Why Blind Faith in AI Fails: ‘They’re Hiring a Colleague, Not an Answer’

After John Silvestro completed his Ph.D. in composition and rhetoric from Miami University in 2017, he entered the job market seeking a faculty position and not knowing what to expect. He was a “nervous wreck” during his first few Zoom and phone interviews.

“I was just kind of winging it,” Silvestro said. “But by the end of it, I was much better because I had heard some of those questions over and over, and I had rehearsed my answers.”

If only there was something at the time that would have prepared him to come up with better answers. Now, there is. With generative artificial intelligence, Silvestro could have popped in the job posting into an AI and asked “What are some questions that a hiring committee might ask for this position?” Then, he could have practiced.

Anyone who is interviewing for a job now would be foolish not to consult AI chatbots, at least to check for blind spots in their preparation, but not to use the technology blindly. In the last two years, controversial apps like Cluely and Final Round AI started providing what are essentially teleprompters that provide real-time answers to questions during job interviews that are not in person.

Silvestro is now an assistant professor of literatures, languages, cultures, and writing at Slippery Rock University, and he’s published research about the use of supportive AI in educational settings.

He said that AI can be a valuable tool for academic job searches, including interview prep to better anticipate questions, but also to initiate application materials.

Starting Point

For those just starting out, AI can help with first drafts of cover letters and CVs.

“AI is really useful for pulling some of your materials together and seeing what those would look like,” Silvestro said. “Your goal should ultimately be to rewrite every single word in there, but it can give you that starting point, because A, the blank page is incredibly terrifying, and B, the job market, especially if [it is] your first time on it, is even more terrifying. You’re, in effect, writing to strangers who you don’t know and you don’t know what they want or what they’re looking for.”

If anything, generative AI can give you coaching and confidence to take the first steps by showing you the path of others who came before you.

“It’s saying, ‘Here’s the average of how other people have done it, and here are the basics of what you want,’ but also the basics of what you want to improve upon and do better,” Silvestro said.

Reaching the Average

Quoting Shannon Vallor in her book “The AI Mirror,” generative AI isn’t designed to be accurate; it’s designed to sound accurate. The AI mirrors that Vallor writes about face backward: “The more we rely on them to know who we are, the more the fullness of our humane potential recedes from our view,” she wrote.

Generative AI gives you a polished reflection of the collective to help you reach the average, but when building your case for hire, you need to go well beyond a vague representation of a typical-sounding candidate. You need an accurate depiction of who you are, and you need to bring to view your unique potential.

You can improve upon the average by using AI to synthesize messaging, refine your voice, and build on a base to reach for greater distinction, but a blind-faith reliance on AI is a regression to the average.

Delaying the Inevitable

You may have heard that the goal of a cover letter or resume/CV is to get invited for an interview, not to get hired for the job. There’s truth to that. But if you’re deceiving whatever screening instrument is in place — human or not — you’re eventually going to have to face a hiring manager or search committee.

Katharine Stewart is a former professor and senior provost who served on more than 30 hiring committees for faculty and academic administration positions before starting her own consulting practice. She coaches her clients to tell concrete stories in cover letters that AI can’t replicate, but even if they pass the first test with average-sounding language, their true selves must emerge.

“One of the ways AI falls apart is if, by some reason, (employers) do get fooled by an AI-written cover letter, when you get into the first-round interview and you can’t tell those stories, or don’t tell those stories, then the voice they fell in love with during the search isn’t your voice,” Stewart said. “That disconnect makes people decide, ‘Oh, this person isn’t who we thought they were.'”

Final Delivery: Colleague, Not an Answer

Who you are going to be once you are hired is not a chatbot. Workplaces already have access to those.

Silvestro considers AI as a valuable tool in the application and interview process, but the end product is a human being who shows up to work. Employers want to know what that person is like.

“AI can be really useful to help you prompt and to help you practice, but it cannot really help you with that final delivery,” Silvestro said. “(Employers) want to know who you’re going to be in meetings, in classrooms, and working on projects.

“They’re hiring a colleague. They’re not hiring an answer.”