Bring together mathematics, increased processing power and big data, and the result is Artificial Intelligence. Although AI has made headlines over the past few years, it is by no means a new topic, the term having been coined in the 1950s. The theory of AI preceded the practice primarily because of two constraints. Early computers were a far cry from today's machines, and data storage was limited in the past. Fast-forward to today, and technological advancement has produced strong computing power at affordable prices, while the world has become flooded with electronic data, thanks to the internet and data capturing instruments.
The human brain can be likened to an electric machine, it collects signals, it processes them and then produces an answer. Not everyone's brain is the same, some people can perform tasks better than others in terms of speed, complexity and robustness. The other problem is that people do not come equipped with an instruction manual, meaning that it is often hard to identify what has influenced their answer (facing a choice between taking cake or ice-cream, a person might choose a cake, not after a careful examination of the millions of relevant factors and their associated objectively-set weights, but simply because that person remembers ending sick after taking an ice-cream).
Computers do not have emotions. They produce the output they have been programmed for (one must acknowledge the bias of the programmers, but at least this could be documented and traced afterward). If the presence of clouds indicates a higher probability of rain, the cloud signal would immediately advise carrying an umbrella rather than wearing sunglasses. Now add all the complex analysis which can be performed when feeding the data of all rain episodes around all the world for the past decades and under various climate conditions. A computer can perform such repetitive and large analysis more efficiently than a human.
The case for widespread use of AI in all aspects of our life is clear. Even if AI is not perfect, the risk-return combination is often superior. However, the pace at which AI will infiltrate our lives will ultimately depend on decisions by society. The recent episode of global IT outage caused by a bug may deter people from making the full shift. Despite emails being in use for many years, physical letters still exist. Despite electronic money and payments have been around for a long time, people still use cash.
Although the increased use of AI is inevitable, its propagation is far from certain, particularly as lobby groups might actively pose barriers. Will there be areas where AI will be outright banned, similar to the ban on human cloning? This explains why the impact on future productivity is hard to pin down, with estimates ranging from modest to extremely high.
AI also raises very serious issues across the globe. Will the divide between the developed and underdeveloped world balloon because the latter are unable to benefit from this new tool? Will the impact on office-based jobs be disproportionate, as white-collar employees are outperformed by computers? At the same time, one must recall that in the Western world, the working-age population is likely to shrink due to ageing, so any opportunity to reduce the required labour resources should be embraced. Ultimately, capital deepening and technological advancement is what can sustain future standards of living, so AI is a bonanza. Even though the first flight attempts by the Wright brothers might have been considered clumsy, this did not hamper our desire to fly. AI promises to be bigger than the industrial revolution, which is a worthy tribute to the visionaries who first explored AI.
The views expressed in this article are those of the author and do not necessarily reflect the views of Bank of Valletta