4 Questions To Help You Translate AI Theory Into Practice
“AI for the Enterprise: The Playbook for Developing and Scaling Your AI Strategy” is a new ebook I wrote for The New Stack with sponsorship from Red Hat and Intel. It walks organizational leaders through how to establish and communicate a cross-organizational AI strategy, and then how to enact and measure it all.
The book balances experimenting with AI for the biggest wins and getting the proven — and yes, often boring — AI use cases in place. It helps engineering teams set up the guardrails that can finally break down data silos and patch up security holes before you consider your agentic AI future. And it teaches you how to measure it all.
This book is intended to help you start work conversations. And, at only about 50 pages — and free, so on-budget! — It makes a perfect office book club subject, because these are topics that everyone in your organization should be talking about.
Get the Conversation Started
Since AI will soon touch all roles in an organization, you should aim to make your book club as cross-functional and cross-departmental as possible.
I had a conversation with Gemini about the book, and together we developed a series of questions that can help guide you through the four chapters. These questions are designed to help anyone translate the book’s concepts into reflections, discussions and then concrete internal plans.
The book offers a framework for building an intentional, top-down (read it before you fight me on it) program. But if you don’t think your C-suite is considering these things, these questions can help empower you and your team to not only do AI in an intentional way that enhances your work, but to help you start the right conversations with leadership.
Thank you for reading and sharing my book! Don’t forget to let me know what you think and what questions it sparks for you and your team.
Chapter 1: Set Your Organization Up for Success
This chapter is all about establishing clear direction, cross-organizational leadership buy-in — not just tech! — and a coherent corporate vision that all staff and stakeholders understand.
Have you read your organization’s AI strategy? Do you even know if your executive leadership has one?
Activity: Write a one-paragraph internal communication that clearly outlines your company’s AI priorities in a way all employees can understand. If you’re not in the C-suite, draft a paragraph you’d like your company’s leadership to write — and share it with them, if you dare!
Discuss your paragraph with other members of your book club. What are the differences? Can you combine your ideas into one even more powerful paragraph? (Spoiler alert: Corporate messaging must position AI as an augmenter of work, an automator of the tedious — not an excuse for layoffs.)
Chapter 1 also discusses the disconnect between engineering leaders and developers regarding AI’s impact on productivity. What concrete, cross-functional mechanisms — like a dedicated experimentation week, a formalized feedback loop and breaking down data silos — could your organization create to close this gap and encourage safe, bottom-up experimentation?
A top-down AI strategy to ensure reliability and security is recommended. Beyond appointing a chief AI officer (CAIO), what are the three specific action items your existing executive leadership team will/should commit to and implement in the next quarter to visibly drive the AI strategy?
Be sure everyone can answer your company’s official stance on using customer data to train AI models and how you will ensure full customer control and transparency in all AI-enhanced products.
Chapter 2: Take the Incremental Approach
The second chapter of the book focuses on placing a big bet on AI solving pervasive, long-standing problems while (most of the time) enacting the proven AI wins.
In this chapter, Red Hat’s Marty Wesley suggests a two-speed AI investment by following the 80/20 rule.
Activity: Each member of your book club identifies and reflects on one proven use case — like internal search documentation, paying down technical debt or customer support response — that you can apply to your own job next week to make it more efficient and less stressful. About 80% of your AI budget should feed such proven use cases.
Then, each book club member should identify a “20% big bet” — a hard, transformative problem, like a large-scale migration — that, if AI could solve, would make a huge difference to your organization. Did you have the same idea or different ideas based on your department? If you are in executive leadership, how would you prioritize the ideas?
For each, discuss the results and talk together about how you will define the measurable success of each of these projects.
Chapter 3: Data Security Backed by Engineering Best Practices
This chapter brings together everything we’ve been recommending for over a decade. It’s just that the urgency changes with AI. As Harness’s Dewan Ahmed recently told me, like a bullet train going past the rail, a lot of AI is automating insecurity at scale.
Activity: Based on the risks outlined in this chapter — data leakage, high-risk systems, security compliance, maybe others that are important to your business or sector — what are the top three high-risk areas in your business that your initial AI policy must immediately consider, vet, define and put guardrails around?
Don’t forget to invite subject matter experts to weigh in on domain- or department-specific AI policies rather than relying on one-size-fits-all rules. For your team or department, name one specific AI use case and the corresponding human-in-the-loop (HITL) process an expert should define to ensure quality and compliance. If you are an executive leader, pair program or embed on these teams and ask them what would help them do their work more efficiently. Experiment alongside them — just make sure they know it’s not a test.
As code rolls out faster than ever, increasingly without a human developer even looking at it, what worries you the most? What are the greatest risks to your reliability, security and stability? When things go wrong, how quickly can you roll things back?
Chapter 4: Future-Proof Your AI Strategy
Seems like every week we have a new winner of the AI GOAT (greatest of all time) award. The reality is that things are changing so fast that no organization can make long-term bets on any AI tooling. Flexibility and measuring everything are the only ways to future-proof in the face of uncertainty. That’s what the final chapter of this book is all about.
The closing chapter introduces the shift to an “agentic future” where AI agents automate workflows.
Activity: Each book club participant should identify one simple, repetitive, pattern-following task in their work that would be a good candidate for an initial AI agent pilot. Once you’ve identified that predictable and boring activity, ask: What observability and context engineering mechanisms must be in place to ensure you can debug the agent’s behavior? Brainstorm together around not only problems but potential agentic solutions.
Then, as you look across your AI strategy and since you can’t improve what you don’t measure, identify two metrics — such as from DORA metrics, the DX AI Measurement Framework or a specific internal cost reduction (please, not lines of code!) — your organization will use to measure the success of its first AI project. Who is the decision owner responsible for determining if the initiative is a success or failure?
This chapter also promotes the value of open source platforms and the hybrid cloud for avoiding vendor lock-in and maximizing agility. How does your current data infrastructure align with this vision, and what is the plan to ensure your AI platform can support the model retraining and redeployment necessary when models inevitably drift?
The final chapter also asks which skills are most in demand in the face of AI. Each team should consider what skills or aptitudes would help it respond to this rapid change. Can that be trained in-house or do you have an opening coming up on your team?
Finally, the book ends by asking you to embrace the opportunity of AI, with caution. Here, I ask you to reflect on its final words:
“AI is an opportunity to take the boring of our day to day and to help everyone do what they do best. AI can be the thing that emphasizes each organization’s differentiation — but don’t believe that AI is that differentiator. Ground your AI enterprise strategy in your people, and then your processes and technology, to get the best out of an AI-enriched future. Good luck!”
Get your free copy of AI for the Enterprise: The Playbook for Developing and Scaling Your AI Strategy today!
The post 4 Questions To Help You Translate AI Theory Into Practice appeared first on The New Stack.
