It’s 2036. A classroom of 123 university students attends a fully virtual session. Holographic avatars fill a beautiful, white, translucent lecture hall, mimicking a meeting space in the Vatican. At first glance, it appears that all students are attending, but that’s an illusion. Sixty-five percent of the avatars are actually AI-powered notetakers, designed to mimic student behavior and generate concise lecture summaries. The remaining 35 percent are operated by students who are alert, engaged, and participating in real time. Professors have begun referring to the AI-generated ones as “empty vessels”: avatars with no conscious presence behind them. The difference is subtle but significant. In this era, engagement is no longer defined by being online, but by being aware.

Using an avatar to attend class without a human behind it is illegal, but the technology is difficult to detect. After the lecture, many students, regardless of how they showed up, rely on AI-generated summaries from the notetaking avatars, feeding them into large language models that appear in their homes through AR contact lenses. Each LLM carries a custom voice, style, and personality the student has designed. With a prompt to reference an award winning essayist persona, the LLM merges that with the student’s minimal input to complete the task. It recites the essay aloud, and the student submits it. All of them pass. Automate and repeat.

As of today, capturing and acting on the best ideas quickly has always been a deeply desired outcome. Executives and educators have spent countless hours studying methods of boosting the human imagination to excel at idea generation and execution. With AI, this becomes amplified. The blank canvas is non-existent. Type a well written prompt, and a world of knowledge materializes at your fingertips in seconds. While many view AI as merely a tool, it fails to capture its true impact; AI serves as our executive idea assistant and collaborator. The way we envision our relationship with it profoundly impacts the outputs we receive.

We are in an era where AI can fuel ideas faster than humans, connect disparate concepts across industries, and polish outputs at speeds we could barely imagine even a week ago! Yet with this new bottomless cookie jar comes a critical question: Will creativity and originality thrive, or will we lose something irreplaceable in the rush for efficiency?

The classroom of 2036 is not a distant dystopia, but a mirror of emerging trends that demand new tools for intentional learning design.In this chapter, I introduce the FINDS framework—a practical blueprint for educators, curriculum designers, and organizational leaders to harness AI thoughtfully. Because the future of the ideas we generate depends not just on AI, but on how intentionally we choose to guide it.

The FINDS framework outlines five proven strategies for using AI to enhance creativity and ideation.

The F stands for fuel. AI gives you quick access to ideas. Lots of ideas. These ideas improve with clear and concise prompt inputs and feedback. With LLMs and AI research assistants, you can try on a multitude of idea “hats” for your projects before taking action on one.

Ignite looks at uncovering originality. AI can help generate novel ideas more quickly by conducting in-depth research into concepts like cross-industry innovation, thereby connecting advancements in two seemingly separate industries into a new discovery.

N stands for nurture. AI can be injected into any stage of a project. 1+1+AI=10™ Chapter Coach GPT, envisioned by Elizabeth Ngonzi to assist in the process of writing this anthology, is a great example of that. To ensure all contributors remained focused on the main goal of this effort, we utilized a GPT to maintain alignment. The GPT nurtured human-written words and melded them into the anthology you’re reading.

D stands for defense. Today, executives and educators can literally tap into the other shoe; for example, AI can simulate dissenting voices—helping teams prepare for resistance, strengthen arguments, and refine ideas in advance. AI is an actor we can put to the test of holding a personality. From persona constructions, we can apply the findings to real life pitches.

And lastly, S stands for share. FINDS looks at easily sharing your work and projects in work environments and not just keeping them. Share looks at how this versatile AI assistant can allow us to grow our vision stronger and faster with stakeholders. Now that we’re familiar with each part of this framework, let’s see it in application.

In 2023, a Coca-Cola marketing campaign invited digital artists to “create real magic” with AI from a partnership with OpenAI.

This platform lets any designer generate original marketing visuals using the Coca-Cola brand’s iconic assets. The goal was to rapidly brainstorm and iterate on multiple ad concepts (ie, fuel), moving from idea to content quickly. By embracing generative AI and the design community, Coca-Cola scaled up its creative ideation process and dramatically increased the velocity of content production. These efforts are being continued with Coca-Cola’s new partnership with Adobe called Fizzion. Fizzion learns how designers work in Adobe Creative Cloud applications like Illustrator. It takes this information and turns it into a StyleID – a machine-readable identity that automatically applies brand rules across formats, platforms, and markets.

Once created, these StyleIDs act as real-time guides, allowing designers to create hundreds of localized assets for a campaign with precision and consistency.

With ignite, Lloyds Banking Group turned to aerospace engineering for inspiration—using AI-powered digital twins to reimagine how a bank tackles climate strategy. Originally used to simulate jet engines, digital twin technology was repurposed to create a live, AI-enhanced replica of Lloyds’ systems and carbon data. This model ran hundreds of what-if simulations across sectors, generating novel decarbonization strategies tailored to its 2030 net-zero pledge. Executives called it “a paradigm shift in problem-solving”—proof that blending ideas across industries can ignite entirely new paths forward, performance data combined with design knowledge.

With nurture, Boeing employed generative AI to refine the design of aircraft components for greater efficiency. Engineers provided an initial design idea accompanied with specific parameters and performance requirements, and the AI algorithm then generated thousands of design variations that met those criteria. This allowed Boeing to identify improved designs quickly. The AI acted as a super-charged refinement tool, taking an existing engineering concept and enhancing it with a dataset of guidelines.

With defends, in 2024, at the University of Pennsylvania’s Wharton School, professors have integrated AI into the classroom as a simulated opponent and mentor.

In one trial, the instructors needed a realistic business negotiation exercise – so they prompted GPT-4 with a short description and let it generate a full negotiation simulation with multiple roles. Wharton has since pivoted its interactive teaching platform so that “AIs are instructors, AIs are mentors, all interacting with each other” to challenge students.

In practice, this means students can debate an AI playing devil’s advocate or engage in case discussions with AI characters, forcing them to defend their ideas and think critically.

Last with share, a landmark study at Fortune 500 firm P&G showed how AI can act as a “cybernetic team member” to boost collaboration. In a large experiment, hundreds of P&G employees were split into product development teams. One with and the other without access to LLM assistance. The results were striking; teams that had an AI teammate were far more likely to produce top-quality solutions (they were 3× more likely to create solutions rated in the top 10% for quality compared to teams without AI).

AI helped bridge knowledge silos – without AI, marketing and R&D specialists on a team tended to stick to their own domains, but with AI, those distinctions disappeared as both types of experts contributed more balanced, integrated ideas.

In other words, the AI created a bridge of ideas and knowledge sharing among team members. By summarizing information, suggesting insights from different domains, and freeing up time, AI allowed the group to share information more openly and focus on creative collaboration.

The FINDS framework layers can also be combined. We can utilize every section of it for a specific project or task. A great use of this is in higher education, where an entire curriculum could be built using FINDS.

Let’s use a fictional example. Jane is a 35-year-old high school teacher looking to develop a strong lesson plan for her math students.

Jane can use fuel to come up with several ideas for supplemental engagement material. She can prompt an LLM to come up with multiple outlines. In fueling multiple ideas for her plan, Jane can mix and match the best parts of each by further refining her objectives, giving the AI more input.

Educators know how metaphors can be a powerful strategy for helping students retain difficult mathematical concepts. Jane can use ignite to tackle her most challenging problem sets by asking it to examine playful metaphor examples that appeal to various learning types. In some ways, the “hallucinating” aspect of an LLM can even help Jane arrive at original concepts.

Jane can use nurture to develop a GPT for students to tackle challenging math sets and problems. To ensure her GPT aligns with school policies and equity standards, Jane includes AI use guidelines from her school district’s new digital ethics framework, embedding rules around plagiarism detection, prompt transparency, and accessibility.

With defends, Jane can take her content plan and give an LLM descriptions on some of her most challenging students, asking for flaws in the lesson plan that may affect their ability to understand and engage with the content. From the discoveries, she can refine and curtail specific examples to those students. As she begins to teach the content, Jane can track how well the strategy is working and modify it in the LLM based on her observations.

And with share, Jane can distribute her lesson plan to other teachers at the school. From that, additional plans can be developed and tweaked to fit the district's prerequisites. By layering the five elements of FINDS, leaders and educators can shape AI collaboration not just for efficiency, but for intentional, human centered innovation.

While FINDS is a framework that promises to augment leaders, executives, and educators in their roles, two points of caution should be considered.

With number 1, leaders should have a baseline of their own iterative strengths to discover the best fit for AI. A great way to uncover this is with a questionnaire looking in on someone’s own views toward ideation and creativity.

Here are some quick and reflective questions you can ask to get a sense of your baseline.

Answer Yes or No:

Once you’ve answered and reflected on these questions, it will lead to revelations on which stage(s) of FINDS you’re likely to leverage the most. For instance, if, on the reflective questions, you find yourself losing momentum close to the end of a project, an LLM or AI project management tool can assist you with rapid ideation to get unstuck.

You can also use LLMs to strengthen your baseline findings. With ChatGPT’s ability to read between conversations, you can ask questions about what it knows about you, your communication style, and this questionnaire.

With number 2, overreliance on AI is happening. Unengaged students in classrooms use AI in their assignments unchecked. AI ghostwritten papers are being handed in without students offering their own input.

If the students of 2036 are to be more than idea delivery avatars, we must teach them how to think. Behind these cautionary flags is a deeper call for action: if AI is becoming a daily collaborator in classrooms and boardrooms, we need policy that matches its influence—guidelines that protect creativity, encourage discernment, and keep ethics in the loop.

FINDS is one way to get there. However, what happens if the majority of young people who start using AI attempt to automate original thought, vs have it augment the skills and knowledge they have spent time acquiring?

It is the responsibility of leaders, executives, and educators to teach and inspire this next generation on how to use and see AI while also encouraging growth in critical and original thought.

All the discoveries that have led to the explosion of LLMs are based on the breakthroughs of neuroscience. What insights from our own minds can we draw on to help the next generation think critically?

There is also one thing AI truly lacks in any great human ideation process: diffuse thinking. Rather than focusing on a defined result, diffuse thinking allows your subconscious to make unexpected connections between disparate ideas. It is the opposite of being “always on”.

From Jeff Bezos to 18th-century painters, diffuse thinking has been hailed as one of the top skills to achieve originality. People often experience diffuse thinking when they go for a walk, have a shower, or stare out their window. Jeff Bezos quotes “real invention, real lateral thinking that requires wandering. You have to give yourself permission to wander.” Diffuse thinking leads us to develop innovative solutions by connecting new and unfamiliar concepts with existing ones from lived experiences.

Typing into an LLM is an active process. It doesn’t allow us to tap into this relaxed state. That’s a human activity. While an LLM might mimic the results of diffuse thinking, it only has access to its library of information. While this is a lot, experience matters.

Striking a balance between fostering the continued development of critical thinking and leveraging the speed, agility, and vastness of knowledge that AI possesses, through the use of the FINDS framework, represents human advancement.

If we want AI to serve creativity—not replace it—we must train the next generation to ideate with intention. FINDS is not just a framework; it’s a mindset.