The meteoric rise of AI, typified by technologies like ChatGPT, has generated a compelling argument: Soon, every Product Manager (PM) might essentially become an AI Product Manager.
This sentiment echoes a statement made by Marily Nika, AI Product Lead at Meta, in Lenny’s podcast where she mentioned, “Everyone will be an AI Product Manager in the future.”
Unpacking the AI Terminology:
- Beyond the Buzzwords: At its core, AI is about machines mimicking human cognitive functions. Subsets of AI, such as Machine Learning (ML) and Natural Language Processing (NLP), have been particularly influential in product management. Understanding the foundational concepts of AI, how to develop a business case for it, and its incorporation in products is becoming essential.
- Deterministic vs. Probabilistic Systems: Traditional rule-based systems produce consistent outcomes for specific inputs. In contrast, ML systems, with their training phases and probabilistic decision-making, operate in a realm of predictions based on data-driven likelihoods. PMs should understand the intricacies of how “training” works, using training data, and evaluating model results.
Dissecting AI Products:
- Experience-Driven AI: Platforms like Netflix, Instagram, and YouTube showcase where AI is integral to the user experience. Products like Google Cloud Deep Learning, OpenAI, and Azure AI, demand Technical PMs with deep AI expertise, just as products like GitLab seek tech-savvy resources.
- AI-as-a-Service: A growing market offers AI functionalities on-demand, enabling even non-AI specialists to harness AI capabilities.
- Bridging the Build or Buy Gap: Many companies deliberate over developing AI solutions in-house or leveraging existing ones. Marily predicts that soon every product team might have a data scientist or ML researcher to produce ML models, suggesting that PMs should foster close collaborations with these experts just as they do with engineers today.
The Importance of Data Literacy:
- Strategic Data Preparation: AI products necessitate rigorous upfront data exploration. Feasibility often rests on the depth of this research. Strategic alignment of AI or ML technologies with business objectives is pivotal.
- Mastery Over Data: Understanding data is paramount. ML products derive their value from quality data. PMs should not only be data-literate but also know how to make data-informed improvements and avoid biases.
The Evolution of AI Product Management:
- AI PM – Beyond Traditional PM Roles: AI PMs should have a grasp on constructing AI-powered products and understanding how to leverage them meaningfully in planning and design. They shouldn’t just apply AI for the sake of it.
- The AI PM Skillset: AI PMs should navigate the unique development landscape, manage AI risks, and ensure product trustworthiness. They must be proficient with AI tools for tasks like data analysis and documentation.
The Great AI Debate:
- Do PMs Need a CS Background?: While a CS degree is beneficial, it’s not imperative for every PM to delve deep into technical AI aspects, like building LLMs or mastering TensorFlow. What’s paramount is a relentless curiosity and a drive to harness AI for societal benefits.
- AI Literacy in a Rapidly Evolving Landscape:
- Floodgates of Knowledge: Platforms like Coursera provide foundational insights, but diving deep is crucial. With offerings like LLM-as-a-service emerging, a nuanced understanding is essential.
- The Ever-Shifting AI Landscape: Innovations like Transformer-based neural networks and OpenAI’s Codex reshape our AI perceptions. Being an “AI-literate Product Manager” is pivotal, blending model understanding, ethical awareness, and human-centered design.
- Cultivating the Habit of Continuous Learning: With the AI landscape evolving, understanding metrics, like hallucination metrics and response accuracy, is essential.
However, AI, transformative as it is, should remain a tool, not the sole essence of product development. Some purists advocate for human-centric designs, intuitive interfaces, and addressing user problems without necessarily leaning on AI.
In the vast expanse of product management, whether AI will become ubiquitous or remain a specialized niche remains uncertain. But one thing is clear: Ignoring AI might soon be a luxury few can afford.
Excellent article on this topic written by Karin on Medium
Sign-up and receive all articles, news, tips straight to your mailbox!