27 November 2014 | C-3PO, Robot Mediator | №11

What if, the next time you walk into a mediation room, there wasn’t a person waiting for you at the head of the table? Instead, a beacon of gold, C-3PO, meticulous humanoid protocol droid of impeccable etiquette, sits in the mediator’s chair.
What if your next mediator was a robot?
Albeit prone to a slight bit of futuristic romanticism, the idea of a robot taking a central role in a legal dispute is not new. Even our own faculty at the University of Ottawa is writing about the possibility of robot judges. As an example of how real such artificial intelligence applications to the legal system are, look no further than Tom Gordon’s Carneades software, which takes human claims and determines whether possible claim supporting statements are justified based on legal principles; the software breaks down each statement into a computerized stream and that is then compared with legal principles rating the argument to score the claim. Currently, computer-based programs support a rapidly increasing share of the legal practice’s workload. Think something as simple as WestLaw – that used to be a clerk, living in the library, pouring over tomes for days on end. And the day is not far off when artificial intelligence will have the capability to assume many of the responsibilities performed by mediators and execute complex, interactive, interpersonal tasks.
But, can robots participate productively in dispute resolution process like mediation? Mediation cannot be subsumed into an algorithm. There is no substantive body of knowledge or established precedents. There is no clear consensus on the objectives or goals of mediation. In mediation, the parties take the stage: solutions to disputes are grounded in norms created by the parties, not legal precedents. Outcomes are assessed against the interests and criteria defined by the parties themselves during mediation. Ultimately, mediation is about the process. Can a robot create the requisite level of trust to facilitate understanding and personal disclosure often necessary to resolve a dispute? Is C-3PO out of a job before we even get started?
These questions imply three considerations. Episode I: what does it take to be a mediator? Episode II: what are the actual capabilities of robots out there that could be suitable for mediation? And episode III: can one and two be reconciled? I think they can. And if I may be more audacious, I think C-3PO sitting in that mediator’s chair might even be inevitable.
What does it take to be a mediator?
A mediator is an impartial and neutral third party who has no authoritative decision-making power. The role of the mediator is to assist disputing parties in voluntarily reaching their own mutually acceptable settlement of the issues in the dispute. Mediation serves various purposes, including providing the opportunity for parties to define and clarify issues, identify interests, understand different perspectives, and explore and assess possible solutions.
There is currently no governing regulatory body for mediators -or even a formal certification process for mediators- in Ontario (or, more generally, anywhere). That is to say, individuals can practice mediation without any credentials. The door is open for a robot mediator to take the chair without any pre-screening; C-3PO would not be prima facie barred from taking the role of mediator.
Despite this vacuum, we have to remember that a mediator is chosen by the parties, and the parties do have certain expectations. While we’re not going to be able to glean the consensus on what abilities are required of mediators from regulations and certification requirements, what we do have is a myriad of non-binding ‘Standards of Conduct’. The American Bar Association’s Model Standard of Conduct for Mediators is one of the earliest and most widely accepted Code, dating back to 1994 and serving as the template for the ADR Institute of Canada’s own Model Code of Conduct for Mediators . The goal of the Code is to “guide the conduct of mediators; to inform the mediating parties; and to promote public confidence in mediation” . To achieve this, mediators have to be ethical and competent. They have to be neutral, avoid conflicts of interest and maintain confidentiality. Though the Code does not provide specifics, it does overall call for mediators that are emotionally intelligent and sufficiently well-versed in the substance of the dispute to be able to provide guidance where needed by the parties.
Generally, mediators are facilitative or evaluative. Facilitative mediators are process-oriented and client-centered; they see their role as enabling effective communication between the parties and helping them identify their interests collaboratively. On the other hand, evaluative mediators are more active: they engage with the parties and even provide their own opinions of possible solutions. Depending on the needs of the parties, a mediator will have to determine which approach is best suited to a particular mediation. There is great flexibility here, and mediators will often fluctuate between these two styles during the course of a mediation.
Ergo, it is possible to identify the capabilities considered essential for effective mediation. Parties choose mediation because it is a private and confidential process; more personal than a more formal method of dispute resolution. Mediators need to be capable of engaging, collaborating and building rapport with human parties taking into account variables such as the personalities of the parties and their emotional circumstances.
Searching for our C-3PO in the land of healthcare robotics
So, we’re looking for an empathetic, personalized, competent robot. We’re looking for artificial intelligence that both behaves intelligently and is intelligent. Turns out, these robots are already out there, serving in hospitals near you.
For empathetic, we have Keepon, a therapeutic robot used with autistic children. Keepon is programmed to interact with children in human way, “by exchanging a variety of social cues, such as gaze direction, facial expression, and vocalization” . Not only do many of the children given Keepon as a companion interact with her to a greater degree than they do with humans, they have also relied on Keepon to facilitate interactions with third parties. This is particularly important: by providing “touch, eye contact, and joint attention”, Keepon and these children are able to establish what has been defined as “an empathetic understanding of each other’s mental states”.
For personalized robot, look no further than the ‘Doctor Kiosk’. This robot -complete with blood-pressure cuff, pulse oximeter, and blood-testing device- visits patients to gather personal and intimate information in an effort to ease the administrative strain on physicians. The Doctor Kiosk has a caring and empathetic bedside manner, augmented with relational dialogue and behaviour. As the patient’s conversation with the Doctor Kiosk progresses, the Doctor Kiosk asks additional questions based on the individual’s personal experiences and history, referring to previous interactions. In its first trials, seventy-three percent of hospital patients said they preferred dealing with the Doctor Kiosk in the initial stages, as they felt it was more attentive to their idiosyncrasies.
And finally, for competent, we have the famous IBM’s Watson. In the words of IBM: “Watson is a knowledge corpus”. Watson is a cognitive system, capable of natural language processing, hypothesis generation and evaluation. Watson has an almost infinite memory, instant recall, and most amazing, Watson learns. After retiring from Jeopardy , Watson has turned his computational powers to cancer research. Watson’s strength is its capacity to analyze huge volumes of data and reduce them down to critical decision points corresponding to treatment suggestions. In other words, to help physicians make “better decisions” . While there is no way that physicians can keep current on all the latest medical breakthroughs, Watson can.
So, the robots are out there and the bar is set. Not only will our forthcoming robot companions be high functioning, equipped with facial recognition and artificial intelligence, they will have be able to understand and reciprocate human-like social cues and communication modalities.
Introducing C-3PO, the robot mediator
By this point, you might be wondering why I brought C-3PO into this. At first glance, C-3PO is not exactly ‘mediator’ material; he doesn’t get a lot of respect from the other players in the Star Wars universe. And yet, despite this, the cast continually subconsciously defer to him to solve their problems; they trust him, a robot, to sort through their human issues rationally.
We’ve seen above that albeit there being no required capabilities, mediators have to be emotionally intelligent and competent in the subject matter of the dispute. For a successful mediation, a mediator will have to engage, collaborate and develop rapport with the parties. A review of the capabilities of artificial intelligence has shown us that empathetic, personal and competent robots are being employed in healthcare, with promising results. Insofar as we have identified the essential capabilities of a successful mediator, Keeton, Doctor Kiosk and Watson do seem to possess those characteristics. In C-3PO, we can see the implementation of these capabilities into one robot. After all, C-3PO is a humanoid protocol robot; his programming is intended to facilitate human interaction.
Robot mediators are possible, and C-3PO is our prototype in this thought exercise.
Obviously, a robot mediator will never the same as a human mediator. But it is precisely because of this that the possibility of robot mediators is an interesting one: robots, being robots, have their own advantages to bring to the table.
First, computational power. Enter C-3PO. C-3PO boasts to be fluent in over six million forms of communication. C-3PO, like IBM’s Watson, can process volumes of data beyond normal human capabilities and instantly recall them when needed. A computer is less likely than a human to overlook byzantine exceptions to exceptions that might arise when applying foreign or domestic substantive or procedural law. Mediation is peculiar in that it has international scope and engages a variety of subjects. For example, in intellectual property mediation, the dispute might go to the technical details of an isolated combinant nucleic acid claim in a gene patent. A robot mediator would be able to engage and collaborate at a deeper technical level than a mediator who is not intimately familiar with the intricacies of the particular dispute.
Another advantage is the intrinsic neutrality of robots. In a study to assess the openness of individuals to engaging with robots, a team of German scientists dropped a robot, ACE, into the streets of Munich. With no global positioning system or map to guide it, ACE had to engage with forty strangers for directions home. The study showed that “the large number of people interacting [arose] from the fact that many of the interactions were started by curious passers-by”. Transposed in the mediation context, parties may find it easier to make certain types of disclosures to an artificial intelligence device. It may be less daunting to confess shameful or intimate information to a nonjudgmental -yet, empathetic- robot. Throughout the original trilogy, Han Solo finds himself confessing more to C-3PO about his emotional and existential conflict than he does even to himself.
Finally, a robot can be many personalities in one. Depending on factors such as the nature of the dispute, the parties, the relationship between the parties, and the type of third-party assistance required, a robot can select a different mediation approach. To use our intellectual property example: a robot mediator could be programmed with the technical-know-how of a very idiosyncratic biotechnology dispute, allowing the robot mediator to facilitate that mediation more effectively.
In sum, the mediator’s chair might be just the fit for C-3PO.
Are we ready for a robot mediator?
Ultimately, mediation is based on the principle of self-determination. Typically, the parties chose to undertake mediation, the parties choose their mediator, the parties give the binding power to any outcome through their voluntary consent. So, whether or not a robot mediator is a feasible prospect, the actual possibility of adopting robot mediators depends on whether or not parties would want their dispute mediated by a robot.
The acceptance of artificial intelligence devices is really a matter of familiarity and trust. Robots already perform functions that are extremely personal and intimate. They provide support and assistance with our day-to-day endeavours, and our occupational and medical concerns. Technology is ubiquitous to our existence. Technology has changed the way we interact. Our thought processes are indivisible from our virtual neurons, our computers, our internet caches.
To be parted from our smartphones, our tablets or our computers paralyzes us. We rely on our phones and our computers for our most private conversations. We find answers to our pressing and intimate questions on Google. We are increasingly delegating our healthcare to robots and, as our comfort and dependence on technology grows, we will eventually demand the same use of technology in mediating our disputes. And what of future generations, who will have grown and lived married to artificial intelligent devices?
Of course, there are many more things to consider if we are really going to open the gate to robot mediators. In this paper, I have deliberately not addressed some key concerns to accepting robot mediators. Chief among them: confidentially and security concerns. If technology is to play a significant role in future mediation, protocols must be established regarding retention, transfer and destruction of data. Self-learning robot mediators that elicit and collect intimate and personal information inevitably convert that information into a digital format and store it for future learning.
Let’s go back to that moment that we walk into the mediation room and see C-3PO. Now, are you really surprised?

 keywords: mediation, robot, c-3po, keepon, doctor kiosk, ibm’s watson, technology