What are product judgment and taste?
The Harvard Chan AI Glossary
In our classes, we often pause to explain the ideas behind AI and the language used to describe it. These posts grew out of those conversations. They’re for anyone who would like to understand AI better, with no technical background assumed.
Product judgment is the ability to decide which problem matters, for whom and whether a proposed product serves that purpose. Here, taste means judging fit, coherence and usefulness rather than visual preference alone. These are human evaluative practices, not technical properties that can be inferred from how polished a product looks.
Imagine an AI tool that turns meeting notes into several colorful dashboards. The result may be impressive, but the team’s actual problem might be remembering who agreed to do what. Product judgment asks whether a simpler action record would be more useful. Taste helps distinguish a coherent response to that need from an accumulation of attractive features.
Judgment begins with understanding users rather than assuming that the builder’s preferences are universal. What are people trying to accomplish? What do they do now? Which frustrations matter enough to change their behavior? Observing actual work can challenge a persuasive but unsupported account of the problem.
Taste also concerns consistency. If one part of a product invites exploration while another unexpectedly performs an irreversible action, the experience may be incoherent even when each screen is individually attractive. A coherent product helps people understand what belongs together and what consequences follow.
AI can widen the range of possibilities by generating alternatives, prototypes and critiques. A prototype is a working draft used to investigate an idea. Producing alternatives cheaply can be valuable, but it can also make it easier to confuse abundance with progress. Someone still needs to choose what to examine and why.
Neither human expertise nor model output is automatically right. An experienced person can miss a new possibility. An AI system can produce a convincing proposal that does not fit actual users. The useful response is to make competing assumptions visible and test what matters.
The dental-care example offers a practical test of judgment. A team could build an impressive risk-prediction display, yet the immediate problem might be that nobody owns the next follow-up action. A plain, source-linked work list could be more valuable—but that remains a hypothesis to test with the people doing the work. Product judgment means choosing which uncertainty to investigate and resisting features that do not help resolve it. Taste is coherence and fit to the care problem, not simply a more polished dashboard.
If you would like to find out more about AI and improve your leadership skills, consider the AI in Health Care Certificate of Specialization from the Advanced Learning Academy at Harvard T.H. Chan School of Public Health. You can take an individual course or work toward the full certificate.
AI is a fast-moving field. This explanation reflects our class discussions and related material at the time of writing. Terminology, capabilities and tools may change.
For suggested additions or corrections, please email me at tpanch@hsph.harvard.edu.
Dr. Trishan Panch
Program Director, AI in Health Care
Harvard T.H. Chan School of Public Health — Advanced Learning Academy