Below is an introduction article for Double Harvest, authored by me (published April 2021 -- written in Japanese). The article aims to convey why this book is compelling and how it can transform the way we think about AI in business—so much so that you’ll want to pick it up and read it.
Why Do We Need “AI × Strategy” Now?
In recent years, the phrases “We need AI!” or “We must pursue DX!” have echoed in every corner of the business world. Yet many professionals still hesitate at the threshold of AI adoption.
They’re asking:
Where do we start?
How exactly can AI help our business?
Will our employees be able to handle it?
These questions often stem from the perception that AI is distant or too technical. There may be no in-house AI engineers or enough know-how to emulate big-case successes. As a result, many companies remain paralyzed or unsure whether AI truly applies to their business.
I countered that AI fails to move forward precisely because we treat it as a technology alone. In Double Harvest, he emphasizes that AI is no longer just a “tool” for automation or cost reduction, but rather a strategic design element—a catalyst for building a system in which you can win and keep winning in the new business landscape.
What Is “Double Harvest”?
The central concept in this book is what Hotta calls the “Double Harvest Loop.” This refers to designing multiple, layered “feedback loops” that leverage AI to create lasting competitive advantage—rather than a one-time efficiency gain.
AI’s unique strength lies in its capacity to learn and improve over time. Simply installing AI to automate a process is not enough. Without an ongoing cycle that feeds it fresh data and captures the insights for sustained improvement, your advantage can fade quickly. That’s why the “harvest loop”—a framework for continuously gathering real-world data, training AI on that data, and reinvesting AI’s output back into your business—is so crucial.
But it doesn’t stop at one loop. By constructing a double or even multiple loops, you can build serious entry barriers that are extremely difficult for competitors to replicate. These overlapping loops strengthen the business, ensuring that every piece of user or operational data further refines the AI, boosts customer value, and fortifies your position in the marketplace.
The Five “End Values” That AI Can Deliver
One of the powerful features of Double Harvest is its straightforward breakdown of “What exactly does AI enable?” Hotta frames this via five End Values:
Increasing Revenue
Reducing Costs
Risk / Loss Prediction
Improving UX (User Experience)
Accelerating R&D
Many organizations jump into RPA or a recommendation engine and see partial success—like a 10% improvement here or a 20% gain there—but then stall. The missing piece is a strategy that positions AI at the core. Instead of seeing AI as a single, standalone tech for cost-cutting, you harness it to continuously collect and refine data that, in turn, spurs new layers of innovation.
Remember “Human in the Loop”
Any talk of AI often stirs worry—What about jobs and accuracy? Hotta dispels the myth that “AI is worthless if it’s not 100% accurate.” The book details frameworks such as Human in the Loop, Expert in the Loop, and User in the Loop to illustrate how humans and AI can collaborate:
The AI covers the bulk of repetitive or low-level tasks at, say, 70–80% accuracy.
Humans fill in the remaining 20–30% where critical judgment is needed, or as an ongoing check.
Because humans correct the AI’s mistakes, the AI’s accuracy steadily climbs—benefiting both productivity and the employee’s work, freeing time for higher-level tasks.
In other words, if the AI has 70% accuracy, that’s not “useless.” It can still take a big load off routine work. Meanwhile, the AI learns from human feedback, eventually boosting accuracy to 80%, 90%, and beyond. This is how you “move from partial automation to a virtuous cycle of continuous improvement.”
Monetization: It’s Not Just About “Building AI” Once
The harvest loop concept does more than power your internal processes—it creates new monetization avenues. One famous example analyzed in the book is Mobileye, whose onboard car cameras capture vast amounts of real-time driving data. As more drivers and vehicles adopt Mobileye, the system continually refines its AI for collision avoidance. Simultaneously, that driving data is fed into incredibly detailed mapping, which fuels an entirely separate line of competitive advantage—giving Mobileye a double loop that helps them dominate the field.
Another case study is the Fave payment app. It started by gathering purchases data (who buys what, when, for how much), harnessing that data to boost personalized offers for customers—and then used those same insights to build “credit scores” for merchants. In partnership with banks, those merchants can secure favorable financing. So Fave fosters further store growth, generating more transaction data, which in turn amplifies both loops. That’s the double harvest effect: weaving different AI loops together so that your position grows stronger in ways competitors can’t easily copy.
A Practical, Step-by-Step Roadmap
One major pitfall in AI adoption is treating it like normal software development with rigid “waterfall” processes. The book offers a nine-step approach to AI implementation—from preparing initial data, to running proofs of concept, to user interface design, to continuous quality checks—helping readers see the entire landscape of an AI project.
Hotta emphasizes the difference between “software dev” and “AI dev”:
AI involves uncertainty in the early stages. You can’t know for sure if you’ll reach 90% or 98% accuracy after X days.
You must iterate and remain flexible, adjusting your goals as you discover constraints and new possibilities.
The harvest loop approach ensures every incremental improvement helps the AI get even better.
He calls this the “Expectation Sandwich” approach: you set minimum required outcomes but also keep an optimistic upper bound for what the AI could achieve. Through short development cycles (iterations), you narrow the gap until you find the sweet spot and gain a workable, continuously improving solution.
The Bigger Picture: Why (Purpose) and How (Strategy) Must Connect
In the final chapters, Hotta reminds us that AI, however powerful, remains a means to an end—and that end must align with your purpose (why you do business). This is especially critical as we move beyond superficial efficiency to creating truly transformative user experiences or even entirely new markets:
Which user experience do you want to deliver, and why does it matter?
How do you bake AI into a blueprint that defends and strengthens that experience?
How can you ensure your business continuously reaps new value from AI to outpace competitors?
The core argument is that AI thrives when wedded to a well-defined purpose—one that gives your team a guiding star through inevitable market shifts. With a clear purpose and a properly designed harvest loop, your business can not only implement AI but sustain a unique advantage that keeps you winning.
Conclusion: Vision and Systems Matter More Than Ever
Double Harvest makes a compelling case that AI is bigger than cost savings or one-off automation. By harnessing feedback loops—where real-world usage generates data that trains AI, which then enriches your offering—it’s possible to achieve long-term, self-sustaining growth. Hotta emphasizes that “AI alone” or “adopting one new technology” rarely leads to victory. Instead, it’s the combination of strategy + AI + human collaboration that sparks ongoing innovation and competitiveness.
For anyone serious about “AI done right,” Double Harvest provides the frameworks, real-world examples, and managerial insights to launch (and keep growing) AI-driven business. Its bold message: Don’t just adopt AI—design your AI strategy to scale, generate continuous data, and power an ever-stronger feedback loop. Done well, this double-harvest system becomes an unassailable stronghold in a rapidly shifting digital age.
If you’re looking to reinvent your business model, excel at DX efforts, or nurture a future-oriented corporate culture, this book is an invaluable, illuminating guide. After reading, you’ll likely find that your vision for AI in your organization has become bolder, clearer, and far more actionable.
Embrace the double harvest. Once you turn that loop on, there’s no limit to how far your business can grow and how long it can keep winning.
Comments