Someone else can give a more comprehensive answer, but here are a couple bullet points:
- Models trained by violating copyright
- AI data centers using tons of water and power, driving up costs for local residents
- AI data center noise pollution
- "Slop" results (uncanny valley pictures, incorrect answers given confidently, etc)
- Companies using AI as an excuse for anti-worker actions (notably massive layoffs)
- Concern over how much of the largest companies valuation is tied to AI hype (because it affects everyone's retirement accounts)