Best AI Resources for Product Managers

Why AI matters for product managers

The single most cited finding in this reading list is also the most useful constraint: in a controlled field experiment with Boston Consulting Group, AI assistance raised task quality by 40 percent for consultants working inside their skill's "jagged frontier," and made performance 19 percentage points worse for tasks outside it, even when the tasks looked similarly hard. Builders Camp's AI Prompting for Customer Discovery bootcamp builds its entire practical challenge around a version of that same discipline: turning a fuzzy assumption into a testable hypothesis, using AI to draft a sharper interview guide, then synthesizing evidence into themes that are still traceable back to a real quote, not a plausible-sounding AI summary.

The bootcamp's own microlessons, delivered across 2 live sessions and 10 self-paced lessons, are explicit that the job is not to skip talking to users. It is to show up to the interview sharper and synthesize the transcript faster, while keeping every insight traceable to evidence a stakeholder could check. That is the same discipline the jagged frontier research points to: know which part of the task AI actually improves, and do not hand it the part where it does not.

Which research explains where AI actually helps knowledge work?

• The Jagged Technological Frontier, Harvard Business School (Dell'Acqua, McFowland, Mollick, and co-authors). The field experiment behind the 40 percent quality gain and 19 point performance drop, run with 758 BCG consultants.
• Centaurs and Cyborgs on the Jagged Frontier, Ethan Mollick. The essay that coined the terms for the two dominant ways people actually integrate AI into their work.

Which primers explain how the technology itself works?

• How Large Language Models Work, Andreas Stöffelbauer, Medium (Data Science at Microsoft). Covers the three training phases behind an LLM in plain language, without requiring a machine learning background.
• Language Models as Agent Models, Jacob Andreas, MIT CSAIL. An academic paper arguing LLMs infer properties of the "agent" that produced a piece of text, which shapes how they generate a response.
• What's the Big Deal with Generative AI?, Cohere. A practitioner-facing explainer on what generative AI actually does differently from prior machine learning.

Where can PMs find ready-to-use AI prompts and use cases?

• Prompt Engineering Guide, promptingguide.ai. A continuously updated reference covering prompting techniques, model-specific guidance, and current research.
• Prompt Library, Anthropic. A curated set of over 60 tested prompts across use cases, each using structured formatting members can copy and adapt.
• The Top Use Cases of AI for Product Managers, Zeda.io. A practical rundown covering feedback analysis, A/B test analysis, and personalization use cases with a concrete conversion-rate example.
• AI for Product Managers: How to Stay Ahead, Userflow. Frames the AI-for-PMs question as two separate skills: using AI tools to do the existing job faster, and building products for a world where AI agents are also users.

How Builders Camp teaches AI for product work

AI Prompting for Customer Discovery is a one-week bootcamp inside the AI Product Expert and Discovery Expert tracks: 2 live sessions, 3 hours taught, and 10 microlessons covering hypothesis-driven discovery, interview planning prompts, and synthesis into decision-ready outputs. Its practical challenge has members run one full loop, from a stated assumption to a synthesized decision, with every theme still tied to a direct quote or observation. See the AI Prompting for Customer Discovery bootcamp for the current syllabus, or the 10x Productivity with AI bootcamp for the broader AI-for-PM-work track.