Artificial intelligence has long had a reputation as a bogeyman threatening to replace jobs and make workers redundant in a relentless pursuit of efficiency and cost reduction. But this seems to be changing, as new research suggests employees are beginning to appreciate the value AI and machine learning can bring to their workplace. However, despite this new enthusiasm from staff, there are still significant hurdles to overcome if AI adoption is to become more widespread.
The survey of office workers in the UK and US, carried out on behalf of SnapLogic by market research company 3GEM, found that 81% believe AI has already improved their own performance at work. A smaller share of respondents, 56%, confirmed that they were already using AI to assist with their daily work projects, but 68% said they hoped their employers would deploy more AI-based technology.
This acceptance of AI as a helpful technology, rather than one that will take jobs away, is also reflected in a recent Juniper Networks survey of 700 people working with AI across different levels and industries, which found that 95% of all respondents believe AI would benefit them if it is applied across their organisations' daily operations, products and services. The enthusiasm across the pool of respondents was borne out by the 88% who said they wanted to use AI as much as possible.
The most important benefits of AI integration into daily work, according to the SnapLogic survey, were increased productivity and efficiency (according to 61% of respondents), followed by improved decision making and accelerating the journey toward getting insights (49%). Overall, respondents pointed to data-related tasks as the main beneficiaries of AI integration – 43% said AI can help them the most in understanding data and how trends and patterns can aid them in decision making. The other two preferred use cases for AI, both at 41%, were moving data from one place to another and accessing data stored in different places across the organisation.
By automating repetitive and technical tasks, AI has also helped employees put greater focus on creative work and facilitated collaboration between and within teams, as well as customer engagement. "As AI is increasingly used to make better decisions and rack up productivity gains, [workers] have gone from tentatively accepting to fully embracing AI," SnapLogic CTO Craig Stewart says, describing this desire for stronger AI adoption within their companies as "a real sea-change".
Is there really a new enthusiasm for AI in the workplace?
Automation has long been associated with the erosion of job security. Indeed, a survey by Boston Consulting Group and online job portal TotalJobs, published in April, found that 35% of UK workers on average had become more concerned about losing their jobs to automation since the start of the Covid-19 pandemic, with anxiety being strongest in the media, technology and financial service sectors.
Katherine Holden, head of data analytics, AI and digital ID at trade body techUK, says that concerns about job loss due to automation are still present. However, she explains, "the reality is whilst we may see some jobs change and decline over the next few years, new roles will also start to emerge". Holden points to the potential benefits of AI augmentation, where AI and existing employees work together to get a task done. She cites a recent report by Software as a Service analytics platform provider Fatehm.ai, which concluded that 2.9% of work in the UK could be augmentable using technologies that are available in 2021. According to the report, augmentation at this scale could result in a 1.3% capacity gain across the whole economy.
What's stopping more companies from adopting AI?
Despite this increased desire from staff for AI systems, there are still major obstacles for UK companies seeking to adopt AI and ML in their operations, including a lack of transparency and accountability. "Companies are aware that they need to do something to make sure that their AI and ML systems are fair and non-discriminatory," says Gina Neff, professor of technology and society at the Oxford Internet Institute at the University of Oxford. "The problem is they don’t know what to do or how to do it or where to put it in their organisational chart."
TechUK's Holden adds that the major stumbling blocks to AI adoption before the pandemic remain a problem, and have "arguably been exacerbated" by the Covid-19 situation. She points to four major issues UK businesses are facing. The first one is democratising access to data as AI research and development needs high quality, readily available training datasets. Holden calls on the government to incentivise and facilitate the sharing and reuse of data across organisations and between sectors. "If we can lead the way in driving effective data sharing between both public and private sector service providers, we can accelerate our ability to develop AI solutions," she says. Another major hurdle before AI adoption is the ability of UK companies to attract and retain the best AI talent from outside the country.
UK businesses have made headway in another important step toward AI adoption – digital transformation has accelerated since the start of the pandemic, but SMEs are facing financial struggles as they seek to implement digital technologies across their operations, Holden notes. She adds that government support is needed to ease AI adoption.
While UK businesses have a lot of work to do if they seek to make it easier for their employees to do their jobs, this doesn't mean that AI hasn't found many other use cases. Holden cites the use of AI to prioritise urgent appointments for patients in NHS hospitals to help clear the Covid-19 backlog as an example. In another instance, Google's Cloud AI Rapid Response Virtual Agent has helped organisations swarmed with customer questions about the pandemic deal with the workload by using customised AI virtual agents to provide 24/7 support.
The humans behind the technology
All these success stories illustrate the promise of AI but they would have been impossible without the humans behind the technology. Professor Neff points to a comparative study of Danish and US hospitals by her team at the University of Oxford, which found that a lot of data systems required new kinds of care work to make sure that the results were fair, ethical and responsive to patients’ and clinicians’ needs. "Behind every AI system is a lot of human work getting the data and the models right," Neff adds.
The good news is that businesses seem to appreciate the importance of a capable workforce to help deliver the benefits offered by AI. Holden points to two initiatives, by Microsoft and PwC, aimed at helping millions of people worldwide build up their digital skills to respond to the needs of the digital economy.