05/14/2026
🛑I am opting out of the recently announced Claude for Small Business solution and the broader "Big AI" ecosystem.
It sounds great on paper: Claude Pro works directly within real-world biz tools like QuickBooks and Google Workspace to automate tasks, leaving users only the final approval.
🚨Beware. We’ve been told Anthropic is the "safe" alternative. But slick branding shouldn't mask environmental vandalism & labor exploitation.
Anthropic and other Big AI companies are treating the South, where I live with my five-year-old son, as a sacrifice zone for energy & water. These inhumane practices need to stop.
Here is why I am making the shift:
❌ The Memphis Betrayal: Anthropic recently partnered with Elon Musk's "Colossus" supercomputer cluster in Memphis. This facility relies on dozens of unauthorized methane generators in a residential neighborhood to meet its massive energy demands. This is a real public health crisis for your neighbors (hi! it's me!) in the South.
Like many major AI providers, Anthropic doesn't share detailed public-facing reports on exact energy consumption data for model training or per-query usage. Whatchu hiding, friends, behind those closed models?
https://lnkd.in/e-jVrbdj
https://lnkd.in/en7W8cck
❌ The "Bazooka" Problem: As James Martin has explained, using a model like Claude Pro to auto-generate monthly P&L exports is "like using a bazooka to swat a fly." The environmental and human harm cost is STILL an unjustifiable hit on resources here in Louisiana and the rest of the country.
https://bettertech.blog/2026/05/09/how-to-quitbigai/
https://savethe.ai/water/
❌ Ghost Work: Massive models still require a workforce of human data labelers in the Global South who are often underpaid and exposed to graphic content to "train" the safety filters we take for granted.
https://lnkd.in/e8ceV3p2
AI Harm-Reduction Workflows: 🛠️
I am committed to learning more about AI Ethics, AI Safety, and AI Harm Reduction to make informed decisions for my business and creative practice.
In an upcoming newsletter, I will share what a realistic AI Harm-Reduction Tech Stack could look like. 📲Subscribe to receive it: www.dcapmedia.com
I am also creating a zine that expands on this post to make resources accessible.
❌ The AI literacy Gap: Many of us don’t need the "most powerful" AI; we can choose less harmful AI tools that are more life-affirming and community-rooted.
+Many of y'all are using AI for things that don't require LLMs or tokens. Here is one example of a FREE no-AI resource for common tasks:
💡delphitools: A collection of small, low stakes tools. No logins, no registration, no data collection.
https://tools.rmv.fyi/
Stay tuned for my zine!
💡Follow these folks at The Distributed AI Research Institute (DAIR) on IG:
https://www.instagram.com/ai_killjoy?igsh=aDYwdnA4MDBjaXRy
https://www.instagram.com/dr.timnitgebru?igsh=MTM4N3lodno5YmNoMw==
A collection of small, low stakes and low effort tools. No logins, no registration, no data collection.
