Data-driven Talent Decisions

The sheer size of many applicant pools and the amount of time and money that companies pour into recruitment is astronomical. Companies like HireVue can narrow the eligible pool at a fraction of the time and cost, and hundreds of companies and organizations – including Hilton, HBO, Boston Red Sox, and Atlanta Public Schools – have signed on. Another added value, according to HireVue, is that there is a lot that a human interviewer misses but that AI can keep track of so as to make “data-driven talent decisions.” The CEO of MYA Systems, another company in the AI-powered recruitment business, described some of the benefits thus: “a 79% reduction in the time it took to fill each position, and a 144% increase in productivity per recruiter who used the technology.”12 But efficiency and effectiveness are not the only drivers. AI programs are called upon to bypass human bias when it comes to evaluating job candidates.

Recall that the problem of employment discrimination is widespread and well documented. Wouldn’t that be a good reason to outsource decisions to AI? And, if so, is the solution to ensure that the algorithm is not exposed to prejudicial associations, so that it avoids replicating our worst tendencies? Not quite, says the Princeton team working on natural language processing mentioned earlier, because exposure to negative biases can be instructive on how not to act – at least for humans. Some warn that completely shielding AI from the “bad” and saturating it in the “good” is not necessarily the way to go. Instead the question is how to code a machine that can vacuum up all the beauty and ugliness from humanity, yet remain without prejudice – an idea we will come back to.

Don't use plagiarized sources. Get Your Custom Essay on
Data-driven Talent Decisions
Just from $13/Page
Order Essay

But for now it should be clear why technical fixes that claim to bypass and even overcome human biases are so desirable, even magical. Ed Finn, director for the Center for Science and the Imagination at the University of Arizona, connects the cultural power we grant to algorithms with a longer genealogy of symbols and sorcery, arguing that “computation casts a cultural shadow that is informed by this long tradition of magical thinking.”13 Magical for employers, perhaps, looking to streamline the grueling work of recruitment, but a curse for many job seekers.