There is a certain level of enthusiasm associated with the concept of artificial intelligence (AI) and its ultimate destination, which is the ‘singularity’: the advancement of machine intelligence reaching the level of human beings, and most probably surpassing it.
As it is a controversial subject in the scientific community, a myriad of intellectuals have voiced their opinions on the issue of its possible benefits. However, there has also been lots of speculation regarding the apocalyptic prospects of AI from the likes of Elon Musk and Stephen Hawking, particularly over how this technology could possibly bring the end of humanity since there is no way of knowing how a super-intelligence devoid of emotions would think and act. There is a possibility that it might turn into a misanthrope for all we know, and from an objective perspective it wouldn’t be a far-fetched stance for an entirely rational creation.
The AI Singularity: Dream or Danger?
But as elusive as the concept of an AI singularity is to the minds of the scientifically inclined, the majority of the population is more interested in the pragmatic consequences of this phenomenon. This is a very legitimate concern, especially now that the prospect of integrating AI into our economy is becoming ever more pressing.
The general sentiment among policy-makers and companies is that AI can have very favourable effects on the economy. For example, Accenture’s report on the issue claims that AI technology can increase the productivity level of 12 advanced economies like Japan, US, Finland and Sweden by a whopping 40%, while economic growth in the US could increase from 2.6% to 4.6% (a figure that has long been stuck around the same mediocre level for a long time now).
In an era during which economic and productivity growth has levelled off and in some places started stagnating, the news of AI’s considerable contribution to the ramping up of these figures may be a breath of fresh air. This revolutionary progress can allow advanced economies to gain much-needed momentum. (Below are the figures for productivity growth of some advanced economies.)
But is all this talk some kind short-sighted, narrowly focused exuberance? Could it be possible that its proponents are not thoroughly thinking about the implications of such unprecedented technology? There is a certain part of the intellectual community that thinks this to be the case. There is a very legitimate concern regarding the possible effects that AI technology could have on unemployment. The estimates range from 6-8% to 20% of the workforce becoming unemployable subsequent to the adoption of AI in a given economy.
There are those who regard the AI era to be very much akin to the Industrial Revolution, making references to its consequences for job creation. However, this comparison remains simplistic when stood up against the pervasiveness and advanced capabilities of AI technology. AI is not only able to automate low-level, low-skill work, but also poses a very significant threat to jobs in the nature of “information processing activities” like decision-making, reporting, communicating and coordinating, fact-gathering, and the supervision of similar activities by subordinates.
Just this year, for example. a Japanese insurance company replaced 34 of its employees with AI. What is under threat of replacement by automation wasn’t only blue collar positions but jobs like insurance, banking, law, translation and even farming.
Replacement or Augmentation?
An optimistic counter-argument against this could be the claim that the integration of AI into the economy would be in the form of capital-labour augmentation instead of absolute replacement. Though this view too has raised some controversy, particularly with the hypothesis that whatever people can do, AI can do just as well – if not better.
So where is the incentive to not replace humans with machines entirely? This view becomes even more relevant with the prospect of the AI singularity. It could eventually reach a point at which there would be no need for high-maintenance, petulant humans involved in the workplace.
Now with this possibility hanging over humanity’s heads, it becomes important to assess the social implications of integrating AI into the economy. It poses a very tangible threat of a total shakedown to the values and structures of many communities. It might very well force people to re-establish their relationship with the concept of work, which for many is self-defining.
Would people be content with that identity taken away from them and their most basic societal values altered by this new technology? It might be too imprudent to assume that the adoption of the AI technology will be a smooth and painless one, as it might even create 21st-century Luddites that violently oppose and reject it.
Another aspect that could become challenging is the political implications of adopting AI. For some, the fundamental problem for the future will not be the creation or maintenance of jobs, but the distribution of wealth created through AI’s replacement of human beings in the workplace. Will the universal basic income be the norm in the future?
How this case of income distribution will unfold is a matter of intense concern because it may very well be the differentiating factor between an extremely unfair and divided society versus a society that prospers tremendously with a fair distribution of wealth. But this type of a system would surely create more dependence on the government for maintaining income equality, whose promise of success looks sombre judging by past and current experiences.
With all these economic, societal and political implications in mind, perhaps it is time to slow down and for the first time look beyond the immediate economics benefits of what is being done with AI, for it may well be the most significant paradigm shift in the modern history of humanity.