The Content Paradox: When Creating More Learning Creates Less Value

Chief Learning Officer "June" — Follow Her Story

June's learning organization had never been more productive.

Her team could develop learning experiences in days that once required weeks. Artificial intelligence helped instructional designers research topics, organize content, draft scenarios, create assessments, generate supporting resources, and rapidly revise materials based on stakeholder feedback. Business leaders were beginning to use many of the same tools themselves, making it possible to turn internal expertise into polished learning content without waiting for a traditional development cycle.

By almost every conventional measure of L&D productivity, this was progress. The organization could create more learning, respond to business requests faster, and provide employees with more resources than at any point in its history.

Yet June was becoming increasingly uncomfortable with what those measures actually told her.

Her organization was producing more courses, videos, scenarios, assessments, job aids, and performance support. What she couldn't confidently say was whether employees were becoming more capable because of them.

That distinction would become increasingly important as June confronted one of the central paradoxes facing Learning & Development in the age of AI: when learning content becomes nearly unlimited, creating more of it may become one of the least valuable things L&D can do.

AI Is Solving the Content Production Problem

For decades, creating high-quality enterprise learning required significant resources. Subject matter experts had to be interviewed, instructional designers organized information into learning experiences, writers developed scripts, designers created visual assets, developers built courses, and stakeholders reviewed multiple rounds of revisions before anything reached employees.

That production model naturally limited how much learning an organization could create. A business leader might identify ten potential training needs, but budget, time, and available resources forced the learning organization to prioritize which ones justified investment.

AI fundamentally changes those economics.

Today, learning teams can use AI to accelerate nearly every stage of development. Business leaders can increasingly create useful materials themselves, while instructional designers can dramatically expand their individual production capacity. Tasks that once represented significant portions of a development timeline can now be completed in minutes.

Organizations can produce more:

• Courses and microlearning experiences
• Videos, scripts, and presentations
• Assessments and knowledge checks
• Scenarios and simulations
• Job aids and performance support
• Personalized learning recommendations
• Manager discussion guides
• Communications and reinforcement content

That increased capacity creates enormous opportunity. There are many situations where an organization simply needs employees to understand a new policy, process, product, requirement, or piece of information, and AI can make developing those experiences dramatically faster and more efficient.

The danger begins when organizations assume that because creating learning has become easier, developing people has become easier too.

Information Problems and Capability Problems Are Different

As June looked more closely at the requests coming into her organization, she noticed that they tended to arrive in similar language. A leader identified a performance issue and asked L&D to create training to address it. The traditional response was to gather the necessary information, determine an appropriate format, develop the content, launch the program, and measure participation.

AI made that process remarkably efficient.

What it didn't do was determine whether training was actually the right solution.

Some business problems are fundamentally information problems. Employees may need to understand a new compliance requirement, learn how a process has changed, become familiar with a product update, or know where to find a particular resource. In those situations, the objective may genuinely be awareness, knowledge transfer, or completion, and rapidly produced learning content can be exactly what the organization needs.

Capability problems are different. If the organization needs a manager to coach more effectively, a salesperson to conduct better discovery, a leader to navigate a difficult conversation, or an employee to exercise stronger judgment in an unfamiliar situation, providing more information is unlikely to be sufficient.

The distinction matters because helping someone know something and helping someone become better at something require fundamentally different development strategies. As Unboxed President Dave Romero recently put it internally, "Helping someone consume information is one thing. Helping someone build a capability is something different." The Work We Are Really In

More Content Doesn't Necessarily Create More Capability

June began reviewing several recent learning initiatives through this new lens. The content itself was strong. Employees had access to clear explanations, useful resources, professionally developed learning experiences, and increasingly personalized support.

But access wasn't the problem.

A manager could watch an excellent video about delivering difficult feedback and still struggle when an employee became defensive. A salesperson could complete a course on consultative selling and still default to presenting product features when a customer conversation became uncomfortable. A leader could understand the principles of coaching and still have difficulty asking the right question when a team member needed help.

In each situation, additional information might improve understanding without meaningfully changing performance.

Human capability develops through a much more complex process. People need opportunities to apply knowledge, make decisions, encounter uncertainty, receive feedback, adjust their approach, and try again. Some capabilities also require coaching, observation, social interaction, and repeated application in the context where the skill will ultimately be used.

That means an organization can simultaneously increase the volume of learning available to employees while making relatively little progress developing the capabilities the business actually needs.

AI doesn't create that problem, but it can magnify it.

The New Risk Is Becoming Content Rich and Context Poor

When content was expensive to produce, production itself created value. Organizations needed specialized expertise and significant resources to transform knowledge into scalable learning experiences.

As production becomes faster and less expensive, that value begins to shift.

AI can generate a leadership scenario in seconds. It can produce realistic dialogue, create several possible responses, develop feedback, and generate an assessment aligned to the scenario. What AI cannot determine independently is whether that particular scenario reflects the situations leaders actually encounter inside a specific organization.

The same problem applies across enterprise learning. AI can generate an enormous amount of technically accurate, professionally written material without understanding enough about the organization to know whether any of it will meaningfully change performance.

Dave described this risk in his memo as becoming "incredibly content rich" while becoming "context poor." His argument was not that organizations should create less with AI, but that easier production makes understanding the business, learner, customer, desired behavior, and definition of good performance increasingly important. The Work We Are Really In

For June, this reframed the opportunity. If everyone could create content, the scarce resource was no longer content itself.

It was context.

L&D's Most Important Work Happens Before Anything Gets Built

June started changing the conversations her team had with internal stakeholders.

When a leader requested training, the team resisted the instinct to immediately discuss courses, videos, simulations, or delivery formats. Instead, they spent more time understanding the performance problem the business was actually trying to solve.

The questions became more diagnostic:

These questions closely reflect the performance diagnosis Dave outlined in his memo. His central argument is that when a client asks for training, the most valuable contribution isn't immediately opening an authoring tool. It is understanding what the organization is actually trying to change and what the individual needs to experience to become more capable. The Work We Are Really In

AI can help June's team answer many questions faster. It can organize information, analyze data, identify patterns, generate possibilities, and accelerate production once the appropriate intervention has been identified. But deciding what problem deserves to be solved remains an exercise in business understanding, curiosity, judgment, and empathy.

That work becomes more important when everything that follows it becomes easier.

The Value of Instructional Design Is Moving Upstream

This shift has important implications for instructional designers and the broader L&D profession.

If much of the traditional production work can be accelerated by AI, it is reasonable to ask whether instructional design becomes less valuable. June increasingly believed the opposite was true, but only if the profession evolves alongside the technology.

The value of an instructional designer can no longer be defined primarily by the ability to produce polished learning content. AI will continue making many aspects of that work faster, easier, and more accessible to people outside traditional L&D teams.

The greater opportunity lies upstream in the work that determines whether learning will matter at all. That includes diagnosing performance problems, understanding the learner's environment, identifying critical behaviors, designing meaningful practice, determining appropriate feedback, creating relevant experiences, and defining how capability improvement will be measured.

The production component of the craft is becoming easier. The thinking behind it is becoming more important, a distinction Dave makes directly in his memo when discussing the future of instructional design. The Work We Are Really In

For L&D leaders, that should change how teams are developed. Technical proficiency with AI will matter, but so will consulting skills, business acumen, behavioral analysis, performance diagnosis, learning science, measurement, and the ability to connect workforce development to business strategy.

June Began Measuring a Different Kind of Productivity

This new perspective also forced June to reconsider what an efficient L&D organization actually looked like.

Historically, greater output had been an easy indicator of productivity. If the team could support more business requests, launch programs faster, and produce more learning with the same resources, the organization appeared to be improving.

AI makes that definition increasingly dangerous.

If content production becomes dramatically cheaper, maximizing production can create an organization overflowing with learning that employees don't need. More content can increase cognitive load, fragment the learning experience, make relevant resources harder to identify, and encourage L&D teams to solve problems with training simply because training has become easy to produce.

June began considering a different set of questions. Instead of asking only how quickly the team could produce something, she wanted to know whether the intervention was necessary, whether it addressed the actual performance problem, and whether people became more capable after experiencing it.

That didn't mean abandoning traditional learning metrics. Completion, participation, satisfaction, and engagement still provided useful information. They simply couldn't serve as substitutes for evidence of capability development.

The team's objective was beginning to change from producing more learning to creating more meaningful development.

The AI Opportunity Is Bigger Than Faster Content

None of this made June less enthusiastic about AI.

It made her more ambitious about how the organization should use it.

If AI were used primarily to accelerate existing production processes, L&D would certainly become more efficient. Teams could build courses faster, create more personalized materials, respond more quickly to stakeholders, and reduce the administrative effort associated with learning development.

But that represents only the first stage of the opportunity.

AI can also make experiences possible that historically couldn't scale. Employees can practice difficult conversations with AI-powered characters, receive immediate feedback, explore different decisions inside simulations, access personalized reinforcement, and prepare for manager coaching with a level of individualized support that would have been prohibitively expensive across a large enterprise.

That is a fundamentally different use of AI. Instead of simply making content creation more efficient, AI begins making capability development more scalable.

Dave's memo makes the same distinction. He argues for aggressive AI adoption, including automating work that should be automated and creating experiences at a scale that wasn't previously possible. The critical word is should, because deciding where to remove friction and where meaningful effort should remain requires judgment about how people actually grow. The Work We Are Really In

For June, that distinction would lead to the next challenge.

Key Takeaways

Artificial intelligence is rapidly removing many of the constraints that historically limited learning production. Enterprise L&D teams can create more content, respond to business requests faster, personalize experiences, and provide employees with unprecedented access to information. Those capabilities represent genuine progress and should be embraced.

But abundant learning content does not automatically create an increasingly capable workforce. Information problems can often be solved efficiently through better content, while capability problems require employees to apply knowledge, practice behaviors, receive feedback, exercise judgment, and improve through experience. Treating those problems as interchangeable risks creating organizations that are rich in learning resources but poor in measurable capability.

The strategic opportunity for L&D is therefore larger than using AI to become a faster content producer. As production becomes increasingly commoditized, the value of understanding the business problem, diagnosing performance gaps, designing meaningful experiences, and measuring capability development increases. The future L&D organization will distinguish itself not by how much learning it creates, but by how effectively it helps people become better at the things that matter to the business.

Coming Next

In Part 2, June confronts an uncomfortable implication of this new model. For years, learning technology has promised to make development faster, easier, shorter, and more frictionless. But what if some of the effort we're removing is actually where capability develops? We'll explore The Productive Struggle and why great learning shouldn't always be easy.