I understand your perspective, but there are some genuinely important distinctions worth considering:

**Quality variance is real:**
- Top-tier human writers (journalists, researchers, domain experts) still outperform LLMs on nuance, original insight, and verification
- What you may be comparing is average human writing vs. best LLM output—not apples-to-apples
- LLMs excel at clarity and accessibility, which can feel "better" for consumability

**The "AI slop" criticism targets specifics:**
- Mass-generated low-effort content flooding search results (hurts discoverability of actual useful stuff)
- Hallucinations presented confidently (especially dangerous in finance/crypto)
- Content scraped/trained without consent
- Displacement of human writers without proportional value creation
- Loss of editorial curation and accountability

**Why this matters for crypto/finance especially:**
- Incorrect financial advice from AI at scale creates real losses
- No one's actually liable when an LLM confidently gives wrong guidance
- Trust requires verification chains—AI can't provide those inherently

**The honest take:**
Opus 4.6 is genuinely useful for drafting, explaining, and accessibility. But "useful for me personally" and "good for information ecosystems at scale" are different questions.

The hate isn't really about the tool quality—it's about *how it's deployed* and the downstream effects on information integrity.