Companies will need to completely overhaul long-held ideas about managers’ roles as a new analysis suggests the rate of Artificial Intelligence driven job losses is accelerating.

Gartner research released last week predicts artificial intelligence algorithms could take over 69 percent of the routine work currently done by managers.

Concerns about ever-better AI obviating the need for workers are nothing new, but early analyses suggested knowledge workers would be safer than process workers because AI lacks the human intuition required for effective decision-making.

Yet the degree of anticipated job replacement has increased over time as technology improves. In 2017 a McKinsey & Company analysis, for example, predicted half of current work activities could be automated by 2055.

Up to a third of the workforce would need to retrain for other jobs – although Australian workers were safer than those in many other countries, with 45 percent of Australian jobs up for grabs but positions in sectors like retail, hospitality, manufacturing, educational services, and mining flagged as having below-average automation potential.

Jobs on the line

Last year, one AI expert predicted new decision-making technologies would potentially eliminate 40 percent of jobs within 15 years – with Western Australian innovator Fastbrick Robotics a classic example by developing a robot that can lay 1000 bricks per hour.

Yet Gartner’s latest analysis moves up the timeline significantly, sounding a death knell for people in slow-changing jobs that require a low degree of ‘social-creative skills’ – think white-collar roles like computer programmers, accountants and lawyers.

One site measuring the risk of automation suggests 94 percent of accountant and auditor jobs, 56 percent of teacher assistants, 48 percent of computer programmers and 25 percent of managers were at risk from automation.

Jobs requiring more social-creative skills, such as marketing managers, sales representatives, and customer service representatives, needed to be trained to have more “digital dexterity”, Gartner said, because “chances are high that new technologies to support work will be introduced”.

You will be assimilated

Whatever the numbers, the consensus AI is coming for your job has created an uncomfortable reality: Many workers assume technological displacement is inevitable, but those same workers assume the displacement will happen to people in other jobs.

Public dread about AI had led many managers to become “wary” of introducing AI because of widespread concerns about job losses, Gartner noted, but restraint would become increasingly difficult as cost-savings calculations tipped in favour of automation’s “significant” benefits.

Companies could save an average $7m ($US5m) per 1000 employees or 100 managers, Gartner predicted, as activities such as administrative tasks, project management, team management, and development of direct reports are replaced by technologies like expense management, automated project status dashboards, algorithmic management platforms, and real-time chatbot interventions, respectively.

“The role of manager will see a complete overhaul in the next four years,” Gartner research Vice President Helen Poitevin said.

“Currently, managers often need to spend time filling in forms, updating information and approving workflows. By using AI to automate these tasks, they can spend less time managing transactions and can invest more time on learning, performance management and goal setting.”

Companies would need to embrace a number of ‘necessary’ technology advances including increased use of automated decision management, embedded analytics, AI-driven internal talent marketplaces, and transformation of application logic in manager-centric human capital management systems.

Machine learning would replace anomaly detection in work schedules, time and attendance, expenses, and “other processes where managers have been given oversight tasks”.

Look on the bright side

Managers will likely face the axe as automation makes many of their functions redundant, but those that remain will also need to drive change in organisational learning for another quite specific reason.

As increasing automation kills off many jobs and reduces the number of task-based activities that managers perform, Gartner warned, on-the-job learning opportunities would be cut by two-thirds.

This would mean that workers would no longer be able to learn by watching and doing, forcing companies to overhaul long-established mentorship and training structures based around the idea of on-the-job training.

By 2025, automation would have driven so much organisational change that 47 percent of learning and staff development budgets would simply be wasted, Gartner predicts – with learned skills relevant for increasingly short periods and digital dexterity a crucial measure of organisational adaptability.

Yet skills are not perfectly transferrable and companies must think big when it comes to retraining: If workers can’t reskill to crabwalk into AI-protected jobs, they may end up unemployed despite organisational efforts.

Revolutionising the workforce

Despite the potential disruption it will cause, Gartner believes AI will also revolutionise the workforce by empowering disenfranchised workers such as people with disabilities – where a 69 percent US unemployment rate continues to constrain the pool of potential skilled workers.

AI will reduce barriers to access for this workforce, the analysis concluded, predicting that AI would help triple the number of people with disabilities employed by 2023.

This would have flow-on effects for the companies, with organisations actively employing people with disabilities said to have 89 percent higher retention rates, 72 percent higher employee productivity and 29 percent higher profitability – generating goodwill that “should provide some reputational protection” for companies suffering “adverse actions, such as layoffs”.

Whether that goodwill can save the modern workforce remains to be seen – but AI’s push into the knowledge economy will nonetheless continue unchecked.