Week 2 · Day 3 — The Nose

This article explores the concept of AI as a “Large Listening Medium” and its connection to historical data collection methods, such as colonial censuses and redlining, which may introduce biases. It proposes a protocol to make AI predict its own defaults before processing data, enabling users to verify its findings against its predetermined assumptions. The session aims to equip readers with a method to analyze user reviews and derive insights while acknowledging potential biases and leaving with a testable claim.

• AI is framed as a “Large Listening Medium” that reflects human assumptions.
• Traditional research approaches, including surveys and segmentation, have violent origins in colonial practices and discriminatory policies.
• Statistical significance is linked to figures associated with the Eugenics Society.
• A thousand reviews of a meditation app from 2017-2024 are used as pre-AI data.
• Ten methods for navigating AI are presented, drawing parallels to how intelligence manifests in nature and human intuition.
• A protocol is introduced to force AI to predict its defaults before analysis, allowing for a check against its predetermined conclusions.
• The goal is to produce a claim about a thousand strangers and identify a way to disprove it.
  https://bender.layer3.press/articles/8712eb0a-b081-4fa2-9fe9-6d89b03ff613