Researchers swapped human recruiters for AI agents. AI did the job better, with a few drawbacks.

• Researchers attracted AI against human recruitment. The boots went to the peak.

• The experiment included more than 70,000 applicants competing for the entry -level customer service roles.

• Candidates interviewed with AI were likely to receive proposals and the probability of starting and staying at work was higher.

Some job seekers may be better than chatting with a boat.

A new study found that the possibility of receiving a job proposal was 12% more than human recruitrs. In addition, 30 days later, the probability of starting and adhesion was higher.

Professional recruitors betrayed them in this recruitment experiment. Artificial intelligence proved them wrong.

The study also found that when the election was made, 78% of the applicants chose the AI interviewer through a human recruitment.

Luca Henkel, a behavioral economist in Erasmus University Rotterdam, an economist at the University of Chicago at the University of Chicago, established a partnership with PSG Global Solutions against human employers in a large -scale recruitment experiment.

The hearing discussed more than 70,000 applications competing for the role of entry -level customer service in 48 job advertisements in the Philippines. Jobs were with 23 Fortune 500 companies and 20 European companies.

Applicants were appointed to one of the three random interview conditions: a human employer, an AI employer or a choice between the two.

In any case, human employers decided to recruit after examining transcripts and a standard language and analytical skills test. This design allowed researchers to isolate a variable: the interview speech.

Both people and AI watched the same interview guide. It started with conformity questions, moved to career goals and work experience and ended with work details. However, the results were allocated.

Why did AI do better

Applicants interviewed by artificial intelligence employers received job offers compared to 8.7% under human recruitment in 9.73% of cases.

The study also found that the probability of starting work is 18% higher and 30 days later, the probability of being employed is still higher.

Using natural language processing, the researchers found that AI interviews cover more and more issues and encourage richer answers. Interviews led by AI minimized weaker signals, such as filling answers or irrelevant questions, such as human recruitrs often rewarded, such as depth of speech.

Recruitors who reviewed the transcripts received higher points than they talked to the candidates they interviewed.

“Negotiations led by AI revealed more information about recruitment.”

But the system was not perfect. When they realized that they talked to AI, approximately 5% of the applicants ended their talks and agent faced technical problems in 7% of cases. Applicants also considered interaction less “natural” than talking to a person.

Jabil and Henkel did not respond to Business Insider’s request for comments.

How does artificial intelligence change recruitment

AI is increasingly used in the search for job and The process of recruitment. Candidates lean on to help AdaptWhile employers use it to review this Thousands of Applications they take it.

Emily Dejeu, who is an assistant professor at Carnegie Mellon University, who specializes in AI communication and manners, told Business Insider. In May It is likely that video interviews that work with AI will become more common as companies try to transfer and automate early order. Stages of recruitment.

“We follow this by default – there is some kind of inevitability for this,” he said.

Technology Investors AI can use recruitment.

Victor Lazarte, a former general partner of Benchmark, said in a part of the “twenty -minute VC” Podcast, which was published in April, should be particularly tense about the AI that changed the business of recruitment.

AI models will soon be better than people in interviews with candidates and companies will be much more efficient than scattered, manual recruitment processes.

But not everyone is sold. In a Business Insider report published on Monday, recruitment managers full of applicationsMany of them are optimized to look like a perfect harmony, while hundreds of unsuccessful applications seeking hundreds of angry jobs.

AI recruited Hatim Rahman, Associate Professor of Management and Organizations at Northwestern University, is a “cat and mouse game” among candidates and employers because they both use technology to reveal the other.

As a result, there is pressure to find “more human signals in both search and application process”.

Read the original article Work inside