We want to understand delusions (fixed false beliefs) and hallucinations (percepts without an external stimulus). Of particular interest is the notion that these aberrant beliefs and experiences occur on a continuum in the general population - from health to illness. One key question that attends such an approach is where to draw the line, between delusions and conspiracy theories for example, beyond which a belief or experience is clinically actionable. Whilst we take an individual symptom approach, we appreciate that delusions and hallucinations co-occur. We are interested in these patterns of symptom co-occurrence, and in particular what they reveal about underlying computational and neural dysfunction. We are also interested in trying to help mollify symptoms. One intriguing idea is that perceptual experiences - including those of oneself and others - may be really helpful for people with psychosis. We think making music and performing it in groups might have preferential access to and impact upon our brains’ prediction making mechanisms. Having an experience of predictability and synchrony with others might be particularly helpful for people with psychosis. We are currently testing this idea with a clinical trial, in collaboration with Musical Intervention, funded by the NIH and the Kennedy Center.