4 october 2019
You want to keep your best employees. You want to avoid absences. You want to increase your productivity. And you have so much data in HR to find out how to face these challenges. Why not use artificial intelligence (AI) to give you the answer you’re looking for?
5 HR experts answered this question at our HRbreakfast about AI and data. And in this blog post we give you Hans Donckers’ (JiGSO) and Jos Blykers’ (IBM) answer, starting with Hans.
> You want to identify people who need support
Challenge: once you’ve hired the right candidate, you want to keep him. You want to prevent absenteeism and increase your employee retention and the only way you can achieve this is by using data.
Solution: JiGSO links existing HR, financial and business data, using AI and machine learning. The company identifies people who behave similarly and which variables have an impact on their behaviour. JiGSO predicts and prescribes. Predict: who you should talk to. Prescribe: what you should talk about.
It’s like checking the weather, says Hans. You don’t want to show up at work soaking wet (at least, that’s what we think). You look at the weather forecast because you know it’s based on data, on facts. The same applies to HR. You want to know if your best employee is thinking about leaving your company. You then check the available HR data because you know it’s based on facts.
Our favourite quote from Hans? The one in which he links ‘bad guys’ to ‘where’s Wally’:
‘We’re not looking for the bad guys. We’re actually helping HR people to identify the people who might need your support, who might benefit from a talk with you. ‘It’s about identifying people. It’s about finding the Wally.’
> You want to take objective decisions
Challenge: you want artificial intelligence to be your trustworthy helper. You want to make better decisions, unbiased decisions and, ultimately, the right decisions. So, AI must behave in the same way humans should behave: always doing the right thing. Easier said than done.
Solution: IBM believes that for people to trust an artificially intelligent system, it needs to behave ethically. Compare it to human interactions: you only want to work with people you trust. IBM identifies 4 main ways to establish trust and we need all four:
- Robustness: protect your systems against hackers
- Fairness: define clear rules and create accurate data
- Traceability: log every step to trace back how you made a decision
- Explainability: explain how the system works. It’s like going to the bank: if your loan got refused, you want to know why.
Our favourite quote from Jos? The one in how hard it is to behave ethically:
‘The challenge is that some sensitive categories we don’t want to discriminate against (race, religion and handicap) are not labelled in our databases (nor would we want to do so). But how then can we formally test against possible unconscious bias?’
There is absolutely no link between breakfast and artificial intelligence. But if you combine them, you get great results: 5 fantastic speakers explaining the value of AI for HR. Do you want to know more about artificial intelligence? Read our first blog post. Do you want to join our events? Subscribe here for our HR events.