The Role of AI in Employment Is Leading to an Endless Cycle of Challenges

Navigating a low-fire, low-hire economy exacerbates challenges. Job listings are limited. Issues with scams and ghost jobs have prompted some states to consider new legislation for protection.

Both job seekers and employers report a growing lack of trust. Candidates invest significant time in their applications without any response. The job application process feels opaque.

On the other hand, employers receive numerous applications that often appear nearly the same. Consequently, they input them into AI rating systems within their ATS to help distinguish between candidates quickly.

Chait remarks that we are trapped in an AI doom loop. “We have this unfortunate scenario where each side faces its own issue. They turn to AI for solutions, but those attempts often exacerbate the problems. Increased reliance on AI leads to greater reliance, benefiting no one. The situation deteriorates with each use.”

Comparing Human Vetting to AI

I engaged with numerous recruiters and HR managers, as well as small business owners involved in hiring. Some acknowledged the use of automated candidate ranking, while others firmly rejected it. This divide isn’t influenced by the organization’s size or the volume of applications; instead, it seems rooted in differing philosophies and cultures.

Kim Jones, vice president of human resources at Toshiba, emphasizes, “We have humans evaluate every application.” While she’s open to applicants enhancing their submissions with AI, she believes it won’t significantly assist in navigating the ATS. The selection process hinges on job criteria, salary expectations, and additional factors such as rehire potential.

Jones notes unwanted AI influences during interview stages. “You catch the pause and perhaps hear typing, leading to a lengthy response,” she states.

Doist, another company that prioritizes human input over technology in hiring, is a small, fully remote organization that recruits globally, resulting in a high volume of applications for each opening.

Nadia Vatalidis, head of people at Doist, shares that her team tested whether ATS rankings were beneficial. They took previously filled positions and analyzed job descriptions alongside saved applicant materials. Their goal was to determine “what would happen if AI shortlisted the same candidates and whether those we interviewed would match those categories,” Vatalidis explains. “In two tests, the hires made didn’t appear on the shortlist.” Although there was some overlap in interview candidates, the new hires—who turned out to be excellent after six months—did not make the AI’s cut.

Using AI Differently

James Jacobsen embarked on his job search five months ago, dedicating as much time to it as a full-time job. As a design professional, he understands that AI doesn’t match his creative capabilities, yet he actively experiments with it, recognizing its value for specific tasks. He utilized Claude and ChatGPT to refine his application documents and align them with job descriptions, but it didn’t lead to any significant progress.

Rather than solely focusing on application quality, he began using Claude to organize and monitor his entire job search. He directed it to sift through job listings, analyze specifications, and log information. He created a detailed scoring system based on factors like job type, seniority, and varying salary expectations for in-person, hybrid, and remote roles. His AI assistant identified the top jobs to apply for. When he declined a position, he noted the reasons so that if it was reposted later, his AI could promptly remind him of his initial disinterest. This approach reduced his search time while helping him discover more suitable opportunities. Although he received some interest, no offers materialized.