Search

ChatGPT is an artificial intelligence chatbot (a computer program that attempts to simulate human conversation) developed by OpenAI, designed to provide conversational interactions similar to humans. It is built on OpenAI's GPT-3.5 and GPT-4 models, functioning by gathering data from internet texts written by people and using computational predictions to respond to user-entered questions and queries. But ultimately, could a machine surpass humans in solving tests or qualification exams? To address this question, this article aims to explore ChatGPT's accuracy levels in the Sufficiency Exam (1st Sufficiency Exam of 2022) and the Technical Qualification of the CFC (2023 edition). The results show that the chatbot achieved scores above 50% in all applied tests, a passing rate for a candidate. This research complements previous articles and demonstrates the application of AI and Natural Language Processing technology in the accounting field, providing relevant evidence for improving accounting education by considering machines as human allies in developing new skills. [ABSTRACT FROM AUTHOR]

This article had as its goal to revisit, through a narrative literature review, different historical moments, from the 15th to the 21st centuries, in which the emergence of new technologies and/or social changes led to new ways of understanding, conceptualizing, and practicing education, until finally reaching the new questions that c/taíóoíssuch as ChatGPT raise, the main one being: to what extent Artificial Intelligence can replace human teachers? It was concluded that, as a tool, ChatGPT and other similar models represent a considerable potential when it comes to being applied as part of educational practices conducted in school environments. It was also concluded that, instead of replacing the teacher, such tools will most likely require a new displacement of their function, in which competence and creativity to formulate complex questions should be highly valued. [ABSTRACT FROM AUTHOR]

The present article aims to discuss and address pertinent issues related to the new possibilities and challenges imposed on mathematical education due to the popularization of artificial intelligence. However, what exactly is AI? Let's reflect on whether today's ChatGPT is simply an updated version of the calculator from forty years ago, the desktop computer known as the "PC," and the internet from earlier phases of digital technology in mathematical education. The text seeks to consider recent news on the subject, as well as a series of simple tests conducted by the authors using this resource, in addition to exploring ethical issues and the generation of problems for ever-changing collectives of humans with digital technologies. The study suggests that we will still face immense ethical concerns, but it will be of paramount importance to consider problems to be solved for this type of collective. [ABSTRACT FROM AUTHOR]

How to enhance projects and initiatives in international partnerships in the area of digital scientific education? This study examines two current digital education plans, the European and the Brazilian, identifying differences and similarities highlighted in an analytical framework. The methodological approach was based on the inductive thematic analysis of grey literature, combined with the documentary analysis of digital educational practices reports from the Europe-Brazil CONNECT open schooling project. The results highlight the relevance of three factors: 1. digital scientific educational opportunities with socio-scientific approaches; 2. participatory science and technology such as design-thinking and inquiry mapping, including emerging hybrid environments with cloud technologies (managers) to artificial intelligence (OpenAI - ChatGPT); and 3. local and global partnerships to foster digital and scientific skills towards the sustainability of life and the planet. [ABSTRACT FROM AUTHOR]

Na argument is put forward against personifying generative artificial intelligence (AI) agents based on natural language models such as ChatGPT and Bard. It is proposed that projecting "humanity" onto this type of artifact is a form of educationally unproductive anthropocentric attachment; Katherine Hayles' model of cognitive assembly (new unconscious) is proposed as na alternative, in which technical (non-conscious) cognitions produce modes of attention-recognition that have moral and political agency, but are not to be confused with human consciousness. The projection of a theory of mind onto the artifact is explained through a post-phenomomenoloical framework and it is warned that neither demonizing artificial intelligence, nor treating these simulacra as de facto intelligence "empowers" the human, but only reenacts progressively more reductionist versions of the idealized humanista human. It is urgent to use in a more transparent way the articulations between consciousness and non-human modes of awareness so that neither students nor AI become less and less "intelligent". [ABSTRACT FROM AUTHOR]

New forms of Artificial Intelligence are gaining more space at schools and among students. ChatGPT stands out for its easy use and, in Chemistry Teaching, this tool has stood out for the numerous possibilities, including problem solving. However, there is still not 100% certainty of the effectiveness of such instrument. This work aims to investigate the potential and capacity of ChatGPT to solve exercises involving stoichiometric calculations and balancing chemical equations. These problems were extracted from a textbook for 2nd year of high school widely used by Basic Education teachers and approved by the National Textbook Program. Due to the research exploratory nature, four questions were selected, two simpler and two more laborious and complex. Furthermore, in case the ChatGPT presented the result incorrectly, some tests were performed to analyze whether a change in the statement would also generate a change in the results. These, by the way, reveal that the chat excels in the language used, as it is simple and easy to understand. On the other hand, it disappointing when solving stoichiometry problems, as it confuses concepts and does not consider all atoms when balancing an equation. In other words, unless it is already balanced correctly, the use of ChatGPT it is not indicated for solving exercises of this type and despite being a promising tool with some qualities, it is not recommended for those who do not have complete knowledge and/or a necessary discernment to detect the mistakes it made and correct them. [ABSTRACT FROM AUTHOR]