Robots acting as a judge, maybe; robots as a jury, probably not. Robots replacing lawyers, at least in some circumstances - a near certainty.
That's the view of University of Auckland Commercial Law Lecturer Benjamin Liu who has looked into the future of artificial intelligence (AI) and robotics and says the fast-rising technology could help determine the destiny of the legal industry.
"The rapid rise of artificial intelligence, which allows machines to learn and become an expert in any field, will pose a big challenge for the legal profession," he says.
Software already exists warning employees of potential legal issues as they work, effectively doing the job of an in-house lawyer; IBM's supercomputer Watson is predicted to pass the bar exams in 2016.
Fortune magazine this year ran a feature on white-collar professions - previously thought not under threat compared to blue-collar roles like car assembly - already being taken over by robots. Lawyers were among them.
The article quoted UK, Deloitte and the University of Oxford predictions that 10 million unskilled jobs could be taken over by robots globally and 47 per cent of US jobs could be automated and taken over by computers by 2033.
New Zealand is not immune - the Future [Inc] report series by Chartered Accountants Australia and New Zealand says white-collar work will be affected, with a predicted 885,000 New Zealand jobs (46 per cent) at risk from automation in the next two decades.
But lawyers? Surely the subtleties of legal argument and the ability to tap into the human condition will be beyond robots. Liu says that will still be true in many cases but AI can replace some of the data-based functions of lawyers.
"I have been practising for a number of years and I know you can't say a lawyer has some magical power," he says. "Lawyers make mistakes and often make decisions based on a few simple steps. While machines also make mistakes, their accuracy levels are much higher."
The question is not so much whether AI will apply in our legal system but what limits should be set: "In many fields, AI will certainly benefit humanity. Self-driving cars are much safer than cars driven by humans, and a robot surgeon can far outperform a human doctor," he says.
"There seems no harm allowing machines to give free legal advice to people who cannot afford a lawyer but it is altogether a different matter if we allow a machine to decide a case.
"Arguably, it is a basic human right to for a person to have their fate decided by fellow human beings, not a machine. We do this already, for instance, with the current jury system; we don't let the judge, a legal expert, make the decision over whether a person is guilty or not in a jury trial, we let common people do that.
"In some areas it could be logical to leave decisions to machines or robots. For instance, the transactional work I used to do - like conducting due diligence in takeover deals or drafting fairly plain vanilla financial documents. That possibly doesn't need a human element but in some other -practice areas, such as family law or matrimonial property, the human element is important."
The Fortune article says new software can already undertake discovery work (the act of sifting through thousands of documents relevant to a court action) and quantitative prediction - the art of researching legal arguments, precedents and even the foibles of a judge.
The magazine also named sports and financial journalists, online marketers, financial analysts and advisors, and surgeons, anaesthetists and diagnosticians as being under threat from computers. Johnson & Johnson's Sedasys system, already FDA approved, can apply low-level anaesthesia in small procedures like colonoscopies at the fraction of the cost of an anaesthetist.
This year, a group of students at the University of Toronto designed a computer programme called Ross, which took full advantage of Watson's super learning power. Ross can answer specific legal questions, in contrast to other legal search programmes that would simply return a list of documents.
"What Watson has achieved is nothing short of amazing," Liu says. "But it still requires human intervention. A sufficiently large dataset has to be labelled to train the computer, which is why this approach is called supervised learning.
"It is becoming more powerful and sophisticated and it might not be a stretch to imagine a computer programme being able to answer almost every question of law after absorbing massive amounts of data - like all the statutes, judgments and journal articles ever written on common law."
Liu says other illustrations lawyers are not the only profession that should be concerned about the effects of automation are in the report Disruptive Technologies: Risks, Opportunities - can New Zealand make the most of them?
"It is a bit scary," he says. "If you don't keep up with technology, you are going to fall behind, very soon."