and if it did, they believed they would see it early and have plenty of time to react. They did not.

As you can see, inertia exists on both sides of the supply/demand divide. Consumers are hesitant to adopt the more evolved, but less well known forms of the activity, and suppliers are hesitant to change, often due to past success.

What you can learn from this

For us, the innocent bystanders blessed with the wonder of hindsight, it’s obvious that using the power utility is a better choice than buying a generator. It’s also obvious that you should do it sooner rather than later. It’s also obvious that IBM, Dell and HP should have gone all in with a compute utility more than a decade ago.

Don’t be IBM. Don’t be Dell. Don’t be HP.

Accept that this evolution will happen, with or without you. Break through the inertia. If you’re a supplier, realize and accept that someone will turn your product into a utility at some point. It might as well be you. If you’re a consumer, keep an eye out for more evolved forms of things that you rely on and adopt them before your competitor. The effort spent on the change process is the best investment you’ll ever make.

If you’re ever in doubt, think of the custom built generators, the off-the-shelf generators, and the power utility. The right path forward is usually clear in that case.

I learned all of this from Simon Wardley. I’ve explained it to others in person enough times now that I wanted to put it in writing and maybe save myself a bit of time.