A long time ago, I told a friend who needed constant consultation, “My brain isn’t big enough for the two of us.”
I said this because I had long grown tired of this person blaming me if my advice or suggestions didn’t pan out. I wanted them to take agency and ownership. I wanted them to make up their own mind and “stand on business.”
But they never did, so we aren’t friends anymore.
What I didn’t know was that this friend wasn’t alone. In fact, scores of individuals outsource their decision-making—not to their nerdiest, most well-read friend, but to an AI chatbot.
Some seek ideas on what to cook for dinner. Some are looking for help managing their household budgets.
Others are asking: “Who should I vote for?”
Over the weekend, NPR’s Weekend Edition Sunday did a story on people consulting AI to decide how to vote in the Midterms.
This means bots from companies like Gemini, Claude, Grok, and ChatGPT are helping people decide how to vote.
According to the Pew Research Center, about 42% of AI chatbot users report doing their research via chatbots. Of the voters NPR spoke to for their story, many said they used chatbots for “down-ballot races with thinner news coverage” or “to make tables comparing candidates for governors’ races and to verify or debunk claims.”
Reading the story sent a literal chill up my spine.
Not because I was surprised. Voters have used search engines and reported information to research ballots and candidates since they’ve been able to do so.
Because I was terrified.
After all, modern journalism has rules. Clear ones, well-defined through generations of molding and study. Anyone can buy a copy of the AP Stylebook, or read up on media ethics and learn.
AI, on the other hand? Which way its chatbots bend depends on the maker.
While it’s not surprising when a conservative-leaning or left-leaning newspaper endorses a candidate, chatbots present themselves as friendly, knowledgeable, and impartial reflections of their users. Meaning, Rupert Murdoch-owned publications like The Wall Street Journal or The New York Post aren’t necessarily writing op-eds to please the public, but to share their thoughts and ideals in an effort to inform or persuade. The New York Times, USA Today, The Daily Beast, and HuffPost often do the same.
But you know these publications, and you often know where they stand.
A chatbot, by contrast, is a malleable sycophant that tells everyone exactly what they want to hear.
According to NPR, “Experts say the properties that make chatbots useful also make them subtly persuasive and eager to please.”
And that can be a problem for information-seeking voters of color, who have been more positively disposed toward large language models such as ChatGPT than white voters are, according to a 2024 Harris Poll. Especially if those voters think they’re getting the unvarnished truth from their AI friends.
OpenAI, the company that owns ChatGPT, has a page dedicated to election information and various safeguards, saying it “monitor(s) bias in our models to keep ChatGPT’s responses politically neutral.”
That sounds nice, but it isn’t the same as having an individual to hold accountable for the information they spread. Users often do not know where a chatbot’s true intentions lie. Is it just to endlessly flatter the preset opinions of the user? Or is it to advance its master’s agenda?
It’s not the best for our Democracy, but neither is much of the mainstream press these days.
Today, the Paramount-Warner Bros. $111 billion merger finally went through, and one of the world’s largest media conglomerates is now under the thumb of one man, David Ellison, while being partially owned by foreign backers in Saudi Arabia, Qatar, and other Middle Eastern countries. This means networks like BET, MTV, CBS, CNN, and dozens of others now fall under one umbrella: Skydance.
And while Ellison has chosen to keep CNN’s current lead, Mark Thompson, in his current seat for now, journalists are still side-eyeing the situation, considering how Ellison’s hand-picked lead for CBS News, Bari Weiss, has all but tanked that division in an effort to remake it as Fox News lite.
While my brain remains not big enough to answer the world’s questions, ChatGPT and media conglomerates like Skydance are more than up for the task. Will they be impartial? Will they be unbiased? Will they simply offer information without confirmation?
That is yet to be seen.
But I wouldn’t bank on the benevolence of anyone who pretends to hold all the answers.
Even if that “anyone” is a chatbot.
