AI adoption can place experienced professionals in the uncomfortable position of becoming beginners again. Leaders should recognize that hesitation may reflect concerns about competence, identity, risk, or professional standards rather than simple resistance. People build confidence when they can start with a real problem, connect new tools to familiar knowledge, practice safely, and learn visibly alongside leaders and peers.
Experience usually brings confidence. We learn the language of our field, recognize familiar patterns, and develop judgment that others trust.
Then a new technology changes the work, and an experienced professional may suddenly feel like a beginner again.
That experience can be uncomfortable. The person who usually has answers may not know what to ask. Skills developed over years can feel less certain. A basic mistake may seem more visible because others expect competence.
AI adoption is often described as a problem of access or training. It is also a problem of identity. People are not simply learning a new tool. They may be renegotiating what expertise means in their role.
I have been a beginner more than once.
When I was 15, I began my first job as a web designer. At home, I had written HTML in Notepad and used free image-editing tools. On my first day, I was placed in front of Microsoft FrontPage and Adobe Photoshop, two programs I had never used, to do a job I had already said I could do.
I did not begin by reading the Photoshop manual from cover to cover. I looked for something familiar. The floppy disk icon still meant save. From there, I could try one tool or menu item, see what happened, and return to the saved version when the result was not what I expected. I used the help menu or searched for the smallest piece of information needed to take the next step.
Years later, I approached generative AI in much the same way. I began using ChatGPT shortly after its public introduction. At first, I used it for small and playful experiments, such as writing a song about my dog or developing a short script from a movie idea. As the technology advanced, I used it to create custom birthday cards and make emails more concise and action-oriented.
Then the field expanded so quickly that trying to keep up with every model, capability, article, and tool became overwhelming. The answer was not to learn all of AI. It was to begin with one problem I actually needed to solve and experiment from there.
That approach still shapes how I think about learning: find the smallest useful next step, connect it to something you already understand, and build confidence through application.
Resistance may be protecting something valuable
When someone hesitates to use AI, it is easy to label the behavior as resistance. That explanation may miss the concern underneath it.
An employee may worry about producing inaccurate work, exposing private information, appearing less capable, losing a valued part of the job, or being expected to do more without adequate support. A professional who has built a reputation on careful judgment may be appropriately skeptical of a system that produces a confident answer without showing its reasoning.
Leaders should treat these concerns as information. Good adoption does not require abandoning standards. It requires helping people apply those standards to a new kind of tool.
Leaders need to learn visibly
Employees take cues from what leaders model. A leader who pretends to have mastered AI can make uncertainty feel unsafe. A leader who shares what they are testing, where a result failed, and how they verified an answer gives others permission to learn responsibly.
Visible learning does not mean careless experimentation. Leaders should be clear about approved tools, protected information, human review, and decisions that should never be delegated. Within those boundaries, curiosity should be treated as part of the work.
Practice must be connected to real work
One demonstration can create interest. It rarely creates fluency.
People learn when they use AI on relevant, appropriately low-risk tasks and compare the result with what they know. They need opportunities to revise prompts, question outputs, notice failure modes, and discuss what the tool changed in the work.
A small peer group can be especially useful. Employees can bring examples, compare approaches, and learn from one another without turning every experiment into a formal project. The goal is not to reward the fastest adopter. It is to build sound judgment across the organization.
Preserve the value of expertise
AI can generate a draft, summarize material, identify patterns, or suggest alternatives. It does not understand the organization, the people affected, or the full history behind a decision unless someone provides that context and evaluates the result.
Experienced employees have an important role in AI adoption because they can recognize when an answer is plausible but incomplete. Their knowledge helps define quality, identify risk, and determine where human judgment creates value.
Becoming a beginner again does not erase what they already know. It gives them a new way to apply it.
Make learning a shared responsibility
Organizations should not tell employees that AI matters and then expect them to learn it after hours. Provide time, relevant examples, accessible support, and room to practice. Recognize progress without creating pressure to perform certainty.
The courage to be a beginner is personal, but the conditions that make it possible are organizational. When people can ask basic questions, test ideas safely, and connect new capability to existing expertise, curiosity can replace fear and learning can become part of the transformation.
