Stephen Zerfas , co-founder and CEO of Jhourney , believes advanced meditation has a feedback problem. Meditation is largely invisible and nonverbal: unlike a swimmer who can see a faulty stroke or a musician who can hear a wrong note, a meditator may not know whether a practice is working or what to try next. Zerfas believes that uncertainty causes many people to plateau long before they experience meditation's deeper possibilities. Jhourney is trying to close that feedback gap, beginning with jhanas, a family of powerful meditative states described in early Buddhism. On its retreats, the company combines intensive practice with two forms of personalized instruction. In structured interviews, students describe what they tried, where sensations arose, how they changed, and what emotions accompanied them. A facilitator pattern-matches across previous students and helps design the next experiment. In “dyadic meditation,” a student meditates aloud while a facilitator suggests where to direct attention next.
Jhourney is exploring how AI could carry both forms of instruction beyond the retreat. For structured interviews, it is developing a tutor grounded in its curriculum that can personalize recommendations; for tacit instruction, it is examining how thousands of hours of facilitator sessions could help an AI coach respond as a person narrates an unfolding meditation. A powerful state can arise within one session, allowing the system to connect a particular suggestion with the practitioner's immediate report.
That learning loop could make high-feedback meditation instruction more reliable, available around the clock, and accessible to parents, workers, and others who cannot spend a week with an expert teacher. Zerfas also believes the underlying experiences are not exclusively Buddhist. He sees parallels in Christian mysticism, Sufi writing, Eastern Orthodox practice, and other contemplative traditions, describing jhanas as universally available features of the human nervous system.
For Zerfas, the larger goal is to move from “altered states” to “altered traits”: a temporary experience of joy, peace, love, or equanimity can become practice for longer-lasting emotional fluency, better decisions, stronger relationships, and a different baseline of wellbeing. AI could make the learning loop around a difficult, historically opaque practice faster, more personal, and available to many more people.
