There’s something disarming about the premise of AI Mozart: an evening of AI-generated compositions brought to life not by a machine, but by human hands at the keyboard. Yet in the performance by Matthew Shiel, technology never felt clinical or cold. Instead, it became the starting point for an expressive musical journey.

From the first notes, the mood settled into a quiet, almost contemplative space in the wonderful surroundings of the Pilgrim Uniting Church, lit only by an array of candles. The program moved fluidly across genres – from the elegance of classical forms to the jaunty syncopation of ragtime and the easy familiarity of pop. But this was far more than a technical showcase. In the ragtime numbers, Shiel’s playing lifted with playful charm; in the classical-inspired pieces, it carried poise and depth; and in the pop-influenced sections, there was a relaxed warmth that drew the listener in.

What truly astonished me was Shiel’s technique. He told the audience that AI-generated scores can often be rhythmically tricky or harmonically unpredictable, yet he navigated every twist with effortless precision. Rapid passages sparkled with clarity and delicate melodic lines were shaped with care and restraint. It felt as though he had lived with each work for weeks, not merely learned it – every phrase carried the confidence and polish of long rehearsal.

It was disappointing that the projections on the rear screen were the same for all styles of music and were repetitive. If, as I assume, they too were AI generated, more variety would have been much better with the images reflecting the type of music. By the end of the performance, the concept of ‘AI music’ seemed almost secondary. What lingered was the artistry – the sense that in Shiel’s hands, AI Mozart became not a novelty, but a genuinely engaging and memorable musical experience: thoughtful, skillful, and quietly captivating from start to finish.