We live in both incredible and absurd times, a paradox where technological marvels coexist with profound inefficiencies. This growing tendency to deploy global systems, leveraging continuous compute across multiple vendors, to solve problems that were already solved locally, or even manually with trivial effort, is a phenomenon that warrants serious discussion. It’s a testament to human ingenuity that such complex systems are even possible, yet it also highlights a potential lack of discipline in applying that ingenuity.

To be clear, my stance is not anti-automation. On the contrary, I am very much in favor of automation and agentic AI. I am actively educating myself with agentic AI courses to keep up with the times, explore new frontiers, and leverage the latest capabilities. In many cases, these technologies are genuinely transformative to businesses and consumers alike. Especially at scale, in repetitive processes, in data-heavy environments, or in cases where accessibility matters, AI agents do unlock real, undeniable value. They can accelerate research, streamline customer service, optimize supply chains, and empower individuals with disabilities, truly enhancing productivity and quality of life.

However, not every problem belongs in that category. And I feel an increasing number of AI-based applications and workflow automations tend to fall squarely into the "shower vent" category—solutions looking for problems, or, more accurately, over-engineered solutions for non-problems. The allure of the cutting edge often blinds us to the simpler, more robust alternatives.

You may think this isn’t an issue: What does it matter if we bring the tech revolution to solve ridiculous tasks, just because we can? The prevailing sentiment often suggests that any automation is good automation, a net positive for progress. But this uncritical embrace carries significant drawbacks and inherent risks, often hidden beneath the surface of convenience and perceived innovation. These risks, if not properly understood and mitigated, can undermine the very benefits automation promises.

Three Risks of Automating Without Discipline

The automate-everything ethos, driven by the sheer capability of modern technology, can lead to substantial challenges across operational, economic, and even environmental and strategic domains.

1. Operational Risk: More Points of Failure, Less Control

The seemingly simple command to turn on a shower vent, while a personal convenience, illustrates a profound operational vulnerability. That single voice command depends on a sprawling, interconnected chain of systems working in perfect sync: your device’s microphone and processing unit, your local Wi-Fi network, your internet service provider, submarine fiber optic cables, Google’s vast global infrastructure, potentially various third-party vendor clouds for specific smart home integrations, and finally, the smart switch itself. If any single layer in this intricate architecture fails—a network outage, a server glitch, an API hiccup, or even a software bug—the entire system fails. The simple act becomes impossible, demonstrating a fragility inherent in over-automation.

This same pattern is emerging, often on a much grander and more critical scale, in agentic AI workflows. These advanced systems are typically built as multistep pipelines, orchestrating interactions across multiple Large Language Models (LLMs), specialized orchestration tools, and numerous external APIs. Each LLM call, each API integration, each data transfer between components introduces additional dependencies and layers of complexity. Debugging such a distributed, heterogeneous system becomes a nightmare, often involving multiple vendor support teams pointing fingers at each other. The more interconnected the system, the more potential points of failure, and the less direct control an individual or organization retains over the end-to-end process. This creates significant technical debt, making maintenance, upgrades, and troubleshooting disproportionately expensive and time-consuming.

Consider another example from my personal life: When my parents acquired their home, they built it as a “smart home,” replete with advanced lighting, climate control, and security systems. It worked great, for a while. Then, a seemingly innocuous “smart light switch” malfunctioned. What would have been a $5 DIY fix with a standard switch transformed into a crisis. The smart home company quoted $1,500 to send a special “smart home engineer” to diagnose and repair what was, at its core, a simple electrical component. This is equivalent to an enterprise needing to hire highly specialized AI engineers and automation experts to support a workflow that, in a non-automated context, could have been handled by a junior, nontechnical person in 10 minutes. The reliance on proprietary, complex, and interconnected systems creates vendor lock-in and dramatically inflates the cost of even minor repairs or modifications, leading to unexpected downtime and frustration. The cost of maintaining an overly complex automated system can quickly eclipse the initial perceived benefits.

2. Economic Risk: Hidden and Compounding Costs

Voice commands and basic AI workflows often feel inexpensive at a small scale. A single query to Google Assistant or a few API calls to an LLM might cost fractions of a cent. However, this perception of cheapness is deceptive, as these operations rely on extensive and paid infrastructure. The costs accumulate through compute resources (CPU, GPU, memory), network egress charges, API call fees (often per transaction or per token for AI models), specialized software licenses for orchestration layers, monitoring tools, and vendor integrations.

In many cases, especially when implemented for those “ridiculous” or low-value tasks, the cumulative cost of automation can quickly approach, or even exceed, the value of the task being automated. Organizations often fail to conduct a thorough cost-benefit analysis before deploying automation, mesmerized by the promise of efficiency. The actual total cost of ownership (TCO) for an automated system includes not only the direct infrastructure costs but also the expenses for development, deployment, ongoing maintenance, monitoring, security, and the highly skilled personnel required to manage these complex systems. When automating a task that occurs infrequently or requires minimal human intervention, the ROI can be negative. We must ensure that we invest in AI and automation where it makes clear economic sense, where the value created demonstrably outweighs the comprehensive costs, rather than automating for automation’s sake. Otherwise, we risk significant capital expenditure and operational budgets being diverted to projects that yield negligible or even detrimental returns.

3. Environmental and Strategic Risk: Scaling Inefficiency

The environmental footprint of our digital world is growing at an alarming rate. Data centers, the physical backbone of our automated future, are voracious consumers of electricity. They create hundreds of millions of tons of CO₂ emissions annually, a figure estimated to grow to a staggering 2.5 billion tons of CO₂ emissions by 2030, according to Morgan Stanley. AI, with its insatiable demand for high-performance computing, particularly GPUs for training and inference of large models, is becoming a rapidly growing percentage of that energy consumption. These are not trivial figures; they represent megatons of CO₂ emissions, a significant contributor to climate change, and the trend is only accelerating.

While each small agentic AI workflow might account for only a few grams of CO₂ emissions, at scale, these individual inefficiencies compound into a massive environmental impact. A million shower vent commands, or a billion trivial data transfers, translate into real energy consumption and carbon output. This “death by a thousand cuts” scenario highlights a critical strategic issue: optimizing for the sake of it, rather than optimizing for meaningful impact. This mindset can mean we often lose focus on solving truly meaningful, high-impact problems that genuinely move the needle for humanity or business.

Beyond the environmental toll, there’s a strategic erosion. By dedicating significant resources—talent, capital, and intellectual energy—to automating tasks that offer marginal gains, we inadvertently divert attention from challenges where AI and automation could be genuinely transformative. This creates an “innovation theater” where the appearance of technological advancement trumps actual strategic value. Furthermore, an over-reliance on opaque, complex automated systems can lead to a deskilling of the human workforce in areas where critical thinking, intuition, and human judgment remain superior. It raises ethical questions about accountability when complex AI systems make decisions without transparent reasoning. The ultimate strategic risk is fostering a culture that prioritizes technological solutions above all else, potentially overlooking simpler, more resilient, and more sustainable alternatives, and losing sight of the fundamental problems we are trying to solve.

The lesson from the shower vent, and countless other examples, is not to reject automation, but to embrace it with discipline, discernment, and a clear understanding of its true costs and benefits. We must ask ourselves: Is this automation truly necessary? What are the alternatives, including manual processes? What are the comprehensive operational, economic, and environmental costs? What is the net benefit, not just the perceived convenience? Only by asking these critical questions can we ensure that our technological progress serves humanity and the planet, rather than merely indulging our capacity to automate everything, just because we can.

Itay Sagie is a strategic adviser to tech companies and investors, specializing in strategy, growth and M&A, a guest contributor to Crunchbase News, and a seasoned lecturer. Learn more about his advisory services, lectures and courses at SagieCapital.com. Connect with him on LinkedIn for further insights and discussions.

Illustration: Dom Guzman

Stay up to date with recent funding rounds, acquisitions, and more with the Crunchbase Daily.