This article was in part inspired by my friend Josh Ginsberg’s piece on Generative AI. Give Josh a read and a follow if you would like. And the image is from Agata of goodstuffnononsense.com. Go check out their work!
In 1940, the McDonalds Brothers opened their first hamburger restaurant. In 1948, McDonalds introduced the Speedee Service System, a process that did away with plates and forks and reduced the menu to handheld foods such as burgers and fries. For Americans with a need for speed and the development of highways and suburbs, McDonalds was convenient and profitable. In 1953, McDonalds began franchising. By 1972, they were global, spreading their efficiency all over the world.
But there was a dark side to the rise of fast food. As McDonalds, Burger King, Wendys, and hundreds of other chains spread across America, American waistlines also spread, blood pressure increased, and obesity soared. In 2004, Morgan Spurlock released “Super Size Me”, a documentary detailing the impact of McDonalds and fast food. This followed Eric Schlosser’s 2001 book “Fast Food Nation: The Dark Side of the All-American Meal”. Numerous studies linked fast food to poor health. An American success story in efficiency was killing Americans.
In response, fast food chains saw a drop in revenue. McDonalds sales lagged throughout the early 2000s until they introduced dollar menus, sparking low-income customers. But for many, the damage was done. Fast food is still considered an unhealthy, albeit convenient, option.
Today our society faces another type of efficient revolution. Not in food, but in thought. Or the outsourcing thereof. Generative Artificial Intelligence (Gen AI) use has soared in the last five years. Outputs from Gen AI have swamped our media in a way that makes the expansion of fast food from 1950 to 2000 look glacially slow. AI slop is everywhere, from emails to letters, wedding invitations to commercials. Students are using AI to write papers, lawyers to write arguments, job applicants to write resumes, musicians are using AI to make music, artists to make art, and authors to write books. It seems everyone is using AI to make everything.
I don’t trust anything created with AI.
AI is driven by Large Language Models. Those LLM drive the outputs of queries. I don’t have the time or the patience to read what is under the hood of every Large Language Model and how it chooses its responses to queries. Nor do I care.
Think about this, if an LLM was only fed text in Yoda-speak, it would only reply like the ancient Jedi master. And that is why we need humungous data centers on every corner, because the LLMs behind popular AI models need to scan through trillions and trillions of gigabytes of data to give users an answer. Because the robot is making the analysis from millions of unknown sources, I don’t trust it.
I compare AI outputs to cake. I could ask a machine to make me a strawberry cake. I can add the strawberries to the LLM, but what else is in the cake? It could be coded to make a strawberry cake out of cement mix and strawberries. That’s a horrible cake.
(Not to mention copyright issues and lawsuits that have resulted from LLMs sucking up very distinct work and providing outputs to users for the financial gain of the AI company, not the original creator.)
According to Gregory Bourne of Tiny Desk Publishing, I am not alone in distrusting Gen AI, especially as a Gen Xer. My generation, the one known for its cynicism and aloofness, is among the most distrusting of AI technology. According to a few reports, Gen X is using AI in the workplace, but reluctant and hesitant to incorporate it in their private lives due to privacy concerns.
As a writer, I believe AI takes the fun out of writing. It removes the joy and wonder of creating. It reduces the need to make mental and creative connections. When I wrote my novels, I often sat with two scenes in my head for days, wondering how I can get them connected. When a middle scene finally arrived in my brain, it was like unlocking a puzzle where suddenly all the pieces have a place. The lightbulb went on and I started writing. I don’t want AI to do the creative lifting and put in puzzle pieces I am struggling with. There is no fun in that to me.
According to Eve Fairbanks’s recent article in The Atlantic, AI writing is like a mass-produced burger, flavorless and boring.
“The tone is bland; individual word choices are baffling; the structure lacks sense; key pieces of the argument are missing; facts are false. Working on AI text, as an editor, is like trying to operate on a body whose skin, muscles, veins, bones, and organs are all compromised. There’s nothing to leave intact, nowhere to begin.”
Burgers created through the Speedee Service System aren’t known for their flavor, they are known for their speed. Unfortunately, speed is a key driver to today’s creators, especially those creating for marketing purposes.
Good art takes time. But when “content creators” need to make “content” to maintain awareness in the algorithm of social media platforms, they query and prompt whatever Speedee Service System they can, lest they fall behind and their posts not reach their desired metrics in views and engagement.
Did you know TikTok recommends posting 5 to 10 times a day to maintain relevancy on their algorithm? Big media organizations are able to do that, especially if they have daily inputs, such as sports highlights. But that is impossible for small creators.
Meeting an algorithm’s demands is not about art. It’s about numbers. Marketing content driven by analytics, driven by ROI and budgets. Financial people don’t care about the quality of art. They only care that the money spent gets the value desired. So be it if a content creator has to go to the fast-food drive-thru 5 times a day to get a mass-produced burger and fries to fill the belly of an ever-hungry algorithm.
When does it end?
While it took 50 years to see the impact of fast food on health, there are already multiple studies on the impact of AI on human thought and learning. Like fast food, the results are not good.
Studies on AI users show less cognitive activity, less creativity, and less thought. Outsourcing decisions is leading to laziness and the atrophy of human decision making. Worse yet, the AI is programmed not to disagree. It is more apt to agree with the user and at the same time, produce content that is homogeneous to other users’ work. In other words, it lacks originality. It lacks flavor. It doesn’t take much brain power to make a hamburger at McDonalds. It doesn’t take much brain power to make art as bland as a Speedee System product.
(No offense to McDonalds workers. My first job was flipping burgers at Mickey D’s.)
Not only is AI ruining our brains, it is also producing content that is hurting the very LLMs the content is derived from. According to a new study by researchers at Texas A&M University, University of Texas at Austin, and Purdue University, Large Language Models can be negatively influenced by attention-grabbing, sensationalist, and shallow information. The results of queries from LLMs fed this “brain rot” content were marked by “worse reasoning, poorer long-context understanding, diminished ethical norms, and emergent socially undesirable personalities”.
The rise of AI programs in our current tech and media spaces is akin to McDonalds incorporating the Speedee System in every restaurant and grocery store in the world while feeding its own workers its own unhealthy fast food. When the great cholesterol calamity occurs, the only survivors will be those who create their own supply chain, breed their own cows, slaughter their own animals, grill their own meat, and make their own burgers. They will be oases of humanity in a desert of Gen-AI slop.
As Fairbanks writes in her Atlantic article,
“Maybe human writing will become like cloth-aged cheese or handloom rugs, an artisanal product created effortfully. Maybe we will come to treasure older writing. Herman Melville, George Orwell, Toni Morrison—all authenticated. Writing like this will be a fossil record for a kind of thought process we buried without realizing it.”
Meanwhile, McDonald’s itself is heavily embracing AI in its Speedee System 2.0, attempting to use automation to get orders out faster. Americans are still partaking in fast food cuisine, reportedly more often than ever before.
Old habits die hard, I guess.
