Since the 19th century when members of the Luddite Movement grew in strength, there have been fears that technology and the devices it creates could take away jobs. The secret oath-based movement of English textile workers, named after the fictive Ned Ludd, destroyed machinery as a form of protest. In more recent years organized labor in the United States has seen robots as a looming threat to job security even for skilled workers.
Modern technology today might be less about replacing workers than retaining them, as artificial intelligence (AI) is leveraged to help predict when employees are considering a move. Some experts argue the technology is so robust, it might predict when an employee is ready to make a move before they’ve made up their mind on the issue.
Earlier this month Ranjeev Mittu, head of the information management and decision architectures branch of the Naval Research Laboratory’s IT division, spoke at a FCW conference on AI. He noted that the lab is working to develop AI tools to comb through data from exit surveys to flag common workplace issues that lead to employees leaving their respective jobs.
The Naval Research Laboratory isn’t the only one looking at how AI can be used to help retain employees. Tech giant IBM has developed AI that can predict which employees may be leaving a job with 95% accuracy, according to the company.
IBM is making full use of AI in its HR department, and is using its big data analysis to help determine what it will take to retain employees, including new skills training, education, job promotions and raises. AI can better predict the reasons why employees may even be thinking of seeking new employment elsewhere, argues IBM. By addressing these issues, IBM can keep its workforce engaged – whether it is adding a new skill or promoting a deserving worker.
“A lot of the techniques used by AI are based on data analysis,” said Jim McGregor, principal analyst at TIRIAS Research. “This data is pulled together over time and AI extrapolates from it.”
AI can be used to monitor an employee from pretty much from time he or she starts with the company.
“On AI and employee management, quite a bit of work is being done in terms of using AI to help analyze new hires,” explained Charles King, principal analyst at Pund-IT.
“For example, IBM offers a Watson Personality Insights API that customers can use to parse the verbal and written comments of potential hires during the interview process,” King told ClearanceJobs. “Some of their vendors are offering AI tools in their ERP processes to flag potential issues that might need a manager’s attention.”
Scanning Social Media
When it comes to where that kind of public facing, predictive media is coming from today, the answer is social media.
“It can help determine things like where someone grew up, are they home sick or do they want to be closer to family?” explained McGregor. “This can paint a vivid picture of someone’s state of mind over days, weeks and even years. This is what a lot of people don’t realize – there is a lot of data out there, and combined with everything else it can tell a lot about you.”
AI can gather this data and find patterns that a human might miss.
“Part of it’s that there is only so much time someone in HR can spend to access data on one person, but AI can pull all this data and then can compare it to models from other people,” McGregor added. “These models become more sophisticated over time. They can be used to look for similarities in two different people while still seeing things that could result in two completely different conclusions.”
Both the size of the data set and its ability to learn over time mean a computer’s predictive analysis capabilities far outpace the average human resources professional. While a person might overlook key facts, AI can determine how the volume of activity has increased or decreased compared to certain deadlines and projects.
“AI can tell if someone isn’t happy,” said McGregor. “It can look for patterns and build a psychological profile that can determine if someone is working under stressful situations and calculate how likely it could be for that person to decide it is time to look for another job.”
Over time, the use of these solutions is likely to increase, added Pund-IT’s King. “It’s just another, potentially valuable tool for of data analysis.”
