Debate over AI in the workforce requires a broad view
Enterprises that are poised to implement AI in the workforce don't have to forfeit human jobs to do so. Read a Dun & Bradstreet analyst's take on the intent of enterprise AI.
It seems every analysis of AI in the workforce boils down to one question: What are the implications of enterprise adoption of AI on job security and safety? Those afraid of AI's place in the future of work say it will automate -- and thus eliminate -- a majority of jobs. Those with a stake in AI say that adding AI to the workforce will revolutionize human employment, not destroy it.
Here's an overlooked aspect of the debate: The aim of AI is not to affect the job market in any way; rather, the goal is to serve or assist humans in achieving a goal. Anthony Scriffignano, chief data scientist at Dun & Bradstreet, advises enterprises to stop thinking of AI in terms of having intent -- whether malicious or benign. It all goes back to simple process automation.
Think by enterprise
Instead of looking at AI's effect on the total global workforce, data scientists, employees and employers need to consider AI locally -- how is it being implemented in specific job functions.
Take, for example, customer service call centers and the implementation of AI chatbots.
"[Chatbots] don't start out with the goal of seeing how many low-paying jobs they can eliminate," Scriffignano said. "The company actually might have an infinite [call] demand, and they're never going to eliminate one job."
"Their metric there isn't how many jobs are eliminated. Their metric is: 'How many more customers can I serve before they hang up?' or 'What's my customer retention rate on known disgruntled customers?'" he continued.
When examining AI and the future of work at the enterprise level, you get a much rounder idea of the potential affects. Certain enterprise functions -- such as call centers, help desks or HR -- will augment the workforce to improve customer experience but likely keep the same level of employees.
Further, there are entire enterprises being created from the increased use of AI.
"There are definitely some lost jobs, but there are also a lot of people training algorithms now. There are a lot of people graduating school learning to code, people that are able to do things in spaces that were untouchable before [AI]," Scriffignano said.
Who gets automated first?
Implementation of AI in the workforce is growing rapidly, contributing to the rising fear that automation will eliminate jobs.
Anthony ScriffignanoDun & Bradstreet
Dun & Bradstreet conducted a survey of 100 business executives at the AI World Conference and Expo that showed 44% of businesses are in the process of deploying AI, with 23% in planning stages. AI is being rapidly adopted. But in the debate over the future of work and AI, implementation means different things to different people. Does a company that has automated its help desk successfully count as total AI integration? If so, then job loss is inevitable.
But when looking at AI adoption in the context of enterprises broadly, we're not seeing widespread loss of jobs. In Dun & Bradstreet's survey, 40% of respondents reported an increase in jobs after AI implementation, 34% reported no change and 8% reported job loss.
That's because AI deployment in enterprises is still a financially motivated process, which means that certain jobs will be automated first, while others may not be impacted. Entry-level jobs that require little or no special cognitive skills or training are likely the first to be automated.
"If you were to look at an industry and say, 'Where are we going to feel this first?' You're probably going to feel it in the unskilled, easily replaceable, repetitive tasks, just like you did with the Industrial Revolution," Scriffignano said.
Amid the hyperbole about lost jobs that has come to feed AI apprehension, Scriffignano sees a brighter future with AI in the workforce, where automation across enterprises delivers more benefits than challenges.
"There are lots of people who have come out and taken a dystopian view that computers are going to ultimately take over the world and wipe us out. Then, there's a more hopeful view, which is that we still -- at least for the foreseeable future -- call the shots and engage where we choose to and AI acts on our behalf," Scriffignano said.