������Ƶ

Written by Ana Stanič, founder of

The whole discourse of AI replacing lawyers and judges is based on a mechanistic understanding of the world.  Already by the beginning of 20^th century physics has revealed to us that we do not live in a Newtonian world. Quantum physics, chaos theory and systems thinking tell us that the world is not one large probabilistic calculation, that may not even be cause and effect in a linear sense and that everything is relational.  And yet the hype about AI dominates the airways and there is talk about AI replacing lawyers, arbitrators and judges. 

AI output is based on complex probabilistic calculations rather than intelligible ‘reasoning’ which in the case of law is above all a matter of judgement, assessment and interpretation. A legal analysis, an award and judgement are made up of a multitude of decision-making factors which cannot and should not be formalised a priori, and are typically based on judgement as to: What the relevant facts are;  Which facts have been proven; What rules apply to the facts; What is the meaning of the rules with regard to the case to be decided; Which source of law should prevail between a range of conflicting sources. Much also turns on the credibility of the witness and experts when giving evidence which is assessed not just by what is being said, but by the tone and the whole demeanour of the person.  

The use of regression models in AI smooths out the small, unexpected, unexplained but meaningful things which are key to ensuring fairness and delivering justice.   

That reasoning involves two processes, one linear and the other holistic,  is explained well by Iain McGilchrist, a psychiatrist, neuroscience researcher, philosopher and literary scholar.  In his book The Matter of Things McGilchrist explains the difference between the left and right hemispheres of the brain and argues that the manner in which they operate is substantially different. It is not that the hemispheres perform different functions, but that they perform these functions in a different way. Drawing on extensive neuroscientific research from the last one-hundred years, McGilchrist argues that each hemisphere offers a unique kind of attention to the world, an attention which brings a certain version of the world into being. Simplifying things for the purposes of this blog the left hemisphere is designed the help us apprehend and manipulate the world to serve our purpose by simplifying it. Whereas the right hemisphere is designed to help us comprehend the world and see it all for what it is. 

According to McGilchrist, we have become entranced by the version of the world brought into being by the left hemisphere and forgotten the insights produced by the right. AI simulates the functioning of the left hemisphere of the brain only- the one that recognises simple patters and reduces things to its parts without being able to see the whole at the same time nor the relational nature between the parts and the whole, nor of the whole with the broader environment.