If the headlines are to be believed, the hype around artificial intelligence (AI) is about to peak. Gartner's 2017 Hype Cycle suggests that—for some AI capabilities at least—we’re already starting down the slippery slope from the "peak of inflated expectations," to the "trough of disillusionment" as early implementations fail to deliver the tangible and widespread benefits that were promised. But how true is this contention? Will AI follow the established route of other hype technologies, or is it something altogether different?
The idea of AI
As a collection of technologies, AI has tremendous potential, which is unlocked and nourished by the growth of big data, cloud storage and unprecedented computer processing power.
But AI is more than a technology; it’s an idea that holds a unique and powerful position in our collective conscious. After all, while the underlying capabilities and some basic applications of AI have been around for decades (let’s not forget that Deep Blue beat Garry Kasparov all the way back in 1996), for most people AI resides firmly in the realm of science fiction; belonging to another time, place or planet, and representing a vision of our greatest hopes and fears for the future.
Today, however, we’re starting to experience AI in reality; and judging by the sheer volume of newspaper headlines on the topic it’s a prospect that we find fascinating and scary in equal measure. The idea of a technology that might sense, understand and mimic our thoughts and actions is an inherently and deeply emotive one. AI inspires excitement and instils unease because it speaks to the very essence of what makes us human.
Learning from the past
Even in its emerging nascent state, we’re beginning to appreciate just how disruptive AI and its related, emerging technologies promise to be. Claims that we’re experiencing a Fourth Industrial Revolution haven’t been exaggerated: AI promises to be a major catalyst in the reorganisation of our industries and societies. Yet the primary narrative around AI is so firmly fixed on the future and all its attendant uncertainty that we risk forgetting the lessons of our past.
The First Industrial Revolution was driven by technology, but it was accompanied by a great "Age of Reform": campaigns and regulations that shaped everything from workers’ rights and education, to the establishment of the National Trust to protect the natural landscape. The organisations, rules and social contracts that were put in place during the 19th century continue to shape the way we live and work today.
The Fourth Revolution will be at least as disruptive as the First, and we’ll need to consider carefully the governance we need to protect the rights, traditions and environments that we value. Moreover, we now have an additional and complex new technical landscape to consider—the vast fields of data and digital identity we have sown and cultivated; data that will be critical to the successful application and advancement of AI. As AI moves from hype to reality, we must find ways of ensuring this data is used in the right way; delivering the benefits we want, while also protecting the privacy of data owners.
Here, at the start of the journey, it’s difficult for us to imagine all the ways in which AI could transform the world, or the many social and moral questions we may need to grapple with; but the parallels are clear—our industrial heritage has much to teach us as we form our technological post-industrial future.
Beyond the hype
As the hype around AI reaches a tipping point, it’s only natural that we enter a new phase. However, AI is not a typical technology, and this won’t be a typical hype cycle. General AI—the possibility that machines will equal, or even surpass, full human intelligence—is something we may experience in the coming decades, or it might continue to present a tantalising mirage on the future horizon.
However, "narrow" AI, applied for a specific purpose, is ready to solve the problems of today. We therefore need to enter a new stage of responsible and purposeful growth; a focused evaluation of where today’s AI can and should be used in the near-term, alongside a more holistic and inclusive debate which embraces citizens, organisations and governments. Moving into this new phase will enable us to harness the unique and transformational potential of AI, while managing the impacts and challenges that accompany it.
The hype may well be reaching its peak, but our conversation on AI is only just getting started.