AI Roadmap for Product Managers

My AI Product Management Journey, Skills, and Milestones

Over a year ago, I began intentionally learning and integrating AI into my product management workflow. What started as simple experimentation has evolved into a complete transformation of how I approach my role.

The next few years will be transformative for product management. Those who embrace AI while growing their core skills will be best positioned to lead the next generation of AI products.

Photo by Joshi Milestoner on Unsplash

I'm sharing my journey - with rough milestones, and skills acquired - to help you navigate your own path to becoming an AI-powered PM. Here's how my AI journey has been progressing, with selective highlights and benchmarks you can use to measure your own advancement.

Must-Have Skills (The New Baseline)

It is 2025, and product managers are among the top roles affected by AI. Here are the things an AI-powered PM should be able to do just to be on par with top performing peers:

1. You generate AI-augmented product docs, release notes, and other routine documentation that previously consumed hours of your week.

2. You use AI to create meeting summaries and craft sharp emails, freeing up mental bandwidth for strategic thinking.

3. You regularly use AI to generate comprehensive user stories and acceptance criteria, with custom templates that reflect your product's specific needs and organizational context.

4. You can generate diagrams for data visualization and communicate trends, ratios, insights with tools like Claude's visualization capabilities or ChatGPT code execution without wasting time in manual tools.

5. You use AI for sentiment analysis and to analyze customer feedback, support tickets, and user behavior patterns to identify trends and opportunities that would be time-consuming to spot manually.

6. You use AI content summarization capabilities of tools like GPT to digest long documents and extract key insights quickly.

7. You use AI to critique and get feedback on UI designs, documentation quality, feature proposals, and new ideas before sharing with stakeholders.

8. You use AI to generate scripts, test data, reports, testing APIs, creating quickstarts, demos, or automating repetitive tasks even if you're not a developer.

9. You know prompting techniques like ReAct and Chain-of-Thought.

10. You know what is an AI agent and understand its advantages over LLM alone.

Good Indicators (You're on the Right Path)

You might not have done these yet, but here are some indicators that you're making the transition towards an AI-assisted PM:

1. You have at least 1 paid AI subscription and know why it's worth paying for premium features.

2. You can articulate the differences between Claude and GPT based on actual usage.

3. You've created custom GPTs or Claude projects that solve repetitive problems in your workflow.

4. You follow at least 5 curated AI news sources (such as this newsletter).

5. You attend AI-focused events or courses and invest time in AI upskilling.

Warning Signs (Time to Course-Correct)

Warning signs that you're falling behind:

1. You haven't completed any AI courses or read any books on the topic.

2. You dismiss current AI by saying "we've been doing ML for 20 years already."

3. You actively discourage team members from using AI tools.

4. You penalize candidates who use AI during interviews as cheating.

5. You still call modern AI "glorified autocomplete" or dismiss it as a bubble.

Advanced Skills (Moving Beyond Basics)

These are signs of product managers having a higher impact thanks to AI:

1. You can write a blog post with some AI in your own voice that passes AI detectors (if you can't create content that sounds authentically like you, you haven't reached this level).

2. You've used AI to chat with at least 5 AI research papers out of curiosity.

3. You maintain your own prompt or context library collection for recurring tasks aligned with your specific needs.

4. You have subscriptions to 2+ paid AI products, have multiple API keys, and have tried out countless other services.

5. You have generated user stories with ChatPRD.

6. You instantly recognize repetitive tasks as AI automation opportunities.

7. You use tools like Perplexity and chat to find information more often than Google.

8. You conduct pricing trends analysis and competitive research with minimal manual effort using AI research tools.

9. You've created a concrete list of threats and opportunities AI brings to your organization.

10. You've convinced your manager to work part time on AI initiatives.

Expert-Level Skills (Reached AI-First Mindset)

At this level, you're educating others in your organization about AI.

1. You've passed 1000+ chat sessions.

2. You are paying for 5+ AI-native products or APIs.

3. You try out new models and features within days of release.

4. You tune your prompts to avoid hitting the limits of paid services.

5. You know exactly where AI can speed up your work and where its blind spots lie.

6. You've mastered parallel prompting and cross-model comparison, criticism, and collaboration.

7. You've built an internal AI community of practice in your org (through comm channels, regular calls, and brown bags in your company) and influencing others.

8. You're presenting your AI insights at conferences (my next goal).

9. You're starting to experience side effects from AI such as losing the ability to write clearly without AI assistance.

10. You aspire to become a Chief AI Officer as your career evolves.