When Alaska software engineer Josh Holbrook was laid off in January, he didn't expect to spend so much time looking for a new job. He certainly didn't see the need to relearn the job search process.
But a few weeks into his investigation, Holbrook realized he was in deep water. Instead of speaking to a human recruiter at a local healthcare facility, he was vetted by an AI chatbot. His resume, created in a technical format popular among academics nearly a decade ago, was incompatible with new automated recruiting platforms. He signed up for his professional service for updates in an AI-friendly format.
“The experience was completely new,” Holbrook told me. “I've never seen anything like that.”
Over the past few years, job seekers have been forced to contend with constant layoffs, a cutthroat hiring market, and days on unpaid assignments. Now you can add AI recruiting systems to that pile. In 2022, the Society for Human Resource Management revealed that approximately 40% of large employers surveyed said they had already implemented AI in HR-related activities such as recruiting. Rick Mistry, a large corporate recruitment consultant, told Business Insider that AI is now being used to create job descriptions, determine applicant skills, power recruiting chatbots, and respond to candidate responses. He said that it is being used for evaluation purposes. His CEO at ZipRecruiter, Ian Siegel, estimated that in 2022, nearly three-quarters of all resumes have never been seen by a human.
Some job hunters are deciding to fight fire with fire and turning to programs that use AI to optimize their resumes and apply to hundreds of jobs at once. But the emergence of an AI vs. AI hiring battle is bad news for everyone. It turns recruiting into a depersonalized process, inundating recruiters and reinforcing weaknesses in the system it was designed to improve. And it looks like things are only going to get worse.
Recruitment automation is not new. After job sites like Monster and LinkedIn made it easy for people to apply for jobs in the early 2010s, companies implemented applicant tracking systems to manage the flood of online applications. Today, most resumes are viewed by software designed to first assess a person's experience and education and rank them accordingly.
Automation has helped ease the burden on busy recruiters, but it hasn't done much. As remote work policies change frequently and the pile of digital resumes grows, the hiring hamster wheel spins faster and faster.
We know AI isn't perfect, but we have to use it because there's pressure from above.
AI is expected to solve this confusion and save companies time and money by outsourcing more of the hiring process to machine learning algorithms. In late 2019, Unilever announced that it had saved 100,000 hours and approximately $1 million in recruiting costs thanks to automated video interviews. Platforms like LinkedIn and ZipRecruiter have started using generative AI to provide candidates with personalized job listings and allow recruiters to generate listings in seconds. Google-backed recruitment tech startup Moonhub features AI bots that explore the internet and collect data from places like LinkedIn and Github to find suitable candidates. HireVue allows employers to have bots with configured surveys conduct video assessments and analyze candidates' personalities. New startups are combining these capabilities into a centralized service that allows companies to put “hiring on autopilot.”
But hiring experts Business Insider spoke to weren't convinced that's all for the best. Many fear that over time, AI will make an already frustrating system even worse and create new problems, such as ghost hiring, where companies are misled into hiring bots masquerading as humans. I am concerned that this may not be the case.
Several veteran recruiters say they haven't incorporated AI into their workflows beyond automatically generating job descriptions and summarizing candidate calls. Tatiana Becker, who specializes in technology recruitment, said software that claims to match resumes to job descriptions lacks the nuanced ability to do more than keyword matching. For example, it is not possible to identify desirable candidates who come from top schools or have a track record of significant promotions. The chatbots Becker's small agency experimented with frequently mismatched candidates with jobs, sometimes ultimately turning candidates away.
This resume matching “may work for applicants for more entry-level jobs,” Becker told BI. “But I would be concerned about using it for anything else at this point.”
Despite the problems, many companies are moving forward. A recruiter at a Fortune 500 company spoke candidly about his company's hiring process on condition of anonymity, saying, “We know AI isn't perfect, but we have no choice but to use it because of pressure from upper management.'' ” he said.
“AI in recruiting, like other industries, is in its very early stages,” said Pallavi Sinha, vice president of growth at Humanly, a startup that provides a conversational AI recruiting platform to companies like Microsoft. Stated. But she predicted it would continue to be incorporated into recruitment.
“AI does not exist to replace human interaction, but rather to make our jobs and lives easier, and this will only become more evident as time goes on,” she said. said. Sinha declined to say how many applications Humanly has processed, but she said the company's chatbot service had “more than 1 million conversations in the last year alone.”
But for candidates, AI was a nightmare. Kerry McInerney, an AI researcher at the University of Cambridge, said AI would increase the workload for applicants, requiring them to complete puzzles and participate in automated interviews to advance to the selection stage. Ta. She argued that it “further alienates the process of depersonalization.”
Holbrook, a software engineer, took to LinkedIn to share his frustrations. “AI is too stupid to recognize transferable skills among technical applicants,” he added. “Even though at work he has worked with C#, he had his resume returned because C# was not listed as a fluent language.” C# and he speak 90% of the same language. I've used it a lot. ”
Daniel Caldwell, a user experience strategist in Portland, Oregon, was perplexed when an AI chatbot sent him a text message asking him to start a conversation about a role he had applied for. At first she thought it was spam. After her interaction, she was left with more questions.
“There was no way to ask the bot questions, and it was a one-way experience,” Caldwell said.
The University of Sussex found that AI video interviews can be disorienting for job seekers, who are unable to act naturally without a reassuring human presence. Additionally, McNerney said assessing a candidate's personality based on body language or physical appearance is “not only reminiscent of the racial pseudoscience of the 19th and 20th centuries, but also completely unworkable.” Her research demonstrated that even things like wearing a scarf or having a bookshelf in the background can change your personality score.
Using AI will only accelerate recruitment failures.
A cottage industry of tools to aid candidate game AI systems has sprung up. For example, a service called LazyApply can apply to thousands of jobs online on your behalf for $250. “Anyone who has to look at more than 50 resumes at once wants to stick toothpicks in their eyes,” says Peter Rafter, who has been in recruiting for about 30 years. Introducing AI “will only accelerate the failure of recruitment,” he said.
Bonnie Dilber, recruiting manager at Zapier, said these services contribute to behavioral problems on both the recruiting and hiring side. When candidates submit hundreds of applications, hiring teams become overwhelmed and applicants feel like they need to submit even more applications to have a chance.
A more pressing problem, Dilbar added, is that these bots often submit poor-quality applications. Mr. Dilber and other recruiters told his BI that some cover letters simply read, “This was submitted by.” [AI tool]Please contact us [person’s email] When BI correspondent Aki Ito tried to apply for a job using AI, the system misjudged her ethnicity, fabricated that she spoke Spanish, and submitted an outdated cover letter.
“Some people accidentally included the entire ChatGPT response,” Hailley Griffis, head of communications and content at Buffer, told BI. “A lot of the unedited AI content has a very unique feel to it, and we’ve seen a lot of that,” said Griffiths, who said he reviews an average of 500 applications for each open role. ” he added.
Recruiters and researchers are also concerned that AI tends to reinforce many of the recruitment industry's existing biases. Researchers at the Berkeley-Haas Center for Equity, Gender and Leadership announced in 2021 that of the approximately 133 biased AI systems they analyzed, approximately 44% exhibited gender bias. Other recent studies have found that AI systems tend to screen out applicants with disabilities and downrank resumes with names associated with Black Americans.
Sandra Wachter, professor of technology and regulation at the University of Oxford, said that unlike humans, algorithms never consider past hiring decisions or correct their mistakes. Always do exactly what you are told.
Wachter, whose team developed bias tests used by companies such as IBM and Amazon, believes AI can be used to make fairer decisions, but to do so, employers must Systemic issues need to be addressed with more comprehensive data and regular checks. For example, Moonhub has human recruiters who vet AI recommendations and speak to candidates who prefer humans over chatbots.
But until AI is sufficiently advanced, humans will remain the most effective recruitment system. Becker, a technical recruiter, says that he uses the app to “get to the heart of a candidate's decision-making process and help them overcome the fear of being left out” and to “deal with counter-offers.” He said humans remain important.
David Francis, vice president of talent tech lab, a recruitment research and advisory firm, said that while “hiring is and always has been a human-centered process,” AI is “a tool, and if used effectively, “In some cases, it is used inadequately.” ”
But when it's used poorly, both candidates and recruiters suffer. Josh Holbrook has been looking for a job for four months, but he still hasn't found one.
Shubham Agarwal A freelance technology journalist from Ahmedabad, India, whose work has appeared in Wired, The Verge, Fast Company, and more.
