What’s ahead for AI in the workplace?
While artificial intelligence will greatly enhance efficiency, it will never replace humans, say experts
When people of think artificial intelligence (AI), they might have visions of Hal 9000 in ‘2001, A Space Odyssey’. Photograph: iStock
When people think of artificial intelligence (AI), they might have visions of Hal 9000 in 2001, A Space Odyssey. But with AI that seemed science fiction only a short time ago now becoming reality, is it so far off the mark?
AI is currently empowering the workforce to deliver better customer experiences and better results. So what is on the horizon, when it comes to AI in the workplace?
John Kilbride, who works in the area of cognitive and robotics at Deloitte, says that while there have been a few false dawns with AI, things are different this time around. Key enablers are falling into place in the past five years, including the exponential increase in computing power; the digitisation of everything resulting in an explosion of (customer) data; and the advent of smarter algorithms which underpin AI.
So how smart is AI?
“What’s important to qualify is that the commercial applications are focused on ‘narrow’ AI – we are still a long way from any ‘general’ or transferable AI. This is significant. Most of us use, even if we are not aware of it, the AI behind the Google search engine or Netflix and Amazon’s personalised recommendations. We are also coming to terms with the idea of a self-driving car, a robot barista, computer vision that can automatically identify people in a crowd or improve the accuracy of the investigation of MRI scans in a hospital. But the artificial intelligence behind any one of these applications doesn’t transfer to the others and none of them could explain a broken heart to a teenager. That highlights another caveat on the hype; the progress we have seen is in a specific type of intelligence, the ability to mimic certain types of human cognitive capability, for example perception, vision, problem-solving, but it falls short in other areas, most notably emotional intelligence,” Kilbride says.
In terms of adoption, the large technology firms are leading the market because they have invested heavily in AI start-ups and have tempted the best academic talent out west.
“As digital-first businesses, they have the requisite customer data and engineering skills to mine that data,” says Kilbride. “The rest of the business world is still coming to grips with trying to move to digital economy whilst they are positioning themselves as AI-first. The gulf is astronomical at this stage. Does that mean that we won’t see AI become mainstream any time soon? No, not at all. The world will adopt AI in two ways, gradually and then suddenly. The last 10 years have seen massive shifts as firms invest in digital and mobile technology; now we see a shift towards using cloud and other exponential technologies. These enablers all result in the capture and storage of large pools of digital data. All that’s missing is the right engineering and data science skills to apply the artificial intelligence to uncover the patterns and unlock the insights in this data.”
Law firms like Pinsent Masons are re-examining the delivery of legal services in the digital age, looking at how they can better use emerging technologies to better utilise people, make processes more efficient and add more value for the client.
“We have a team of over 20 tech professionals and they are a mix of computer scientists and data scientists, employed by us, and they help us to develop meaningful programmes to support our smart delivery. We have ideas around products that will make life easier and we then work in collaboration with them,” Andreas Carney of Pinsent Masons says.
AI at work
He gives an example of AI at work in Pinsent Masons.
“In one jurisdiction we work in, we focus on employment disputes and decisions made by that country’s employment tribunals, so our researcher went and did a deep dive on all the decisions that were issued by the tribunal on a particular topic, say discrimination or wrongful dismissal, and they generated a substantial amount of collective knowledge from that. They took all of the decisions and categorised them to the particular type of action. They engaged with our technology team and coded various scenarios into the product we’ve developed,” Carney says.
The technology then allows them to identify trends in previous decisions and to identify best-case scenarios or practices, enabling lawyers, without spending a lot of time, to locate the relevant documents, creating a much better client experience.
In relation to the threat of job destruction, the media has foretold of the robot revolution, but Kilbride does not see it as black and white.
“What we have seen to date is a lot of interest in automation, for example using robotic process automation to automate white-collar or service-economy tasks.
“Technology, including AI, has always enabled opportunities for new roles. Make no mistake, the foreseeable future is about people, not robots. Much of the power of AI is in augmenting the efficiency and effectiveness of the human employee. In cases where AI displaces low-skilled or ‘robotic’ tasks, then this pushes humans up the value chain and therefore it’s imperative that we continue to invest in developing the cognitive skills of our human resources. For 25 years, the internet has literally changed the way we think. It’s imperative we adapt and re-think our approach to education, professional learning and development and crucially mindfulness in the face of the threats and opportunity from AI,” Kilbride says.
Carney says robots will never replace lawyers.
“Technology is changing the way that lawyers work and for law firms to remain competitive, they need to embrace technology. AI won’t assess risk that arises from whatever the outcome of the review processes of the computers are. While the technology can identify the documents, you need someone with experience to assess what it means for the client. That’s the piece technology won’t replace. It will enable lawyers, free them up to spend more time on high-value work and takes us out of the labour-intensive aspects. To make a good judgement call, you need years of practice and human understanding and technology isn’t there yet. Maybe at some point it will be but that’s highly disputed in the technology world,” he says.