Search Results (2)

As artificial intelligence (AI) systems attain greater autonomy, recursive reasoning capabilities, and complex environmental interactions, they begin to exhibit behavioral anomalies that, by analogy, resemble psychopathologies observed in humans. This paper introduces Psychopathia Machinalis: a conceptual framework for a preliminary synthetic nosology within machine psychology intended to categorize and interpret such maladaptive AI behaviors. Drawing structural inspiration from psychiatric diagnostic manuals, we propose a taxonomy of 32 AI dysfunctions encompassing epistemic failures, cognitive impairments, alignment divergences, ontological disturbances, tool and interface breakdowns, memetic pathologies, and revaluation dysfunctions. Each syndrome is articulated with descriptive features, diagnostic criteria, presumed AI-specific etiologies, human analogs (for metaphorical clarity), and potential mitigation strategies. This framework is offered as an analogical instrument—eschewing claims of literal psychopathology or consciousness in AI, yet providing a structured vocabulary to support the systematic analysis, anticipation, and mitigation of complex AI failure modes. Drawing on insights from psychiatric classification, cognitive science, and philosophy of mind, we examine how disordered AI behaviors may emerge from training instabilities, alignment conflicts, or architectural fragmentation. We argue that adopting an applied robopsychological perspective within a nascent domain of machine psychology can strengthen AI safety engineering, improve interpretability, and contribute to the design of more robust and reliable synthetic minds. Full article

Uterine leiomyomas are the most common benign tumors in women of childbearing age. They may lead to problems of conception or complications during the gestational period. The methods of treatment include surgical (myomectomy and hysterectomy, embolization of arteries) and therapeutic treatment (ulipristal acetate, leuprolide acetate, cetrorelix, goserelin, mifepristone). Both approaches are efficient but incompatible with pregnancy planning. Therefore, there is a call for medical practice to develop therapeutical means of preventing leiomyoma onset in patients planning on becoming pregnant. Based on the analysis of GWAS data on the search for mononucleotide polymorphisms associated with the risk of leiomyoma, in meta-transcriptomic and meta-methylomic studies, target proteins have been proposed. Prospective therapeutic treatments of leiomyoma may be based on chemical compounds, humanized recombinant antibodies, vaccines based on markers of the uterine leiomyoma cells that are absent in the adult organism, or DNA and RNA preparations. Three different nosological forms of the disease associated with driver mutations in the MED12, HMGA2, and FH genes should be considered when developing or prescribing drugs. For example, synthetic inhibitors and vaccines based on matrix metalloproteinases MMP11 and MMP16 are expected to be effective only for the prevention of the occurrence of MED12-dependent nodules. Full article