In the previous blog posts, we have explored how machine learning and artificial intelligence (AI) are at the heart of the next industrial revolution towards an algorithmic economy. We discussed how these technologies reshape businesses by automating up to 99% of all operational decisions.
A recent report by McKinsey (“The age of analytics – competing in a data-driven world”) highlights the opportunities still to be seized and maintains that deep learning is “the coming wave” of innovation. For example, only between 30-40% of value has been captured in the retail sector so far according the report, highlighting that the major barriers to further growth are Data Science know-how as well as siloed data within the retailers’ IT systems.
Other major obstacles for retailers are, according to McKinsey, that senior management is not getting sufficiently involved in bringing their business forward as well as securing internal leadership for data and analytics projects. Reluctance to invest in new technologies and solutions at scale is seen as a further barrier to advance the business. In other words, companies – retailers in particular – have the future in their own hands, but it requires boldness and vision within each retailer’s senior leadership team to embrace the next wave of innovation.
Artificial Intelligence - more than a hype
However, just blaming senior management for their apparent lack of leadership and vision isn’t quite the whole story. Machine learning, deep learning, and artificial intelligence are the focus of many, many blog posts and news articles at the moment – as other topics have been before and others will follow in the future. For us “in the trade”, it’s obvious that embracing data-driven decisions at the core of each business will bring significant benefit to any company, but from an outside perspective it’s hard to discern what is a hype fuelled by buzzwords and what is an emerging new technology which will bring long-term benefits. After all, hypes come at go and are even recorded regularly in Gartner’s hype cycle – chasing after each of them clearly isn’t beneficial to bringing a company forward. The lack of talent doesn’t help as there are few AI experts to start with and those with deep knowledge in machine learning and AI, combined with a thorough understanding of a distinct and complex vertical such as retail, are few and far between. Who should C-level managers and board members ask?
What’s more, automated data-driven decisions directly tap into tribal fears that machines will take over our world. It begs the question: How will our new relationship with machines look like, in particular in the business world? Is it “humans and machines”, or rather “humans vs. machines”, or even “humans or machines”? Do humans even come first in this comparison or is it more like “machines without humans”? Undeniably, automation means that jobs currently done by humans will change, sometimes so significantly that they will be hardly recognisable from today’s point of view. Bill Gates even suggested in a recent interview with Quartz that robots should pay taxes, at least temporarily, to slow the rate of automation until we as a society have figured out how to deal with the changes in employment. Fears that artificial intelligence may turn out to be hostile like Skynet in the science fiction series Terminator are not as unfounded as one might hope, following recent research from Google’s AI company DeepMind (Wired). When resources are scarce, AI turns out to be quite “human” in its behaviour and punishes a competitor, much like in the classic prisoner’s dilemma and other social science experiments used to study human behaviour. However, this also illustrates that in the end such behaviour is part of evolution. Our understanding of society must change and integrate AI into the understanding of how society works, including both humans and AI.
Automation creates new jobs
Automation itself is hardly new, starting with automated looms in the 18th century and Ford’s assembly line in the early 20th century. Jobs have changed significantly and many new occupations have emerged which were only possible because automation brought us forward to the next level. The Atlantic calls this “the automation paradox”: “When computers start doing the work of people, the need for people often increases”. A recent study by Quartz looked at the occupations in the US in the last 60 years. They found that of the 270 jobs listed, only 5 have been removed from the census because they have been made obsolete by technology (e.g. telegraph operators) and only 1 has fallen victim to automation (i.e. lift boy). A further article by Quartz, aptly titled, “The optimist’s guide to the robot apocalypse”, uses Amazon, known to be a rigorous data-driven company, as a showcase to illustrate that their headcount has more than tripled in the last 2-3 years, despite – or even because of – automating what can be automated.
The accelerating development of artificial intelligence poses significant challenges for companies ahead. Rather than just jobs being lost to automation, tasks will change, and new ones will arise. Change management becomes more important than ever as employees need guidance and training to adapt to a rapidly evolving workplace. Domain knowledge will continue to play a crucial role in making business decisions. Maybe the CEO of Deloitte, Cathy Engelbert, summarised it best in a recent post on LinkedIn when asked by her 15-year-old son about jobs and automation, she replied, “Don’t worry—I’ve never met a machine with courage and empathy.” She then continued, “We’ll still need those in the new economy. To be sure, technology will change what we do.”