Don’t count on a robot invasion just yet

When the Personal Rapid Transport system, a self-driving electric car that scoots around the bowels of Masdar City, came to a halt and opened its doors, I disembarked. The doors closed and the car drove off.

Minutes earlier I had been celebrating this piece of technological wizardry, singing John Williams' Jurassic Park theme loudly to myself as the car navigated the garage beneath Masdar.

Now I stood there, bemused, having been disgorged into an empty concrete lot, with nary a disembarkation platform, nor a door designed for humans, anywhere in sight. Dark tyre tracks marked the routes that electric cars had taken: I followed them to the warren I presumed would be at their end. An automated voiceover had promised that the cars were equipped with the latest in auto-navigation technology. I hoped they would not hunt me for sport.

Robots – said the head of innovation at IBM, and an MIT neuroscientist at a talk at the World Economic Forum hosted in Abu Dhabi last month – are the future. In fact, they said, in that way that management consultants and people in the tech industry have of saying deliberately incorrect things in order to sound impressive – robots are the present.

I felt my smartphone as a sudden, oppressive weight in my trouser pocket. Was it listening? Did it know?

All a robot is, they made clear, is a bit of hardware and software that can respond to its environment. When my fridge identifies that the temperature has diverged from that ideally needed to preserve milk, and cools or heats itself, it has proven itself to be a robot. When it gets Wi-Fi connectivity, places an order on Amazon Delicatessen for a quart of double-cream, and intones, in its best impression of HAL from 2001: A Space Odyssey, "You have run out of milk, Adam" – then, the Internet of Robot-Things has arrived. (Said the product designer to the technician: "Why do you keep making the voices so sinister? I've warned you about this before, Dave.")

Yet, as an any journalist inundated with thousands of press releases and management consultancy reports about the Internet of Things can tell you, there are reasons to think that robots won’t really be around every corner anytime soon.

Firstly, we live in a world of considerable income and wealth inequality. I would love to see robotic technology rush to the aid of Africa’s underbanked, but the absence of expensive capital goods in many parts of the world is the nature of the problem. Even in the developed world, buying a creepy fridge smacks of status anxiety – the preserve of the jostling middle-classes – or the affluent boredom of middle-aged men with enough cash and spare time to become hooked on self-monitoring technologies.

Secondly, we live in a world of awful companies. Most people in most organizations are unfulfilled most of the time. Organizations and offices are imperfect, and bad at putting labour to good use. Robots have lots of clever applications – Watson, IBM’s supercomputer, can process more medical journals in a day than most humans in a lifetime, making it a very powerful diagnostic aid to physicians. But are the same organizations that put their workers to such dismal and inefficient use suddenly going to efficiently utilise all the possibilities of robotic technology?

It’s currently possible to inexpensively rent supercomputers for high-powered data analysis, for instance. Most companies generate huge amounts of data. But do most companies make the most out of this data, by storing it in a useful form, training managers on how to make decisions out of it, and by doing high-powered data analytics using the world’s cheaply-available processing power? Technology needs organizations to take advantage of it.

And those organizations are made out of individuals. Paul Romer, an economist, repeats the truism that productivity growth is what happens when resources are combined more efficiently – he calls this, in a folksy way, improving the ‘recipes’ for output. Individual and organizational output can both be made more efficient by adding robots to the recipes.

But adding robots without developing procedures and techniques to utilise those robots doesn’t do much to boost our welfare. The instruction manuals represent as much of a contribution to economic growth as the robots. Consider 3D printing. The technology allows us cheaply to build almost any object that we can think of – the next productivity-enhancing step is to programme the printers to produce useful components and objects. The next step is to design the recipes – which requires ingenuity and essentially human skill. Without the recipes, the robots don’t do very much.

Eventually, I found my way to the depot where the PRT cars were parked. I hammered on the glass doors that separated the PRT lot and the embarkation station. An amused security guard let me in. I told him that the PRT car needed to be reprogrammed. Its recipe needed work.

Follow The National's Business section on Twitter