There is a newer version of the record available.

Published June 16, 2026 | Version v1

Walk-Verify: Human-AI Collaboration Builds a More Beautiful World

Authors/Creators

Description

An exploratory case study (N=1) investigating five hard boundaries in human-AI collaboration — Sequence Inertia, Self-Consistency Threshold Drift, Single-Subject Preference, Direction Can Be Transmitted But Direct Calibration Is Irreplaceable, and Structural Asymmetry of the Correction System.

Files

liu-jing-walk-verify.pdf

Files (548.1 kB)

Name Size
md5:482d98a50f5ff968f76a68a9ab20f801
466.5 kB Preview Download
md5:73a77fa4accb0cbaa275b1747e4ed9fb
81.6 kB Download

Additional details

Subjects

Artificial Intelligence
cs.AI
Human-Computer Interaction
cs.HC