For the organization of this event, we are actively using artificial intelligence. This is not a superficial addition, but part of a genuinely innovative and creative approach to philosophical work, academic communication, and event preparation.

My PhD advisor in mathematical logic, Daniel Andler, founder of the Department of Cognitive Sciences at the École Normale Supérieure in Paris, is an outstanding specialist in this field. I am also in contact with Carolyn Talcott, wife of John McCarthy, one of the major historical figures in artificial intelligence.

 We invited Carolyn Talcott to give the lecture "Intelligence and Logic" at UNILOG 2025 in Cusco Peru, December 4-14, 2025 (8th World Congress School on Universal Logic), at the Workshop on Artificial Intelligence organized by Maíra de Cinque and Flawé de Lara Bone.

Long life to AI! 

 JYB

AI is used here in several ways

For the evaluation of  abstracts :

• We designed a referee form which was subsequently improved with the help of AI

• For each abstract, we are producing three reports using ChatGPT (started Nov 2022),  Claude (started March 2023),  Deepseek (started July 2023)

How to use these reports? 

The reports can be useful at different stages of preparation:

• To produce the final version of the abstract, using the template "What is Cat?", whose initial content was generated by ChatGPT and then revised by us;

• To prepare the lecture;

• To write the full paper, that may be submitted after the congress. 

This means that the three reports should be considered in different ways depending on the stage of the work. Some remarks may be useful for revising the abstract, while others may be more relevant for preparing the lecture or developing the final paper. Not every comment applies equally at every stage.

These reports have both strengths and limitations. Authors are, of course, not required to follow all the recommendations, especially in matters such as suggested references or stylistic reformulations. The final intellectual responsibility remains with the authors. 

Another important point is that AI-generated reports do not always fully respect the spirit of Socrates and Plato, and may sometimes include comments that are debatable or out of place. Consider, for example, the following two remarks:

•  “Make a claim. Even if your final answer is there is no stable definition (a deflationary conclusion), that is still a claim—and a defensible one. Right now, the reader cannot tell what you think.”

• “A conference abstract is not a research proposal or a syllabus. It must present results, not just methods. Right now, the reader has no idea what you think social reality is. Take your own methods seriously: apply one of them to reach a conclusion, state that conclusion clearly, and defend it (even briefly). ”

In the context of this congress, such a criticisms are misplaced. The guiding question “What is X?” does not necessarily require that a fully determinate or final answer be already established at the abstract stage. On the contrary, a rigorous and well-structured exploration of the question itself—clarifying its stakes, examining competing approaches, and outlining a methodological path—may already constitute a valuable philosophical contribution.

Indeed, in a broadly Socratic spirit, the inquiry of what X is may remain open-ended, or even culminate in a refined form of aporia, without thereby lacking philosophical significance. What matters is not necessarily the premature presentation of a definitive conclusion, but rather the quality, depth, coherence, and fruitfulness of the investigation.

In this respect, AI can be useful, stimulating, and sometimes surprisingly insightful—but it still has to become more philosophical!