Taste Is Not Preference
Taste has become an increasingly common word in technology, often used as shorthand for good aesthetics or an instinct for what is cool.
I mean something more specific.
Taste is the learned ability to judge what is right in relation to an aesthetic logic, a particular context, and an intended outcome.
Preference says, “I like this.”
Taste says, “This belongs here.” “This is almost right.” “This should come later.”
The core beneath taste is discernment: the ability to perceive meaningful differences and understand which of them matter. In music, this might mean recognizing that a track is beautiful but wrong for the moment, or that the direction is right but the transition came too quickly.
Taste can feel instinctive, but it is really a compression of lived experience into judgment. Thousands of hours of listening, cultural context, memories, trusted friends, previous rooms, and repeated experiences of what worked and failed become distinctions we can feel before we can explain them.
It is socially learned but personally expressed. We inherit ways of hearing from friends, lovers, DJs, artists, scenes, and cities, then internalize and recombine those influences into something distinctly our own.
Taste is also contextual. A track is not right in isolation. Its meaning depends on the person, the room, what preceded it, where the experience is going, and what the choice makes possible next.
And taste is not only a record of who we have been. It contains an aspiration toward who we are becoming. A system that perfectly predicts our past behavior may personalize well while confining us to a statistical portrait of our previous selves.
This is the distinction between revealed taste, what has consistently felt right, and directed taste, what we are trying to move toward.
This is also where Oto departs from conventional recommendation. Most music recommendation systems learn what someone is likely to consume next. At Oto, we're more interested in how someone exercises discernment to shape an experience.
Language makes that discernment legible. Someone might begin with “make it warmer,” hear the result, and clarify that they want the same rhythm and energy, but a softer emotional register. Through that loop between hearing, judgment, articulation, and direction, the system develops a richer understanding of the person’s taste while the person becomes more precise about their own.
The goal is not to automate taste. It is to make human discernment actionable.
The previous era of music technology solved access to sound. Sound is now abundant, even overwhelming. The next will be defined by better systems for expressing judgment, intention, and direction.
That is what we are building at Oto.