A hot topic on the internet right now is the “OK Boomer” movement. The phrase “OK Boomer” acts as a disdainful retort, often used to dismiss and mock those who are perceived as being out-of-touch, narrow-minded, or outdated, particularly baby boomers. OK Boomer is a form of generational warfare. But who or what is a boomer? Although “boomer” refers to the baby boomer generation, “OK Boomer” has developed into an insult that refers to a very specific older mindset. A mindset that uses particular rhetoric and diction. I found this sort of verbal distinction very interesting, so I trained an artificial intelligence neural network to “detect boomers.”
The AI’s initial training set was 10,000 public profiles on Facebook by people who supported Donald Trump in the 2016 election, and were above the age of 54. I also hand-selected a few hundred profiles that I declared to be “super-boomer.” After I trained the AI using these profiles, I gave my college friends an app where they could speak into their phones, stating their opinions on a variety of topics; the program would determine if what they were saying was “Boomer” or “Not Boomer.” For example, if they said, “Climate change is based on fake science,” or “Why the hell are my coupons for cashews expired?!” the AI would determine “Boomer.” However, if they said, “Dreamers should be protected” or “I just got a victory royale in Fortnite with the squad,” the AI would determine, “Not Boomer.” The user could then input if they agreed or disagreed with the algorithm’s output. This user feedback was then be fed back into the machine learning algorithm, and the AI would further “learn” how to detect a “boomer.” After crowd-sourcing hundreds of responses, the AI could confidently detect who is a boomer about 95% of the time.
The neural network classifies boomers by not only their language, but the syntax, punctuation, and keywords they use in their speech. It turns out, Boomer speech patterns are actually very distinct. Words like “fake,” “immigration,” “wall,” or “climate change,” followed by tell-tale pauses, or adjectives determining acceptance or failure, would trigger a label of Boomer or Not Boomer. Some phrases like “tuckered out” would also trigger. After being trained with thousands of nodes, the AI would pick up on the slightest boomer type language. Now that I had a neural network that could detect the use of boomer-like language of all types, I applied it to real-world cases. I first began running it on a variety of web-pages. The program would web scrape all the text it possibly could from public domains of web pages and run groups of sentences through the AI. Interestingly, there is a pretty strong correlation between the most conservative companies and how much “Boomer” language they use. I generated a list of websites that I thought excreted both boomer and millennial energy.
Twitter acts as a modern-day platform for powerful political figures to speak their minds to their audience. With the 2020 political election coming up, I thought it would be interesting to see which politicians speak in “Boomer” like language. To do this, I created a simple program that grabs all the tweets a politician makes from their Twitter account, and then ran each one through the neural network. The program then takes the average of all the tweets and produces a verdict and a confidence level. Although political jargon tends to be more Boomer in general, it is interesting to see whether the intelligence struggles to place a candidate or nails them down with certainty.
The AI began as a joke. But it became something much more. The landscape of potential for this type of classification seems to be infinite: music, novels, or even chat rooms. It seems to unlock outrage, interest, and curiosity. When moving through the complicated political sphere of the twenty-first century, taking time to consider which candidates rely on traditional rhetoric and express beliefs consistent with the Boomer generation may further inform who or what you support. You might find yourself exclaiming, “Ok, Boomer” more often than you’d think.