Almost every article printed about robots and jobs starts and ends the same way. Here’s a recent example from the FT. The first hallmark of the genre, previewed in the headline ‘Poor education leaves emerging markets vulnerable to automation shock’, is a dire prediction of job losses – in this case in the developing world, where ‘the replacement of workers by machines threatens two-thirds of jobs’, according to a UN report. Then, as always – ‘As always, the only answer is education.’

Of course, prediction and ‘solution’ vary slightly. The looming job loss can be in particular sectors, countries or continents, and the answer can be training or other preparation, or, more frequently nowadays, some form of universal basic income. But both diagnosis and are cure are characterised by the same infuriating mixture of fatalism and complacency.

Past technological surges have always ended up creating more jobs than they destroyed, albeit in unpredictable areas, the argument runs; all we can do now is to tempt employers by giving them more skilled, willing and flexible workers. There are better and worse variations on this argument – for instance, this by Tim Harford is fine. But what most share is the unquestioned assumption that the only half of the equation that can be operated on or influenced is the offer – the workers. So the FT article above is more about Indian and African education than employment. As for the demand for workers – well, it’s just what demand will be.

But this is pushing on a piece of string. What if companies don’t want to employ people? After all, it’s not ‘the whirlwind of automation’ or even ‘machines’ that create or eradicate jobs. It is investment decisions made by human beings in company boardrooms. And those decisions do not take place in a vacuum.

A recent White House report on AI, automation and the economy underlines that ‘Technology is not destiny… The direction of innovation is not a random shock to the economy, but the product of decisions made by firms, governments, and individuals’. As Brian Arthur shows in his excellent book on the nature of technology, the latter is an integral part of the evolving economy, both shaping and being shaped by it. In some areas, such as medicine, the avenues science pursues are already economically determined – cures for first-world conditions are more lucrative than those for poorer countries. In others, management motivations decide how discoveries, once made, are diffused. Technologies such as the internet, voice recognition, touchscreen, and GPS – all developed in the public sector – could have been combined in countless different ways, or none at all: it took an inspired Steve Jobs to bundle them together into the familiar shape of the iPhone. The platform economy the iPhone then made possible is a further techno-economic evolutionary twist.

How the platform economy evolved, and why it is now the go-to model for every budding start-up, is in turn in part the result of developments in other socio-technological areas, in this case the company and management. In 2014 Martin Wolf wrote in the FT: ‘Almost nothing in economics is more important than thinking through how companies should be managed and for what ends’. He went on: ‘Unfortunately, we have made a mess of this. That mess has a name: it is “shareholder value maximisation”. Operating companies in line with this belief not only leads to misbehaviour but also militates against their true social aim, which is to generate greater prosperity’.

This is the first time in history that one of the great technological spurts has taken place when companies are being operated in line with this anti-social belief: that is, under a regime where one stakeholder is supposed to maximise its returns at the expense of the others, including society, and where the most widely taught and practiced version of strategy is largely about preventing other stakeholders from eating the shareholders’ lunch.

Now put today’s gig and task-based economy in perspective. It hasn’t suddenly popped up at random from the blue. It is just the latest step in a process of corporate dis-employment which began in the 1980s. Simply put, under shareholder value employees are costs to be minimised like any other. So responding to their new incentives, managers began to pass up employment-creating initiatives they would have undertaken in the past in favour of cost-minimising measures to benefit shareholders.

The downsizing and outsourcing trends initiated then have expanded steadily to the present day. Collateral damage was first lifetime employment, then defined-benefit pensions, then corporate responsibility for career. In recent years automation and AI have further eroded the full-time permanent employment bond, with a corresponding upturn in the growth of freelance, short-term and zero-hours contracting.

Enter in 2007 the iPhone.

There are many ways the smartphone could have been used for economic and social gain – including the enabling of a real sharing economy. But in the labour-historical context there is an inevitability about executives driven by shareholder value deploying it to dismantle the last element in the traditional employment regime: the job. With labour a commodity to be contracted as easily as any other, one of the traditional justifications for the traditionally shaped company collapsed, and with it the last link between corporate growth and employment. At best new-generation companies are job-neutral (Instagram: $19bn in value, 19 employees); at worst, like Uber, they are job killers.

The blunt truth is that companies today have no intention of employing people unless they absolutely have to (which explains why the Silicon Valley titans are hypocritically rallying to the cause of the universal basic income). Next down the road: driverless cars. In this situation, expecting better qualifications to improve employment chances is a bit like hoping that faster, stronger horses would stave off the advance of the combustion engine.

So do we sit passively while employment continues to wither until it becomes the preserve of a privileged few?

Or do we decide to act on what we can still influence, the demand side of labour?

Here are some suggestions.

The first step is for governments to reinstate employment as a central plank of economic policy, as it was up to the 1980s when it was hastily abandoned as politicians let themselves be persuaded that markets know best. Tax policies should be adjusted accordingly. Companies that fail to pay a living wage should not be considered for state contracts.

Employment policy should go alongside a root-and-branch rethink of ‘how companies should be managed and for what ends’, to requote Martin Wolf. Far surpassing the timid and irrelevant tweaks envisaged by Theresa May, the aim would be to prise companies from the grasp = of short-term shareholders (and executives) and restore them to their proper mission of generating greater prosperity for society.

Individuals also have their part to play. Strikingly, Gallup surveys show that globally what more people want than anything else – more than security, family and peace – is a full-time job with a pay cheque. OK: their responsibility then is to prepare themselves to engage with the employment they want as both workers and citizens – including in the responsible trade unions that governments (and companies) should foster as counterweight to the current corporate dominance.

Employment is one of the time-bombs left un-defused by the failure to reform business-as-usual after the financial crash of 2008. It would be better to act now than to wait for a chance detonation to set it off.