How AI is reshaping the product role | Oji and Ezinne Udezue

Sep 7, 2025 Episode Page ↗
Overview

Aji and Ezine Udezwe, product leaders with 50+ years of combined experience and authors of "Building Rocket Ships," discuss the evolving PM role, the impact of AI on product development, essential skills for PMs, and key lessons learned from their careers.

At a Glance
17 Insights
1h 18m Duration
15 Topics
6 Concepts

Deep Dive Analysis

Initial Thoughts on PM Role in the AI Era

Evolving Role of Product Managers: Changes and Constants

PMs as Bottleneck with AI Accelerating Other Functions

Defining and Identifying Sharp Problems

The Shipyard Model for Product Development

Essential Skills for PMs in the AI Era

Importance of Humility and Teachability for PMs

Hands-on Learning and Building with AI: Oji's Smart House Project

Company Strategies for Succeeding with AI Adoption

Key Product Lessons from 50+ Years of Experience

The Importance of Simplicity and Opinionated Design

Communication as a Critical Element of Strategy

Intentional Career Management for Product Managers

Ethical Responsibility in AI Product Development

Introduction to the Book: Building Rocketships

Sharp Problem

A sharp problem refers to an old, core customer need that, if reimagined with new technology, can be improved three to five times (or even 10x) or have its cost reduced 10x. Solving such problems makes the solution incredibly compelling and profitable, helping founders avoid aimless product development.

Shipyard Model

The shipyard model describes a product development approach that embraces 'controlled chaos' through constant communication and high skill, similar to a bustling shipyard. It involves a six-capability team (PM, engineering, design, user research, data/ML/AI, product marketing) working in continuous collaboration to solve problems, moving beyond traditional stand-ups.

AI at the Core vs. AI at the Edge

AI at the edge means 'sprinkling' AI (like LLMs) onto an existing product's codebase at various connection points or user interfaces. AI at the core, conversely, involves fundamentally looking at the problem space and workflows, then using AI as a central capability to re-transform how the customer's problem is solved, potentially shrinking the traditional codebase.

Humility/Teachability (for PMs)

This concept emphasizes the importance for Product Managers to admit what they don't know and be willing to learn from anyone, regardless of their own seniority or the teacher's experience. It is crucial for continuous adaptation and survivability in the rapidly changing AI landscape where existing playbooks are constantly being rewritten.

Agency/Ownership (for PMs)

High agency means being able to see opportunities and take initiative to execute on them without needing permission, acting like an owner. It's about being a 'thermostat' that changes the room's temperature rather than a 'thermometer' that merely measures it, driving forward motion and impact.

Evals (for LLMs)

Evals refer to the critical skill of evaluating the results produced by Large Language Models (LLMs) to verify their intelligence and ensure they haven't 'hallucinated' or gone off-track. This goes beyond mere prompt engineering and involves creating true assessment mechanisms to validate AI outputs.

?
How is the role of a Product Manager changing in the AI era?

PMs are increasingly freed up to focus on deeper customer insights, orchestrating living software with complex feedback loops, and ensuring data literacy and ethical guardrails. The build process is accelerating, shifting PM focus more to defining sharp problems and supporting go-to-market.

?
What is a 'sharp problem' and why is it important for product success?

A sharp problem is an old, core customer need that, if reimagined with new technology, can be improved 3-10x or have its cost reduced 10x, making it incredibly compelling for customers. Focusing on sharp problems helps avoid aimless product development and increases the likelihood of success.

?
What is the 'shipyard' model for product development?

The shipyard model describes a six-capability team (PM, engineering, design, user research, data/ML/AI, product marketing) that embraces 'controlled chaos' through constant communication and collaboration to solve problems, rather than relying on traditional stand-ups and static processes.

?
What are the most important skills for Product Managers in the AI era?

Key skills include curiosity and humility (being teachable and open to learning from anyone), high agency and ownership (taking initiative and acting like a thermostat), and technical skills like understanding data organization, writing evals for LLMs, and tweaking models for performance.

?
Why is it important for PMs to get hands-on with AI technology?

Getting hands-on, such as writing code (which is now closer to architecture and English), converting PRDs to prototypes, and building personal projects, helps PMs adapt to faster build cycles, understand new toolsets, and stay sharp in a rapidly evolving technological landscape.

?
How do successful companies adopt AI differently from struggling ones?

Successful companies recognize AI as a core capability to fundamentally re-transform problem-solving (AI at the core), rather than just 'sprinkling' it on existing products (AI at the edge). They also focus on building specific, specialized AI solutions before attempting broader, multi-model intelligence.

?
What are some of the biggest lessons learned from 50+ years in product management?

Key lessons include focusing on sharp problems, prioritizing simplicity and opinionated design, relentlessly communicating the 'why' of strategy, and continuously learning from customer observation rather than just what they say.

?
How can PMs ensure they are truly understanding customer needs beyond what customers say?

PMs should prioritize ethnographic research and observation, watching what customers *do* rather than just listening to what they *say* they want. This helps uncover the underlying drivers and true intimacy with customer problems, which AI tools alone cannot fully provide from transcripts.

1. Focus on Sharp Problems

Prioritize building solutions for ‘sharp problems’ – old, core needs that are still difficult to the point that a 3-10x improvement or cost reduction would be compelling to customers. This avoids ‘drunken startup building’ and increases the likelihood of success.

2. Embrace Controlled Chaos (Shipyard Model)

Adopt a ‘shipyard’ team model, a six-capability team (PM, engineering, design, user research, data/ML/AI, product marketing) that fosters controlled chaos, high skill, and careful communication. This structure allows for orchestrated progress in a rapidly changing, ‘weird’ world, especially with AI.

3. Get Hands-On with AI Tech

Become super hands-on with AI technology by writing code (now architecture and English), converting PRD writing into prototype writing, and calling API interfaces yourself. This helps PMs adapt to faster build processes and avoid becoming a bottleneck.

4. Cultivate Humility and Curiosity

Develop humility and curiosity to admit what you don’t know and be willing to learn, even from those less senior. This ’teachability’ is crucial for survivability in careers and for building important products in an era where blueprints are constantly changing.

5. Develop High Agency and Ownership

Cultivate high agency and ownership, acting as a ’thermostat’ to change the room’s temperature rather than just a ’thermometer’ measuring it. This means seeing opportunities and taking action without waiting for permission, moving from a position of strength rather than fear about AI taking jobs.

6. Master Data Literacy & AI Evals

Gain a high-level understanding of how data is organized and leveraged in the AI world, and develop the skill to write effective ’evals’ (evaluations). This helps verify AI model results, constrain hallucination, and understand how to combine and tweak multiple models for better performance.

7. Build Personal AI Projects

Pick a passion project that touches on things you need to learn, like automating your house or creating a personalized outfit recommender. This provides a motivated, specific problem to solve, helping you learn and get hands-on with AI tools and concepts like quantization and fine-tuning.

8. Communicate the ‘Why’ Relentlessly

Never spend too much time communicating the ‘why’ behind your strategy to activate the entire organization. Understand that different people adopt information at different rates, similar to crossing the chasm, and continuous communication helps ensure everyone understands and aligns with the purpose.

9. Prioritize Simplicity and Opinionated Design

Strive for simplicity and clarity in product design, especially for ‘distracted brains’ in 2025 and beyond. Have the courage to be opinionated and make decisions, creating the most compelling, simplest experience rather than offering too many confusing options, which hinders learning and adoption.

10. Observe Customer Actions, Not Just Words

Go beyond customer interviews and what customers say they want; focus on what they do. Engage in ethnographic research, observing their actions and understanding the ‘why’ behind them, as this provides deeper, more accurate insights than relying solely on AI-processed transcripts of interviews.

11. Rethink Products with AI at the Core

For companies, fundamentally rethink problem spaces and workflows by using AI as a core capability, rather than just ‘slathering’ it onto existing products at the ’edge.’ This often involves shrinking existing codebases and allowing LLMs to become central to solving customer problems, leading to revolutionary changes.

12. Specialize AI Solutions First

Instead of trying to build a massive, broad AI solution that does it all, focus on building very specific, specialized AI solution sets. Then, create a connective tissue or multi-model solution to tie them together, as this approach has proven more successful in practice.

13. Take Chances on Dynamic User Experiences

Experiment with dynamic user experiences that personalize to the customer, moving beyond the chat interface as the ‘final boss’ for AI UX. Recognize that GUIs exist for a reason and static experiences are insufficient for the evolving nature of AI products.

14. Integrate Ethics into Product Development

As PMs, recognize the ‘ordinance level’ power of AI and take responsibility for the ethical implications of what you build. Lead on considering the human race when creating new digital products that give superpowers, avoiding the ‘I don’t care’ attitude seen in some past tech developments.

15. Understand Strategy Fundamentals

Learn the fundamentals of product strategy, including sources of competitive advantage (e.g., intellectual property, economies of scale/scope) and the various growth levers beyond just product. This knowledge empowers PMs to direct their company’s future compellingly and advance their careers.

16. Be Intentional About Your Career

Hold a clear intention and visualize your next career steps, allowing this desire to drive your progress. Being able to see and chase where you want to be is a powerful motivator for managing your own career path.

17. Learn All the Time, Confidently

Continuously learn, understanding that there is more knowledge outside your brain than inside it. Embrace being ‘stupid’ for a second (in the sense of not knowing) while maintaining confidence in what you have accomplished, allowing you to grow without being in a ‘crouch’.

The problems are still the problem. The customer is still the problem is still the customer and their pain or sharp problems still exist as they have.

Ezinne Udezue

Humility is teachability and teachability is survivability if you're thinking about careers.

Oji Udezue

Being able to visualize and chase the thing that you see about where you want to be is so powerful.

Oji Udezue

Don't generate. Generate first and let it refine. That's the major trick.

Ezinne Udezue

It's not what they say they do. It's not what they say they want. That's not the true intimacy you want with a customer. You actually want to observe them as much as possible and understand the why behind the actions they're taking.

Ezinne Udezue

The problem you focus on is probably the most predictive of your success.

Oji Udezue

You will never, never spend too much time communicating the why to your organization over and over and over again.

Ezinne Udezue
over 50 years
Combined product leadership experience of Oji and Ezinne Total experience of Oji and Ezinne Udezue in product leadership.
more code in the last one year than in the last 10 years
Oji's coding frequency in the last year compared to the previous decade Reflects Oji's increased hands-on engagement with technology due to AI.
six
Number of capabilities in a 'shipyard' team Refers to the distinct capabilities (PM, engineering, design, user research, data/ML/AI, product marketing) required, not necessarily six individual people.
three to five times, maybe 10x
Desired improvement for a 'sharp problem' To make a solution compelling enough for customers to adopt.
10x the cost (one tenth)
Desired cost reduction for a 'sharp problem' To make a solution compelling enough for customers to adopt.
about 5%
Percentage of early adopters for a new strategy within an organization Refers to the initial segment of an organization that quickly grasps and adopts a new strategy.
10
Number of growth levers discussed in the book 'Building Rocketships' Fundamental concepts for product strategy.
seven
Number of competitive advantage levers discussed in the book 'Building Rocketships' Fundamental concepts for product strategy.