Most companies report measurable benefits from artificial intelligence (AI), with a group of high performers showing how best to apply the technology. However, a great deal of work remains within organisations to scale its impact, manage the risks, and retrain the workforce.
These are the headline findings from the new Global AI Survey from management consultants McKinsey & Co. The report, which gathered responses from over 2,600 professionals, shows a near 25 percent year-on-year increase in the use of AI in standard business functions, with a jump in companies using AI across multiple areas of their business.
Sixty-three percent of respondents claim revenue increases within specific business units from AI adoption, with respondents from high performers nearly three times likelier than those from other companies to cite gains of more than 10 percent.
Forty-four percent of respondents report cost savings from AI adoption in the functions where it has been deployed, with high performers more than four times likelier to say it has decreased business units’ costs by at least 10 percent.
McKinsey defines a high performer as a company that has adopted AI in five or more business activities, has seen an average revenue increase of five percent or more from AI adoption, and average costs fall by the same amount.
Cost decreases are most commonly found in manufacturing and supply chain management deployments, says McKinsey, with savings on the factory floor from optimising yields, more efficient energy use, and improved throughput. Revenue increases from adopting AI are most often reported in marketing and sales.
Retail has seen the largest sector increase in AI adoption, with 60 percent of respondents saying their companies have embedded at least one AI capability in one or more business units, up 35 percent year on year.
Overall, the results show that a leading pack of AI adopters are attaining “outsize business results from AI”, widening the gap between power users and laggards, says McKinsey. High performers are using the technology to drive value across the organisation, mitigate risks associated with it, and retrain workers – the roadmap that other companies should be following.
Despite media hysteria about job losses to AI, robotics, and automation, AI adoption has had a modest effect on the workforce sizes over the past year. However, roughly one-third of respondents say they expect AI adoption to lead to a workforce decrease in the next three years. One-fifth anticipate an increase in staff numbers.
In 2018, the World Economic Forum hit the headlines with a report saying that 75 million jobs would be lost to the global economy due to the adoption of AI, automation, robotics, and other Industry 4.0 technologies. That report predicted that 133 million new jobs would be created by the same route, thanks to productivity improvements and the birth of new companies, services, and products – a net gain of 58 million human roles, with a focus on data analysis and transferable skills.
Not everything in the AI garden is rosy or clear cut, warns McKinsey, and best practices are essential to capturing value at scale. These include: aligning business, analytics, and IT leaders on the value of AI across each domain; investing in talent; and ensuring that both business and technical teams have the skills necessary for successful deployment.
Supporting strategic aims is essential, says McKinsey. Seventy-two percent of respondents from high performers say their companies’ AI strategy aligns with their corporate strategy, compared with just 29 percent of respondents from other companies. Similarly, 65 percent of high performers have a clear data strategy that supports and enables AI, compared with just 20 percent of other companies.
But one area where most organisations are falling down is in recognising and mitigating against risk in AI: a troubling finding, given the high risk of AI automating systemic or historic bias. The report says: “A minority of companies acknowledge most AI risks – fewer mitigate them.
“Despite extensive dialogue across industries about the potential risks of AI and highly publicised incidents of privacy violations, unintended bias, and other negative outcomes, the survey findings suggest that only a minority of companies recognise many of the risks of AI use. Even fewer are taking action to protect against the risks.”
Fewer than half of respondents (41 percent) say their organisations comprehensively identify and prioritise AI risks, suggesting that nearly two-thirds of companies have a policy of ‘AI now, worry about consequences later’. That needs urgent redress.
McKinsey’s findings broadly mirror trends identified in other recent surveys of Fourth Industrial Revolution technologies, including a Capgemini report on smart manufacturing that was published earlier this month. However, that report found much higher failure rates in deployments at scale.
This, then, appears to be the repeating pattern in Industry 4.0 programmes: a leading pack of high performers – typically around 10 percent of sample size; much bigger challenges of scale and risk than managers had anticipated, and insufficient workforce training to bridge the gaps. Arguably, these are signs of widespread tactical deployments.
The longstanding advice must therefore remain: always apply these technologies strategically in support of business aims; look for measurable impacts; augment human expertise and skills rather than replace them; train and re-skill your human teams; consider the reputational and financial impacts of ignoring bias, especially if you are forced to reveal the methodology behind decisions; and don’t jump onboard the hype train expecting an easy ride to cost savings.
In other words: make your business smarter, more agile, and productive, not simply less expensive to run – and avoid letting human expertise ebb out of the enterprise. At the end of the day, your staff investments are the critical ones.