More Buzzwords Won't Help: The Long History of DevOps Failures
summary
Argues that repeated DevOps-era failures come from automating technology without changing work design: 1950s British coal studies show mechanization lowers productivity, safety and morale while cross-functional teams raised all three; GM spends $45 billion on automation in the 1980s and gains only one percentage point of market share; modern efforts (Puppet, Kubernetes, serverless, platform teams) repeat the same mistake. Recommends joint optimization of social and technical systems, expects a J-curve dip in performance while teams learn and integrate changes.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:08Defining DevOps failure
- 4:17Speaker background
- 8:55Deming’s principles and hero culture
- 10:39Socio-technical systems theory
- 16:14Why paradigms fail
- 19:25Deming’s impact on Japan
- 23:18Automation pitfalls
- 26:09What orgs want from DevOps
- 27:12J-curve of change
- 28:38Why buzzwords spread
- 31:47Org charts and tribalism
- 33:55Systemic solutions needed
- 34:16Closing remarks and Q&A teaser
related talks
25mDevOps: Why do we Keep Messing Up?
11mDevOps Pitfalls, Beyond The Technology
29m5 ways to screw up your DevOps Transformation
17mDevOps Leadership in practice. True stories
26mThe Timeless way of DevOps
44mContinuous Stupidities - Myths of DevOps
33mWhat Got Us Here Won’t Get Us There: A Story of Transformations
27mThe curious case of the DevOps engineer and other things ...