Summary of Facts, Concerns, and Links about AI:
AI is environmentally unsustainable:
- A single large data center can consume as much water daily as a medium-sized town to cool their computers. U.S. data centers consumed 17 billion gallons of water in 2023, with projections indicating this could quadruple to 68 billion gallons by 2028 as AI workloads intensify cooling demands.
- Data centers could soon account for up to 40% of local electricity consumption. ChatGPT’s 300 million weekly users collectively consume 621.4 MWh daily, equivalent to powering 35,000 U.S. homes annually. Data centers now consume more energy than entire nations, and AI workloads are responsible for a rising share of that demand. By 2027, AI electricity use could match Argentina’s entire grid consumption. By 2030, AI could consume as much electricity as 22% of U.S. households each year. Today, AI demand already rivals Ireland’s national grid.
- 60% of AI workloads operate on non-renewable energy consumption.
- More than 40% of the world’s data center capacity is located in the US.
- AI’s national emissions already surpass the aviation industry’s carbon footprint and is climbing fast.
- Not only creating more pollution and more global warming, it creates a cycle that results in less energy and water for more important things, like food and drinking water. For example, the Colorado river only now has 57% less water in it than usual because of less snow in Colorado and Montana, and provides water to 40 million Americans; billionaire tech companies are buying water rights (Just ChatGBT to the tune of 39 billion gallons of water daily) and get priority over towns, who were already running short before, and the situation will only compound as more people use AI.
When AI talks to itself we will no longer be in control:
Ex-Google CEO Eric Schmidt shares a blunt warning about AI: pull the plug.
When AI agents Talk to AI Agents
Weaponry and surveillance with no human oversight:
Recently, Claude (Anthropic) refused, but ChatGBT (OpenAI ) was quick to agree (to the tune of $200 Million) to sign a deal with the Pentagon that supports mass civilian surveillance and AI operated weapons that make decisions about when to deploy without any human oversight. (short clip about that)
Breaking up with ChatGPT and move to Claude – how to take your data with you
Transferring from ChatGBT to Claude
Other issues:
There will be an estimated loss of 100 million white collar jobs. Your conversations and data are no longer private. Teens in large numbers are turning to AI “companions.” Currently there is no regulatory oversight; what legislation is needed? How will Ai effect democracy? Warfare?:
Watch Bernie Sanders summarize these issues
There are many more issues:
- Simulated reality now can make things look true that are not. This is already being used for intentional misinformation and propaganda, as well as entertainment.
- There is a lack of transparency among AI corporations. The current stance from current American leaders is to give them carte blanche with no guardrails.
- AI is changing education, doing kids’ homework and writing papers for them. Learning is getting outsourced. Plagiarism has become easier and more difficult to stop.
- It is difficult to fully disengage with AI because it is woven into every update on your computer or phone.
- etc!!!
All this is troubling, and snowballing very quickly. I advocate trying not to use AI, or at least switching to Claude, which as a company seems to have more of a conscience. For environmental reasons alone, I think we should all side step this as much as possible.
Educate yourself and others. Demand safeguards from your legislatures. Switch companies or better yet, abstain. And be alert to your duty to pray about this!
