The AI 'myth'

Future net job loss as a result of automation is much overstated.
Robots 4 x 3

Artificial intelligence (AI) could significantly transform the way we work, but new research finds the phrase is a hijacked term, and notably absent from real workplaces.

Instead, businesses are adopting more modest automation technologies, which are apparent across numerous industries, applications and shared services. 

The findings are published in a new book, co-authored by Professor Leslie Willcocks of LSE’s Department of Management and Professor Mary Lacity of the University of Missouri-St Louis, providing a more nuanced picture of automation in the workplace.

“Our research for this book reveals the incoherence and exaggeration surrounding the application of AI and cognitive automation to work,” says Professor Willcocks.

“AI is notably absent from the major organisations we studied across Europe, USA and Asia Pacific, but has almost become an umbrella market term for the whole area of robotic process and cognitive automation,” Professor Willcocks adds.

Building on two previous studies, Robotic Process and Cognitive Automation: The Next Phase, presents findings from 85 case studies from a range of organisations which are using automation. Over 600 automation providers, clients and analysts were surveyed as part of the study.

Robotic process automation adoptions – the automation of tasks previously performed by humans using rules to process structured data – are found to be accelerating, maturing, and scaling in global enterprises.

Although an immature market, more advanced cognitive automation – based on machine learning, data visualisation technologies, natural language processing and fast powerful computing – is experiencing exponential growth which the researchers anticipate will continue for the next five years. 

But the authors question whether AI applications that get computers to replicate how the mind works are anywhere near ready for major workplace deployment, and suggest they are unlikely to be so for decades.

Contradicting the picture of perfectible technologies and fast seamless deployments to scale, Professors Lacity and Willcocks provide insights into the challenges businesses face, and use case studies to establish action principles that mitigate automation risks.

The book outlines 32 emerging management practices which organisations can adopt in order to gain significant business value from investments in service automation technologies.                     

The researchers also tackle popular misconceptions regarding the threat of automation to the job market. “Future net job loss as a result of automation is much overstated”, says Professor Willcocks.

“The deployment of these technologies will take a lot longer than many assume. Meanwhile shortfalls in productivity, dramatic increases in the amount of work we have to do, and skills shortages – amongst other factors, for example job creation – are being seriously underestimated.”

Behind the article

Robotic Process and Cognitive Automation: The Next Phase by Professor Leslie Willcocks and Professor Mary Lacity is published by SB Publishing.