Every Product Manager Needs to Become an “AI Product Manager”

In 2020, while working at Google Assistant, I was introduced to Google’s AI model Meena, and was completely awe-struck by how it could generate language to converse back and forth with me instead of just giving canned replies. I had never heard the term “Generative AI” before, but as I dove into the field, I knew that it was going to fundamentally shift how humans interact with technology. These changes are already happening — as a Product Manager, here are three that I’m most excited about:

1: Non-Deterministic Systems Turn "Buses" Into "Taxis"

On the bus, everyone gets the same stops - the same predetermined experience. In a taxi, each user puts in their own destination, and the driver creates a personalized journey for them. Similarly, products will evolve from being deterministic as they are today, to non-deterministic, allowing for a range of different experiences for every user. We see this today in places like our individualized social media algorithms or Amazon recommendations, and we’ll continue to see more and more personalized experiences. 

2: Connecting the Dots through AI-Driven Data Analysis

Companies are filled with billions, if not trillions of internal data points. Every email, meeting, document change, CRM update, etc. represents the collective knowledge and historical decisions of a business. There is enormous opportunity hidden in this data, from optimizing processes to understanding an individual’s strengths and weaknesses to predicting outcomes. 

3: Supercharging Productivity in Everyday Tasks

As a “future of work” startup founder and product manager, I spend a lot of my day thinking about the inefficiencies of working in a modern team. Whether that’s breakdowns in communication that result in more meetings and more emails, or spending hours writing a weekly status update, the workday is full of inefficiencies that add up to massive amounts of employee burnout and lost productivity that costs companies millions of dollars. Generative AI has the potential to eliminate these inefficiencies and create a workday that is empowering and fulfilling. 

Product Management Has Already Changed — Now Product Managers Need to Catch Up

Source: McKinsey 

To make the future we just talked about a reality, we need to usher in a new generation of product managers with a new set of skills. Being a product manager for a non-deterministic product requires a different approach than for a deterministic one. It requires a new framework for developing product strategy, a new way of thinking about product requirements, and an understanding of a new kind of technology. 

There are three key areas all Product Managers will soon need to focus on:

1 Understand AI Models:

To effectively harness the power of generative AI into value-creating products, PMs need to deeply understand the abilities and limitations of the models. They need to become intimately familiar with what kind of tasks they’re good at, and bad at, and understand how to pull the various levers at their disposal to get the most out of them. The best way to gain this knowledge is through hands-on experience working with models such as GPT-4 by building apps, practicing prompt engineering, and developing a well-informed point of view. 

2 Identify Strategic Opportunities: 

Just because someone changed their website from a .com to a .ai or slapped a chatbot into their product does not mean they have a good product strategy (or any product strategy). In a frenzy to satisfy leadership or jump on the buzzy new bandwagon, we’ve seen an explosion of AI products and features - many of which are terrible. Product managers need to understand where the right opportunities are that are generated (no pun intended) by evolving technology,       what the associated risks are, and create thoughtful product strategies accordingly. 

3 Leverage DataI for Next Generation User Experiences

Tomorrow's PMs will need to understand how to unlock the value of the unique datasets their products generate, and how to harness valuable, usable, and unique data to train or fine-tune their AI models. Proprietary data not only gives products a competitive edge but also provides unique insights to their customers that no other product can. Whether it’s building AI products to create an ideal travel itinerary, review legal contracts, or optimize SEO, the unique data that teaches AI models how to fulfill these tasks is invaluable to set one product apart from the herd. 

Want to start today?

I partnered with Product School to help create the groundbreaking Artificial Intelligence Product Management Certification (AIPC). This program provides hands-on instruction to help students learn the skills and knowledge required to build the next generation of products, from how the technology works to how to create a product strategy for this new era.