AI recruiters are a long way off

9 November 2019  |  Rich M

In theory, AI and frontier technology should make finding the right candidates easier. In practice, we have reasons to be sceptical, for now at least.

Firstly, The Paradox of Choice is not a contested topic amongst social scientists. The more options we are presented with, the harder and slower a decision becomes. This raises the point that drowning in candidates is equally as frustrating as a drought of candidates, more is not necessarily better.

Current systems for eliminating candidates are at best scientifically suspect and at worst discriminatory. Daniel Kahneman has proved this in his work regarding how we evaluate people for certain jobs. We tend to have a mental model of what the successful candidate looks like and the system of elimination is based on these mental models.

Secondly, data from LinkedIn suggests roughly 70% of the labour market are passive job seekers. Currently, recruitment innovations simply mean that more people will be tempted more of the time. If Company A is seducing talent away from its competitors using this technology, you can bet your competitors are doing it to you and in turn it becomes a tautological game of attraction, selection and attrition.

Broadly speaking, the innovations behind attracting and recruiting talent are good news in the long run. However, we are a long way from accurate and scientifically validated ways of streaming people into jobs and companies that match who they are as a person.