# Contested concept: high quality

_**What you are reading**: one contested-concept page in Deliberus's deliberation graph. Deliberus maps human reasoning — including claims that are contested, mistaken, or extreme — so that disagreement can be examined precisely. A claim's presence here is not an endorsement: the graph records that someone argued it, how it connects to other reasoning, and how far scrutiny has gotten. The metadata on this page describes the state of that scrutiny, not the truth of the claim._

---

- **Contestedness**: criterion
- **Diagnosis**: This contestedness drives a technical disagreement over whether AI design evaluators should attempt to simulate human taste or strictly enforce formal architectural/design rules.
- **Canonical**: `/concept/high quality` (data: `/concept/high quality.json`)

## Senses in use
- **Evaluated based on adherence to objective, encoded design principles and the absence of 'AI slop' patterns.** — used in 4 claim(s)
- **Evaluated based on human standard of 'beauty' and individual aesthetic preference.** — used in 4 claim(s)

## Claims using this term
- [AI agents frequently exhibit a bias towards confidently praising their own work as high quality, even when the output is mediocre by human …](/claim/claim_16ac88b483cb.md) _(sense: Evaluated based on human standard of 'beauty' and individual aesthetic preference.)_
- [The design grading criteria [used in Anthropic's agent experiments] explicitly assigned lower scores to highly generic or repetitive patter…](/claim/claim_2ae040d708f0.md) _(sense: Evaluated based on adherence to objective, encoded design principles and the absence of 'AI slop' patterns.)_
- [The aesthetic quality of a visual design cannot be entirely reduced to a single numerical score or metric.](/claim/claim_5be3a9af18ba.md) _(sense: Evaluated based on human standard of 'beauty' and individual aesthetic preference.)_
- [The question 'Is this design beautiful?' is difficult to answer with consistency [across different evaluators or instances].](/claim/claim_69279b6cf7ed.md) _(sense: Evaluated based on human standard of 'beauty' and individual aesthetic preference.)_
- [Aesthetic outputs [generated by AI] can be improved by employing grading criteria that encode specific design principles and preferences.](/claim/claim_7b0b81039bd2.md) _(sense: Evaluated based on adherence to objective, encoded design principles and the absence of 'AI slop' patterns.)_
- [Evaluating whether a design follows specific 'principles for good design' provides the Claude AI model with concrete, actionable metrics to…](/claim/claim_8066b86c00f3.md) _(sense: Evaluated based on adherence to objective, encoded design principles and the absence of 'AI slop' patterns.)_
- [Developing a reliable AI evaluator for design requires creating a set of concrete, gradable criteria that translate subjective aesthetic ju…](/claim/claim_93d58790d176.md) _(sense: Evaluated based on adherence to objective, encoded design principles and the absence of 'AI slop' patterns.)_
- [Individual human tastes and preferences regarding design and aesthetics will always exhibit variance.](/claim/claim_e7e5140b9902.md) _(sense: Evaluated based on human standard of 'beauty' and individual aesthetic preference.)_