How AI is changing the way we work

The long-term impact of the robot revolution on the human workforce has been overstated: the truth is artificial intelligence is changing jobs, not stealing them 


The coronavirus crisis has given rise to fresh concerns about automation in labour markets. With people locked down and social distancing still in place, many businesses around the world have scaled up their investment in artificial intelligence (AI). 

AI’s ability to identify and learn from data patterns, and translate them into useful technologies, has proven to be indispensable for many organisations, from healthcare providers to delivery subscription services, for example, in responding to the pandemic.

Business applications of AI over the past 12 months have ranged from those designed to increase productivity and yield, through to customer-service functions.

Robots have been rolling in to sanitise UK and US hospital corridors and deliver crucial supplies such as blood samples. Cincinnati/Northern Kentucky International Airport has enlisted mechanical floor scrubber Neo to whizz around and clean its floors every night, freeing up staff to make sure high-touch areas of the terminal are spotless. Following its $300-million acquisition of so-called decision engine Dynamic Yield in 2020, McDonald’s has also got in on the AI action, trailling drive-through systems chaired by smart speakers.

Goldman Sachs, L’Oréal and Procter & Gamble are among those that have bet big on less-robotic looking AI tools to meet the challenges posed by a move away from physical consumerism to digital, including technology that predicts people’s financial outlook and “try-before-you-buy” product-matching systems.

However, AI’s business impact will outlast COVID-19. Recent data from Grand View predicts the market will be worth $390 billion by 2025, marking 46.2 per cent growth since 2019. McKinsey is showing an appetite to invest in machine learning, with C-suite leaders who are seeing strong results from AI adoption gearing up to boost their spend over the next three years and develop their own solutions in-house.

There are implications for humans, though. According to a World Economic Forum forecast, in a shift likely to worsen inequality, half of all work tasks will be handled by machines by 2025. A separate report from the Massachusetts Institute of Technology projects AI could replace as many as two million more workers in manufacturing alone by 2025.

This shift, coupled with an economic downturn, is paving the way for fresh concerns from employees about how many jobs will be lost to automation, especially for those in the services, electronic and manufacturing industries. In the United States alone, a little over a quarter (27 per cent) of all workers say they are worried the job they have now will be eliminated within the next five years as a result of new technology, robots or AI.

However, AI’s automation effect may well be overstated. Analysis using data from the European Centre for the Development of Vocational Training reveals that, in the European Union, the share of jobs facing a very high risk of being automated by new digital technologies is close to 14 per cent. However, what two in five EU jobs do face is a high probability of “substantial transformation”.

So, while AI might change how humans work, the reality is it’s unlikely to replace most jobs entirely. And, although “robophobia” is on the up, there are several examples of how machine learning is already positively impacting employees across a host of industries.

AI’s role in supporting the workforce

Dr Samer Al Moubayed, chief executive at Furhat Robotics, a conversational-AI social robotics startup that builds tech designed to interact with humans in a natural, fluid way, has overseen the conception of the firm’s namesake robot that can speak, show emotions and maintain eye contact.

The past few years have seen Moubayed work with the likes of human resources company Tengai to develop a robot that autonomously performs job interviews, scores the interview according to an established framework and summarises the output for a human recruiter.

Furhat Robotics is also collaborating with Deutsche Bahn to place intuitive travel assistants in train stations that can answer questions related to departure times, delays and more in over 35 languages.

Moubayed is firm in his belief that such AI investments can support humans in their day-to-day jobs, working in harmony with the workforce instead of against it. Robotics, he argues, will open up more job opportunities.

“Looking at innovations like Tengai, though the solution might reduce the need for recruiters, it won’t remove them from the process entirely,” he explains. “It will also increase the need for robot operators.”

The same applies to Deutsche Bahn travel assistants. “Long term, it might lessen the need for hiring multiple customer service agents to cover a broad range of languages, but it will require developers and translators to a higher extent,” says Moubayed.

The AI expert cautions that no one is in a position to understand the lasting net effect automation will have on the workforce. But, done right, many of the risks associated with automation can be mitigated in a similar fashion to previous industrial revolutions, through upskilling staff and also upholding a sense of social welfare.

“Businesses should focus their AI investments on increasing the average quality of life for workers and customers, rather than simply relying on automation as a tool to cut costs,” he says.

Identifying human gaps, instead of filling them

Marketing giant Publicis Groupe credits the use of its internal AI application Marcel for helping it save 2,000 jobs during the pandemic. As a result, it plans to spend more heavily on the tool, having seen an encouraging impact on its workforce.

The smart intranet works by assigning relevant briefs and projects to Publicis’ 80,000 staff. As ad execs everywhere from Tokyo to Toronto shifted to remote working at the outset of 2020, Marcel hosted an internal job mobility platform that allowed people to change agencies, move to different markets and stretch their skillset. 

Suddenly, as some Publicis-owned agencies and markets found themselves working harder and faster to respond to COVID, others were experiencing a slowdown as clients in the hardest-hit sectors, like travel, were incapacitated. 

Marcel enabled those overstretched agencies to post jobs and requests, making use of talent that had time on their hands. 

Dan Murray, Marcel’s chief executive, says as well as automating the mundane, the tech is enabling the business to identify human gaps for things robots can’t do, like think creatively or communicate effectively. 

“It’s allowed our staff to focus on things that might take a higher-level thought process or more creative mind. It’s also opened opportunities we might not otherwise have had,” he says.

The business is now plotting a lot more additional uses for Marcel’s future, based on AI and machine learning, that will connect people and allow staff to better flex their creative muscle.

Improving the working experience

Tech behemoth Cisco is leveraging AI, machine learning and smart data to enhance the quality of remote working for its 75,000 staff, from engineers to sales reps.

With 96 per cent of employees currently working away from the office and a hybrid working model here to stay for the long-term, it has been using deep-learning algorithms to power an “empathetic, inclusive, secure, work-from-anywhere” experience, says senior director of people and communities Gianpaolo Barozzi.

Solutions include use of intelligent video so people can move around freely in their workspace and still be available in meetings, as well as real-time meeting transcription and automated debriefs.

An AI tool, dubbed People Insights, is also providing individuals and teams with analytics around personal and collective work practices.

“This means staff can make better use of their time and take care of their peers,” says Barozzi. “The 21st-century world of work is about connecting people to people, wherever they are, safely and securely. Productivity is no longer a sole matter of efficiency, it’s increasingly achieved through inclusive collaboration, adaptability to unforeseen change and the effective leverage of collective strengths. AI is helping our teams in all these regards.”

With most jobs set to involve some form of AI by 2025, he stresses education is key to ensuring staff are comfortable with these types of technologies, rather than fearful of them.

Cisco’s standing as a tech firm means its people are largely open to being early adopters, but his advice to other businesses is clear. “Fears are real, so first of all we have to treat them with empathy. We need to be transparent and clear about the opportunities, while not underestimating the risks and challenges,” says Barozzi.

“It’s time to move beyond the paralysing utopia-dystopia dichotomy: as always the future is ours to shape. Digital has already turned from a tool to the environment where our lives happen. Humans and artificial agents will increasingly work together, enhancing and amplifying each other’s capabilities.”

How AI helped Publicis save 2000 jobs

In the immediate months after Covid-19, Publicis Groupe’s revenues declined by 13%. The French advertising network, which owns agencies such as Saatchi&Saatchi, quickly implemented a €500m cost-cutting drive that included a recruitment freeze, elimination of freelance work and cuts to general expenses.

As the business considered the impact of the pandemic on its workforce, however, AI emerged as an unlikely saviour. Its Marcel management tool, aimed at reinventing how the agency communicates internally, was launched in 2018, but in the crisis has truly come into its own.

“We woke up one day and instead of having 80,000 employees we had 80,000 offices,” says Carla Serrano, chief strategy officer of Publicis Groupe. She reveals that Marcel’s global roll-out was accelerated as the company realised the importance of connecting staff with relevant briefs and each other. “We didn’t know it would end up being the saviour of so many people’s jobs,” she says.

The key driver was Marcel’s ‘Gigs’ function. Pre-pandemic it hosted an internal job mobility platform that allowed people to change agencies, move to different markets and stretch their skillset. Amid the crisis, Gigs pivoted to use algorithms to match individuals with briefs based on their skillset and allocating resource to clients where it was most needed.

Key to its has been educating staff on the platform and ensuring the digital experience runs smoothly. Publicis design teams have focused on making the platform as “digestible and personalised” as possible, with work ongoing around the user experience.

Prior to 2020, Serrano admits she had reservations about AI’s impact on the workforce, but is now “bullish” about the benefits it can bring. “It’s immensely benefited our people and culture, as well as learning and discovery. We can use AI for good. There are things we need to be cautious of – like privacy or bias – but we’re certainly learning a lot more about it.”