Counselor, have you passed the Turing test?
The continuous advances being made in cognitive computing leave us wondering whether a time will eventually come when we are no longer able to tell if a particular legal argument has been developed by a human being or by a machine.
One example is the case of the conversational bot Eugene Goostman which, in 2014, managed to fool 33% of judges into believing that they were speaking to a 13-year old Ukrainian boy, thereby passing the Turing test and proving right Alan Turing’s prediction that machines would one day be able to convince over 30% of judges that they were human. And then there is the astonishing victory of the IBM computer Deep Blue, in 1997, over chess Grandmaster Garry Kasparov.
Although new technologies offer obvious benefits in the area of legal practice—it is clear that data bases and data processing systems have taken over completely from the classic legal manuals and law reports which were once consulted in libraries, since they offer immediate access to the information – in the face of achievements such as those referred to above, we cannot help but wonder whether the day will come when the reasoning will be done by computer programs, and lawyers, instead of applying their knowledge to the particular case in hand, will merely execute the legal decisions reached by these cognitive computing systems.
Bearing in mind that predictive coding now outperforms human capacity by far, what lawyer would be so bold as to ignore the conclusion reached by a machine?
Unlike the human brain, however, an algorithm is predictive rather than intuitive, meaning that there are aspects in which it will be difficult to replace it. The capacity of humans to interact socially, to perform creative tasks, or to adapt to new situations suggests that the work of a lawyer cannot be fully automated, since only a real-life legal practitioner can defend a position in a court case, where anything can happen and where the human factor and contact with reality are essential. What is more, if the profession of the lawyer were capable of being automated, then that of the judge would be too.
But would we really be willing to leave justice in the hands of machines? How would a computer program have responded to the dispute brought before King Solomon, between two women who both claimed to be the mother of a child.
As long as there continue to be tasks which only the human brain is capable of undertaking, it appears that there is still hope for lawyers. If the world depicted in Ridley Scott’s Blade Runner becomes a reality, we may find ourselves using the Turing test to identify replicant lawyers…. but we haven’t got to that stage yet.
Garrigues Labor and Employment Department