Over the last ten years, I have struggled with a Facebook motto: Go Fast and Break Things.
It reminds me too much of the Great Gatsby quote:
They were careless people, Tom and Daisy—they smashed up things and . . . then retreated back into their money . . . and let other people clean up the mess they had made.
The process “go fast and break things” does not describe a method for creating value; it represents a process for an adrenaline high, an excuse to do whatever you want.
Let me ground this into a real-world example. Suppose I have a system, and I want to radically re-imagine what the system can do.
There are two paths.
The first is to create a brand new system, that is entirely incompatible and breaks everything. What do I mean by everything? Any system is part of an ecosystem of tools and operations, and people that interact with the system. When you break the system, you are breaking that web of relationships and interactions. The net outcome is a radical change of that web.
So why do it? Well, because the cost of that change is borne by the people who use the system, not by the people who built the system. The more powerful the market position, the easier it is for the entrenched system to break things.
For example, Facebook used to break APIs all of the time. And that was okay for Facebook, because their captive audience, had no choice but to change to use the new APIs. The consumer-owned the full cost of the disruption.
The second approach is to evolve the system in a way that doesn’t break anything. In this model, instead of forcing the world to adapt to your system, you figure out how to integrate into their world.
Intel is an excellent poster-child for both of those. The first time was when they delivered the Pentium. At the time of the 486, there was a bunch of RISC processors like MIPS, Alpha, SPARC, that those of us in the field thought had a legitimate chance of dethroning Intel because the CISC core was intrinsically slower than RISC systems. But getting off of Intel meant breaking things. Instead, Intel did something that was a surprise to the casual followers of the industry, they embedded a RISC-like core into their processor, preserving the CISC instructions. By choosing not to break things, they won the CPU core wars.
Ironically, a little later, Intel pursued a foolish – in retrospect – strategy of the Itanium. The thing about the Itanium was that at the time there was no 64 bit Intel processor. Intel planned to move away from the x86 instruction set to go to a new kind of instruction set. The switch was highly disruptive. And AMD delivered what the market wanted, a 64-bit x86 processor and achieved a huge market opportunity.
In both cases, the winner deliberately chose to break as little as possible and add value in a way that did not disrupt the consumers of the technology.
To me, that is the best kind of engineering.
Leave a Reply