Enough with AI FOMO: Go Slow-Mo, Says Domo CDO

For months, the drumbeat has been relentless. Every keynote, every LinkedIn post, every earnings call seems to declare the same urgent message: adopt artificial intelligence now, or risk total irrelevance. This pressure creates a low-level hum of anxiety that follows professionals from the C-suite to the intern desk. The fear is real, but according to Chris Willis, chief design officer and futurist at the data platform company Domo, the response to this pressure is often misguided. He proposes a different approach, one that embraces an ai slow mo philosophy.

The Unspoken Resentment Against AI Hype

Willis recently spent time in San Francisco, a city that serves as the global epicenter for companies like OpenAI and Anthropic. Walking those streets, he found himself puzzled by a collective silence. He wondered why more people weren’t openly frustrated with the way this technology has been thrust upon the workforce. “Why aren’t people more resentful that these companies have pushed this technology upon them and now everyone is feeling a tremendous amount of anxiety?” he asked in a recent interview.

He points to surveys showing that nearly everyone, from executives to entry-level staff, feels a ticking clock. The narrative suggests careers are on the line if one does not become an AI expert overnight. This manufactured urgency creates a culture of fear, which Willis argues is a terrible foundation for genuine progress. “Fear,” he stated plainly, “is not a durable strategy for innovating.”

Beyond the billboards along the US 101 corridor celebrating AI’s promise, there is a darker undercurrent. Movements like Stop AI and Pause AI, along with isolated incidents of protest, signal a deep existential dread. But Willis is less concerned with the philosophical debate about human obsolescence and more focused on the immediate, practical damage this fear-based marketing is doing to businesses today.

AI Models: A Product Without a Specification

Willis identifies a core problem that fuels much of the confusion. He describes modern large language models as a product without a spec. In traditional product development, teams define the target user, the specific problem being solved, and—crucially—what the product will not do. The feature specification for a typical LLM, however, is dangerously broad: “It’ll do anything for anyone, anyway, anyhow, in any language.”

This lack of boundaries makes strategic planning nearly impossible. When a tool claims to solve every problem, it often fails to solve any single one reliably. This ambiguity places an enormous burden on company leaders who are pressured to innovate with a technology that remains poorly understood. The result is a flurry of activity that looks impressive but delivers little substance.

Organizations are spending heavily on AI subscriptions and API access, expecting innovation to simply materialize. Willis notes that this is not how innovation works. You cannot buy a powerful engine, bolt it onto a vehicle you don’t understand, and expect to reach a destination safely. The missing piece is a clear, human-defined strategy.

The Impatience Problem, Not an Innovation Problem

What company leaders actually face, according to Willis, is an impatience problem. The pressure to act immediately leads to what he calls “AI theater.” Companies feel compelled to show stakeholders that they are doing something, even if that something lacks a clear purpose. This performative adoption creates a veneer of progress while the underlying business processes remain unchanged.

This impatience manifests in a phenomenon known as “tokenmaxxing.” This term describes the behavior of buying access to AI models and encouraging employees to use them as much as possible, simply to justify the expense. Willis explains that this approach feeds the narrative of active engagement but fails to move the needle on actual business outcomes. Employees might feel personally more productive, generating summaries or drafts faster, but the company’s bottom line remains stagnant.

Tokenmaxxing is a convenient way to show activity without demonstrating strategic value. It avoids the hard work of identifying where AI can actually make a measurable difference. The effort is spent on volume, not on impact.

Treating AI as a Solution, Not a Tool

The deeper strategic error, Willis argues, is that companies treat AI itself as the solution. They look at a business challenge and ask, “How can we apply AI to this?” instead of starting with a more fundamental question: “What specific process needs to change, and what is the best tool for that job?”

This subtle reframing is critical. AI is an incredibly powerful engine, but without the proper context, it can accelerate a business in the wrong direction. Willis uses a vivid analogy: if you do not understand your existing workflows and automations, you risk installing a very powerful engine that drives your business faster, but with the lights off, at night. You move quickly into unknown territory without seeing the obstacles ahead.

A real-world example illustrates this point. The financial technology company Klarna famously replaced a large portion of its customer service team with an AI chatbot. The move generated headlines and was celebrated as a victory for automation. However, reports later indicated that the company quietly reversed course, rehiring human agents to handle complex queries. The lesson is clear: no customer ever just wants to talk to your chatbot. They want their problem solved efficiently. Sometimes a human is the best tool for that job.

Starting with Business Needs First

Willis suggests a more grounded approach. Instead of setting moonshot goals for enterprise-wide AI transformation, leaders should start with something simple. They should look at a spreadsheet. What repetitive, manual process lives there? Perhaps it is reconciling invoices, checking for discrepancies, or updating inventory logs.

This is where the ai slow mo methodology shines. By slowing down, you can identify a narrow, well-defined task. You can then test whether an AI model can perform that task reliably, with verifiable accuracy. This creates a small, measurable win. It builds confidence and institutional knowledge without the risk of a catastrophic failure.

Understanding where human judgment is required is another key step. Some decisions can be automated and verified by a simple rule. Others require context, empathy, and ethical reasoning. Failing to distinguish between these types of tasks invites serious problems. A model might process a refund request correctly, but it cannot understand the nuance of a frustrated customer’s personal situation.

The Coming Reckoning with AI Budgets

Willis predicts that a major reckoning is coming for corporate AI spending. Chief Financial Officers are beginning to ask hard questions. They want to see a return on investment, not just an increase in token usage. When the budget cycle arrives, leaders who invested in AI theater will struggle to justify the expense. Projects that were approved based on hype will be scrutinized for their actual contribution to the bottom line.

This financial pressure will force a shift. Companies will need to do the hard work of understanding how AI may or may not be useful for a desired outcome. The era of buying a subscription and hoping for magic is ending. The new era demands discipline, measurement, and a clear line between cost and value.

For example, a marketing department might spend thousands of dollars per month on an AI content generator. If that content does not drive traffic or conversions, the tool is a cost, not an investment. The CFO will notice. The conversation will move from “we are using AI” to “what measurable business metric did this AI improve?”

You may also enjoy reading: Arkansas vs Texas Tech Prediction: Stats Preview for Revocruit Rematch.

Building Durable, Deployable Systems

The result of the current rushed approach is a graveyard of proof-of-concept projects. These are small experiments that work in a controlled demo but fall apart when exposed to real-world data and scale. They lack the durability, trustworthiness, and security required for enterprise deployment. Willis argues that taking an ai slow mo approach actually increases the likelihood of successful scaling.

By moving deliberately, teams can build systems that are robust. They can implement proper testing, validation, and human oversight. They can document the process so that it can be repeated and improved. This methodical pace is the opposite of the frantic “move fast and break things” mentality that has dominated tech culture for years.

A slow approach also allows for better risk management. AI models can produce biased, incorrect, or harmful outputs. Rushing a model into production without safeguards is a liability. Taking the time to audit outputs, establish guardrails, and create escalation paths for failures is not a luxury; it is a necessity for any responsible organization.

Practical Steps for an AI Slow Mo Strategy

For leaders feeling the heat to act, Willis offers a practical roadmap. It begins with a single, honest assessment of your current processes. Map out the workflows in your department. Identify the bottlenecks, the repetitive tasks, and the points where errors frequently occur. Do not look for AI solutions yet; simply understand the landscape.

Next, pick one small, low-risk task. It should be a task where failure is not catastrophic. A good candidate is a data entry validation step or a simple categorization task. Test whether an AI model can perform this task better, faster, or cheaper than the current method. Measure the results rigorously.

Finally, build a feedback loop. How do you know the AI is performing correctly? What happens when it makes a mistake? Who is responsible for correcting it? These operational questions are more important than the technical ones. A successful AI deployment is 20% model and 80% process, governance, and human training.

The Role of Human Judgment

One of the most overlooked aspects of the AI conversation is the irreplaceable value of human judgment. Willis emphasizes that understanding when a human needs to be in the loop is critical. Some decisions require empathy, ethical consideration, or an understanding of context that no current model possesses.

Consider a loan application. An AI can process financial data and flag risky profiles based on historical patterns. But it cannot understand that an applicant had a medical emergency that temporarily impacted their credit score. A human loan officer can make that judgment call. The AI serves as a powerful assistant, not the final authority.

This is the essence of using AI as a tool. The goal is not to replace humans but to augment their capabilities. By offloading routine, high-volume tasks to machines, humans can focus on the complex, creative, and relational work that drives true value.

Escaping the Fear Cycle

The current market is driven by fear of being left behind. This fear is manufactured by companies that benefit from rapid, uncritical adoption. Willis encourages leaders to step off that treadmill. The world will not end if you take three months to properly evaluate a use case. In fact, you will likely be ahead of competitors who rushed and failed.

The ai slow mo philosophy is not about rejecting technology. It is about respecting it enough to use it wisely. It is about understanding that a powerful engine requires a skilled driver. It is about building a foundation that can support long-term, sustainable growth rather than a short-term spike in activity.

Willis’s message is ultimately optimistic. By removing the pressure to perform AI theater, companies can actually achieve meaningful innovation. They can build systems that are trustworthy, scalable, and aligned with their core business goals. The path forward is not to sprint blindly into the unknown, but to walk deliberately, with eyes wide open, examining each step before taking the next.

The snail on the coins may seem slow, but it is moving steadily forward. In the race for genuine business transformation, that steady pace may be the winning strategy.