Although a Western visitor to Bhutan can in theory travel independently, in practical terms a guide is necessary in order to negotiate the language and culture, as well as to be granted access to many of the places one might wish to visit. As a result, for most of visitors the experience is heavily mediated by a guide. If our experience of visiting is anything to go by, guides can be friendly, knowledgeable and adaptable - a good thing, and integral to the experience. A good guide is essential when you're visiting dzongs. Dzongs are strategically placed fortified monasteries and the ones we visited in October were very well preserved or in some cases carefully, even lovingly, restored. Decoration, particular in terms of wall painting, is exquisite and uses a distinctive regional colour palette. But iconographic interpretation is complicated, sometimes even challenging the advanced knowledge of an experienced local guide - ours was both knowledgable and experienced. I was, however, surprised to see our guide using AI to solve one particularly knotty problem. This intrigued me and from that point onwards I started using AI, too - sometimes to check and at other times to add to my own knowledge. I found that ChatGTP was pretty knowledgeable about some quite remote areas of Bhutan, and reasonably reliable in commenting on iconography. On one or two occasions I felt like I’d run into a wall, as if I’d found the edge, the furthest reach of AI, where I knew more or at least knew that there was more to know. It was like being in a large room in the half-light and then suddenly, unexpectedly touching the wall. But the way in which an AI app can become entangled with people on the move, with niche representational practices in remote places is impressive. I quickly realized that we really did need a guide, but found myself wondering how AI might change what you need a guide to do and even what a guide needs to know in the first place, and by extension what any of us need to know. The digital materiality of devices, the infrastructure of connectivity, the labour of production and of software engineering, the more-than-human flows of energy and resource all seem to coalesce in the interpretive work I've mentioned – but yet it seems as if this coalescence alters the place you’re in, how it is seen and how it is experienced. Everything is placed in a new kind of relation – phone, guide, dzong, knowledge, movement – the list is endless. Should we worry?
