Search

The goal of this research is to analyze various arguments related to the (im)possibility of ascribing the status of moral agents to artificial intelligence. According to traditional ethical theories, any individual capable of distinguishing right from wrong is considered a moral agent. However, the accelerated development of artificial intelligence has opened up the question of whether the status of moral agents is reserved exclusively for the human species or whether we can apply some form of moral agency to modern technologies as well. The currently available answers to this question can be divided into three thematic units: arguments based on emotions, arguments based on autonomy, and arguments based on happiness maximization. After analyzing the advantages and disadvantages of each of the presented points of view, the arguments will be tested on the example of ChatGPT, currently the most widespread form of generative artificial intelligence., Cilj ovog istraživanja je analiza različitih argumenata koji se odnose na (ne)mogućnost da se veštačkoj inteligenciji pripiše status moralnih delatnika. Prema tradicionalnim etičkim teorijama, moralnim delatnikom se smatra svaka individua koja je sposobna da razlikuje ispravne od neispravnih postupaka. Međutim, ubrzani razvoj veštačke inteligencije otvorio je pitanje da li je status moralnih delatnika rezervisan isključivo za ljudsku vrstu ili se neki oblik moralne delatnosti može primenitii na savremene tehnologije. Trenutno dostupni odgovori na ovo pitanje mogu se podeliti u tri tematske celine: argumenti zasnovani na emocijama, argumenti zasnovani na autonomiji i argumenti zasnovani na maksimizaciji sreće. Nakon analize prednostii mana svakog od izloženih stanovišta, argumenti će biti testirani na primeru ChatGPT­-ja, trenutno najrasprostranjenijeg oblika generativne veštačke inteligencije.

Cilj ovog istraživanja je kritička analiza Kvajnovog shvatanja ontoloških obaveza, prevashodno u kontekstu njegovih teza o apsolutnoj istini i subdeterminaciji. Koristeći se interpretativno-kritičkim metodom, nastojaćemo da pokažemo da je Kvajnovo fizikalističko određenje ontologije posledica njegovog usvajanja naturalizma, ali i holističkog pristupa problemima saznanja. Sledeći metod pojmovne analize, videćemo da Kvajn pojam „fizike” koristi na dva različita načina. Fizika je u užem smislu nauka o materiji, dok u širem smislu označava globalnu teoriju o spoljašnjem svetu. Analogno pojmu „fizike”, analiziraćemo uži i širi smisao koji Kvajn pripisuje pojmu „jezika”. Zastupaćemo tezu prema kojoj su jezik i fizika (shvaćena u užem smislu) ontološki paralelni i da oni, zajedno sa ostalim naučnim teorijama, čine globalnu teoriju koja tek u svojoj celini postaje nosilac ontoloških obaveza. Tvrdićemo da Kvajnovo tumačenje ontologije svoj potpuni oblik zadobija u njegovim poznijim radovima, gde uvodi tezu o apsolutnoj istini kojom se garantuje da ontološke obaveze globalne teorije neće biti međusobno protivrečne. Konzistentnost ontoloških obaveza globalne teorije biće dovedena u pitanje u kontekstu rasprave o istinosnom statusu teorija koje podležu subdeterminciji. Naime, Kvajn se dugo kolebao u pogledu toga da li su u ovim slučajevima obe teorije istinite (ekumenski odgovor zasnovan na empirizmu) ili samo jedna od njih (sektaški odgovor zasnovan na naturalizmu). Oslanjajući se na istorijski i interpretativno-kritički metod, kao i zaključak o globalnoj teoriji kao nosiocu apsolutne istine, tvrdićemo da naturalizam u pogledu ontologije ima prednost nad empirizmom i da samo jedna teorija može biti istinita, uprkos tome što su obe potkrepljene empirijskim svedočanstvom., This research aims to critically analyze Quine's interpretation of ontological commitment, primarily in the context of his theses on absolute truth and underdetermination. Using the interpretative-critical method, we will try to show that Quine's physicalist rendition of ontology originates from his adoption of naturalism, along with his holistic approach to the knowledge problems. Following the method of conceptual analysis, we will see that Quine uses the term ‘physics’ in two different ways. In its narrow definition, physics is the science of matter, while in a broader sense, it represents a global theory about the external world. Analogous to the term 'physics', we shall analyze the narrow and broad sense that Quine attributes to the term 'language'. We shall uphold the thesis of ontological parallelism between language and physics (understood in its narrow sense) and claim that they, together with other scientific theories, form a global theory that only in its entirety becomes the reservoir of ontological commitment. We will argue that Quine's interpretation of ontology takes its final form in his later works, as he introduces the thesis of absolute truth, which guarantees that the ontological commitment of the global theory does not contain contradictions. The consistency in the global theory's ontological commitment will be challenged in the debate on the truth status of theories subject to underdetermination. Quine wavered for a long time as to whether, in these cases, both theories are true (an ecumenical response based on empiricism) or only one (a sectarian response based on naturalism). Relying on the historical and interpretive-critical method, as well as the judgment that the global theory is the bearer of absolute truth, we will argue that naturalism has an ontological precedence over empiricism and that only one theory can be true, despite both of them being supported by the empirical evidence.

In his famous article “Epistemology Naturalized” (1969), Quine established a novel take on the position of epistemology in philosophical and scientific discourse. According to his views, epistemological questions are a subset of psychological questions, and psychology in itself is a branch of natural science. Thus, epistemology, as understood in the Quinean sense, threatens the very idea of its normative aspects, as natural science is empiristic and, as a result, relies on purely descriptive claims. Hence, the following question arises: Does the naturalized account of epistemology entail the rejection of epistemic norms? In this research, we explore the three possible answers to this question and argue there is a sense of normativity in Quine’s naturalized epistemology, but only insofar as we are willing to accept his imperfect notion of the truth.

The role of mass media in the dissemination of political propaganda has been widely explored; however, this research focuses on more sublime political messages that are conveyed through films, particularly blockbusters that target relatively broad audiences. Since the beginning, Hollywood movies have served as a reflection of viewpoints that have been socially acceptable at a given moment. Nowadays, filmmakers and movie studios often use the method of mild self-deprecation as a way of legitimizing broader political messages. In other words, they criticize themselves to avoid much harsher criticism from the audience. We use the recent examples of Barbie and Oppenheimer to explore the current trend in strategies - such as limited self-awareness and pseudo-self-criticism - that are used to justify far more glaring issues with capitalism and (American) imperialism.

In this paper, we reconsider the notion of epistemic trespassing and attempt to explore possible scenarios in which it could lead to positive outcomes in scientific research and information dissemination. As we will point out, some of the significant discoveries in the history of science would not have been possible were it not for the epistemic trespassers, whose shift in paradigm changed the approach to specific issues for the better. Furthermore, we will present instances where individuals, often labeled as 'trespassers' - those who venture into fields outside their expertise - can be considered trustworthy and reliable sources of testimony. Consequently, epistemic trespassing turned out to be essential for the advancement of science. Epistemic trespassing in our age is virtually inevitable - whether it be from a political perspective (to preserve democratic values) or due to purely practical reasons (i.e., searching for information online). There are interdisciplinary fields where a strictly negative characterization of epistemic trespassing would question their methodology since these disciplines rely on experts from various fields. For these reasons, we advocate for a more charitable approach to epistemic trespassing and greater public education to help them distinguish between correct and false information.

Široka rasprostranjenost generativne veštačke inteligencije, pre svega velikih jezičkih modela poput ChatGPT, otvorila je niz rasprava o uticaju ove tehnologije na obrazovanje. Paralelno s time, novi razvoji na polju diskriminativne veštačke inteligencije navode nas da očekujemo porast u izradi personalizovanih planova za učenje koji će zadatke prilagođavati potrebama i interesovanjima pojedinačnih učenika, kao i razvoj algoritama koji će pomagati nastavnicima pri ocenjivanju učenika. Ipak postoji rizik da će nove tehnologije perpetuirati postojeće predrasude. Kako bi deca i mladi bili zaštićeni od neproverenih informacija i zloupotrebe podataka, potrebno je da u razvoj edukativnih softvera budu uključeni i roditelji i nastavnici. Iako će ova inkluzivnost dovesti do veće odgovornosti svih onih koji su involvirani u obrazovanje dece i mladih, dugoročna prednost će biti stvaranje bezbednog i pouzdanog virtuelnog prostora. Principi inkluzivnosti i pravičnosti značajni su i na globalnom nivou. Budući da se većina softvera zasnovanih na velikim jezičkim modelima trenutno razvija i proizvodi u zemljama kapitalističkog centra, postoji rizik od nametanja kulturnih vrednosti zapadnih zemalja na ostatak sveta, kao i stvaranja još većih razlika u dostupnosti kvalitetnog obrazovanja. Iz ovih razloga, već danas je potrebno da razmišljamo o inputu za buduće tehnologije, kako bismo sutra mogli da ih implementiramo na odgovoran način koji neće ostavljati prostora za diskriminaciju pojedinaca i grupa., The widespread availability of generative AI - large language models such as ChatGPT in particular - has opened up discussions about the impact of new technologies on education. On the other hand, novel developments in discriminative AI have led us to anticipate an increase in personalized studying programs that will adapt assignments to each student's needs and interests. Furthermore, we can expect the development of algorithms that will assist teachers in evaluating student progress. However, there is a risk that advanced digital solutions could perpetuate existing prejudices. To protect children and youth from exposure to unverified information and data mismanagement, we need to involve parents and teachers in the development process of educational software. Although this inclusion will result in greater responsibility for all those involved in children and youth education, it will contribute to the long-term advantage of a safe and reliable virtual space. The principles of inclusion and fairness are also significant at the global level. Since the vast majority of LLM-based software is currently developed and produced in the countries of the capitalist center, there is a risk of imposing Western cultural values ​on the rest of the world. Moreover, it may lead to further disparities in the availability of quality education. For these reasons, we need to start thinking about the input for future technologies today so that tomorrow, we can implement them responsibly and in a way that will not leave room for discrimination of individuals and groups.

In his famous article "Epistemology Naturalized" (1969), Quine established a novel take on the position of epistemology in philosophical and scientific discourse. According to his views, epistemological questions are a subset of psychological questions, and psychology in itself is a branch of natural science. Thus, epistemology, as understood in the Quinean sense, threatens the very idea of its normative aspects, as natural science is empiristic and, as a result, relies on purely descriptive claims. Hence, the following question arises: Does the naturalized account of epistemology entail the rejection of epistemic norms? Several philosophers have taken up the task of providing the answer to this question. Kornblith (1993) asserts that Quine's epistemic norms are merely an instrument in the specific field of natural science and, thus, that he gives up the requirement for epistemic justification of beliefs. However, Foley (1994) seems to paint Kornblith's interpretation of Quine's epistemology as somewhat reductionist and argues that Quine has never intended to reduce the whole of epistemology to psychology but only to convey the continuum between epistemology and the rest of the science. Roth (1999) claims that Quine's project of naturalized epistemology entails nothing short of explicit normativity insofar as the Quinean epistemologist relies on standards and norms of scientific inquiry. Houkes (2002) shifts the perspective of the discussion once again and claims that epistemological techniques cannot be understood independently of human nature and that Quine's understanding of them as a mere “technology” in the truth-seeking process is inherently misguided, as it would entail intentionality which Quine rejects. I shall argue that the acceptance (and the rejection) of any of those answers will depend not only on the interpretation of this specific project of epistemology naturalized but also on the variations in understanding Quine's take on naturalism in general. Q

In our paper “How to Fight Linguistic Injustice in Science: Equity Measures and Mitigating Agents” (2023), we discuss the challenges that non-native English speakers face in communicating their scientific findings. We explored several obstacles they/we encounter, with the most severe leading to epistemic injustice. In his critical reply “Do Scientists Need Mitigating Agents and Who is Responsible for Providing Them? Musings on Vučković and Sikimić’s ‘How to Fight Linguistic Injustice in Science’“, Vitaly Pronskikh raises several riveting questions that need further exploration, which inspired us to approach this topic once again and to further explore the consequences of the solutions we propose., U našem radu "How to Fight Linguistic Injustice in Science: Equity Measures and Mitigating Agents", razgovaramo o izazovima sa kojima se, pri komunikaciji naučnih otkrića, susreću istraživači kojima engleski nije maternji jezik. Istraživale smo nekoliko prepreka sa kojima se oni/mi suočavamo, među kojima su najozbiljnije one koje dovode do epistemičke nepravde. U svom kritičkom odgovoru "Do Scientists Need Mitigating Agents and Who is Responsible for Providing Them? Musings on Vučković and Sikimić's 'How to Fight Linguistic Injustice in Science: Equity Measures and Mitigating Agents'", Vitali Pronski postavlja nekolicinu uzbudljivih pitanja koja iziskuju dalje ispitivanje. To nas je inspirisalo da se još jednom vratimo ovoj temi i da dodatno preispitamo dalje posledice rešenja koja predlažemo

While having English as a shared language within the scientific community contributes to more efficient communication and peer reviews, it also puts non-native speakers in a disadvantaged position. Moreover, we argue that some concepts are inherently untranslatable and, thus, the pluralism of languages can prevent the loss of unique concepts. We use Quine's thesis on the indeterminacy of translation to demonstrate the way the knowledge of marginalized scientists gets lost. We analyze the consequences of linguistic injustice in science and propose two measures for overcoming it: practicing epistemic equity and introducing mitigating agents.

Though a common language of science allows for easier communication of the results among researchers, the use of lingua franca also comes with the cost of losing some of the diverse ideas and results arising from the plurality of languages. Following Quine's famous thesis about the indeterminacy of translation, we elaborate on the inherent loss of diverse ideas when only one language of science is used. Non-native speakers sometimes experience epistemic injustice due to their language proficiency and consequently, their scientific insights get marginalized. Thus, it is important to epistemically include the results of all researchers independent of their native language. As a solution, we promote epistemic equity and inclusion both on the individual level and on the level of the scientific community. Epistemic equity means that researchers who suffer disadvantages because of their language skills get support from the rest of the scientific community that will compensate for their disadvantage and at the same time facilitate their epistemic inclusion. This can be achieved through the introduction of mitigating agents - the individuals and organizations that ought to serve as a communication bridge between individual researchers and the scientific community.