Extended Intelligence II
Doing this exercise—where we didn’t just learn about LLMs but actually stepped into the role of the people behind one—felt genuinely important. It shifted my relationship with AI from something I passively use to something I actively question and understand. That shift alone made the process feel worthwhile, especially in thinking about how we might engage with these systems more thoughtfully in the future.
Artificial Intelligence is everywhere right now, so embedded in our everyday lives that we’ve almost stopped noticing it. Yet with this growing presence of LLMs comes an underlying paranoia. As with anything new, there’s a mix of excitement and discomfort—it’s powerful, intriguing, but also very #unknown#.
Our group chose to build a fortune-telling LLM. Like most ideas, it began as something much larger: a camera that could read facial pointers and generate a printed, fortune-cookie–style output. Due to time and technical constraints, this idea had to be let go. What stayed, however, was the question of intimacy. A fortune teller needs a personal touch—something that makes the interaction feel connected to the person standing in front of it. This led us to body temperature as the input. By placing a finger on the temperature sensor, the LLM calibrates both the environment and the individual, and generates an output in response. In doing so, the system reflects not just data, but the subtle negotiation between the user, the machine, and the space they share.