Future of Jobs: Job Interview with an Artificial Intelligence

New Year, new job? In an interview, applicants and companies find out whether the potential new employee has the necessary knowledge for the position and fits well into the team. At the latest since Covid-19, many of these conversations have been conducted digitally – but what if suddenly there is no recruiter sitting across from you, but you have to answer questions from an artificial intelligence (AI)? Dr. Martin Fladerer explains what’s behind it.

Recently, Jakub Cichor explained whether robots could be our bosses in ten years. Today we are starting one step earlier – and are looking at application processes using artificial intelligence: Dr. Martin Fladerer works at the Chair of Research and Science Management at the Technical University of Munich. Together with colleagues, he is engaged in the research project “New Job, New You” with the personality development of people in professional life and accompanies people who have recently changed their job.

Dr. Fladerer, how have job interviews changed as a result of digitization?

Nowadays, more or less by default, job interviews are conducted by means of video telephony. One trend is the use of algorithms in these video conversations, which use pre-selected criteria to assess the appearance of a candidate. This information is used to support selection processes. In extreme cases, the conversations are already fully automated: you talk to an “artificial intelligence” instead of an HR manager. And in the end, the algorithm makes the decision who advances in the process.

For applicants, such an interview is certainly associated with uncertainty in the first step. How does such a conversation take place and how do you prepare well for it?

As a rule, the applicant receives an access code to a platform and is guided through the interview there on the basis of pre-programmed questions. The answers are recorded and automatically evaluated by algorithms. What exactly is analyzed is usually unclear to the applicant. In principle, three sources of information can be used for analysis: firstly, video-based data such as facial expression, secondly, verbal content, e.g. the use of keywords, and thirdly, verbal modulations, i.e. how to say things.

A background research on the technology used can help to understand what the algorithm attaches importance to. The platform developers are different here – and to be honest, the technology is far from reliable and valid.

An important point in job interviews is whether the applicant fits into the team and, of course, could also imagine working with the people he/ she gets to know. This is not possible in such a digitized process. What are the consequences?

For this reason, in my opinion, an automated conversation can only ever be one component in the selection process, e.g. to capture specific skills. At a certain point, personal contact is essential for both sides. For companies, there is a risk of being perceived as impersonal through the use of these tools. Applicants describe, for example, that they do not feel that they are seen as a human being. Therefore, it is important for the organizations to create transparency in the use of the technology and to involve human contact persons despite all digitization.

What does the future of job interviews look like – will we all be interviewed by a robot in 10 years?

The use of digital technology is progressing in all areas of life. Improvements in the ability to analyze human facial expressions, gestures and speech will not be an exception here. At the same time, we must protect ourselves from the glorification of “artificial intelligence”. The methods are by no means mature so far: AI is still programmed by people who encode their assumptions and prejudices into the software, which then reproduce them. So I hope that AI will be seen and used in selection interviews as a source of information among others, and not as the only true one.

Dr. Martin Fladerer

About “New Job, New You”

The project investigates the development of people after a change of job over a period of three years. For this purpose, the participants will be interviewed at an interval of three to six months with the help of an online questionnaire or personal interviews. The project is funded by the German Research Foundation (DFG) and carried out in cooperation with the TUM Institute for Lifelong Learning. It is headed by Prof. Dr. Claudia Peus and Assoc. Prof. Dr. Armin Pircher Verdorfer (University of Amsterdam) with the collaboration of Dr. Martin Fladerer and Clarissa Zwarg. Further information on the research project can be found on the website of the Chair of Research and Science Management.