Generative AI Usage: Where does it start... and end?
- 5 hours ago
- 6 min read
Updated: 4 hours ago
By: Gabriel Crawford-Paree
AI is increasingly present in our online lives, and the allure of adopting it to make our lives (and campaigns) easier as we try to promote our values can be compelling. This isn’t something we can take lightly. The truth is we, as Democrats, should be extremely cautious about using generative AI if we use it at all. This is true for anyone, but I believe it is important for the party of intellectual honesty, the party of labor, the party of marginalized communities, and the party of environmentalism in particular.
A little background
I do want to level set a bit what exactly I’m referring to, as it can get messy trying to decipher what people mean by AI these days. AI in the abstract can be just about anything from basic spell check to the newest model from OpenAI.
What I’m concerned with is generative AI.
This is any AI system which can take an input and generate a “new” or “novel” output. When I say AI moving forward, this is what you should think of.
Credibility is also important (more on that in a moment), so I want to assure anyone reading this that 0% of this has been written by AI apart from whatever Google Docs uses for spell check. I’m also not just someone with an opinion, I’m an electrical engineer by training with a minor in computer science, including multiple courses using various pre-2020 AI systems. I use AI models sparingly in my day to day work, primarily to support creating test scripts and as one avenue for early stage research. I’ve used it to help edit technical write-ups for clarity/brevity (having suggestions given, not rewriting it directly).
I’m not an AI absolutist, I’m a moderately well-informed AI skeptic who wants to see our party do the right thing in a shifting landscape.
Don’t Lose Your Voice to AI
One major risk to consider in using AI is your voice and credibility. Why should people listen to what you have to say? It’s increasingly difficult to get people to hear what you’re putting out; I’m facing that myself in trying to create a social media presence for my school board campaign.
While it may be tempting to just have AI make content for you, is that really you? If someone discovers you’ve used AI rather than make it yourself, will they think less of the message? Worse yet, will they trust the thing you write tomorrow? Credibility is a precious resource, and if you let AI speak for you, you hand your credibility over to it along the way.
I can say that personally, I stop reading things from people I come to find out are generating them with AI. That’s not even to mention the fact that leaning too hard on AI “slop” leaves you looking identical to the rest of the crowd doing the same thing. In a world increasingly mired in the same-ness of AI imagery and writing, a real human voice stands out.
Even if an AI output is “good enough” for your use case, you also run the risk of losing your own skills. This isn’t an immediate effect, but if we lean too hard on it for any particular task for too long we may lose the mental muscle to do it ourselves. The APA has a great article on the topic1.
The crux of the issue is this: if we blindly use AI because it’s “easy”, we may wake up one day and realize we’ve regressed in our thinking, writing, or creativity. Maybe I don’t have to write this whole thing myself (and believe me, I’m an engineer not a writer so I don’t really want to) but I would be doing myself a disservice if I allow myself to cease writing altogether. Human thought and creativity are valuable, don’t take yours for granted.
AI May Seem Shiny, but it Doesn’t Come Free
Looking outside ourselves, use of AI has an impact on the world around us. AI use doesn’t come free or even cheap, it harvests very real resources from the very real natural world. Microsoft has been restarting and building new fossil fuel power plants entirely to run its data centers2 . AI data centers require massive amounts of water to operate, and are threatening to be built using the water supply in our own backyards with the water supply from Eagle Creek3.
This isn’t to say we should use absolutely no AI, but we should strongly consider whether a given use is truly worth the emissions it costs. I don’t run a tractor for an hour to get a couple weeds out of my yard, and I don’t use AI for simple/unnecessary tasks either.
I don’t run a tractor for an hour to get a couple weeds out of my yard, and I don’t use AI for simple/unnecessary tasks either.
The original sin: Theft
Up there with the environmental impact is what I believe to be the original sin of the current AI models: intellectual and artistic theft. This isn’t an issue with the underlying technology itself, the technological leaps of modern AI really are amazing and could be used for the good of humanity. This also isn’t a claim around actual legal culpability, as we all know legality and morality are not one and the same.
The ugly truth is that ALL of the major AI models are built on scraped and stolen works by journalists, authors, artists, and anyone who put anything on the internet. AI models require immense amounts of data for what’s called “training”. Over millions of cycles of showing this data to the model, it is refined to reproduce things that look like, but are not exactly like, the data it was shown in the first place. This was done without the consent of the people who generated the data, and now it is threatening the livelihoods of those very people. Even the New York Times is in the midst of a lawsuit over use of its copyrighted works4. Regardless of the legal outcome, it cannot be ignored that an ethical boundary was violated.
Some things are more important that ease
This doesn’t even cover the implications on future works and those who would make them. One of the reasons it’s been generally accepted that generative AI use for programming is (mostly) okay is that although work was taken for training, those who had their work taken are broadly the ones using the tool.
That’s not true for writers and artists.
We need to think long and hard about where we stand on the dignity of labor. There’s a good reason we as democrats frequent union establishments over sweat shops, and it isn’t because it’s “easier” or “more efficient”. Some things are more important than being easy, and we need to support those who have had their works taken, if not by paying for their services at least by not allowing the tech companies who stole from them to profit off of it.
And margalized communities?
AI also threatens marginalized communities more than those of us in more privileged positions, myself very much included. This is true up and down the spectrum. Those of marginalized races who have already been impacted by redlining and other discriminatory practices tend to have less “valuable” neighborhood properties as a result.
Where do you think the AI data center is going?
It’s certainly not (as of this writing) going up in downtown Fishers, Carmel, Noblesville, or Westfield. How about those in economically challenging positions like our farmers? The impact on them of a diminished water supply will be much more harmful. How about the art and music that tends to arise from from queer and non-white communities? Drowned out by a wave of generative slop. Young people trying to get into the workforce? Sorry, but the entry level jobs have been outsourced to AI.
Democrats Need to Take a Stance
I do believe AI is with us now, and we can’t be luddites (though reading up on them, the Luddites weren’t exactly wrong). I also truly believe that the technology could be revolutionary and improve society in any number of ways. However, like nuclear power or the internet, the goodness or badness of technology comes from our application of it. Unfortunately, AI is being used to undermine creative labor, burning more fossil fuels than we would have otherwise, and the underlying structures the models are built on are rotten to the core.
We must demand better, and we are complicit if we mindlessly allow ourselves to fall in line with the companies ravaging our communities and environment. Democrats need to take a strong and unwavering stance not against the technology, but against these companies, and we cannot do so if we rely on them to do our thinking for us.
References
Gabriel Crawford-Paree
Gabriel Crawford-Paree is a lifelong resident of Hamilton County, and has set down roots in Westfield. He is running for Westfield-Washington School Board At-Large in 2026. As a firm believer in public education, someone invested in the long term health of WWS for his future family, and problem-solving engineer he is the right person with a unique perspective not currently represented on the school board.
Comments