Post by account_disabled on Mar 8, 2024 22:45:11 GMT -5
There is no consensus on the definition of artificial intelligence (AI). It will not be the same for a technician, for a jurist or for a philosopher. Some general definitions are usually used, such as the one established by John MacCarthy in his article «What is Artificial Intelligence» or the one used in the proposal for a European Regulation on Artificial Intelligence (which in turn has evolved quite a bit). From the different approaches it seems that a common denominator can be drawn: the ability to learn, reason and perceive by reference to how human beings do. Within this very broad field, the purpose of this article is modest: to briefly propose some reflections on How the leap from discriminative AI to generative AI affects the world of law, understood one and the other in the following way: Within this very broad field, the purpose of this article is modest: to briefly propose some reflections on How the leap from discriminative AI to generative AI affects the world of law, understood one and the other as follows: Discriminative AI systems: they are those oriented to Insights of information received and the classification of that information into predefined categories. In other words: they “discriminate” data according to certain predefined criteria.
Generative AI systems: they are those aimed at the generation of new content (texts, designs, images, etc.) based on certain starting information. We will start from a parallel with the methods of human reasoning: deduction and induction. La deduction It goes from the general to the particular, starting from general premises to reach a specific conclusion. The induction, on the other hand, attempts to reach general conclusions from particular premises. From another point of view, Discriminative AI is based on a deductive process: from long bodies of data tries to reach a conclusion for a caso concrete. On the contrary, Generative AI focuses on the creation of new content, such as text, images, music and videos based on a specific request. That USA Phone Number is, these are systems that follow an inductive process. This purely human capacity, associated with a strong component of creativity, and that is why the glimpse on the horizon is so groundbreaking. Until recent advances in generative AI systems, most AI systems used in the legal field were based on deductive processes (discriminative AI). These systems have been of great help in tasks such as the management and filing of large volumes of documentation to carry out legal audits, review, classification and systematization of databases or assistance in drafting contracts through the automated identification of precedents of clauses.
Although these tasks have made the work of lawyers easier, we are currently at a turning point that can change the world of law. Generative AI systems, especially those based on large language models (LLM), are trained to perform text analysis and writing tasks. Transferring these characteristics to the world of law, it is evident that generative AI models capable of “understanding” potentially have a transformative effect on the way lawyers work and, in general, the world of law. As a result of the above, we currently find multiple generative AI solutions based on LLM capable of helping lawyers in their tasks, such as: Analysis of legal documents (sentences, contracts, regulations, etc.) to, for example, summarize relevant aspects of the document or identify legal risks. Assistance in negotiations that help lawyers prepare for a negotiation or a hearing, reproducing opposing roles and presenting arguments for each of the parties.
Generative AI systems: they are those aimed at the generation of new content (texts, designs, images, etc.) based on certain starting information. We will start from a parallel with the methods of human reasoning: deduction and induction. La deduction It goes from the general to the particular, starting from general premises to reach a specific conclusion. The induction, on the other hand, attempts to reach general conclusions from particular premises. From another point of view, Discriminative AI is based on a deductive process: from long bodies of data tries to reach a conclusion for a caso concrete. On the contrary, Generative AI focuses on the creation of new content, such as text, images, music and videos based on a specific request. That USA Phone Number is, these are systems that follow an inductive process. This purely human capacity, associated with a strong component of creativity, and that is why the glimpse on the horizon is so groundbreaking. Until recent advances in generative AI systems, most AI systems used in the legal field were based on deductive processes (discriminative AI). These systems have been of great help in tasks such as the management and filing of large volumes of documentation to carry out legal audits, review, classification and systematization of databases or assistance in drafting contracts through the automated identification of precedents of clauses.
Although these tasks have made the work of lawyers easier, we are currently at a turning point that can change the world of law. Generative AI systems, especially those based on large language models (LLM), are trained to perform text analysis and writing tasks. Transferring these characteristics to the world of law, it is evident that generative AI models capable of “understanding” potentially have a transformative effect on the way lawyers work and, in general, the world of law. As a result of the above, we currently find multiple generative AI solutions based on LLM capable of helping lawyers in their tasks, such as: Analysis of legal documents (sentences, contracts, regulations, etc.) to, for example, summarize relevant aspects of the document or identify legal risks. Assistance in negotiations that help lawyers prepare for a negotiation or a hearing, reproducing opposing roles and presenting arguments for each of the parties.