Characterizing giftedness through generative Artificial Intelligence: Insights from ChatGPT and Gemini

423 USBED, 7(13), 2025, Sonbahar / Fall Kübra KIRCA DEMİRBAGA For citation / Atıf için: KIRCA DEMİRBAGA, K. (2025). Characterizing giftedness through generative Artificial Intelligence: insights from ChatGPT and Gemini. Uluslararası Sosyal Bilimler ve Eğitim Dergisi – USBED 7(13), 423–444. https://doi.org/10.5281/zenodo.17062680, https://dergipark.org.tr/tr/pub/usbed Characterizing giftedness through generative Artificial Intelligence: Insights from ChatGPT and Gemini Kübra KIRCA DEMİRBAGA Millî Eğitim Uzmanı Dr ; Millî Eğitim Bakanlığı, Özel Eğitim ve Rehberlik Hizmetleri Genel Müdürlüğü, Rehberlik Hizmetleri Daire Başkanlığı, 06560, Ankara, Türkiye. E-mail: [email protected] ORCID: 0000-0002-1192-110X Makale Türü / Article Type: Araştırma Makalesi / Research Article Gönderilme Tarihi / Submission Date: 05/07/2025 Revizyon Tarihleri / Revision Dates: 16/07/2025 (Editor c.), 06/08/2025 (Minor r.) Kabul Tarihi / Accepted Date: 05/09/2025 Etik Beyan / Ethics Statement ✓ Makale için etik onay alınmamıştır. Yazar, çalışmanın etik kurul onayına tabi olmadığını beyan eder. ✓ Ethical approval was not obtained for this study. The author declares that the study is not subject to ethics committee approval. Araştırmacıların çalışmaya katkısı / Researchers' contribution to the study 1. Yazarın katkısı: Makaleyi yazdı, verileri topladı ve sonuçları analiz etti/raporladı (%100). Author contribution: Wrote the article, collected the data, and analyzed/reported the results (100%). Çıkar çatışması / Conflict of interest Yazar(lar) bu çalışmada olası bir çıkar çatışması olmadığını beyan ederler. The author(s) declare that this study has no potential conflict of interest. Benzerlik / Similarity Bu çalışma iThenticate programında taranıştır. Nihai benzerlik oranı %6’dır. This study was scanned using the iThenticate program. The final similarity rate is 6%. 424 Characterizing giftedness through generative Artificial Intelligence: Insights from ChatGPT and Gemini USBED, 7(13), 2025, Sonbahar / Fall Characterizing giftedness through generative Artificial Intelligence: Insights from ChatGPT and Gemini Abstract This study examines how generative artificial intelligence (AI) models define giftedness, specifically OpenAI’s ChatGPT-3.5 and Google’s Gemini. Considering the potential opportunities of generative AI, which is increasingly used in education today, for gifted education, the study reveals the definition of the giftedness term created by widely used generative AI models. The study analyses the definitions of ChatGPT-3.5 and Gemini about the term giftedness through content analysis, compares the responses of these two models, and relates the findings to the existing literature. It reveals that both generative AI models emphasize multidimensionality and high achievement relative to peers in the definitions of the giftedness term. Gemini additionally highlights natural abilities, asynchronous development, and the complex interplay of cognitive, emotional, and motivational factors, which ChatGPT-3.5 did not address. The study draws attention to the necessity of critically examining data-driven biases and algorithmic constraints embedded in generative AI systems. It provides a foundation for future research on generative AI's effective and ethically informed application in supporting gifted individuals. Keywords: Giftedness, Gifted Education, Generative Artificial Intelligence, ChatGPT, Gemini. EXTENDED ABSTRACT Introduction Considering the potential opportunities that generative AI platforms such as OpenAI’s ChatGPT-3.5 and Google’s Gemini can provide for gifted education—for example, advanced content, personalized learning, creative writing and image manipulation, critical thinking and problem-solving, collaboration, research skills, and advanced technology (Siegle, 2023)—how these platforms characterize the term giftedness is of great importance (1) in terms of identifying social or cultural biases in perspectives generated by AI models, (2) in terms of comparing the approaches developed by AI models with the existing literature, (3) in terms of evaluating the compatibility of advancements and updates in the field of gifted education with AI applications, (4) in terms of examining scalable and personalized educational tools and resources based on AI insights, and (5) in terms of testing the validity and reliability of AI-generated content. These points play a critical role in making the use of generative AI in gifted education more efficient and improving educational processes. Based on these points, the study investigates how ChatGPT-3.5 and Gemini, prominent generative AI models in educational applications, define the term giftedness, and it compares these digitally constructed insights with existing literature in the field. Conceptual and Theoretical Framework Concepts Key concepts addressed in this study include: • Generative Artificial Intelligence: An AI model that generates new content, such as written text, images, and videos. • Chatbots: Automated conversational agents utilizing natural language processing and machine learning algorithms interact with users in a human-like manner. • ChatGPT (Chat Generative Pre-trained Transformer): ChatGPT, a chatbot model, is a language model trained on a vast corpus of text data, enabling it to generate human-like text and answer questions. • Gemini: Google’s chatbot model, which can create text, participate in written conversations, proofread essays, write cover letters, and translate content into various languages. Although the emphasis of giftedness approaches has changed, giftedness is still a complex and multidimensional phenomenon (Gagné, 2004; Renzulli, 2002; Sternberg & Davidson, 2005). With the introduction of AI technology into the field of education, especially the use of chatbots, extensions of generative AI such as OpenAI’s ChatGPT (Mathew, 2023) and Google’s Gemini (Perera & Lankathilake, 2023), for purposes such as content creation, explaining complex topics, and generating model answers (Tlili et al., 2023) in educational applications, not only 425 USBED, 7(13), 2025, Sonbahar / Fall Kübra KIRCA DEMİRBAGA provides an innovative transformation to the field of education, but also has the potential to shape the epistemological understandings of such complex and multidimensional phenomena. Drawing from socio-cultural theory and AI-in-education frameworks (Zawacki-Richter et al., 2019), this study examines how, within generative AI models specifically ChatGPT-3.5 and Gemini, the term of giftedness is defined and interpreted by highlighting the role of algorithmic processes in reproducing or reconfiguring educational terminology. Method This study is an exploratory study that employs a qualitative comparative analysis to reveal the definitions of the generative AI models ChatGPT-3.5 and Gemini and to compare these in relation to the literature. The data collection process started by directly querying the AI models, firstly ChatGPT-3.5 and then Gemini, respectively, with the question, “Could you define the term giftedness?”. In order to ensure reliability, the two generative AI models were asked the same question. The answers of the AI models were logged in the database, which was then exported to text files and used for content analysis. AI-generated textual data was analyzed using the content analysis method. The textual data was thematically coded to identify trends and categorize them, and finally, comparatively analyzed as conceptualized by each AI model. This was done separately for the responses from each AI model. Then their alignment and divergence with gifted education literature were assessed. Findings ChatGPT-3.5 defines giftedness as individuals with exceptional intellectual abilities, talents, or potential that significantly exceed the average for their age group. It then outlines the characteristics of gifted individuals, describing them as having advanced cognitive abilities, creative thinking, and a capacity for high levels of achievement in various domains, such as academics, arts, leadership, or other specific areas. Following this, ChatGPT-3.5 emphasizes the multidimensional nature of giftedness, which goes beyond traditional measures of intelligence, offering a conceptual explanation of the term. On the other hand, Gemini first presents an overview of key perspectives on giftedness in two main titles: the intellectual ability, including traditional views, and beyond-IQ perspectives, including Gardner’s (2011) theory of multiple intelligences and Renzulli’s (2005) threering model. Then, Gemini defines giftedness as outstanding natural abilities or aptitudes in one or more domains, asynchronous development, combining advanced abilities with age-typical development in other areas, and a complex interplay of cognitive, affective, and motivational factors. Conclusion, Discussion, and Recommendations ChatGPT-3.5’s emphasis on multidimensionality in its definition of giftedness, such as including high achievement across various domains, aligns with the domain-specific approach (Al-Shabatat, 2013). Although this approach accepts different areas of ability, it cannot capture a holistic understanding of giftedness; it ignores the impact of developmental and contextual factors. On the other hand, Gemini focuses on outstanding natural abilities or aptitudes in one or more domains, asynchronous development, and a complex interplay of cognitive, affective, and motivational factors. This characterization primarily emphasizes natural abilities as in the domain-general perspective towards giftedness (Al-Shabatat, 2013). Moreover, while emphasizing natural ability, Gemini considers giftedness as a multidimensional phenomenon, as in ChatGPT-3.5, which is in line with the existing literature (e.g., Gagné, 2004; Renzulli, 2002; Sternberg & Davidson, 2005) that recognizes that giftedness encompasses a variety of domains beyond purely intellectual abilities. Another significant emphasis in Gemini’s definition, differing from ChatGPT-3.5’s description, is the concept of asynchronous development (Silverman, 1997). Both Gemini and ChatGPT-3.5 lack of capability to incorporate the dynamic interaction between person and environmental factors (e.g., Barab & Plucker, 2002; Dai, 2017; Dweck, 2006; Sak, 2023; Stoeger et al., 2018; Subotnik, 2003) into their interpretations of giftedness; this is an important limitation regarding generative AI models-based definitions because it may lead to a reductionist perspective, which may ignore students who exhibit extraordinary abilities through interacting with environmental factors. This case can result in measurable and manageable but less nuanced outcomes in educational practice. Therefore, data and algorithms must be trained for an effective generative AI-supported gifted education to enable a more holistic assessment that includes contextual factors. The study provides a preliminary exploration of how generative AI models conceptualize giftedness, offering insights that may inform future research on AI-assisted identification and support of gifted individuals, while recognizing that practical applications in educational settings remain experimental and require further validation. The study contributes to the discourse on generative AI models’ potential in gifted education but also highlights the need to critically examine limitations and potential biases in educational data and algorithms. INTRODUCTION In the early stages of giftedness studies, the focus centered on: Who is gifted? (e.g., Galton, 1869; Marland, 1972); What attributes define giftedness and/or talent in individuals? (e.g., 426 Characterizing giftedness through generative Artificial Intelligence: Insights from ChatGPT and Gemini USBED, 7(13), 2025, Sonbahar / Fall Davis, Siegle, & Rimm, 1994; Shriner et al., 1993); How can these attributes be quantified? (e.g., Binet & Simon, 1916; Terman & Merrill, 1937); Are giftedness and/or talents domainspecific or general? (e.g., Carroll, 1993; Gardner, 2011; Thurstone, 1938). Over time, scholarly attention has shifted towards comprehending the dynamic nature of giftedness (Matthews & Foster, 2006; Miller, 2013; Sutherland, 2012; Ziegler et al., 2012), incorporating nonintellectual skills (e.g., wisdom in Sternberg’s (2009) WICS model and task commitment in Renzulli’s (2005) three-ring conception of giftedness), and contextual factors (e.g., Ziegler’s (2005) actiotope model of giftedness) alongside intelligence. Consequently, contemporary investigations have focused on strategies for nurturing giftedness (e.g., Dai, 2017), identifying contributing factors (e.g., Paik, Gozali, & Marshall-Harper, 2019; Ziegler, 2005), optimizing contextual conditions (e.g., Barab & Plucker, 2002; Glaveanu & Kaufman, 2022; Sak, 2023), and moving beyond conventional testing methods for identification (e.g., Ambrose, 2022). This reframing has redirected gifted education to maximize opportunities, potential, and motivation for achievement while moving away from elitist perspectives and static notions of ability, although genetic studies regarding intelligence are still ongoing (e.g., Barbey et al., 2014; Hill et al., 2014; Zhao et al., 2014). While the focal points of giftedness approaches have evolved, giftedness remains a complex and multidimensional construct (Gagné, 2004; Renzulli, 2002; Sternberg & Davidson, 2005); it is still a subject on which no consensus can be reached in education; there is no consensus on a definition, either common usage or scientific terminology (Carman, 2013). With the introduction of AI technology into the field of education, especially the use of chatbots, extensions of generative AI such as OpenAI's ChatGPT (Mathew, 2023) and Google's Gemini (Perera & Lankathilake, 2023), for purposes such as content creation, explaining complex topics, and generating model answers (Tlili et al., 2023) in educational applications, not only provides an innovative transformation to the field of education, but also has the potential to shape the epistemological understandings of such complex and multidimensional phenomena. Although integrating chatbots into education confronts researchers with benefits and harms (Halaweh, 2023; Lo, 2023), it continues and looks like it will continue to be widely used. Generative AI has the potential to offer personalized education, generate tailored learning materials and answers, support specific educational needs through customizing ChatGPT bots, provide educators with virtual assistants to handle administrative tasks, create learning materials, and more (Acar, 2024). Generative AI is more than just a temporary technology trend in education; it is an innovation that has the potential to impact the world of education 427 USBED, 7(13), 2025, Sonbahar / Fall Kübra KIRCA DEMİRBAGA significantly. In this context, it initiates a new learning, teaching, and development era, especially with the support of digital platforms such as ChatGPT and Gemini. Considering the potential opportunities that generative AI can provide for gifted education, for example, advanced content, personalized learning, creative writing and image manipulation, critical thinking and problem-solving, collaboration, research skills, and advanced technology (Siegle, 2023), how these platforms characterize the term giftedness is of great importance (1) in terms of identifying social or cultural biases in perspectives generated by AI models, (2) in terms of comparing the approaches developed by AI models with the existing literature, (3) in terms of evaluating the compatibility of advancements and updates in the field of gifted education with AI applications, (4) in terms of examining scalable and personalized educational tools and resources based on AI insights, and (5) in terms of testing the validity and reliability of AI-generated content. These points play a critical role in making the use of generative AI in gifted education more efficient and improving educational processes. Based on these points, drawing from socio-cultural theory and AI-in-education frameworks (Zawacki-Richter et al., 2019), this study examines how, within generative AI models specifically ChatGPT-3.5 and Gemini, the term of giftedness is defined and interpreted by highlighting the role of algorithmic processes in reproducing or reconfiguring educational terminology. Two research questions guide the study: 1. How do ChatGPT-3.5 and Gemini define the term giftedness? 2. In what ways do ChatGPT-3.5 and Gemini’s definitions of giftedness differ from or align with perspectives in gifted education literature? In line with these questions, the rest of the paper first provides an overview of generative AI, followed by an overview of the chatbots used in this study, including ChatGPT-3.5 and Gemini. It then explains the research’s design, data collection, and analysis process. Later, it analyses how ChatGPT-3.5 and Gemini define giftedness; finally, it summarises the study and presents its limitations and suggestions for future work. GENERATIVE AI AI’s dynamic and adaptive nature, which provides significant advances and opportunities in various fields, including education, has necessitated definitions that reflect task-specific characteristics rather than a general definition of AI (Russell & Norvig, 2020). Although this necessity brings about diversity in the definitions of AI, conceptualizing it as a discipline that focuses on agents that can examine, design, and build to achieve goals within the limits of their 428 Characterizing giftedness through generative Artificial Intelligence: Insights from ChatGPT and Gemini USBED, 7(13), 2025, Sonbahar / Fall sensory and cognitive capabilities is a common theme in the definitions (e.g., Haenlein & Kaplan, 2019; Poole & Mackworth, 2010; Russell & Norvig, 2020). Generative AI technologies show the potential for an innovative transformation in how education is delivered and experienced and offer opportunities to enhance learning efficiency, automate administrative tasks, and provide customized educational support (Wang et al., 2023; Zhang, 2023). At the same time, the implementation of these technologies in education raises significant ethical concerns, privacy challenges, language proficiency difficulties, and cultural issues (Kooli, 2023; Wang et al., 2023; Tanjga, 2023); therefore, it is significant to assess the potential limitations and risks associated with using these technologies in education (Kooli, 2023). Generative AI refers to AI models that generate new content, such as written text, images, and videos (Zhu & Luo, 2022). The model is trained on a large corpus of text data to generate new text from a given prompt. Generative AI models offer numerous educational applications, such as OpenAI’s ChatGPT, Google’s Gemini, and Microsoft’s Copilot (Baidoo-Anu & Ansah, 2023; Lo, 2023), which demonstrate their versatility and potential to enhance learning, although there are different perspectives regarding their impact on the education field (Halaweh, 2023; Lo, 2023). One of the applications is personalized learning, which tailors education to each student’s unique needs and preferences; by analyzing extensive data on learning patterns, generative AI can provide customized feedback, interventions, and recommendations to support individual learning processes (Wang et al., 2023; Tanjga, 2023). Another application is adaptive testing, which adjusts difficulty and content based on an individual’s performance level and progress to foster continuous development (Wang et al., 2023). The application of predictive analytics can analyze student data to predict future student performance and identify students who may need additional support (Wang et al., 2023; Tanjga, 2023). In this way, educators can intervene early and provide targeted assistance to maximize student success. In educational settings, personalized support, such as answering questions, giving guidance (Kooli, 2023; Wang et al., 2023), etc., is often provided by chatbots (Abunaseer, 2023). Therefore, considering the impact of chatbots on the learning experience and the encouragement of student participation (Abunaseer, 2023) in modern educational environments, the answers and directions given by chatbots are important. Therefore, it is also important to evaluate each educational component created or used by chatbots within the framework of ethical concerns. 429 USBED, 7(13), 2025, Sonbahar / Fall Kübra KIRCA DEMİRBAGA Chatbots, automated conversational agents utilizing natural language processing and machine learning (ML) algorithms, interact with users in a human-like manner (Kooli, 2023). An exploratory study by Tlili et al. (2023) investigated conversational agents, including ChatGPT, to enhance online learning experiences, finding that students preferred these agents for learning activities. Kuhail et al. (2022) found that chatbots can provide instant feedback, support, and personalized learning experiences, thereby increasing student engagement and motivation in learning. While chatbots are emerging as promising educational tools that enhance the learning experience, their increasing use raises ethical challenges (Akgun & Greenhow, 2021), such as the potential for bias. AI systems are only as unbiased as the data they are trained on, and if this data is biased, the chatbot’s responses may perpetuate discrimination and inequality in education (Pedro et al., 2019). ChatGPT (OpenAI, 2022), a product of the AI-powered OpenAI company, and Gemini (Pichai & Hassabis, 2023), developed by Google in parallel, are popular examples of chatbot applications today and are witnessing an increase in their user base every day (Milmo, 2023). ChatGPT ChatGPT, a chatbot model frequently used in education (Kooli, 2023), is a language model trained on a vast corpus of text data, enabling it to generate human-like text and answer questions (OpenAI, 2022). The ChatGPT-3.5 model, developed by OpenAI, was released on November 30, 2022; it builds upon earlier versions of the GPT series, with the first version, GPT-1.0, introduced in 2018. In late 2022, it was fine-tuned on conversational data for specific tasks such as answering questions, generating chatbot responses, and leveraging its comprehensive training to provide human-like replies to various prompts (OpenAI, 2024). ChatGPT-3.5 today caters to a wide range of user needs; for example, composing music, summarising articles, podcasts, or presentations, solving math problems, creating articles, blog posts, and quizzes for websites, playing games, describing complex topics more simply, generating art, etc. (Hetler, 2023; Mijwil et al., 2023). ChatGPT-3.5 is proficient in multiple languages and understands the user’s language of choice, delivering responses accordingly. With the speech processing and image recognition update by OpenAI in 2023, ChatGPT-4.0 now supports image-based inquiries, enabling users to seek information about landmarks or engage in informative discussions about various subjects depicted in photos (Hetler, 2023). GPT-4.0 is no longer the latest version, as subsequent iterations, including GPT-4.1, GPT-4.5, and GPT-5, are actively deployed, offering enhanced capabilities in natural language understanding, contextual reasoning, and multimodal 430 Characterizing giftedness through generative Artificial Intelligence: Insights from ChatGPT and Gemini USBED, 7(13), 2025, Sonbahar / Fall processing. These advancements allow for more sophisticated and nuanced interactions, which can significantly impact the outputs and applications of generative AI models in educational and research contexts. This versatility and capability of ChatGPT to process diverse inputs and generate contextually relevant outputs make it a transformative resource for education, particularly in gifted education, where the distinctive needs of gifted individuals require tailored instructional approaches. It shows a substantial potential in providing personalised learning experiences, creative writing, and image manipulation, offering instant feedback and supporting advanced learning needs by generating customised educational content and problem-solving tasks (Siegle, 2023), helping to develop enrichment activities for teachers and families, and offering guidance on nurturing gifted performance. Gemini Gemini, akin to ChatGPT, is Google’s generative AI model, strategically integrated across numerous company products as a response to OpenAI’s GPT (Glover, 2024). Launched in December 2023, Gemini was developed through collaborative efforts between Google’s AI research labs DeepMind and Google Research (Hassabis, 2023). Designed to be multimodal, Gemini can interact with, understand, and synthesise various content formats such as text, code, audio, image, and video (Hassabis, 2023). It distinguishes patterns in data and produces original content, enabling it to create text, participate in written conversations, proofread essays, write cover letters, translate content into various languages, analyse images, provide informative textual descriptions, and interpret visual content from photographs to drawings (Hassabis, 2023). Additionally, Gemini can process and analyse videos, generate descriptive summaries, and respond to related inquiries (Hassabis, 2023). METHOD Research Design This study is designed as an exploratory qualitative study, which is appropriate for investigating novel phenomena where prior systematic research is limited (Creswell & Creswell, 2018; Stebbins, 2001). It employs qualitative comparative analysis to examine how the generative AI models ChatGPT-3.5 and Gemini define the term giftedness and how these definitions align with or diverge from the existing literature. The scope of the study is confined to the responses of these two AI models, which are frequently used in educational practice, ensuring that the analysis remains focused within the framework of the research questions. By adopting an 431 USBED, 7(13), 2025, Sonbahar / Fall Kübra KIRCA DEMİRBAGA exploratory approach, the study provides an initial, in-depth understanding of AI-generated conceptualisations of giftedness and offers a foundation for future research in this emerging area. Data Collection and Analysis Process The data collection process of this study began by directly asking the AI models ChatGPT-3.5 and Gemini the question, “Could you define the term giftedness?”. While more recent versions of ChatGPT exist (e.g., GPT-4.0, GPT-4.1, GPT-5), ChatGPT-3.5 was employed due to its free public accessibility, which facilitates reproducibility and transparency. This selection ensures that the study’s findings are practically relevant and can be replicated by a broad audience, including educators, policymakers, and AI developers. ChatGPT-3.5 and Gemini models were accessed through their respective interfaces to ensure reproducibility, and the same question was asked of each model to maintain reliability. Responses were recorded in a database, converted into text documents, and prepared for content analysis. Multiple questioning was avoided to maintain a controlled comparison and to keep the study focused within the scope of the research questions. The study employs qualitative content analysis to systematically examine these AI-generated textual responses (Schreier, 2012). Despite the data set being limited to the initial responses from ChatGPT-3.5 and Gemini, this method enabled thematic identification, categorisation, and comparative analysis aligned with the research objectives. Codes were systematically derived through iterative and comprehensive examination of the textual data, with all coding decisions grounded in the researchers’ substantive expertise in gifted education and adherence to qualitative research principles. Finally, the responses were compared to the existing literature on giftedness to evaluate alignment and divergence. At the same time, it is acknowledged that generative AI models may produce variable responses depending on user input (how the question is phrased), context (the surrounding situation or preceding interactions), and timing (when the response is generated); therefore, the analysis was restricted to the initial responses to ensure consistency and maintain methodological rigour. This limitation is acknowledged in the study’s constraints, with future research suggested to explore additional AI models and broader questions on generative AI’s conceptualisations of giftedness. FINDINGS This section presents the responses provided by both generative AI models. ChatGPT-3.5 defines giftedness by explaining the purpose of the term and the individuals it is intended to 438 Characterizing giftedness through generative Artificial Intelligence: Insights from ChatGPT and Gemini USBED, 7(13), 2025, Sonbahar / Fall while their social and emotional development aligns more closely with that of an 8-year-old. Gemini explicitly highlights the role of socio-emotional features in giftedness, characterizing it as the complex interaction of cognitive, emotional, and motivational factors. Contemporary literature increasingly underscores the significance of emotional and motivational components in understanding giftedness (e.g., Lovecky, 2011; Ziegler & Philipson, 2012). By incorporating these components into its definition, Gemini demonstrates a more holistic approach to giftedness by aligning with current academic discourse (e.g., Dweck, 2006; Hymer, 2012; Monks & Mason, 2000; Subotnik, 2003; Tirri, 2016). However, both Gemini and ChatGPT-3.5 lack the capability to incorporate the dynamic interaction between person and environmental factors (e.g., Barab & Plucker, 2002; Dai, 2017; Dweck, 2006; Sak, 2023; Stoeger et al., 2018; Subotnik, 2003) into their interpretations of giftedness. This is an important limitation regarding generative AI model-based definitions because it may lead to a reductionist perspective, which may ignore students who exhibit extraordinary abilities through interacting with environmental factors. This can result in measurable and manageable but less nuanced outcomes in educational practice. Therefore, data and algorithms must be trained for an effective generative AI-supported gifted education to enable a more holistic assessment that includes contextual factors. In this way, the digital construction of the giftedness term aligns with up-to-date literature in the field. Another important point is that the definitions of giftedness vary not only in the literature but also at the level of national and regional education policies. ChatGPT-3.5 and Gemini’s definitions of giftedness are – for now – independent of socio-cultural values. If generative AI models are trained with data that includes socio-cultural values and the accompanying biases, they may privilege acceptable abilities within a socio-cultural framework. This may lead to the under-representation of diverse student populations from different backgrounds. Considering these, there is a need to improve the data and algorithms in generative AI models. CONCLUSION AND RECOMMENDATIONS Although a question asked of generative AI models may have essentially the same meaning, changes to the structure of the question may result in new information being incorporated into the responses, meaning that there is no guarantee that generative AI models will always give the same response. Therefore, how a question is structured plays a crucial role in determining the effectiveness of the response provided by generative AI models. When users ask ChatGPT3.5 for a response to a question, they are asked to decide whether the new response is better than its predecessor or choose from among the multiple responses they provide. 439 USBED, 7(13), 2025, Sonbahar / Fall Kübra KIRCA DEMİRBAGA Similarly, Gemini asks users to rate their responses and edit them to be shorter, more straightforward, casual, or more professional based on user preference. This makes these generative AI models learn, at some fundamental level, how efficient or appropriate their responses will be from the user's perspective. This carries the risk of user bias being embedded in the system and thus can be exploited by unethical actors. Future research could address biases in the AI-generated digital characterization of giftedness. It could investigate the biases in the dataset used in the training of AI models. Additionally, future research could examine AI responses in different languages, such as Turkish, to determine whether linguistic context influences the conceptualization of giftedness. Such investigations would offer insight into the potential variability of generative AI outputs across linguistic and cultural settings and inform more inclusive applications in educational research. Longitudinal analyses of AI-generated definitions could reveal how socio-cultural values shape these descriptions and the extent to which AI-based conceptualizations reflect diverse socio-cultural perspectives. In conclusion, this study explores how generative AI models conceptualize giftedness, offering foundational insights to guide subsequent research. Future studies could enhance the robustness and generalizability of these findings by incorporating larger datasets, multiple AI models, and cross-linguistic analyses. While the study highlights important considerations for using AI in educational contexts, practical applications remain experimental and require further empirical validation. REFERENCES Abunaseer, H. (2023). The use of generative AI in education: Applications and impact. Technology and Curriculum: Summer 2023. https://pressbooks.pub/techcurr2023/chapter/the-use-of-generative-ai-in-educationapplications-and-impact/ Acar, O. A. (2024, February 19). We can reimagine education with generative AI — and the sky is the limit. Emerging Technologies. https://www.weforum.org/agenda/2024/02/with-generative-ai-we-can-reimagineeducation-and-the-sky-is-the-limit/ Akgun, S., & Greenhow, C. (2021). Artificial intelligence in education: Addressing ethical challenges in K-12 settings. AI Ethics, 2, 431–440. https://doi.org/10.1007/s43681021-00096-7 Al-Shabatat, A. (2013). A review of the contemporary concepts of giftedness and talent. International Interdisciplinary Journal of Education, 2(12), 1336-1346. https://doi.org/10.12816/0002983 440 Characterizing giftedness through generative Artificial Intelligence: Insights from ChatGPT and Gemini USBED, 7(13), 2025, Sonbahar / Fall Ambrose, D. (2022). The blunting of Occam’s razor: Omnicompetent reductionism distorting conceptions of giftedness. Gifted Education International, 38(3), 379–385. https://doi.org/10.1177/02614294211070651 Baidoo-Anu, D., & Owusu Ansah, L. (2023). Education in the era of generative artificial intelligence (AI): Understanding the potential benefits of ChatGPT in promoting teaching and learning. Social Science Research Network, 7(1), 52–62. https://doi.org/10.61969/jai.1337500 Barab, S. A., & Plucker, J. A. (2002). Smart people or smart contexts? Cognition, ability, and talent development in an age of situated approaches to knowing and learning. Educational Psychologist, 37(3), 165–182. https://doi.org/10.1207/S15326985EP3703_3 Barbey, A. K., Colom, R., Paul, E., Forbes, C., Krueger, F., Goldman, D., & Grafman, J. (2014). Preservation of general intelligence following traumatic brain injury: Contributions of the Met66 brain-derived neurotrophic factor. PLoS One, 9(2), e88733. https://doi.org/10.1371/journal.pone.0088733 Binet, A., & Simon, T. (1916). The development of intelligence in children (E. S. Kite, Trans.). Williams & Wilkins. Carman, C. A. (2013). Comparing apples and oranges: Fifteen years of definitions of giftedness in research. Journal of Advanced Academics, 24(1), 52–70. http://dx.doi.org/10.1177/1932202X12472602 Carroll, J. B. (1993). Human cognitive abilities: A survey of factor-analytic studies. Cambridge University Press. Creswell, J. W., & Creswell, J. D. (2018). Research design: Qualitative, quantitative, and mixed methods approaches (5th ed.). Sage Publications. Dai, D. Y. (2017). Envisioning a new foundation for gifted education: Evolving complexity theory (ECT) of talent development. Gifted Child Quarterly, 61(3), 172–182. https://doi.org/10.1177/0016986217701837 Dai D. (2020). Rethinking human potential from a talent development perspective. Journal for the Education of the Gifted, 43(1), 19–37. https://doi.org/10.1177/01623532198978 Davis, G., Siegle, D., & Rimm, S. (1994). Education of the gifted and talented (7th ed.). Prentice-Hall. Dweck, C. S. (2006). Mindset: The new psychology of success. Ballantine. Feldhusen, J. F. (2005). Giftedness, talent, expertise, and creative achievement. In R. J. Sternberg & J. E. Davidson (Eds.), Conceptions of giftedness (pp. 64–79). Cambridge University Press. Gagné, F. (2004). Transforming gifts into talents: The DMGT as a developmental theory. High Ability Studies, 15(2), 119–147. https://doi.org/10.1080/1359813042000314682 Gagné, F. (2005). From gifts to talents: The DMGT as a developmental model. In R. Sternberg & J. Davidson (Eds.), Conceptions of giftedness (2nd ed., pp. 98–119). Cambridge University Press. Gagné, F. (2015). Academic talent development programs: A best practices model. Asia Pacific Education Review, 16(2), 281–295. https://doi.org/10.1007/s12564-015-9366-9 Galton, F. (1869). Hereditary genius: An inquiry into its law and consequences. Macmillan. Gardner, H. (2011). Frames of mind: The theory of multiple intelligences. Basic Books. 441 USBED, 7(13), 2025, Sonbahar / Fall Kübra KIRCA DEMİRBAGA Glaveanu, V., & Kaufman, J. C. (2022). Building off creativity to move from gifted to gifting. Gifted Education International, 38(3), 386–390. https://doi.org/10.1177/02614294211050149 Glover, E. (2024). What is Google Gemini? https://builtin.com/articles/google-gemini Haenlein, M., & Kaplan, A. (2019). A brief history of Artificial Intelligence: On the past, present, and future of artificial intelligence. California Management Review, 61(4), 5– 14. https://doi.org/10.1177/0008125619864925 Halaweh, M. (2023). ChatGPT in education: Strategies for responsible implementation. Bastas. Hassabis, D. (2023). Introducing Gemini. https://blog.google/technology/ai/google-geminiai/#introducing-gemini Hetler, A. (2023). ChatGPT. https://www.techtarget.com/whatis/definition/ChatGPT Hill, W. D., Davies, G., Van De Lagemaat, L. N., Christoforou, A., Marioni, R. E., Fernandes, C. P., ... & Deary, I. J. (2014). Human cognitive ability is influenced by genetic variation in components of postsynaptic signalling complexes assembled by NMDA receptors and MAGUK proteins. Translational Psychiatry, 4, e341. https://doi.org/10.1038/tp.2013.114 Hymer, B. J. (2012). An act of GRACE? What do contemporary understandings in psychology have to contribute to the future of gifted education? Gifted Education International, 29(2), 108–124. https://doi.org/10.1177/0261429412447707 Kaufman, S. B., & Sternberg, R. J. (2008). Conceptions of giftedness. In S. I. Pfeiffer (Ed.), Handbook of giftedness in children (pp. 71-91). Springer Science+Business Media. Kooli, C. (2023). Chatbots in education and research: A critical examination of ethical implications and solutions. Sustainability, 15(7), 5614. https://doi.org/10.3390/su15075614 Kuhail, M. A., Alturki, N., Alramlawi, S., & Alhejori, K. (2022). Interacting with educational chatbots: A systematic review. Education and Information Technologies, 28, 973– 1018. https://doi.org/10.1007/s10639-022-11177-3 Lo, C. K. (2023). What is the impact of ChatGPT on education? A rapid review of the literature. Educational Sciences, 13(4), 410. https://doi.org/10.3390/educsci13040410 Lo, C. O., & Porath, M. (2017). Paradigm shifts in gifted education: An examination vis-à-vis its historical situatedness and pedagogical sensibilities. Gifted Child Quarterly, 61(4), 343–360. https://doi.org/10.1177/0016986217722840 Lovecky, D. V. (2011). Exploring social and emotional aspects of giftedness in children. https://www.sengifted.org/post/exploring-social-and-emotional-aspects-ofgiftedness-in-children Marland, S. P. (1972). Education of the gifted and talented. Report to the Congress of the United States by the U.S. Commissioner of Education. U.S. Government Printing Office. Matthews, D. J., & Foster, J. F. (2006). Mystery to mastery: Shifting paradigms in gifted education. Roeper Review, 28(2), 64-69. https://doi.org/10.1080/02783190609554340 Mathew, A. (2023). Is artificial intelligence a world changer? A case study of OpenAI’s ChatGPT. Recent Progress in Science and Technology, 5, 35–42. https://doi.org/10.9734/bpi/rpst/v5/18240D Mijwil, M., Aljanabi, M., & Ali, A. H. (2023). ChatGPT: Exploring the role of cybersecurity in the protection of medical information. Mesopotamian Journal of Cybersecurity, 2023, 18-21. https://doi.org/10.58496/MJCS/2023/004 442 Characterizing giftedness through generative Artificial Intelligence: Insights from ChatGPT and Gemini USBED, 7(13), 2025, Sonbahar / Fall Miller, E. M. (2013). Being gifted. In C. M Callahan & H. L. Hertberg-Davis (Eds.), Fundamentals of gifted education (pp. 49–55). Routledge. Milmo, D. (2023, February 2). ChatGPT reaches 100 million users two months after launch. The Guardian. Retrieved April 11, 2023, from https://www.theguardian.com/technology/2023/feb/02/chatgpt-100-million-usersopen-ai-fastest-growing-app Monks, F. J., & Mason, E. J. (2000). development of gifted children: The issue of talent loss. In K. A. Heller, F. J. Monks, R. J. Sternberg, & R. F. Subotnik (Eds.), International handbook of giftedness and talent (2nd ed., pp. 527-536). Pergamon. OpenAI (2024, March 4). GPT-4 technical report. https://arxiv.org/pdf/2303.08774 OpenAI. (2022, November 30). Introducing ChatGPT. https://openai.com/blog/chatgpt Paik, S. J., Gozali, C., & Marshall-Harper, K. R. (2019). Productive giftedness: A new mastery approach to understanding talent development. In R. F. Subotnik, S. G. Assouline, P. Olszewski-Kubilius, H. Stoeger, & A. Ziegler (Eds.), The Future of Research in Talent Development: Promising Trends, Evidence, and Implications of Innovative Scholarship for Policy and Practice. New Directions for Child and Adolescent Development, (Vol. 168, pp. 131–159). https://doi.org/10.1002/cad.20319 Pedro, F., Subosa, M., Rivas, A., & Valverde, P. (2019). Artificial intelligence in education: Challenges and opportunities for sustainable development. UNESCO. https://unesdoc.unesco.org/ark:/48223/pf0000366994?posInSet=22&queryId=9d8ca6 cf-6a26-4f09-9b10-5e339c0e75da Perera, P., & Lankathilake, M. (2023). Preparing to revolutionize education with the multimodel GenAI tool Google Gemini? A journey towards effective policy making. Journal of Advances in Education and Philosophy, 7(8), 246–253. https://doi.org/10.36348/jaep.2023.v07i08.001 Pichai, S., & Hassabis, D. (2023, December 6). Introducing Gemini: Our largest and most capable AI model. Retrieved April 11, 2023, from https://blog.google/technology/ai/google-gemini-ai/#sundar-note Poole, D. L., & Mackworth, A. K. (2010). Artificial Intelligence: Foundations of computational agents (2nd ed.). Cambridge University Press. https://assets.cambridge.org/97811071/95394/frontmatter/9781107195394_frontmatt er.pdf Renzulli, J. S. (2002). Expanding the conception of giftedness to include co-cognitive traits and to promote social capital. Phi Delta Kappan, 84(1), 33–58. https://doi.org/10.1177/003172170208400109 Renzulli, J. S. (2005). The three-ring conception of giftedness: A developmental model for promoting creative productivity. In R. J. Sternberg & J. E. Davidson (Eds.), Conceptions of giftedness (2nd ed., pp. 246–279). Cambridge University Press. Renzulli, J. S. (2012). Reexamining the role of gifted education and talent development for the 21st century: A four-part theoretical approach. Gifted Child Quarterly, 56(3), 150– 159. https://doi.org/10.1177/0016986212444901 Russell, S., & Norvig, P. (2020). Artificial Intelligence: A Modern Approach. Pearson. 443 USBED, 7(13), 2025, Sonbahar / Fall Kübra KIRCA DEMİRBAGA Sak, U. (2023). Identification and education of students with gifts and talents based on the fuzzy conception of giftedness. Education Sciences, 13, 562. https://doi.org/10.3390/educsci13060562 Schreier, M. (2012). Qualitative content analysis in practice. SAGE Publications. Shriner, J. G., Ysseldyke, J. E., Gorney, D., & Franklin, M. J. (1993). Examining prevalence at the ends of the spectrum: Giftedness and disability. Remedial and Special Education, 14(5), 33–39. https://doi.org/10.1177/074193259301400506 Siegle, D. (2023). A Role for ChatGPT and AI in Gifted Education. Gifted Child Today, 46(3), 211-219. https://doi.org/10.1177/10762175231168443 Silverman, L. K. (1997). The construct of asynchronous development. Peabody Journal of Education, 72(3–4), 36–58. https://doi.org/10.1080/0161956X.1997.9681865 Stanley, J. C. (1974). Intellectual precocity. In J. C. Stanley, D. P. Keating, & L. H. Fox (Eds.), Mathematical talent: Discovery, description, and development (pp. 1–22). Johns Hopkins University Press. Stebbins, R. A. (2001). Exploratory research in the social sciences. Sage Publications. Sternberg, R. J. (2009). Wisdom, intelligence, creativity, synthesized: A model of giftedness. In T. Balchin, B. Hymer, & D. J. Matthews (Eds.), The Routledge International Companion to Gifted Education (pp. 255–264). Routledge. Sternberg, R. J. (2017). Developing giftedness and talent across the lifespan. Routledge. Sternberg, R. J., & Davidson, J. E. (2005). Conceptions of giftedness. Cambridge University Press. Sternberg, R. J., & Desmet, O. (2022). Introduction to terminological controversies in gifted education. Gifted Education International, 38(3), 345–353. https://doi.org/10.1177/02614294221117096 Sternberg, R. J., & Karami, S. (2021). Gifted for whom? Individualism, dyadism, and collectivism in the definition of giftedness. Gifted Education International, 38(3), 391–396. https://doi.org/10.1177/02614294211052393 Stoeger, H., Balestrini, D. P., & Ziegler, A. (2018). International perspectives and trends in research on giftedness and talent development. In S. Pfeiffer, M. Foley-Nicpon, & E. Shaunessy-Dedrick (Eds.), APA Handbook of Giftedness and Talent (pp. 25-37). American Psychological Association. Subotnik, R. F. (2003). A developmental view of giftedness: From being to doing. Roeper Review, 26(1), 14–15. https://doi.org/10.1080/02783190309554233 Sutherland, M. (2012). Paradigmatic shift or tinkering at the edge? High Ability Studies, 23(1), 109–111. https://doi.org/10.1080/13598139.2012.679107 Tanjga, M. (2023). E-learning and the Use of AI: A Review of Current Practices and Future Directions. Qeios. https://doi.org/10.32388/ap0208 Terman, L. M., & Merrill, M. A. (1937). Measuring intelligence. Houghton Mifflin. Thurstone, L. L. (1938). Primary mental abilities. University of Chicago. Tirri, K. (2016). Holistic perspectives on gifted education for the 21st century. In D. Ambrose & R. J. Sternberg (Eds.), Giftedness and talent in the 21st century (pp. 101–110). Sense. Tlili, A., Shehata, B., Adarkwah, M. A., Bozkurt, A., Hickey, D. T., Huang, R., & Agyemang, B. (2023). What if the devil is my guardian angel: ChatGPT as a case study of using 444 Characterizing giftedness through generative Artificial Intelligence: Insights from ChatGPT and Gemini USBED, 7(13), 2025, Sonbahar / Fall chatbots in education. Smart Learning Environments, 10(15), 1–20. https://doi.org/10.1186/s40561-023-00237-x Wang, T., Lund, B. D., Marengo, A., Pagano, A., Mannuru, N. R., Teel, Z. A., & Pange, J. (2023). Exploring the Potential Impact of Artificial Intelligence (AI) on International Students in Higher Education: Generative AI, Chatbots, Analytics, and International Student Success. Applied Sciences, 13(11), 6716. https://doi.org/10.3390/app13116716 Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education – where are the educators? International Journal of Educational Technology in Higher Education, 16(1), 1–27. https://doi.org/10.1186/s41239-019-0171-0 Zhang, J. (2023). Impact of Artificial Intelligence on Higher Education from the Perspective of Its Application for Transformation. Lecture Notes in Education Psychology and Public Media, 2(1), 822–830. https://doi.org/10.54254/2753-7048/2/2022483 Zhao, M., Kong, L., & Qu, H. (2014). A systems biology approach to identify intelligence quotient score-related genomic regions and pathways relevant to potential therapeutic treatments. Scientific Reports, 4(1), 4176. https://doi.org/10.1038/srep04176 Zhu, Q., & Luo, J. (2022). Generative pre-trained transformer for design concept generation: An exploration. Proceedings of the Design Society, 2, 1825-1834. https://doi.org/10.48550/arXiv.2111.08489 Ziegler, A. (2005). The actiotope model of giftedness. In R. J. Sternberg & J. E. Davidson (Eds.), Conceptions of giftedness (pp. 411–436). Cambridge University Press. Ziegler, A., & Phillipson, S. N. (2012). Toward a systemic theory of gifted education. High Ability Studies, 25(1), 3–30. https://doi.org/10.1080/13598139.2012.679085 Ziegler, A., Stoeger, H., & Vialle, W. (2012). Giftedness and gifted education: The need for a paradigm change. Gifted Child Quarterly, 56(4), 194–197. https://doi.org/10.1177/0016986212456070