I finally watched The Matrix last week. It’s a thought-provoking movie that asks three disturbing questions:
1. Does anyone call what Keanu Reeves does “acting”?
2. Has there ever been a stupider premise than the human body as the ideal source of electrical energy?
3. Does Moore’s Law make the movie’s basic premise inevitable?
We’ll leave the first two questions to Roger Ebert. Before we dig into the third …
Last month I asked what you envisioned as the center of your network, the mainframe or the PC. In other words, is the point of your network to connect terminal devices to the systems that drive them, or is it to connect employees to the resources they need to do their jobs?
The e-mail and forum exchanges on this question surprised me. Most unexpected was that nobody proposed putting processes in the center, even though the process view of the enterprise dominates consulting circles these days. The correspondents who proposed an “acentric” perspective also caught me off-guard, since to me acentrism means no design focus.
What bothered me the most, though, was how many respondents told me the enterprise “never stopped running on the mainframe.” This contingent disputed my assertion that a company’s work is performed by individual human beings, and that companies succeed or fail one person at a time.
On reflection, this isn’t a question of who is right – the question is which perspective is the most useful. With the mainframe in the middle you’d divide work into three categories: Data preparation, where people and feeder systems massage data into processable formats; The Work, which is what host applications do; and exception-handling, which is what people do with system outputs (since the system does The Work, it only reports the exceptions it can’t handle).
With Process in the middle, both humans and information systems fulfill roles in the company’s core processes, performing well-defined tasks that transform inputs into outputs.
Both of these perspectives can be useful. I’ve designed and implemented quite a few successful applications based on the systems-centric view myself, and as mentioned, the process-centric perspective currently dominates business design.
When you put the employee in the middle, though, several good things happen. First, you reduce overhead. Every time one employee hands work to another, entropy happens – work goes into managing the transfer of work rather than getting the work itself done. With a human-centered view you’ll organize resources so work stays on a single desk until it’s done.
Second, customer relationships will improve. When one human being owns each piece of work, the company has a chance of looking less like an impersonal machine that answers all requests with, “We can’t do that – it violates our procedures.”
To understand the third benefit, let’s revisit the basic idea behind The Matrix – that eventually we’ll all be slaves to one or more artificial intelligences. Just thirty years into the future, Moore’s Law will have clicked over twenty times, so computers will be one million times more powerful than they are today. One million.
No matter what the cognitive task, computers will be better at it than you are, so if the mainframe is in the middle, you’ll be working for it. Likewise for process-centered work – computers, being far more capable than humans, will do all the interesting stuff. (In the movie version, we’ll do nothing but cheap manual labor. Fortunately, Microsoft will have written the operating system and our heroes will take back the world when the blue screen of death happens.)
If humans are in the middle, we may have a cable going into our skulls (although I sure hope wireless technology has progressed more by then) but it will be to augment our abilities, not to boss us around.
Okay, this is the stuff of a summer movie, and your choices today will neither save nor destroy the world two decades from now. My point is to illustrate the third benefit of putting humans in the middle of your system designs – you’ll help make your company a better place to work.