Have you read a financial report by the Associated Press recently? Chances are that a non-human brain conceived and wrote it. AP — one of the biggest news agencies in the world — publishes more than 3,000 articles every quarter using an content-automation programme, Wordsmith. This ‘robot’ runs on technology from US-based big data analytics and language generation company Automated Insights.
Employment is the first casualty in the automation revolution. Many experts in the field of communication and media have sounded alarm already, about machine writers replacing human journalists. Authors Thomas Davenport and Julia Kirby disagree. In Only Humans Need Apply the American writers argue that artificial intelligence (AI) and automation that it powers can in fact help humans augment work and enhance productivity. The key is to embrace change (AI) the right way, and not to panic.
But panic is the default mood for most while discussing how AI is going to change the world. There are doomsday forecasts on advances in AI. Physicist Stephen Hawking, for one, last year said efforts to build thinking machines could wipe out humanity. The subject has inspired an assembly line of authors churning out literature on how automation and AI could make or break our future. Rise of the Machines: A Cybernetic History by Thomas Rid and Rise of the Robots: Technology and the Threat of Mass Unemployment by Martin Ford are two good examples.
Davenport and Kirby understand the risks arising out of automation, which has just entered its third era. This period brings both “promise and peril”. The authors cite an Oxford University study that says automation can replace nearly half the American jobs. That’s a big number. As of today, there are more than 125 million full-time workers in the US and if automation can impact half of this crowd (that’s more than 60 million), that’s serious business.Augmenting life
Davenport and Kirby like to look at the brighter side. The good news is, they write, new cognitive technologies will help solve many important societal and business problems. To give an example, now your local doctor can easily access global expertise online and find healthcare solutions at the speed of a click. Big Data solutions make correlations and decision making easier and coupled with AI they can revolutionise healthcare, finance, mass media and even art. But the obvious fallout is job losses. And this time, those losing their jobs will not be ticket collectors or even farmers, but blue-collared knowledge producers, leaving a huge army of “citizens of zero economic value.”
Can we avoid this? Well, that’s a big ask. But there are ways we can circumvent this or put the changes to good use, . Davenport and Kirby finds an interesting pattern in the way humans will deal with their skills in this era: we will step up, step aside, step in, step narrowly and step forward. Step up is where people use computers or artificial intelligence to step up their skills, be it decision-making or knowledge production. Next, they step aside by opting for jobs machines cannot do. By stepping in, they will act as links between the man and the machine. Stepping narrowly is where some people will focus their skills so narrow and sharp that they attain great expertise in some super special areas so that automating them won’t make economic sense at all.
Finally, stepping forward is about “creating new cognitive technology solutions for the rest of the world to use”. Davenport and Kirby say this category of work is not a big slice of the job market now, but soon it will. This mainly includes software vendors and companies developing their own systems. This demand will be huge and almost all software company will want to hire people who can assist them in stepping forward. And “no one will ever go broke automating the intelligence” of the knowledge worker.
That’s hope. But the authors do not bother to discuss the viability and practicality of these enhancements. For one, they don’t seem to ignore the fact that today’s workers function in a system that does not care enough about enhancing skill. Few corporations would want to do that at the cost of their balance sheets. Most skill enhancements happen on the margins. So in all likelihood, they would see automation as a way to pep up profits that creating a culture of improved workmanship.Law needs intelligence
To be fair, Davenport and Kirby ask businesses to augment human labour with technology, but they forget that only the very highly skilled work force would become beneficiaries of such a process. The rest will be forced to work under situations that favour only the companies — such as long working hours, reduced or no social security benefits, and unexpected termination. But this can be controlled by policy changes. States can introduce legislations that can encourage or if need be coerce corporations to augment, not necessarily automate, work and create a level-playing field for all kinds of workers.
But most lawmaking bodies are incapable of tracking, analysing and acting upon such changes with the fastidiousness and alacrity they demand. They always function slow and hence miss the bus on most occasions.
Take India as an example. The country is witnessing a boom in e-commerce thanks to advances in technology. Still, we have very few regulations and laws that understand the impact of such technologies on society. Here Davenport and Kirby make sense; they highlight the need for AI or data analytics to enter the myriad avenues of public policy making. Administrations armed with tools of AI and meticulous, fast and clear analysis of societal trends can foresee future and formulate rules that will help societies augment work and welfare. In such an environment, there will be little to worry about, say, robot journalists entering the newsroom.