A recurrent theme at DevOps Days and in discussions: Is the DevOps community focusing too much on automation? Rather than the more difficult organizational changes that everyone recognizes are needed?
We humans are tool-makers so our fascination with tooling is well-founded and clear. I am captivated by new tools and tooling myself. But it’s deeper than that: I think we transform and improve the excellence of everyone’s life when we automate and improve conditions in the best possible ways. We can raise it up for everyone.
Sometimes we get massively confused by the role of our tools — like the dominant view of automation from last century found in the Charlie Chaplin “guy in a cog” MODERN TIMES movie and reflected brutally in the US by the managerial nightmares institutionalized in Henry Ford and Alfred Sloan’s sort of thinking. For too long, the assembly line has been the focus, the humans around it of interest only as long as they serve the line.
This human night is slowly being addressed by the dawn of Lean, of Gembi Kaizen, of DevOps, of DRiVE Motivation thinking, of a present rid of the managerial assumption that “staff needs to be motivated otherwise they’ll do nothing”.
It reminds me though of the other, better, more enlightened, humanist perspective IMO provided by the tech titan Steve Jobs. And I think time will show the correct view. I was vividly reminded of this last weekend when I watched STEVE JOBS: THE LOST INTERVIEW. He was still running NeXT Computer (although he had pivoted to a software only business model by 1995, when the interview was recorded).
[On a related note, I own an original NeXTstation, and I had to apply for a bank loan to buy it . . . it was so expensive. And I got a lot of grief from friends and colleagues about why in the world had I bought a NeXT. And I learned to reply to these scoffers what I saw when I used it: that in ten years, computers will maybe not be called NeXT but this is what our computers will be like. And today I’m on my MacBook running OSX, which is basically NeXTStep.]
So Steve Jobs had this story about being a young boy and he read this article in Scientific American about a study of the efficiency of locomation — measured by kilocalories per kilometer — with a wide-ranging sampling of animals and birds. And the condor won. Humans came in very mid-list. Not so impressive being human apparently.
Then someone doing the study was smart enough to also measure a human riding a bicycle. And that blew away all other contenders, orders of magnitude better than the condor, totally off the charts.
That’s the human awesome at work.
Tools dramatically amplify our innate abilities and this is a wonderful thing. And I agree with Jobs the computer is the most wonderful tool yet. But it’s possible for humans to lose focus and to become too tool-oriented. For the human to be a tool rather than use a tool (I’m pretty sure the pun is intended…). As discussed in Mark’s Excerpt #3, culture is a big deal for humans and we have a lot of corporate cultures where the humans work for the machines and that is wrong.
Contrary to Henry Ford and Alfred Sloan thinking — that has unfortunately and viciously dominated American business for nearly a century and still does at many businesses — lasting innovative business value is not created by humans working for the machines.
Mark Burgess: I mean, tools are important and we can talk about culture, but of course culture can also be formed around tools. Tell the iPhone guys that culture is not formed around tools and, you know, [that] makes no sense.
So tools, clearly important. The point that I’ve mentioned about monitoring, I think, was a good one which was that — in the past we’ve had very different technologies that don’t really meet in the middle. We’ve had deployment technologies and we’ve had monitoring technologies and they’re two independent systems.
So look at the technologies over the last 20 years and [it’s] a lot of alarms and monitoring… waking humans up to work for the machine.
You know, change my nappy — this kind of alarm that wakes people up and make humans work for the machine.
I actually call it — I can be fairly rude about monitoring systems — I say that a lot of monitoring systems at least are really vanity software […] Sys admins look at these tools and they say, “Mirror, mirror on the wall, who is the smartest sys admin of them all? Look how beautiful my system is. Look how clever I am for making this line be flat or go down or up,” — or whatever it’s supposed to do. And they look into this mirror all the time, looking for something to happen and that’s wasting people’s time, right?
Making a human do this mind-numbing work, instead of focusing on strategy, instead of focusing on improvements, on — so I call it vanity software but it sells really well, because we like to feel good about ourselves.
But what DevOps is trying to do, I believe, and in a great way is to say: guys, think about the business value. Don’t think about those tools.
Note: I’m going on a bit of vacation so I won’t be publishing an excerpt next Wednesday.