Who shall benefit from AI

A long time ago, (seems like the start of a fairy tale, it is a tale though not by a fairy!), there was the industrial revolution. It started in UK, spread to Europe and eventually to the US and it was chracterized by stable and efficient manufacturing. There was the discovery of steam and using it in production, machine tools that were developed, factories became more mechanised. As a result of all of this, output increased and thus population increased. Often compared to adoption of agriculture, it is widely regarded as the biggest event in human development history.

An economist, however, while delving into details pointed out a conundrum. While industrial growth happened, population growth happened, economic growth also happened, there was no commensurate increase in standard or living for workers for even increase in wages. In fact since the cost of food increased with wages remaining stagnant, inequalities increased too. And thus was born the concept of Engel's pause. Introduced by the economist, Robert Allen, who named the period after German philosopher Friedrich Engels.

If you want a good reading on this, Wiki does a fair explanation and economists of the time gave their own views on the matter. English scholar Thomas Robert Malthus proposed that technical progress would increase demand for labour, but this would be offset by an increase in population. German economist Karl Marx believed that technical progress, enhanced by a large amount of labour, would reduce demand for labour and prevent steady wage growth. (This and more on the Wiki page!)

This concept has come into discussion forums a little more nowadays since there are a lot of folks who do believe that AI could follow a similar path. Geoffrey Hinton first alluded to it in an interview/talk. And that started a whole lot of other conversations.

The crux of this discussion being that as the positive effects of AI become visible, the benefits would accrue to a tiny minority and the vast majority especially in the near term would be stagnating. In fact, anecdotal evidence suggests that in the Philippines and in the call centres in the Philippines, AI has shown that a great deal of productivity has been achieved but that has not changed the lives of the employees themselves, in fact they may be working more than before. Stanford actually released a paper on this. They called it Canaries in a coal mine. The phrase “canaries in a coal mine” is a metaphor for an early warning signal of danger. It comes from an old mining practice where coal miners once carried live canaries underground because the birds are very sensitive to poisonous gases like carbon monoxide and if the canary became ill or died, it signaled that the air was unsafe before the gases reached levels that could harm the miners and these miners could then move to safer places.

Specifcally for AI, this has the potential for a much bigger impact that what has been seen in the past for other technologies. And that is why we need deliberate action to ensure that AI does not widen the economic gap. Unfortunately, most regulations thus far have failed to take any action in this regard.