search.noResults

search.searching

saml.title
dataCollection.invalidEmail
note.createNoteMessage

search.noResults

search.searching

orderForm.title

orderForm.productCode
orderForm.description
orderForm.quantity
orderForm.itemPrice
orderForm.price
orderForm.totalPrice
orderForm.deliveryDetails.billingAddress
orderForm.deliveryDetails.deliveryAddress
orderForm.noItems
Human capital


present, the data which led to that biased result can be de-weighted or removed.”


Voices of dissent


Though AI evangelists abound, there are those who feel less than comfortable with its current role in recruitment. Among them is Alex Engler, AI data and democracy fellow at the Brookings Institution in Washington DC, who studies the implications of AI and emerging data technologies on society and governance. He is far from a luddite, but he does have concerns about the proliferation of algorithms across the hiring process, particularly when it comes to NLP, facial expression analysis and bias. “AI is completely unsuited for most interpersonal tasks,” he believes. “For instance, it does not really understand human language, despite sometimes appearing to. We know that AI models that transcribe speech to text can be less accurate for minorities, and people with accents or regional dialects. We also know that natural language models – whether learned on proprietary datasets or the internet – can exhibit many biases, for instance against women and people with disabilities.”


What Engler is calling for is balance – exploiting what AI does well while accepting its failings. “AI might be able to find meaningful patterns in interview answers or résumés, though it would be best to use relatively simple and interpretable AI models, and then have a person check to ensure they are sensible and fair heuristics,” he explains. “Since AI does not really understand these answers, and is only making correlations, we should use this approach sparingly.” Parker accepts this, noting that HireVue carries out rigorous testing for bias related to age, gender, or ethnicity throughout the process, and re-trains and re-tests its models. He has also shown a willingness to respond to concerns about the most controversial aspects of AI recruitment tools – facial expression analysis. With more people working remotely – a trend sharply accelerated by the pandemic but undoubtedly here to stay – video interviews are more common than ever. Amid the upheaval, some recruiters sensed the potential for AI technology to analyse a candidate’s facial expressions and body language to assess personality traits.


The backlash was swift. In 2019, the AI Now Institute published a paper calling for facial analysis in recruitment to be banned on the basis that emotional states indicated by facial expressions might be the result of many factors that an AI engine has no way of exploring.


“There is no compelling evidence that facial analysis can consistently reveal information about someone’s emotional condition, nor do I think it’s possible to tell anything meaningful about their personality or intellect,” says Engler.


Finance Director Europe / www.ns-businesshub.com


Having initially included facial analysis as part of its platform, HireVue later removed that capability. “The goal of HireVue’s assessment models is to correlate an applicant’s interview to how they would perform in a specific role,” Parker says. “Our own research concluded that for the significant majority of jobs, the predictive power of language has increased greatly, and our algorithms do not see significant additional predictive power when additional visual data is added.”


“This is definitely progress,” believes Engler. “Many people, myself included, are working to help the public understand the limits of AI. Facial analysis for job interviews is a clear example of a task that modern AI simply cannot do, and it can only possibly be unfairly disadvantaging people.”


Finding equilibrium


Engler believes we are a long way away from meaningful AI analysis of a video job interview, not least because of the frightening amount of data such a process would require.


“Even if it were possible, to do that would entail so much data collection I’m not sure that’s a world I want to live in,” he emphasises. “How we use AI affects how we incentivise data collection – saying it’s OK to use AI for something inherently creates an enormous demand for data about that thing, which can be quite dangerous.”


Though limited in scope, AI-enabled platforms can nevertheless reduce human workloads and speed up the hiring process, which is obviously appealing for many businesses. They will undoubtedly play a central role in helping recruiters match candidates with opportunities suited to their individual skills and experience.


“AI might be able to find meaningful patterns in interview answers or résumés, though it would be best to use relatively simple and interpretable AI models.”


Alex Engler


Thanks to AI, candidates will be able to minimise the time spent applying for multiple roles, and recruiters will more quickly identify a better pool of candidates. It is essential, however, that both groups understand where AI is effective and where it fails – and that it is not a suitable alternative in every case. AI has its limitations and platform developers such as HireVue have shown themselves willing to recognise where their technology adds value and where it does not. Before investing in AI, recruiters should know precisely what they need the technology to do – and should make sure a human hand is steering the process of talent acquisition. ●


90%


The decrease in time to hire that HireVue clients benefit from, a 16% increase in new hire diversity, and a 131% return on investment. HireVue


47 55%


The percentage of companies in the US using predictive analytics in HR in 2020, compared to 52% in Germany and only 18% in


China. Mercer


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52  |  Page 53