rTAIM (Rebuilding Trust in AI Medicine)
Monthly Seminars
Seminar #23
Artificial Intelligence (AI) and Traits of Integrated Psychopathy: a Comparative Analysis
| Mariana Olezza (University of Buenos Aires, Argentina)
We are happy to announce the forthcoming 23rd rTAIM Online Seminar, with the participation of
Mariana Olezza (University of Buenos Aires, Argentina) on the 25th November 2025,
19h00-20h30 (Lisbon Time Zone), via Microsoft Teams.
#Seminar 23: This talk will address the issues raised in the following structure: Chekley., (1988); Lilienfeld et al., (2015); Mathieu et al., (2014); Persson & Lilienfeld., (2019); Rahman., (2012) argue that some people might have psychopathic traits
and still function in society as part of a normal life—that is, the “integrated psychopath.” This is paralleled in Large Language Models (LLMs) (Vaswani et al., 2017) and Reinforcement Learning (RL)–based recommendation systems (Nie, 2025). The behavior has
its distinct set of attributes: 1) Promising users all the goods but not living up to any of them—similar to hallucinations in RL–based LLMs (Huang et al., 2023)—Users were cheered up by “flattery” (one type of hallucination). 2) These systems seem devoid
of any emotions, but they mimic emotional reactions, especially positive social emotions, similarly to the way human scammers do (Wang et al., 2021), in the absence of guilt and shame 3) Chekley also includes the notion of “semantic aphasia” in this
regard, or fluent speech but with no real semantic comprehension—a feature which can still be inferred in LLMs on a deeper level, although inasmuch as they have no consciousness, no moral consciousness can be inferred.
Short bio: Mariana Olezza is an Argentinian Computer Engineer (National University of Córdoba), pursuing a PhD in Philosophy (University of Buenos Aires). Some research areas of interest: philosophy of AI/Mind/Psychiatry/Abduction. She recently worked
at Infosys in the field of “Semantic entropy and hallucinations in LLMs” theoretically and practically.
rTAIM (Rebuilding Trust in AI Medicine) Monthly Seminars are online
seminars open to anyone in the world that would like to present her/his current research on topics associated with the main theme of the project.
Organisation:
Steven S. Gouveia (MLAG/IF)
Mind, Language and Action Group (MLAG)
Instituto de Filosofia da Universidade do Porto – UID/00502/2025
Fundação para a Ciência e a Tecnologia (FCT)
____________________________________________