And utterly consumed with sharp distress,
…why should we toil alone,
We only toil, who are the first of things,
…Why should we only toil, the roof and crown of things?
– The Lotos-Eaters, Alfred Tennyson
Replace the ‘we’ in Tennyson’s lines with ‘you’, and you’ll get the gist of the choric song that Big Tech has been reciting in the last decade. For, just like Homer’s land of the lotos-eaters, the world of automation and AI seems to hold the prospect of a life free from confusion, drudgery, and uncertainty aloft, but full of improvement, abundance, and dreamful ease.
To add to this illusion, a doyen of the Indian IT industry recently claimed that AI could drive productivity, solve unsolvable problems, and lead to economic growth. Much more revealing than this trope of a tech panacea, however, was his insistence that in those instances where greater precision could lead to better outcomes – like driverless cars – humans should allow tech to replace them.
Could his optimism for AI be as sound as his scorn for a sub-70-hour work week? It depends. The devil now hides in a double-edged definition. Always comfortable with ambiguity – the envy of aspiring CEOs – he nestles in the nebula of a ‘better outcome’.
In an imaginary market where one owner operates a fleet of 50 taxis with 25 drivers, self-driving cars will solve the labour deficit. Productivity will improve, an ‘unsolvable’ problem will be resolved, and the market will experience economic growth, even if returns accrue to none but the owner.
However, in a real market dominated by AI, exogenous resources could be neutralised to maximise efficiency. All 50 taxis may be made driverless, and to sustain profitability, market concentration may become necessary to sustain only a few – if not one – beneficiary owner(s).This is because ‘assistive’ tech is geared towards improving business outcomes, not human limitations. This outcome centres on supercharging the functions of a process – creation, design, development, production, usage – and if, consequently, it can make a human agent more effective, well and good. If not, and if human agency is undesirable or dispensable, just as well.But let’s not blame AI tools alone. It is often meek complaisance that drives dispensability more than the tender ministrations of a machine. When the first wave of artisans agreed to relinquish their way of life to support new capitalists at the start of the Industrial Revolution, they also tacitly consented to surrender individual creativity, critical reasoning, and judgment in the interest of the greater good.
Since then, every technological revolution has seen an ever-greater sacrifice of the individual’s mental contribution at the altars of mechanisation, standardisation, formalisation, and automation. This has taken its toll on everything but greater work hours, uniformity, output, and profit. Nowhere has it resulted in an impartial and equitable distribution of benefits, and nowhere has it universally encouraged the capacity to think, identify one’s talent, upskill, make decisions, and change course.
Now, the sponsors of AI encourage frequent engagement so that users can train it, correct it, improve it, augment it, and perfect it. Undoubtedly, this will reap benefits in the short term – flawlessly streamlined processes, better-informed decision-making, augmented customer-service resolution times, and optimised resource allocation for enhanced returns. Doubtless, some will continue to flourish.
But it’s hard to ignore the long-term consequences for many. If AI is to succeed, it must become more independent, pervasive, and profitable – read, sort, and interpret data faster; formulate a decision to act more precisely; show consistently profitable growth. Humans, zombified on a diet of convenience and instruction and unable to adapt fast enough, are bound to become more and more redundant.
In Homer’s The Odyssey, only Ulysses’ forcefulness freed those who had eaten the lotos-flower from their apathetic lethargy to sail home. The remaining island-dwellers, helpless in their bovine contentment, were never heard of again.
Of course, AI could be another software programme that needs human agency to work. If so, the doyen is prescient – the fad will fade, and humans will live to work another day.
