📕The AI playbook I wish I had 2 years ago

Hi ,

After leading 100+ real-world AI projects across industries, I’m thrilled to finally share something I’ve been working on for the past year, my new book: 📕Do More With Less: The AI Playbook for Amplifying Talent and Output.

This book is a no-nonsense, field-tested guide for anyone in an organization - from individual contributors to CEOs - who wants to harness Generative AI to boost productivity, streamline workflows, and drive real results.

If you're wondering where to start with AI at work - or how to scale beyond ChatGPT experiments - this book gives you the playbook.

Inside, you'll find:

• A simple 4-step framework for rolling out AI in your org

• Dozens of examples from real deployments

• Templates, prompts, and pitfalls to avoid

• A one-page plan to align your team.

I also cover:

• A bottoms-up explanation of how Generative AI works

• The emerging shape of jobs, based on the strengths and weaknesses of Generative

Whether you're in innovation, operations, R&D, or IT, this book will help you and your team do more with less - without the hype.

Available May 31st

While the book was designed for print, we’ve created a special reflowable Kindle edition that preserves the core content. The Kindle edition is available now for a limited time at just $0.99.

The paperback and hardcover editions will be available at launch.

I wrote this for leaders like you- people who want to move fast, act smart, and stay grounded in what actually works.

I’ve included the first chapter from the book - an executive summary - below.

Thanks for being part of this journey.

Best,

Dino Gane-Palmer
CEO, PreScouter & Auxee
Author, Do More With Less

P.S. If you have questions or thoughts on the book, I’d love to hear them. Just hit reply.

Do More With Less: The AI Playbook

Book Executive Summary

Why organizations have struggled with Generative AI.

Recurring intellectual workflows - such as drafting project reports, monitoring competitor pricing, or triaging customer complaints - consume hours of staff time each week. These mind-numbing activities are exactly the kinds of tasks Generative AI should be able to alleviate. Yet many organizations struggle to get Generative AI working for use cases such as these, where time and money is already being spent and the potential for tangible ROI best exists. Instead, Generative AI use in organizations is largely limited to the use of chatbots.

In the meanwhile, every day, we see headlines about astounding new AI capabilities. From mimicking human conversation to generating code and videos - these new developments seem miraculous and hard to keep up with.

But at its core, all Generative AI capabilities are powered by a simple concept: autocomplete. Think of how Gmail offers you suggested completions, with the press of tab key accepting the completion. That’s Generative AI in action. 

The underlying technology for these completions - the Transformer algorithm discovered in 2017 - has simply been scaled over the past few years to auto-complete a wider variety of data types, in larger quantities at faster speeds.

Generative AI models have gone from auto-completing text conversations, to auto-completing “thinking” steps before providing a response to auto-completing “tool calls” to run web searches and run code before providing a response. 

We can expect in the future to have our heads spun by Generative AI models autocompleting large, sophisticated software applications, full movies, 3D worlds, and much, much more.

Nevertheless, there have been no fundamental breakthroughs since 2017.

Ultimately, these Generative AI models are taking some input and autocompleting vast amounts of output that traditionally may have taken even months or years to create. 

The same Transformer architecture from 2017 continues to be scaled and fine-tuned, but the technology’s core limitations - limitations that matter deeply in real-world business settings - remain unaddressed:

• Generative AI models only develop true “understanding” during multi-month training cycles, unlike humans who learn in real time.

• They rely exclusively on public data, leaving them blind to the nuanced, undocumented expertise within organizations.

• Their outputs are non-deterministic, often varying unpredictably even when given the same prompt.

Most demos are amazing because there’s no real context required. When you ask a model to generate an image of a cat astronaut riding a unicorn, there are no constraints. It's a creative autofill.

In the real world, business problems don’t live in a vacuum. They require deep context: background documents, historical data, constraints, goals, formats, and organizational nuance. 

Almost no progress has been made in systematically shaping the context and inputs provided to the Generative AI models for them to easily take on real world work tasks. 

Indeed, the pursuit by big tech companies towards shaping this technology as a new species - through chatbots and autonomous agents - has distracted everyone from the truth: the entire value of the output of a Generative AI model is dependent on the inputs provided to the model.

What is needed, to apply this technology to real-world problems, is then to build systems that shape the input, so the model has everything it needs to “autocomplete” that use case successfully. 

Thus far, building such systems has been left to the domain of software development teams, but this need not be the case. Anyone with Excel can build a system to which they can offload several hours of recurring work each week.

How to get Generative AI working 

While Generative AI models can produce impressive demos, integrating them into real operational workflows requires serious work. It’s not just about the model - it’s about the system around it. That system must feed the AI with the right context and knowledge, using prompts refined through iteration, and include infrastructure to guide, govern, and scale its use.

This book is about how to build that system.

This book focuses on applying Generative AI to internal, recurring activities - those that either drain valuable time and energy from your team or fail to be executed with consistent quality. When implemented well, AI becomes a force multiplier: enabling staff to move beyond repetitive tasks and become high-leverage experts, using AI to amplify their impact. The result? Doing more with less.

In this book, I’ve distilled what we, at PreScouter, have learned from working on hundreds of client projects each year. Our cross-industry perspective gives us a unique vantage point to see how Generative AI can be applied to real business use cases. The 8-Hour AI Playbook presented in this book distills these insights into a step-by-step playbook for successful Generative AI implementation:

• Identify: Pinpoint high-friction, recurring workflows where employees spend time on manual research, synthesis, or repetitive work. These are prime candidates for AI augmentation.

• Investigate: Understand the task deeply by studying how people currently perform it, what information they use, and what outputs are valued. This insight is critical to guiding how the AI system is built.

• Instruct AI: Design prompts and data flows that guide the AI to perform the task. Conduct testing to ensure the AI reliably mimics the thinking process of the expert whose work it is amplifying.

• Introduce: Deploy the solution into daily work through onboarding, SOP updates, and training. Ensure teams understand when and how to use the AI tool, and track performance to drive adoption and improvement.

Though much of this process may seem like common sense, the reality is in the nuance. Each chapter includes real-world tactics, strategies, and even email templates we’ve used in the field—because execution is where success happens.

Historically, expert knowledge workers were constrained by their time. Their output was limited by the hours they could dedicate to each task. But partnered with an AI system, such as the one you’ll build through this book, that’s changing. AI does for knowledge work what the industrial revolution did for physical labor: shifting jobs from that of doing the work to instead designing the work for machines to undertake.