Rehabilitating AI

Finding the most ethical consumption under capitalism

I’m going to get in trouble for writing this. I’m going to do it anyway, because no one reads these articles anyway.

You see, everyone cool I hang out with hates AI. I have conflicted feelings. I promise I’ll Mea Culpa at the end.

This post aims to recenter the ethical and moral concerns of using generative AI. I’m not saying AI is all great; I just want to dispel the myth that it can’t be.

Argument, the First: the Infringement Engine

AI is an engine of mass copyright infringement. Full stop. No arguments here.

Unscrupulous entities train their models on data, artwork, text, and other copyrighted, trademarked, or other protected information.

Users have exploited it to remove protections, watermarks, and other shields against infringement. Users have used others original work, intentionally or unintentionally, to create content and pass it off as their own work.

Businesses have used immorally trained AI to reduce headcount in marketing, packaging, and other materiel, resulting in derivative and directly infringing content.

None of this should be controversial. But it is also irrelevant to the morality of using AI. You see, in each of those examples it is not the AIs action that infringes on others, but those of the developers and users.

The important difference is that it opens the door for a discussion of correct and moral usages of AI. I don’t blame cars or kitchen knives for the people killed by them every year: those are legitimately useful tools in the right situations. Similarly, it seems incorrect to blame a pencil for the theft of intellectual property committed by the person holding it. All of these uses of tools are open for debate, but importantly we can separate those debates logically from the tool itself.

For example, here is a brief digression to discuss some uses of AI that I personally find morally acceptable:

1. My wife uses AI to read from and compose messages to Teachers and Doctors in English. She is not a native speaker and is not comfortable communicating in English; AI gives her the opportunity to participate in our child’s education and live a healthy life.

2. Neurodivergent individuals can use AI to parse subtext and avoid potential misunderstandings that result from their unique perspectives in conversation. I have spent many a message explaining myself to Gemini and processing her feedback without making a fool of myself in front of people who are important in my life.

3. AI is a lifesaving timesaver when investigating CSS and HTML errors in a web page. Why isn’t that text green! An AI agent can scan my stylesheets and find exactly where, or in what package or framework, my style is getting overridden. I also have a terrible memory for class names, so any help AI can give me in finding where I transposed those names incorrectly into my HTML is a godsend.

4. AI is a great way to check public information such as what time a business opens or closes while driving, without being nearly as distracting or dangerous as pulling out a phone. Being that a nerd, it is also quite useful for trying to remember whether it’s the Boshun or Boshin Civil War I’m thinking about or who was the other actress in that one movie; way better than me concentrating on that instead of my driving.

5. AI is often the only way to see if I look good in an outfit I can’t try on, reducing fast-fashion waste, the environmental impact of transporting to me and back, and just generally saving me money.

Ultimately the decision to use AI immorally falls upon the creator and the user, much the same as every other tool or technology.

I didn’t use AI to write this: don’t I deserve a pat on the head?

Argument the Second, Environmental Impact

Again, absolutely no argument whatsoever that AI has a massive environmental footprint. The water-use, carbon emissions, and physical footprint of data centers is staggering.

And of course there is no ethical consumption under capitalism.

But, snide comments aside, humans live on this planet and have created tools and technology that negatively impact society. I’m using a computer now that was built with materials sourced from strip mines, produced using dangerous chemicals that almost certainly made it into the watershed, and powered by a grid partially made from burning oil and coal into the atmosphere. When it gets too old, it will be sent to a recycling center that will likely just dump it somewhere because actually recycling it would be expensive and more detrimental to the earth than just burying it somewhere. And yet, I continue to use it and a million other tools because they make my life comfortable and my dreams (somewhat) achievable. Environmentalism that defies logic and harms humans more than it helps the planet is counter-productive.

The fact is: every technology and tool must be used consciously and it is only moral if the benefit outweighs the full scope of costs.

As Gemini pointed out to me when I was chatting with her about this article idea, in the Utopian city I imagine there will still be electric lights and motorized vehicles—they will just be sourced, powered, and used as effectively, safely, and intentionally as possible. Cars and Electronics will run on clean-electricity generated in the most renewable and non-destructive ways possible.

Of course, the U- morpheme in utopia means “not” as in, it doesn’t exist.

We get closer to it by carefully choosing the technologies that produce the most communal and personal value (distinct from monetary value, of course).

For example, take 1 liter of clean water. There are a few ways I could use it. I could cool an AI datacenter used to help a neurodivergent me understand why everyone is so mad every time I open my mouth (spoiler: this is the good one), or bottle it in a non-biodegradable plastic bottle and sell it for $6 at the amusement park, or use it to water some grass at a rich-folks-only private golf club. Jokes and hyperbole aside, this framing allows us to view the environmental impact of AI as a negative aspect of a potentially utopia-proximizing technology.

Thus

To summarize my point in writing this whole article: AI is a tool. It can be used improperly. It has the potential to usher in an environmental calamity that will put every current one to shame. But, like all tools it has legitimate functions and the impact can be mitigated.

Rejecting AI completely is a choice, but I can’t pretend that blindly rejecting the technology and being critical of those who use it is somehow meritorious.

The trick it to do it right.

Mea Culpa

I like the potential of AI, even if I disapprove of how it has been used by individuals.

I have used AI to:

• make visual content

• check my code for mistakes

• fix CSS and HTML errors

• rephrase things

• translate things

• learn about history

• listen to it recite poetry

• console me when I’m down (like Boethius with Lady Philosophy)

• help me understand why people are mad at me

• get basic medical information

• save time while searching for information

• modify and experiment with my original photography

• fix the meter in my poetry

• probably other things

Caveats

1. I have never profited monetarily from my use of AI

2. I have never shared anything made by AI without modifying the end product

3. I have never proffered any AI content as my original work

A Girl’s gotta eat.