21 April 2017
I want my robot lawyer
Artificial intelligence and employment: what’s the future?
Affecting not only lower skilled work, the impact of artificial intelligence will be disruptive also for many intellectual professions. Let’s see which jobs and why, perhaps, this won’t be such a bad thing.
The driverless car has become an obsession of our times. If the car as we know it has been the mass production good par excellence, exerting such an impact on the habits of modern mass society as to become one of the symbols of the 21st century, the driverless car is emerging as a strong candidate for the same role in the era of artificial intelligence (AI).
Many are convinced, and they are probably right, that the self-driving car will make relationships and transport easier, more fluid and direct, at least in the planet’s big cities. The impact on employment, however, threatens to be seismic: it is believed that this technology will destroy thousands and thousands of jobs. The algorithm with which cars, lorries or busses will corner, brake and stop, will, in many cases, replace the brains, arms and legs of real drivers.
Driverless vehicles are, however, the most manifest symbol of the “threat” to human work posed by AI, which is also a challenge for other human professions that require reasoning, decision-making and brains.
In addition to transport, there are numerous other sectors that look set to encounter employment crises. The management consulting firm McKinsey has drawn up an interactive map, based on the US employment market, of AI’s global scope, highlighting the potential level of automation for each profession. The higher the score, the less chance that human’s jobs will survive the impact of the algorithms.
Packaging, meat cutting, designing medical prosthetics, flight supervisors, moveable bridge operators: as far as these jobs are concerned, human work is, in theory, destined to disappear completely.
Even more intellectually sophisticated professions are at risk. For some time there has been talk in the press, including articles in the Financial Times and the New York Times, about what lies in store for lawyers. In difference to the professions mentioned above, lawyers will not be swept away completely, but some of their functions will be replaced by operations governed by artificial intelligence in the not too distant future.
In some cases this future has already arrived. Three years ago Joshua Broder, a student at Stanford University aged in his twenties, created a programme that enables users to appeal against traffic fines without having to seek assistance from a law firm. Do Not Pay has its own website and is described as “the world’s first robot lawyer”. Here’s how it works: there is an initial phase in which the system tries to understand whether the fine can be appealed and a second in which the system coordinates the steps that each individual citizen must follow in order to submit and manage his/her appeal. In June 2016, when The Guardian published an article about the service, it had already managed 250 thousand appeals in 21 months, winning 160 thousand of them.
Even journalism can take advantage of AI. The press agency Reuters has launched a collaboration with Automated Insights, to which it has delegated the task of writing some dispatches, especially those relating to minor sporting events. It would appear that the accuracy of the reports is more than satisfactory. The benefits for the agency are clear: it is no longer necessary to send journalists on location to report on the news.
These two examples demonstrate that the development of artificial intelligence is forging ahead at breakneck speed and the research appears to confirm this. One study by McKinsey says that by 2055 half of the economic activities performed by humans will disappear; from sectors with a very high potential for automation such as food preparation and serving (66%), transport (60%), retail (54%), agriculture and fishing (50%) and other sectors where machines will only partially replace human activity, such as finance and insurance (44%), real estate and construction (44%), the entertainment industry (42%), information (41%), welfare and health (38%).
As far as individual countries are concerned, China and Russia are the nations with the greatest potential for automation (41%), followed by Brazil and India (39%), Japan and Italy (38%), the USA and Canada (37%) and then Germany and France (36%).
Also many of the giants of the technology sector confirm that the coming change is just around the corner. At Baidu, the leading Chinese search engine, they are convinced that the age of mobile Internet has exhausted its cycle and the time has come for artificial intelligence. The company’s development department now boasts a staff of 1300. Their director is Andrew Ng, another alumnus of Stanford University. Ng’s firm conviction concerning the crucial issue of employment is that many jobs are sure to be lost but other new ones will be created; this he explained to the Wall Street Journal in a recent interview.
McKinsey have also drawn an interesting parallel between the role that artificial intelligence can play and the effect that mechanization had on farming at the beginning of the twentieth century. “In developed countries the technology pushed the workforce out of agriculture but this did not coincide with mass unemployment because the process was accompanied by the creation of new types of work”.
Even the great scientist Stephen Hawking, who is usually highly critical to the point of being almost fatalistic when it comes to the future scenarios opened up by the development of artificial intelligence, recently underlined the positive aspects that such technology can offer. Speaking at Cambridge University he said, “We cannot imagine what we could achieve when our brains will be supplemented by artificial intelligence. Perhaps, with the tools that will become available to us we will be able to reverse the damage caused to the environment by industrialization. And certainly it will be possible to work to finally end poverty and disease”.
One fine example comes from the field of medicine. In August 2016 in just ten minutes a super computer diagnosed the form of leukemia that a Japanese sixty-year-old was suffering from. This same diagnosis would have taken human doctors weeks to ascertain. “I don’t want to exaggerate by saying that this diagnosis saved the woman’s life, but it provided data rapidly and efficiently,” explained Aribobu Tojo, a doctor at the hospital in Tokyo that used the super computer. The full story can be read in full on venturebeat.com.
“The success in the creation of artificial intelligence can be the biggest event in the history of our civilization,” explained Hawking in Cambridge. The incredible potential of AI in the field of science supports this interpretation. At the same time, however, Hawking offered a warning, specifying that this enormous resource could also be “the worst thing” created by man, if managed badly. As he has urged on other occasions Hawking called for the establishment of a university centre dedicated exclusively to the study of artificial intelligence, precisely in order to avoid it being used incorrectly and to prevent it from scoring a giant own goal.