moral surrender, the environment, and generative AI

Last week, I blogged about how the purported inevitability of generative AI gets used to sidestep moral concerns about it. Earlier this morning, I shared a link to a story from The Verge that illustrates that perfectly, and so I wanted to write just a little bit more about it.

First, let’s quote some more from Jacques Ellul, whom I referenced in the last post (and whom I’ve just been referencing a lot in general recently). Right before the passage I quoted in last week’s post—a passage from his 1980 essay The Power of Technique and the Ethics of Non-Power, he had this to say:

The reality, however, is the insidious ethics of adaptation, which rests on the notion that since technique is a fact, we should adapt ourselves to it. Consequently, anything that hinders technique ought to be eliminated, and thus adaptation itself becomes a moral criterion.

It’s remarkable—especially for a blurb written decades before the creation of generative AI—how well this predicts what Trump said earlier this week about increasing energy production in the U.S.:

“You know, we need to do the AI, all of this new technology that’s coming on line,” Trump said on Tuesday during a signing ceremony for all four executive orders. “We need more than double the energy, the electricity, that we currently have.”

We need to do the AI, and therefore we need more than double the energy. These aren’t just Trump’s words, either. The article from The Verge quotes Eric Schmidt—former CEO of Google—as using near identical language, and OpenAI has used a similar argument to say that they need to be able to scrape copyrighted works if America is going to stay on top of the AI races.

As Ellul puts it, “technique is a fact,” and therefore, the rest is all a necessity. No need to ask ourselves if the cost of energy production is too high to pay, we just gotta adapt, adapt, adapt.

• Jacques Ellul
• morals
• ethics
• generative AI
• Donald Trump