Just between us, have you ever used AI to help you complete a school assignment? Have you ever used AI for something other than schoolwork? As many students know, AI can do a lot for us. It can construct study guides, summarize text, and even write essays. It can also stalk your mother. 

  Generative AI (Gen AI) refers to AI models that are capable of text and/or image generation, like ChatGPT, Claude, Copilot, and Google Gemini. People commonly refer to these systems simply as “AI,” but Gen AI doesn’t encompass all that AI is capable of. For the sake of concision, and my sanity, I’ll only be looking at the Generative AI model Gemini in this piece. 

I am not a Gemini user myself, but a friend of mine made a concerning discovery last week. If you type a name and location (for example, “Brooke Snyder EMU”), within seconds Gemini compiles every piece of information about that person on the internet. I asked my friend to search my mom’s name, and Gemini provided a lot of information about her place of work, including the address and her office phone number. All the information that Gemini provides is public information pulled from sites that feature the person’s full name, such as LinkedIn. 

Now is this any different than simply Googling someone’s name? I argue that, yes, it is. First and foremost, Gemini makes the “investigation” process much quicker. I Googled my own name along with “EMU,” and counted about four links that I would have to click on, scroll through to find my name, and extract and process the information from in order to obtain all the points that Gemini provided me within about four seconds. With all that in mind, who wouldn’t want to use AI in place of the Google search engine?

Me. I wouldn’t. But I do feel obliged to tell you that I did talk to Gemini in order to understand its capabilities for this piece.

One misfire I noticed during my brief Gemini interaction was that it credited me as a Weather Vane Editor-in-Chief. I am not currently, nor have I ever been, an Editor-in-Chief, but there was another student named Brooke Snyder who previously served in this role. Gemini mistakenly condensed myself and this other Brooke Snyder into one person with a patchworked resume of accomplishments. This is the dilemma of the incredible speed by which AI processes information – it can just as quickly become misinformation.

Errors like these are called “hallucinations,” or misinformation presented as a fact. Gen AI works by predicting the likelihood of words appearing in a certain order, kind of like a distorted infinite monkey theorem. AI collapses infinity to answer your queries on schoolwork, sports, or whatever other thinking you outsource to it. 

There are a lot of arguments for and against AI right now, but the one that most speaks to me is that AI could start to become a replacement for human thinking and connection. I already see it happening: I take information from Google’s AI search overview without even checking its sources. My friend jokes about asking Gemini a question about an acquaintance instead of speaking to them herself. But it doesn’t have to be this way. Once we have more clearly defined AI’s place, we can step back and use the time it saves us for more human things, like creating art, connecting with one another, and experiencing our world.