Beyond the AI doom cycle: finding the productive middle ground

A recent AI Daily Brief episode identifies the 'AI Doom Cycle' — the emotional whiplash from skepticism to mania, through job-loss panic, and finally to a more grounded reality. This cycle can distract businesses and lead to poor, reactive decisions. The most productive conversations and actions around AI happen when the panic subsides, allowing for a focus on specificity, constraints, and practical application. For business operators, understanding and managing this cycle is key to successful AI adoption.

The AI Daily Brief podcast recently coined the term “AI Doom Cycle” to describe the predictable emotional and narrative arc that follows major AI developments.

Instead of a linear progression, the public and business conversation tends to swing between extremes. The episode argues that the most valuable work happens only after the initial waves of hype and fear have passed.

The four stages of the AI doom cycle

The cycle isn't formal, but it generally follows a pattern that will feel familiar to anyone tracking the technology.

From theory to reality

The podcast points to several real-world examples that feed this cycle:

• Executive Whiplash: High-profile leaders, like Citadel's Ken Griffin, can publicly pivot from skepticism to massive investment, amplifying the sense of mania and urgency.
• Enterprise Friction: The headlines suggest instant adoption, but the reality for most large organisations is that integrating AI into legacy systems and established workflows is slow, complex, and fraught with friction. This friction, while frustrating, is part of the shift to a grounded reality.
• Conflicting Signals: The market receives mixed messages simultaneously. While some companies announce AI-driven layoffs (fuelling panic), leaders like NVIDIA's Jensen Huang and OpenAI's Sam Altman discuss the immense compute and infrastructure requirements, highlighting the very real constraints on widespread deployment.

For business owners and operators, getting caught in the AI Doom Cycle is not just a distraction; it's a significant business risk.

The cost of reactivity

Reacting to the peaks of mania and troughs of panic leads to poor decision-making. You risk:

• Wasted Investment: Pouring money into ill-defined projects during the mania phase, driven by a fear of missing out rather than a clear business case.
• Damaged Morale: The constant swing between "AI will make us 10x more productive" and "AI will replace our jobs" creates anxiety and uncertainty among your team, hindering the very productivity you seek.
• Strategic Whiplash: Shifting priorities every time a new model is announced or a dire prediction is made prevents you from making steady, incremental progress on a coherent AI strategy.

The productive plateau

The real opportunity lies in the 'Grounded Reality' phase. This is where the headline-chasing stops and the practical work begins. In this phase, the conversation shifts from what AI could theoretically do to what we can specifically and profitably make it do today.

This is particularly relevant for the adoption of agentic AI. The concept of autonomous agents that can perform complex tasks is a powerful driver of both mania and panic. However, their practical, reliable implementation depends entirely on understanding their constraints.

The best AI conversation starts when panic gives way to specificity, constraints, and agency.

The friction of enterprise adoption is not a sign of failure, but a necessary filter. It forces organisations to justify projects with a clear return on investment, integrate tools with existing security and data governance policies, and train staff to work with the technology, not just be replaced by it. Resisting the cycle allows you to operate in this productive zone consistently, regardless of the media narrative.

As a business leader, your role is to act as a buffer against the cycle, keeping your team focused on value creation. This requires a deliberate, methodical approach.

1. Educate your leadership team. Make them aware of the Doom Cycle pattern. When you can name the phenomenon, you're less likely to be swept up in it. Agree to use a shared, sober framework for evaluating AI news.

2. Filter the signal from the noise. Don't let your strategy be dictated by social media or tech headlines. Assign a person or a small, cross-functional team to be the organisation's 'AI sensor'. Their job is to track developments and report back on what is actually relevant to your specific business problems and strategic goals.

3. Start with problems, not solutions. The most common mistake is asking, "How can we use agentic AI?" The better question is, "What is our most time-consuming, repetitive, and error-prone internal workflow?" Once you have a clear, well-defined problem, then you can evaluate whether an AI tool or agent is the right solution.

4. Run small, scoped pilots. Before trying to automate an entire department, identify a single, narrow task. For example, can you use an AI tool to categorise and route the first 10% of customer support emails? Can you automate the generation of weekly sales report summaries from structured data? Run a pilot with clear metrics for success: time saved, accuracy improved, or cost reduced. This builds internal knowledge and demonstrates value without massive upfront risk.

5. Create your own build, buy, or borrow analysis. You don't need to build a foundational model or a complex agent from scratch. For most businesses, the answer will be to:

   • Buy: Adopt off-the-shelf SaaS products with AI features.
   • Borrow: Work with an external consultancy or agency for a specific, high-value project.
   • Build: Only consider custom development for capabilities that are truly core to your competitive advantage and cannot be bought.