IBM WATSON HAS dabbled in all kinds of work from editing highlights video at Wimbledon to diagnosing cancer risks. Despite those two items demonstrably not belonging on the same human CV, IBM's plucky AI has a new HR-themed string to its bow: knowing when employees are about to quit.
At CNBC's Work Talent + HR Summit, IBM CEO Ginni Rometty revealed that the company AI has got so adept at detecting employee satisfaction that it's now in the "95 per cent accuracy range."
Rometty wouldn't be drawn on what data points Watson consults, but Googling the spelling of "curriculum vitae", alongside extended periods on LinkedIn would be our first clue. Nobody has ever visited LinkedIn recreationally, after all.
Watson's "predictive attrition program" is used to retain talent, because as Rometty says, "the best time to get to an employee is before they go." The company estimates that the early interventions will have saved it nearly $300m in retention costs.
One department that finds itself, uh, less retained is HR itself. Rometty said that the adoption of technology, cloud services and other modern improvements means that the size of the HR department has been reduced by 30 per cent. The remaining workforce is said to be better paid though, so Watson probably doesn't need to keep too close an eye on them.
Of course, this isn't the first time Watson has dabbled in company matters. Last summer, it was revealed that the AI is getting better at analysing data points to assess which employees should be looking at new responsibilities, promotions and raises, with a 96 per cent accuracy rating compared to HR assessments.
That could just mean the HR department was off, of course. Maybe that's why it's shrunk so much. µ
'Some of us like the misery'
That'll surely affect its credit score
Smarter than your average pair of smart glasses