It’s pretty much impossible to interact with media without seeing yet another hot take on generative AI.* A significant percentage of the LinkedIn posts from marketing and business folks extol the virtues of AI for streamlining your inbox, boosting your efficiency and profits, reducing administrative burdens and overhead, or some other absolute rubbish.
It doesn’t matter what benefit there is to you if the use of the technology is inherently unethical.
Ethical Issue #1 – Theft
There is no ethical use of generative AI when your program is built on stolen content. In order to function, generative AI needs to be trained on a massive amount of the type of material it is expected to replicate. For writing, each program has its own large language model (LLM). Every general use generative AI currently available has used the intellectual property of authors, journalists, fan writers, and bloggers to build these LLMs without permission, credit, or payment to those creators. The intellectual property being used is protected by copyright, and this use is theft.
Text generative AI is a gigantic plagiarism machine. Using it means you condone theft.
Ethical Issue #2 – Substitution
There is no ethical use of generative AI when its intended purpose to avoid paying skilled professionals for necessary work. It’s being promoted as cheaper than hiring writers, editors, photographers, and other professionals (and right now so many people are using it free of monetary charge). It’s especially galling that generative AI companies are using these creators’ works (without compensation) to simply replace them.
About two years ago, I attended a panel discussion on LinkedIn that was allegedly going to cover the ethics of using generative AI. In the end, the panelists just spent all their time blathering about how generative AI was not going to take our jobs away, there would be no mass layoffs, and AI would simply help us get through our email and the tedious tasks that didn’t take advantage of our true skills. Even at the time, I knew this was garbage.
The market is flooded with skilled and talented writers and editors who used to have great jobs. The bulk of the writing and editing jobs posted right now offer entry level pay, and most of them involve training those LLMs, so continued training of our replacements. I had a thriving freelance writing and editing business for six years before Chat GPT came along, and now I’m having to embrace other options, most of which pay less. I can no longer use decades of intentionally honed skills to make a living. The careers generative AI is expected to overrun are expanding beyond creatives, so your job could be next.
Text generative AI is a massive workforce displacement machine. Using it means you condone wage suppression and don’t value the people or skills necessary for the work you need done.
Ethical Issue #3 – Environment
There is no ethical use of generative AI when your program has significant negative environmental impact. This technology hits multiple environmental domains, water use, electricity, and waste.
An LLM needs the computing capability of an entire data center, and these run hot. Generative AI requires an obscene amount of fresh water to keep its computing systems cool enough to run. Current systems are estimated to use over 41 gallons of water per second. The exact amounts of water drawn are often concealed from the public through privacy agreements. This is a massive warning flag; the only reason to hide this information is because it is objectionable. Even without the definite numbers, we have seen how this water consumption strains municipal water supplies. Residents in some Texas communities faced heavy water restrictions this summer, restrictions that limited their bathing, while AI data centers freely consumed water without any oversight.
We’re seeing the construction of data centers all over the United States. This power-hungry generative AI infrastructure exceeds what most power grids can accommodate without expansion. This generally results in price hikes to nearby communities and a need for more fossil fuels. It is estimated that a single generative AI text prompt will use more electricity than a household refrigerator uses in the entire day. In addition to the end user electrical needs, the training of an LLM requires a massive amount of computing power and electricity. There is no true finish line for LLM training; between development and continuous updates, training is unending.
The final environmental impact of generative AI is waste. The machines running these systems require routine replacement. The need to stay at the highest possible end of computing and the large scale of this equipment means these data centers will overload our already critical e-waste management processes.
Text generative AI is an enormous waste of resources. Using it means you prioritize unnecessary convenience or profit over the well-being of the planet and its inhabitants.
Generative AI is actively harmful to creators, workers, and the planet. There is currently no ethical use of this technology. If you’re using generative AI for your work, you’re signaling to your customers and the public that you are an unethical hack. If you don’t have the skills to do something, hire a professional.
*Reminder, analytical AI and generative AI aren’t the same. While there is no ethical way to use generative AI at this time, analytical AI can serve valuable purposes in many fields and isn’t mired in the same ethical issues as generative AI.
Need sources for some of this info? Great! Thank you for being a responsible netizen.
- All gen AI is inherently theft – James O’Sullivan
- Authors’ Class Action Lawsuit Against Open AI Moves Ahead – Publishers Weekly
- Theft is not fair use: Artificial Intelligence companies have a copyright problem and they know it – Stanford University
- AI could deplete drinking water sources around Illinois, Midwest – Chicago Sun Times
- AI Drinks Water – aidrinkswater.com
- AI’s water problem is worse than we thought – HEATED
- ChatGPT needs to ‘drink’ a water bottle’s worth of fresh water for every 20 to 50 questions you ask, researchers say – Business Insider
- Environmental Impact of Generative AI: Carbon and Water Footprint – Cal State
- Explained: Generative AI’s environmental impact – MIT News
- Generative AI’s environmental costs are soaring – and mostly secret – Nature
- How Much Water Does ChatGPT Consume? – Inside Tech World
- Texas AI centers drain water supply as residents face drought restrictions and shower limits – News Target
- Texas Data Centers Use 50 Billion Gallons of Water as State Faces Drought – Newsweek
The Write Mann, LLC 
1 thought on “There is Currently No Ethical Use of Generative AI”