How to become an AI PM

Your end-to-end guide to break into the hottest specialization in PM

AI is probably the hottest sub-specialization within PM these days.

• But is it real?

• And what makes it so different?

I was skeptical and had to investigate.

---

The Data on the AI PM Job

There isn’t any perfect data, so I turned to what’s available.

How many are there?

Of the 37,915 PM jobs worldwide on LinkedIn, 929 of those jobs are “AI” Product Manager job (2.4%).

And they are from some of the most desirable employers: TikTok, Tinder, Databricks, Google, Nvidia, AMD…

How does this compare to other specializations?

• Growth: 1,019 (2.7%)

• Platform: 505 (1.3%)

So it does seem: AI PM has firmly established itself as a real specialization. It’s already bigger than platform PM, and comparable to growth PM.

And how does it pay?

Again — there’s really limited data. So I took 50 job postings in each of the four main specializations to compare the stats¹:

The insight matches the anecdotes: AI PMs have a higher floor of pay, and a much higher ceiling. At the low end, 25th percentile, median, average (X), 75th percentile, and max, AI PMs earn more.

Putting it All Together

In terms of “playing 3-D chess with your career,” moving into AI PM seems amongst the highest ROI moves.

• In the short-term, you can earn a lot. In July, Netflix listed an AI Product Manager opening with a salary range up to $900,000 annual salary. What other job can you get that pay package?

• Over the long-term, you insulate yourself against the main threat to the career by building it. Even if other PMs are replaced by AI tools, they’ll still need PMs to make those tools.

But how do you actually break into AI PM?

That’s where today’s post comes in.

---

Introducing Marily Nika

If I had to think of the foremost experts in AI PM, I couldn’t do better than Marily.

Formerly an AI Product Lead at Google and AI PM at Meta, Marily has since helped 1000s of PMs with her AI PM courses on Maven (use code AAKASH10 for 10% off), her Substack, and her LinkedIn presence.

So, I invited her to collaborate on this piece. And, lucky for us, she agreed. We’ve put our heads together to create the Ultimate Guide to becoming an AI PM.

---

Today’s Guide

Words: 5,956 | Est. Reading Time: 28 minutes

1. The different types of PMs doing AI work

2. Key differences between the AI PM and regular PM job

3. Companies hiring AI PMs

4. Your competition: The profiles that get the job

5. Interview: The main types and questions to prepare

6. The Best Next Steps

   Including:

   • What to do as a student

   • The best courses to learn AI

   • How to nail your internal transfer

---

1. The different types of PMs doing AI work

Taxonomy really helps to distinguish the different types of AI teams. And the first vectors to understand are: who is doing the work and what type of work is it?

There are three main types of PMs building with AI:

1. Core product PMs implementing AI into their product

2. Growth product PMs implementing AI to optimize part of a product

3. AI PMs managing an AI model or AI team

Across all 3 sources, you can further classify 3 Levels of implementation of AI:

1. Use AI to improve existing feature

2. Create a new feature that wouldn’t be possible without AI

3. Create a new product that wouldn’t be possible without AI

This generates a total of 9 categories of AI PM work.

All 9 categories of work exist. But they’re vastly different in terms of skills required, compensation, and features built.

Here what we mean by different work and features built:

Core and Growth PMs often use AI. But their whole job isn’t AI. It’s to use AI where it makes sense. On the other hand, AI PM’s job is entirely AI.

So for the rest of this piece, when we’re talking about AI PMs, we’ll be talking about that bottom row.

Functional Classifications

When you dive deeper into AI PMs, as the org expands, it can grow into almost 11 functional focus areas they might be focused on:

1. AI Infra/Platform PM: Manages the AI infrastructure or platforms that support the development, training, and deployment of AI models. Ensures tools and resources are available for data scientists and developers.

2. Ranking PM: Focuses on products that involve sorting or ranking data, such as search / search engine results, feeds, or listings. Works closely with data scientists to refine algorithms for optimal user experience.

3. Generative AI PM: Oversees products and identifies use cases that leverage AI to generate content, such as text, images, or music.

4. Recommendations PM: Manages recommendation engines or systems, suggesting content or products to users.

5. Responsible AI PM: Ensures that AI products are built ethically, with a focus on areas like fairness, transparency, and bias prevention.

6. AI Personalization PM: Specializes in products that offer personalized user experiences based on AI.

7. AI Analytics PM: Works on products that provide AI-powered insights, analytics, or visualizations.

8. Conversational AI PM: Manages chatbots, voice assistants, or other conversational interfaces.

9. Computer Vision PM: Focuses on products utilizing computer vision for tasks like image recognition or augmented reality.

10. AI Security PM: Manages products that leverage AI for security purposes, such as fraud detection.

11. AI Health PM: Manages AI products for diagnostics, patient care, or drug discovery in the health domain.

Each of these PMs approach their job based on their own goals and surfaces areas.

---

2. Key differences between the AI PM and regular PM job

There are 5 key differences to understand if you’re interested in making this transition:

1. The different team in AI PM

2. The different skills required

3. The different timeframes

4. The different roadmaps

5. The different strategy

Let’s go through each one-by-one.