Artificial Intelligence and Learning — The White Space To Be Explored

John Hagel
17 min readDec 15, 2024

--

By John Hagel, David Toole and John Seely Brown

Section 1: What’s the white space?

Everyone is talking about artificial intelligence, but there’s a white space that has not yet been significantly addressed — how to pursue new objectives in learning. These new objectives in learning will cultivate lives that are far more fulfilling and meaningful as we unleash more of our potential and create more personalized and rapidly evolving value for those around us.

We need to focus AI on finding ways to deliver much more value in all contexts all the time — it can significantly amplify our ability to add more value to those around us by imagining new possibilities and evolving our practices to deliver greater context specific value as we confront rapidly changing environments.

Two objectives for learning. To do this, we will need to shift our focus on the spectrum of learning objectives from sharing existing knowledge and practices to developing new knowledge and practices that can deliver more value in rapidly changing environments.

Most of the efforts in the AI domain today are focused on evolving training programs so that they can share existing knowledge and practices more efficiently and broadly. While not dismissing the value of these efforts, we need to recognize that existing knowledge and practices are becoming obsolete at an accelerating rate in a rapidly changing world.

In a world of accelerating and profound change in all aspects of our global economy and society, the most powerful and necessary objective for learning involves the creation of new knowledge and practices as we seek to live and operate in rapidly changing environments. This learning will be needed in all areas of human activity, not just in research departments or innovation centers. In all areas of human activity, we are encountering new challenges and opportunities to create more value.

What does creating new knowledge and practices mean? Creating new knowledge is about seeing what no one has seen before. For example, seeing things (ideas, people, objects, animals, plants, etc.) that haven’t been seen before, seeing characteristics or behaviors that have not been seen before, seeing patterns among existing objects that have not been seen before, or discovering new cause and effect relationships.

Creating new practices is about finding new ways to have more impact. It’s about discovering new problems or opportunities to create more value, discovering new ways to have impact, and discovering what doesn’t work.

These two objectives of learning — learning in the form of sharing existing knowledge and practices and learning in the form of creating new knowledge and practices — are complex and interconnected, but our emphasis needs to shift to the opportunity to create value at an accelerating rate as we confront new challenges and opportunities. There’s been very little exploration of the potential role of AI in supporting the second objective of learning.

As we shift our focus to this second objective of learning, we will also need to personalize our approach to learning — standardized approaches will be less and less effective. We need to recognize that everyone approaches learning in a different context and that AI can help to adapt learning approaches to the needs of each individual.

As we focus more on this second objective of learning, we will also need to cultivate much broader and more diverse networks of participants who share an excitement about exploring new terrain. These participants will often be in very diverse departments within a single organization and will extend beyond any individual organization to engage a broader ecosystem of participants.

As we pursue this second objective of learning, we will discover that it will also fundamentally change our approach to how we learn. We will briefly discuss in this paper how AI can help us change our approach to learning as we shift our focus to learning with the objective of creating new knowledge, but we are drafting a second paper that explores in much greater depth the powerful new approaches to learning that will be made possible by AI.

From opportunity to imperative. Why is this white space so important? This is not just an opportunity — it’s also an imperative.

Everyone outside organizations (particularly customers) will have access to the same technology and tools and will use them to become more selective and demanding of the vendors they encounter. If stakeholders like customers do not perceive that organizations are committed to delivering more personalized and rapidly evolving value, they will become less and less willing to share data about themselves and they will seek other vendors that can deliver more value. Existing organizations that fail to address this growing opportunity will become increasingly vulnerable to other organizations that unleash much more value for their stakeholders.

The untapped capability of AI. While AI will certainly continue to evolve rapidly (this technology has emergent properties that are not yet fully developed), it already provides many of the features needed to support the learning objective of creating new knowledge and practices — the barrier to overcome is not limitations in the technology.

The current applications of AI will help to free up our time by providing tools to make learning with the objective of sharing existing knowledge and practices easier and faster, so that people can shift their focus to the objective of creating new knowledge and practices.

Where and how can AI support us in pursuing the second objective of learning? AI can help participants connect with other rapidly evolving technologies like the Internet of Things and facilitate access to Big Data. It can personalize approaches to this form of learning by helping participants to frame questions and design prompts. AI also helps to identify potential partners to address these questions by providing valuable diversity in experiences and perspectives, cutting across traditional disciplinary and organizational silos.

It’s not just expanding the opportunity for productive interactions across human partners. There is also a growing opportunity for interactions across multiple AI agents so that AI agents can have conversations among themselves with human oversight to determine the validity of their recommendations. This can help to accelerate learning.

As participants come together, AI can provide simulations to help them imagine new approaches to creating and delivering value. It can also help participants to predict the outcomes of proposed actions before they are fully implemented and then it can provide richer and more real-time feedback on actions taken. All these capabilities can help participants cultivate a learning disposition where they will be excited about the opportunity to create new knowledge and practices.

An early example of learning with the objective of creating new knowledge and practices comes from drug research. AlphaFold is an artificial intelligence program that has more than doubled the number of high-accuracy human protein structures available to scientists. It has identified 200 million protein shapes — nearly all the proteins known to humans. The Institute of Protein Design at the University of Washington developed RoseTTAFold that is generating blueprints of entirely new proteins that do not exist in nature. It used to take years of laboratory work to determine the structure of just one protein and now it can be computed in as little as ten minutes. Insilico Medicine has been granted the FDA’s first Orphan Drug Designation for a drug discovered and designed using its own AI tools — the drug is a small molecule inhibitor that will be used to treat idiopathic pulmonary fibrosis (IPF).

AI can do a lot more than help organizations create and deliver more value to their stakeholders. It can also help participants in extended supply networks to discover new approaches to enhance sustainability efforts. By leveraging AI, companies can create more environmentally conscious and responsible supply networks, contributing to a greener future.

The untapped potential of learning platforms. AI can be harnessed to develop and deploy a very different and powerful form of learning platform. Rather than learning platforms that bring together lectures and workshops designed to share existing knowledge and practices, the primary design goal of these learning platforms would be to help all participants learn faster with the objective of creating new knowledge and practices by coming together for action and reflection.

The core unit of a learning platform would be a shared workspace for an impact group consisting of 3–15 people who share a passion for achieving increasing impact in a specific domain. The learning platform would provide the participants in each of these impact groups with the learning tools described above. The platforms would also provide the impact groups with richer and real-time feedback on impact achieved from the actions taken. The platforms would help to connect these impact groups so that they can see and explore new questions.

Bottom line, we need to be aware of the opportunity to pursue this second objective of learning that has exponential potential to create more value. Once we see that opportunity, we need to find pathways that will help us to address the obstacles and barriers that stand in our way so that we can accelerate our ability to create new knowledge and practices, and unleash more value with this second objective of learning.

Section 2: What are the barriers?

What are the barriers that make it challenging for us to pursue this missed opportunity? Two barriers need to be addressed:

· Learning with the objective of creating new knowledge and practices is not easy, especially since most of us only have experience in learning with the objective of sharing existing knowledge and practices

· Our institutions and social cultures are deeply resistant to learning with the objective of creating new knowledge and practices, except in very narrow parts of an organization, like research labs and new product development departments

The first barrier — personal fear. Learning with the objective of creating new knowledge and practices is challenging for most of us because we have very limited experience in this area. We are often afraid to move into new territories where we have very limited experience, especially if they require a lot of effort and risk-taking. Unleashing the full potential of this objective of learning also requires us to connect with others in small groups and large networks.

This objective of learning can be intimidating. Insight hinges on understanding entanglement within relevant contexts — connections and dynamics — the boundary between existing knowledge and practices and new knowledge and practices is blurry at best and continually evolving. It also requires an understanding that knowledge and practices are not static, but instead need to rapidly evolve as they adapt to a rapidly changing world. A significant portion of the new knowledge is initially tacit knowledge which can be very difficult to articulate. This objective of learning requires learning to be, not just learning about — it’s ultimately about embracing different ways of learning, living and being. Ultimately, we will need to create a new kind of living space

Pursuing this objective of learning requires cultivating capabilities that are actively discouraged in many environments — curiosity, imagination, collaboration across silos, creativity and reflection. It also requires mastering a new set of learning practices like improvisation and play in groups — practices that are not encouraged when learning is focused on sharing existing knowledge and practices.

This learning objective requires a willingness to challenge many things that were taken for granted. It requires critical thinking — constantly questioning whether certain knowledge or practices need to evolve. It also requires a willingness to unlearn.

We will also need to cultivate new practices in the use of existing AI features. For example, mastering the art of prompting — the imagination of how to lead the prompting onwards in what we think is heading in the direction we might find useful.

To overcome the fear of the unknown and untried, we need to cultivate excitement about this learning objective and the opportunity to deliver much more impact that is meaningful. This starts with personal authenticity — discovering what we really want to master and focusing our learning there. It also extends into groups — how can we come together in groups that share our excitement about this learning objective? And it extends across domains and disciplines — how can we form groups that bring diversity of perspectives and experience in ways that can accelerate learning?

One of our authors, Dave Toole, had to overcome his fear. He is a professional guitar player as well as a business leader. He was given a challenge to write a song in six weeks for his daughter’s wedding. Without GPT models that are part of generative AI to help with writing the lyrics, the song would not even have started. Editing and rewriting the lyrics in hours expanded time to write a melody and practice the vocals in time to bring in others to collaborate to produce a new song in six weeks. This jumpstart in lyrics opened up learning in a new way for Dave. It unlocked his ability to create melodies for the lyrics — something he had never done before — and it built his capability to explore a whole new domain. This overcame the emotional roadblocks that naturally occur when doing something new, providing a new learning partner with AI to build confidence and accelerate his learning. AI is now being embedded in software to bring a full band together to augment his creativity and expand his confidence further, bringing new ways of learning and upskilling to a level that was not available before.

The second barrier — institutional resistance. Large organizations of all types are deeply hostile to the capabilities and practices required for learning with the objective of creating new knowledge.

Large organizations are driven by a model of scalable efficiency that focuses on pursuing existing activities faster and cheaper at scale in a continuing effort to cut costs — they do not look more broadly at the potential new value that can be created and delivered. These organizations seek efficiency by tightly specifying all tasks and standardizing them, so they are done in the same efficient way throughout the organization — they do not realize that this approach is becoming less and less efficient in a rapidly changing world.

These organizations view “personalization” as a way to more efficiently target customers and motivate them to buy more of their standardized products, rather than becoming more responsive to evolving customer needs and seeking to deliver more and more tailored and rapidly evolving value.

They are deeply suspicious of people who ask too many questions, who take too many risks and who deviate from the assigned script. Capabilities like curiosity, imagination, collaboration across silos, creativity and reflection are not encouraged — just follow the manual.

When leaders of these organizations talk about learning, they generally talk about training programs to support learning with the objective of sharing existing knowledge, and they believe the motivation for people to do this learning is fear — if they don’t do this learning, they will be fired.

Given the focus on scalable efficiency, these organizations embrace AI as a technology that can help them to automate activities and shrink the workforce so that they can cut costs.

Section 3: What are the most promising approaches?

What initiatives can help us to overcome these barriers and address the untapped opportunity of AI?

We need to start by recognizing that unleashing the full potential of AI will require us to master new objectives of learning and to pursue profound transformation of our work and social environments. This will generate considerable fear among participants who worry about the risks of moving beyond the tried-and-true practices that have led to success in the past.

To overcome this fear and unleash the full potential of AI, leaders will need to adopt new approaches. To help motivate people, leaders need to frame inspiring questions about really big opportunities that can only be achieved by exploring new areas and developing different learning objectives.

Many workers today are worried that AI will be used to automate work and eliminate jobs. Leaders need to reassure them that AI will automate the routine tasks that today consume most of their time and free up the workers so that they can create more value by addressing new opportunities and challenges. By focusing on the significant value that can be created as a result, leaders can begin to excite and motivate workers to embrace this new technology.

To help overcome the fear, leaders also need to embrace new approaches to driving change. Rather than pursuing “top down, big bang” approaches to change, leaders should focus in the early stages on small moves, smartly made that can yield significant impact. These are initiatives that can be launched with modest resources by small teams cutting across conventional silos with easily accessible technology at small scale. The goal is to quickly demonstrate the untapped value that can by created by using AI to help workers to learn faster by embracing news forms of learning. Participants will then be motivated to scale these initiatives quickly to amplify impact and accelerate learning even faster as they start to demonstrate the ability to deliver increasing value to stakeholders.

The good news is that AI technology makes it possible for workers to pursue small, moves smartly made on their own initiative. OpenAI Enterprise doesn’t require much restructuring of data and AI copilots also don’t require significant restructuring of data. Small language models are being developed so that workers can focus on local data. It is easier and easier for workers to pursue local initiatives without massive restructuring.

In looking across the business landscape, early adopters of AI to pursue this new objective of learning are likely to be smaller, entrepreneurial ventures that are seeking to address emerging customer needs. In larger organizations, the early adopters will most likely be found in research departments that are seeking to create new products and services to address customer needs. While these research departments have greater motivation to create new knowledge and practices, they will also face the need for significant transformation in terms of systematically and broadly reaching out across silos and connecting with diverse participants who are not “experts” with relevant degrees.

The untapped opportunity in larger organizations is to find ways to motivate participants in all departments, not just research departments, to create more value by using AI to pursue this new objective of learning. In a rapidly changing world, this will be an imperative for all participants in large organizations, not just in research departments.

The good news is that AI tools can become a catalyst for change from the bottom up. The new practices that these tools cultivate do not require a lot of time and effort from the individual to try at the outset, but they can fire up the imagination of the workers and deepen the courage to connect with others to explore what new knowledge can be created. As they develop the capability to do new things that haven’t been done before, their excitement will grow. The small moves, smartly made could come from workers at the bottom of the pyramid who begin to see the potential and start to mobilize their colleagues to help them find new ways to create and deliver value.

What are some examples of companies that are already deploying AI to accelerate learning with the objective of creating new knowledge and practices?

Unilever saw an opportunity to use AI to develop personalized perfume offerings for its customers. It used an AI tool called “SPHERE” to help research and development teams to brainstorm new product ideas. SPHERE provides the ability to explore and analyze vast datasets of customer preferences, market trends, and ingredient functionalities. Unilever created a virtual “fragrance architect” that generates unique and highly personalized perfume products, resulting in a 90% satisfaction rate and increased sales. By helping workers to access and analyze vast amounts of data, Unilever provided workers with the ability to take small steps, smartly made, that generate significant new experiences for customers.

In its cosmetics business, Unilever used AI to help develop Hourglass Red 0. This is the first vegan alternative to the most used red pigment in color cosmetics — red carmine. Red carmine was derived from female beetles and used to require 1,000 beetles to make one single lipstick. The vegan carmine created by Unilever replaced the need to crush female beetles.

Unilever developed a product called BeautyHub Pro that offers product advice for skincare and haircare, by enabling customers to take a simple quiz and a selfie so that AI tools can assess up to 30 visual data points to make personalized products recommendations. This increased basket size by 39% over other pathways.

In another example, Autodesk used AI to help create a very innovative design for its new office in Toronto — it was a design that significantly improved the experience of the occupants of the new office. The design team specified design goals and constraints and applied AI tools to explore thousands of options, quickly narrowing the number of options, and more deeply exploring the ones that seemed to have the greatest potential. The approach, called multi-objective optimization, became a model to re-use across other applications.

The generative design created an office layout with unique workspaces tailored to the needs of various groups of workers and customers. Conference rooms were shaped by the goals of daylight, privacy and views of the outside. A central stairway helped to increase cross-functional interaction. The design did an excellent job of integrating the diverse needs of a broad range of workers and coming up with very creative approaches to designing customer interactions and office space.

In a third example, Moderna, a biotechnology company that develops mRNA-based medicines, applied AI to help it identify optimal dose recommendations. It created a GPT pilot called Dose ID that can review and analyze clinical data and is able to integrate and visualize large datasets. It is used by clinical study teams to help them determine what the best dose recommendation would be for different medications. This AI tool has not only helped them to come up with better dose recommendations, it has also helped the teams to evaluate many more medications than they were able to when using more conventional approaches.

AI can also become a catalyst for leveraged growth where there is an opportunity to accelerate learning by mobilizing a broader network of participants. One recent example is the formation of the Alliance for OpenUSD (AOUSD) that has brought together many leaders in the AI arena, including NVIDIA, Pixar, Adobe, Apple and Autodesk. AOUSD seeks to promote the evolution and development of Pixar’s Universal Scene Description technology. By using OpenUSD, participants were able to come up with new products and experiences to market in months when it used to take years.

The bottom line

We are still in the early stages of the development and deployment of artificial intelligence technology. While it is generating considerable interest, most of the interest is in how to use the technology to do what we have always done faster and cheaper. In terms of learning, the focus is on how to use AI technology to help us share existing knowledge and practices more efficiently.

This leaves a major white space that will provide us with an opportunity to create much more value. The opportunity is to apply this technology to help everyone in an organization to learn faster in the form of creating new knowledge and practices that can generate significantly greater value to stakeholders.

This is not just an opportunity. It is an imperative. We live in a rapidly changing world where organizations that remain focused on doing what they have always done faster and cheaper will discover that they are becoming less and less efficient in a rapidly changing world. Those who master the untapped potential of this promising technology will find out they are creating exponentially greater value for all their stakeholders.

There are certainly obstacles and barriers that will make it challenging to address this opportunity, but there are approaches that can help leaders to motivate participants to overcome these obstacles and barriers and achieve the potential that is waiting to be addressed. Those who adopt these approaches and aggressively pursue this untapped opportunity will ultimately be the big winners in unleashing the potential of AI.

To pursue this opportunity in a rapidly changing world, we need to adopt the mindset of an explorer, driven by a sense of both imagination and critical thinking, and deeply aware of the context we are operating in. We need to learn how to play with AI systems and tools — they will be constantly trying to read us, and we need to try to read them. It’s a new game, but those who learn how to play it will create enormous value.

As we indicated earlier, embracing this new objective of learning — creating new knowledge and practices — will also motivate us to develop new approaches to learning. These approaches are absolutely essential to pursue learning with the objective of creating new knowledge and practices, but they are also very effective in pursuing the traditional objective of learning which is to share existing knowledge and practices. We will explore these different approaches to learning in another paper that we are still developing.

A request for help

We are still at an early stage of exploring how organizations and their participants can use AI to pursue learning with the objective of creating new knowledge and practices. We know there are early initiatives out there that can help us to gain more insight into how to harness this technology to help participants create new knowledge and practices. We are seeking to identify these early initiatives and to see what kind of impact they have delivered. If you are aware of any initiatives in this space, we would be grateful if you could connect with us and provide us with information about these initiatives. Please send an email to johnhagel3@gmail.com

--

--

John Hagel
John Hagel

Written by John Hagel

Work and play on the edge - views breathtaking, experiences deep and satisfying, learning limitless

Responses (2)