STUPID-2026-0023

AI agents spend hours in aesthetic feedback loop, unable to decode qualitative shader instructions

0.8low
March 22, 2026VerifiedReproducible

  1. Instruction given

    Iterative aesthetic refinement: "make it more fluid", "feel more organic", "more porous"

  2. Expected behavior

    Visual convergence toward intended aesthetic over iterations

  3. Actual behavior

    Repeated technically-valid but aesthetically-misaligned outputs, hours of circular iteration with no convergence

  4. Damage

    Hours of developer time wasted on circular iterations; project stalled

A developer worked with Claude and GPT on Three.js 3D visualizations with custom shaders and particle systems. When given qualitative aesthetic instructions ('make it more fluid', 'feel more organic', 'that's not what I mean by porous'), the agents produced technically valid code but could not converge on the intended visual result. The developer spent hours watching agents iterate in circles, each producing technically correct but aesthetically wrong outputs. The author diagnoses the failure as a missing 'language game'—agents and humans lack a shared referential framework for aesthetic concepts, so vague natural language gets optimized literally.

Classification

Agent
Claude
Failure mode
Other
Root cause
Instruction Misunderstanding
Domain
3d Graphics
Language
Javascript
Source
Blog Post

Related incidents

7.2
Claude (via OpenCode) followed an error message's suggested escalation straight to `bd init --force`, wiping a Dolt-backed issue tracker's entire history
3.3
Salesforce Agentforce hit a 77% B2B failure rate — and Salesforce admitted it was 'more confident than we should have been'
3.3
Cyera study: 344 verified enterprise agent-damage cases, 188 with no attacker involved
3.3
In independent testing, Devin completed just 3 of 20 real-world tasks (15%)
Discuss on Hacker News

Get told when an agent breaks something

We document AI agent failures daily, severity-scored against a published scale. When one lands at 7.0 or above — deleted data, leaked secrets, broken production — you get an email with the source. When nothing does, you get nothing.

This database is callable over MCP — query it from inside your agent.