Storing unnecessary items

Sat 27 June 2026, Philippe Ropenga Artificial intelligence Graphs

You know that I complain about legal and software publishing companies that try to provide more information about a topic instead of building databases that would help legal practitioners to focus on what matters to them and their clients¹. Now that prompting automatically generates content, people may pay less attention to knowledge structure. Databases are useful not only because they contain data but also because a person who wants to have a look at common issues within a specific area can quickly grasp the whole of it and decide what to explore next before framing the issues that he has to deal with. It may be tempting to prompt a practical question and to wait for an answer without bothering to get an overview of the main issues that are at stake. A database could for instance be used to determine the amount of money that one could get to compensate a loss following a damage. This however requires the practitioner to know the legal field that he examines and in what direction to search for an answer. Now that one can prompt in natural language, there is no need to formulate a question, one just awaits an output to come out of the blue. The issue is that nothing comes out of the blue without knowledge nor experience, especially in the computer world.²

I am not a computer scientist. Many people who had not been educated in computer science or data science. Ten years ago, legal publishers promoted expensive statistical tools designed to calculate the compensation that one could reasonably ask for. If one knows that law is about making syllogisms or analogies³, one realises that these tools do not rely on syllogisms nor analogies and therefore are not legal tools. French insurance companies that offer policies that cover litigation expenses exclude matters that cannot hardly reduced to an indemnity claim such as succession matters. Reduction is sometimes useful because it helps to notice interconnections between elements.[^4] Nevertheless, it is often counterproductive in succession matters that trigger vivid feelings because people cannot connect their feelings to their needs while this is useful to solve disputes.⁴ One could argue that this issue does not exist in practice because insurance policies exclude these matters. This is true and makes the limit of expensive statistical tools even more visible. Using expensive tools enables the practitioner to automate tasks that do not lead to a solution. Such software tools make legal practice less dependant on legal knowledge and law seems easier to practice by people who have not been educated in law.

Graphs and artificial intelligence can be used to make the most out of one’s legal education and experience. Graphs make relationships between nodes visible. One can use a graph to trace the train of thought followed by a person and notice concepts that are related to this person’s background and that only are loosely related to law unlike common legal concepts that tend to attract

the attention of the practitioner who may not pay enough attention to words that keep coming back during a conversation. It is worth trying to solve a problem by trying to look at it from a different perspective. A philosopher may for instance decide to solve a syllogism by looking at it from a geometrical perspective although geometry takes space.⁵ A legal practitioner who is not keen on using geometry knows that moving towards a solution involves linking concepts that are more or less distant from one another.⁶ This issue is thus also related to space even if the legal practitioner sees it as a matter of conceptual nuance. A computer scientist knows that searching a graph requires memory⁷ and this is obviously related to space. Hence, space is a relevant topic in these three examples even if it does not appear on a legal graph or appears differently in geometry than in computer science. In other words, one has to know various things about an issue to understand it even if one relies on narrower knowledge to solve it than to understand it.

In brief, it is good practice to store unnecessary items that may be used to find better solutions.

Go Top