Your Company Doesn’t Do Machine Learning Operations. It Does Theater.
Okay, one more on this and then I’m done, I promise.
MLOps theater persists because it satisfies the organization’s dominant incentive structures. The actors are rational. The script is flawed. Teams perform maturity they have not achieved because the performance itself is what the system measures and rewards. The underlying operational capability matters far less than the visible signals of progress that move up the management chain.
Why the Curtain Stays Up: A Three-Act Incentive Analysis
For Leadership: Signaling Transformation Beats Building Capability
The executive-level incentive structure rewards declarative statements over measurable outcomes. A board slide declaring “Enterprise MLOps Platform Deployed” closes a perceived capability gap in a single presentation. It demonstrates responsiveness to industry trends and positions the organization as technically sophisticated. The alternative, a multi-year investment in operational infrastructure with delayed and incremental visibility, carries presentation risk with no immediate signaling value. Leadership is not irrational for choosing the former. The system penalizes honest timelines and rewards confident declarations. Theater is the economically optimal choice when stakeholders demand proof of transformation but lack the technical depth to evaluate actual capability.
For Teams: Visible Launches Outweigh Invisible Reliability
Performance reviews, promotion committees, and quarterly planning cycles all optimize for countable, attributable output. Shipping a new model feature generates clear, countable metrics that fit neatly into performance reviews and quarterly reports. Building the automated retraining pipeline that prevents future model degradation creates none of these signals. Its success is measured by the absence of problems rather than the presence of achievements. Because the system rewards visible launches over invisible reliability, teams rationally prioritize work that produces legible signals of productivity. Theater provides those signals at a lower cost than operational excellence.
For Individuals: Role Boundaries Reduce Accountability
Theater provides clear, defensible role definitions. Theater provides clear, defensible role definitions with explicit handoff points. Each party executes its specialized function within narrow boundaries and transfers responsibility at the interface. No single role owns the complete system. When the system fails in production, accountability diffuses across the handoff gaps between roles. No single actor is responsible for the end-to-end outcome because the script was never written that way. The system rewards staying within lane boundaries. Attempting end-to-end ownership means accepting accountability for failures in domains outside your formal expertise and control. Theater is safer.
An Antidote is to Rewrite the Incentives
Changing the performance requires changing what the organization measures and rewards. Leadership must signal that operational health metrics carry more weight than launch announcements. Performance reviews must credit engineers for the reliability work that prevents incidents, not just the feature work that generates them. Role definitions must create accountability for end-to-end outcomes, not just specialized subtasks. When the incentive structure rewards operational capability over performative signaling, theater becomes both unnecessary and unrewarded. For a structured framework to audit your current system and define new operational metrics, The MLOps Reality Gap: A Practical Maturity Assessment provides the foundation.


Thanks Hodman, this is such a great reminder to those of us working in institutions and companies that are obsessed with virtue signalling rather than actually developing effective strategies and governance for how we should be using machine learning.
This article comes at the perfect time. You've really nailed this. It makes me think of Pilates – it's so easy to just look like you're doing a pose perfectly, rather than truly engaging the core. Your point about signalling transformation over actual capability resonates. Frustraiting, but your incentive analysis makes perfect sense.