002 Jump In: Learning AI Without Waiting to Feel Ready

🎧 Listen to this episode on Apple Podcasts, Spotify, or your favorite platform.

There is a strange kind of pressure hanging over product management and UX design right now.

Everywhere we turn, AI is part of the conversation. AI tools. AI workflows. AI strategy. AI productivity. AI risk. AI creativity. AI replacing jobs. AI creating new ones.

I feel like I’m standing next to a fancy Vegas pool, watching everyone else talk about jumping in.

At some point, reading about the water is not the same as getting in.

In this episode, J and I discuss the overwhelming news cycle and buzz around AI and the temptation to keep reading articles instead of starting to use AI.

Exposure is not the same as learning.

We can attend the conference session. Watch the YouTube video. Save the LinkedIn post and hope we remember to read it. Nod thoughtfully at the phrase “AI transformation.”

But learning only begins when we can do something differently than we could before.

“Exposure is when you’re presented with information… learning is when you actually are able to do something differently than you were able to do before.” – J Schuh

That may be one of the most useful ways to think about AI adoption in product and design work.

Because right now, many of us are highly exposed to AI, but we are still learning how to use it well.

Exposure Feels Productive. Learning Feels Messy.

Many of us know the feeling of consuming information and mistaking it for progress.

We read the book.
We attend the webinar.
We bookmark the tutorial to watch someone else demonstrate the workflow.

And to be fair, exposure matters. It gives us language and awareness. It helps us understand what is possible.

Exposure does not automatically become capability.

J compared this to learning how to swim. You can watch hours of YouTube videos about breathing, stroke mechanics, kicking, and turns. You can understand the theory. You can probably explain the steps.

When you jump into the water, everything changes.

The body has to learn what the mind has only observed.

Using AI works the same way.

We can know about ChatGPT, Claude, Gemini, Copilot, Midjourney, Leonardo, and every new tool wandering into our feeds with a promise to change everything. Knowing they exist is not the same as knowing how to use them in a product workflow, a UX research process, a design critique, or a strategic planning conversation.

For product managers and UX professionals, the shift from exposure to learning might look like:

• Using AI to review a user story for ambiguity
• Asking AI to identify gaps in product documentation
• Summarizing research themes without exposing sensitive data
• Generating alternate ways to frame a stakeholder problem
• Testing whether a design concept is directionally clear
• Exploring multiple solution paths before going to a design team
• Creating reusable prompts for recurring product tasks

The value is not in trying AI once. The value is in building enough repetition that we begin to understand where it helps, where it struggles, and where our own judgment needs to intervene.

AI Learning Requires Permission to Be Bad at First

Learning almost always feels uncomfortable at the beginning.

J used the example of a child learning an instrument. Those first sounds coming from the bedroom are rarely beautiful. They are squeaky. Strange. Inconsistent. Alarming to nearby pets.

No one expects mastery on day one.

And yet, many organizations seem to expect teams to adopt AI with immediate polish.

Companies are hearing that AI will transform operations, accelerate delivery, unlock insights, and reshape workflows. Leaders want results. Teams are told to ramp up quickly, but the people doing the work are still figuring out what meaningful AI usage even looks like. This is the dreaded learning curve.

For product and design teams, AI adoption is not just a tooling decision. It is a capability-building process.

If organizations want people to use AI well, they need to make space for:

• Trial and error
• Prompt refinement
• Workflow testing
• Security and privacy guidance
• Shared examples of good usage
• Honest discussion of bad outputs
• Time to reflect on what worked and what did not

Learning AI is not simply about access to tools.

It is about developing judgment through use.

Product Managers Can Start Small

We do not need to wait for perfect confidence before we begin.

Just start.

For product managers, starting small might be the most realistic and sustainable path.

I’ve been learning AI tools on a personal computer for my own projects, because many of us work in organizations where AI use is limited inside company systems, often for good reasons. Privacy, security, intellectual property, customer data, and regulated industries all matter. That means responsible experimentation requires care. Companies are still deciding how much AI to enable within their ecosystems as they weigh the risks.

We need to look for safe ways to build skill. A product manager might use AI to:

• Turn a generic problem statement into clearer discovery questions
• Ask for risks or assumptions in a fictionalized user story
• Practice explaining a product concept to different audiences
• Generate a checklist for evaluating a feature idea
• Summarize public documentation
• Explore possible user pain points for a non-sensitive scenario
• Improve clarity in communication or meeting notes

The goal is not to outsource product thinking. The goal is to strengthen it.

AI can help us slow down the right parts of our thinking while speeding up the repetitive parts of our work.

Designers and Creatives Need Room for Ethical Complexity

J and I also chatted about a very real tension in the creative community.

AI image tools and large language models (LLMs) have raised serious concerns around scraped artwork, creative ownership, royalties, environmental impact, and the future of creative labor. Those concerns should not be brushed aside with a cheerful “just adapt” sticker slapped over them.

It’s ok to see AI as necessary for career growth while still wrestling with the ethical implications.

We can recognize that AI is becoming part of product and design work while still asking hard questions about how it was built, how it is used, and who is impacted.

For designers and artists, J described AI as useful for direction, ideation, and inspiration, but not necessarily for the final craft. He shared that AI-generated visuals often require human refinement, correction, and artistic judgment before they become usable in real-world scenarios.

AI can help generate options, but options are not outcomes.

Human practitioners still need to decide:

• Is this useful?
• Is this usable?
• Is this ethical?
• Is this accessible?
• Does this reflect the actual problem?
• Does this respect the people affected by the work?

AI may accelerate creative exploration, but it does not remove responsibility from the humans shaping the final experience.

AI Is Not a Traditional Tool

J pointed out that AI behaves differently than the tools many of us are used to.

Traditional tools are predictable. If we click a specific Photoshop command or use a specific spreadsheet formula, we expect the same output every time.

AI is different.

It can surprise us. Misunderstand us. Overextend. Hallucinate. Drift. Add the metaphorical dog back into the picture even after we clearly said, “Please remove the dog.”

Working with AI is less like operating a predictable machine and more like collaborating with an unpredictable thinking partner. Sometimes that partner is brilliant. Sometimes it is confidently wrong. Sometimes it brings you an idea you did not ask for but might actually need. Sometimes it wanders into flights of fancy based on unknown and undeclared source material.

For product and design professionals, that means AI fluency is not just tool fluency. AI fluency includes:

• Knowing how to provide context
• Knowing when to start a clean chat
• Knowing how to constrain outputs
• Knowing how to evaluate accuracy
• Knowing when to ignore the response
• Knowing when the unexpected idea is worth exploring

The skill is not simply prompting.

The skill is judgment.

From Polishing to Ideating

During our discussion, I realized I had mostly been using AI to polish work rather than ideate. That is probably true for many of us.

It is natural to start with polishing because the use cases feel safer and clearer:

• Improve this paragraph
• Make this user story clearer
• Summarize this information
• Find gaps in this draft
• Help me communicate this better

The next layer of AI learning may be using it to explore:

• What patterns are visible in this research?
• What are three possible strategic directions?
• What assumptions might we be making?
• What risks should we consider?
• What would a designer, developer, or customer support person notice?
• What alternate solution paths exist?

Product management and UX design are not just output roles, they are sensemaking roles.

If AI helps us see more possibilities, ask better questions, or prepare more thoughtfully for cross-functional conversations, it can become more than a productivity tool.

AI can become a strategy amplifier, but only if we stay actively involved.

Better Inputs Help Teams Use Their Creativity

The conversation also connected AI learning back to collaboration with designers and developers.

We can use AI to strengthen user stories and documentation so developers have clearer requirements and can use their creativity to solve implementation challenges.

A well-formed product artifact does not eliminate creativity. It creates the conditions for better creativity.

The same applies when product managers bring ideas to designers. The goal is not to hand over a finished design and say, “Build this.” The goal is to communicate the problem, the direction, the context, and the constraints clearly enough that the designer can make the solution better.

AI can help us prepare for those conversations, though it should not flatten them.

Product managers, designers, and developers all bring creative judgment to the work. AI should help clarify the handoffs, not replace the partnership.

Questions for Reflection

1. Where are you currently only exposed to AI, and where are you actually practicing with it?
2. What recurring part of your product, design, or communication workflow could become a safe AI experiment?
3. Are you using AI mostly to polish existing work, or are you also using it to explore better questions and ideas?

Key Takeaways

• AI exposure creates awareness, but AI learning requires practice.
• The first experiments with AI may feel awkward, inconsistent, or disappointing. That is part of the process.
• Product managers and UX designers can use AI for both polishing existing work and expanding strategic thinking.
• AI can accelerate ideation, documentation, and creative exploration, but human judgment still shapes the final value.
• The professionals who build AI fluency will likely be the ones who experiment consistently, not the ones who wait until they feel perfectly ready.
• We can recognize that AI is becoming part of our daily work while still asking hard questions about how it was built, how it is used, and who is impacted.

Final Thoughts

Learning AI does not require instant mastery. It requires curiosity, experimentation, and the willingness to be awkward at something before it becomes useful.

For many of us in product management, UX design, product leadership, and creative work, the most important step may be the simplest one: stop waiting to feel perfectly ready.

The water might feel cold at first.

Jump in anyway.

– Metsy
Co-host, Pixels & Priorities

Stay Connected

🎧 Listen to new episodes every Tuesday on Apple Podcasts, Spotify, or your favorite podcast platform. Please leave us a rating, submit a review, and share with a friend. It helps more people find the show and join the conversation.

Connect on LinkedIn: Metsy Rose | J Schuh | Pixels & Priorities