AI's PR Problem: Flat Wages, Soaring Capital, and Public Backlash
David at DSHR's Blog argues that AI's public relations problem stems from a fundamental disconnect: while capital owners have captured nearly all productivity gains since 2000, they now expect workers to embrace AI—a technology that threatens jobs, privacy, water resources, and democracy.
Key Data: Flat Wages vs Soaring Capital
Paul Campos reports that median real wages have been flat for 25 years across all education levels:
- High school degree only: $968 (2000) → $980 (2025) in CPI-adjusted dollars
- Bachelor's degree only: $1,587 (2000) → $1,580 (2025)
- Bachelor's degree or higher: $1,705 (2000) → $1,747 (2025)
During the same period, the inflation-adjusted S&P 500 went from $1,394 (Jan 2000) to $6,688 (Sep 2025)—a 380% increase. The post argues stock owners have captured all productivity gains, leaving workers with nothing.
Why the Public Resists AI
The post lists documented harms the public associates with AI:
- Nudifying images, leaking passwords, increasing utility bills
- Consuming water resources, tax breaks for data centers
- Forcing 401Ks into an AI bubble, eventual bailout
- Eliminating jobs, destroying democracy, accelerating global warming
- Encouraging murder/suicide, autonomous drone targeting
The post notes that 49% of Bay Area billboards advertise AI companies, and all ads shown while watching critical videos were from AI companies.
Context: Laws Against Data Centers
This backlash is translating into legislation—laws against data center construction and permitting are passing, and tech CEOs get booed at commencements.
📖 Read the full source: HN LLM Tools
👀 See Also
Agent Infrastructure for SMB Operations: A White Paper from a QSR Operator-Turned-Builder
A 16-year QSR operator published a white paper arguing for a missing infrastructure layer between generic AI chat and vertical SaaS dashboards, with 8 skills on ClawHub, 1,500+ downloads, and one live deployment outside QSR.
RTX 4090 vs H100 for Fine-Tuning Llama-3-8B: A Cost-Performance Comparison
A developer tested fine-tuning Llama-3-8B on both an RTX 4090 and rented H100 instances. The 4090 setup cost $2,000 upfront and took 24 hours, while H100 rental cost about $80 and completed in 4 hours.
Qwen3.5-122B-A10B-MINT-MLX runs smoothly on M5 Pro with 64GB RAM
A user reports successful local deployment of the Qwen3.5-122B-A10B-MINT-MLX model on an M5 Pro with 64GB RAM, achieving 39.58 tokens/sec generation speed with specific VRAM allocation commands.
Anthropic releases Claude Code Remote Control for mobile development
Anthropic has launched Remote Control, a feature that lets Claude Code users control their local development sessions from mobile devices. Available initially to Claude Max subscribers, it requires version 2.1.52 and uses a QR code to sync sessions.