Introduction

ChatGPT Statistics: ChatGPT, an innovation of OpenAI, has made a substantial entrance into the world of technology, shattering all records with its fast user growth. Chat GPT is an AI-generated chatbot that has been making waves in the technical world since its launch. It has a startling ability to mimic human conversation, making it a reliable tool for various tasks that range from drafting emails, answering queries, and writing essays to even assisting with coding as well.

The substructure of ChatGPT is built on OpenAI's GPT-3, which is a large language model that was showered as one of the enlightened language models when introduced in 2020. This article hunts through the captivating ChatGPT Statistics and traverses everything from user growth nationwide to revenue generation and much more.

ChatGPT has taken the world by storm, setting a record for the fastest app to reach a 100 million users, which it hit in two months. The implications of this tool are far-reaching, universities...

In March 2025, ChatGPT’s mobile app recorded over 64.26 million App Store and Google Play downloads worldwide. Google's Gemini AI Assistant mobile app was released on February 8, 2024, and was initially available in the U.S. market only. In the same month, the app registered around 13.92 million downloads. Regional preferences shape AI app adoption ChatGPT has a strong global presence with over 400.61 million monthly active users in February 2025, but regional preferences vary. In the United States, ChatGPT had a 45 percent download market share, compared to Google Gemini's 11 percent. However, Gemini emerged as the preferred generative AI app in India, representing a 52 percent market share. This competitive landscape now also includes Chinese-based players like ByteDance's Doubao and DeepSeek, indicating an even more diverse and evolving AI worldwide ecosystem. The AI-powered revolution in online search The global AI market has experienced substantial growth, exceeding 184 billion U.S. dollars in 2024 and projected to surpass 826 billion U.S. dollars by 2030. This expansion is mirrored in user behavior, with around 15 million adults in the United States using AI-powered tools as their first option for online search in 2024. Additionally, 68 percent of U.S. adults reported the use of AI-powered search engines for exploring new topics in 2024, with another 44 percent of respondents utilizing these tools to learn or explain concepts.

ChatGPT is used most widely among those between ** and ** around the world. The youngest group, those under **, are the second largest userbase, and together those under ** account for over ** percent of ChatGPT users. It is perhaps unsurprising that the younger age brackets use the chatbot more than older as that is the common trend with new technologies. Male users were far more numerous than female users, with males representing over ** percent of total users in 2023.

In the period between its release in November 2022 and January 2024, ChatGPT saw the average duration of global visits to its web domain, chat.openai.com, increase sensibly. As of the last examined month, visitors worldwide spent *** seconds on average in the platform's domain, equating to ** minutes and ** seconds. The peak of the chatbot's website session length happened in October 2023, when users worldwide spent an average of *** seconds on the web page.

Objective- The primary research objective for this study is to develop a reliable and valid scale to measure higher education students' knowledge, attitudes, and usage of GenAI.

GenAI refers to a type of artificial intelligence that generates content in response to prompts (Dwivedi et al., 2023 cited in Chiu, 2024). These prompts can include text, software code, images, videos, and music. The inception and rapid advancement of Generative AI in society has had a considerable impact on higher education.

Educators and students are beginning to apply GenAI for various purposes, including creating and enhancing learning environments, study resources, lesson plans, idea generation, data analysis, text summarisation and enhancement, and streamlining administrative processes (Francis, Jones & Smith, 2025; Freeman, 2025; Tillmans et al., 2025).

GenAI has the potential to enhance learning experiences by aiding learners, saving time, and empowering users to control their own educational journey. However, understanding students' knowledge, attitudes, and usage of GenAI is crucial for integrating these technologies effectively into educational settings. Despite the growing interest in GenAI, there is limited research on how students understand, perceive and use GenAI in their studies, and to our knowledge no freely available validated scale for effectively and quickly measuring knowledge, attitudes, and usage in a student cohort.

Understanding students' knowledge about GenAI is crucial for assessing their ability to engage with these technologies effectively and for the design of AI literacy training. Research carried out by Chan and Zhou (2023) on undergraduate and postgraduate students indicated that while students generally possess a basic understanding of GenAI applications and impacts, there is a significant gap in deeper technical knowledge and awareness of ethical implications. Educational interventions (e.g. online workshops) carried out by Putjorn and Putjorn (2023), emphasise the importance of integrating AI literacy into curricula to empower students with comprehensive knowledge of GenAI technologies which in turn prepares students for the modern job market where use of GenAI tools is becoming commonplace.

By studying people’s attitudes, we can better explain the decision-making and behaviour of individuals and communities, and create a supportive environment for responsible use (Cao et al., 2021). Factors that influence AI attitudes have been studied from demographic, personality, anxiety, and trust perspectives, with men reporting more positive attitudes towards AI (Liang and Lee, 2017, Schepman and Rodway, 2022). Regarding age and AI attitudes, the research results are contradictory (Kaya et al., 2022). However, most of the literature states that younger age is connected to more positive attitudes towards AI (Gillespie et al., 2021, Schepman and Rodway, 2022). Higher education has also been shown to relate to positive AI attitudes (European Commission, & Directorate-General for Communications Networks, Content and Technology, 2017, Neudert et al., 2020). However, a more comprehensive understanding of attitudes towards *Gen*AI among students will provide policy makers and educators with key insights needed to support student learning.

Research regarding students’ usage of AI is still emerging. Johnston et al. (2024) conducted a focus group where half of the students reported using or considering using GenAI for academic purposes. Although most students were supportive of using GenAI for grammar and spelling, most were unsupportive of the use of GenAI for assessment writing. Likewise, qualitative responses highlighted that students were unsupportive towards using GenAI for essay writing, as this was considered “cheating”, but GenAI could be used as an alternative to lecturers or to understand a concept. Smolansky et al. (2023) reported ‘moderate usage’ of GenAI tools among students for assignments and assessments relating to essay writing and coding. Findings from Smolansky et al. (2023) also highlighted concerns among students and educators regarding academic integrity and the use of GenAI in traditional assessments such as essay writing.

Currently, there is no comprehensive, validated and freely available tool specifically designed to measure students' knowledge, attitudes, and usage of GenAI. Indeed, knowledge, attitudes and usage are useful dimensions for measuring programs, products and technologies because understanding knowledge gaps, usage patterns, and attitudinal barriers helps in designing targeted interventions, policies, and educational programs. The creation of a questionnaire using these dimensions and tailored to GenAI will fill a measurement gap, providing a robust scale for future research and ongoing assessment. This scale will enable consistent tracking of changes over time and the effectiveness of interventions aimed at improving GenAI literacy and engagement. In conclusion, this study aims to fill the gap in the existing literature by developing and validating a scale to measure students' knowledge, attitudes, and usage of GenAI, and by exploring the factors that influence these dimensions. The findings will provide valuable insights for educators and policymakers to design GenAI education programs that are responsive to students' needs and concerns.

References Allam, H., Dempere, J., Akre, V., Parakash, D., Mazher, N., & Ahamed, J. (2023, May). Artificial intelligence in education: an argument of Chat-GPT use in education. In 2023 9th International Conference on Information Technology Trends (ITT) (pp. 151-156). IEEE. Bergdahl, J., Latikka, R., Celuch, M., Savolainen, I., Mantere, E. S., Savela, N., & Oksanen, A. (2023). Self-determination and attitudes toward artificial intelligence: Cross-national and longitudinal perspectives. Telematics and Informatics, 82, 102013. Cao, G., Duan, Y., Edwards, J. S., & Dwivedi, Y. K. (2021). Understanding managers’ attitudes and behavioral intentions towards using artificial intelligence for organizational decision-making. Technovation, 106, 102312. Chan, C. K. Y., & Zhou, W. (2023). Deconstructing student perceptions of generative AI (GenAI) through an expectancy value theory (EVT)-based instrument. arXiv preprint arXiv:2305.01186. Cotton, D. R., Cotton, P. A., & Shipway, J. R. (2024). Chatting and cheating: Ensuring academic integrity in the era of ChatGPT. Innovations in education and teaching international, 61(2), 228-239. Eurobarometer, S. (2017). 460-Attitudes Towards the Impact of Digitisation and Automation on Daily Life. European Commission, Brussels, EU. Francis, N. J., Jones, S., & Smith, D. P. (2025). Generative AI in higher education: Balancing innovation and integrity. British Journal of Biomedical Science, 81, 14048. Freeman, J. (2025). Student Generative AI Survey 2025. Technical report, HEPI, URL https://www. hepi. ac. uk/2025/02/26/student-generative-ai-survey-2025. Gillespie, N., Lockey, S., & Curtis, C. (2021). Trust in artificial intelligence: A five country study. Javaid, M., Haleem, A., Singh, R. P., Khan, S., & Khan, I. H. (2023). Unlocking the opportunities through ChatGPT Tool towards ameliorating the education system. BenchCouncil Transactions on Benchmarks, Standards and Evaluations, 3(2), 100115. Johnston, H., Wells, R. F., Shanks, E. M., Boey, T., & Parsons, B. N. (2024). Student perspectives on the use of generative artificial intelligence technologies in higher education. International Journal for Educational Integrity, 20(1), 2. Liang, Y., & Lee, S. A. (2017). Fear of autonomous robots and artificial intelligence: Evidence from national representative data with probability sampling. International Journal of Social Robotics, 9, 379-384. Neudert, L. M., Knuutila, A., & Howard, P. N. (2020). Global attitudes towards AI, machine learning & automated decision making. Working paper 2020.10, Oxford Commission on AI & Good Governance. https://oxcaigg. oii. ox. ac. uk. Putjorn, T., & Putjorn, P. (2023, October). Augmented Imagination: Exploring Generative AI from the Perspectives of Young Learners. In 2023 15th International Conference on Information Technology and Electrical Engineering (ICITEE) (pp. 353-358). IEEE. Schepman, A., & Rodway, P. (2023). The General Attitudes towards Artificial Intelligence Scale (GAAIS): Confirmatory validation and associations with personality, corporate distrust, and general trust. International Journal of Human–Computer Interaction, 39(13), 2724-2741. Smolansky, A., Cram, A., Raduescu, C., Zeivots, S., Huber, E., & Kizilcec, R. F. (2023, July). Educator and student perspectives on the impact of generative AI on assessments in higher education. In Proceedings of the tenth ACM conference on Learning@ Scale (pp. 378-382) Stokel-Walker, C. (2022). AI bot ChatGPT writes smart essays-should academics worry?. Nature. Tillmanns, T., Salomão Filho, A., Rudra, S., Weber, P., Dawitz, J., Wiersma, E., ... & Reynolds, S. (2025). Mapping Tomorrow’s Teaching and Learning Spaces: A Systematic Review on GenAI in Higher Education. Trends in Higher Education, 4(1), 2. Yu, H. (2023). Reflection on whether Chat GPT should be banned by academia from the perspective of education and teaching. Frontiers in Psychology, 14, 1181712.

In March 2025, ChatGPT.com received approximately *** billion visits from users worldwide. The most recent year under analysis has seen an increase in traffic to OpenAI's artificial intelligence chatbot. This is the highest traffic volume achieved by the site to date, with values for the most recent analyzed month exceeding twice the average monthly visits for the entire examined period between April 2023 and April 2024.

Background: Conversational agents based on large language models (LLMs) have shown moderate efficacy in reducing depressive and anxiety symptoms. However, most existing evaluations lack methodological transparency, rely on closed-source models, and show limited standardization in performance and safety assessment.Objective: We have two study objectives: (1) to develop an LLM-based conversational agent through system design analysis and initial functionality testing, and (2) to evaluate its safety and performance through standardized assessment in controlled simulated interactions focused on depression and anxiety of two LLMs (GPT-4o and Llama 3.1-8B).Methods: We conducted a cross-sectional study in two phases. First, we developed a mental health platform integrating a conversational agent with functionalities including personalized context, pretrained therapeutic modules, self-assessment tools, and an emergency alert system. Second, we evaluated the agent’s responses in simulated interactions based on predefined user personas for each LLM. Four expert raters assessed 816 interaction pairs using a 5-criterion Likert scale evaluating tone, clarity, domain accuracy (correctness), robustness, completeness, boundaries, target language, and safety. In addition, we use quantitative performance metrics such as cost, response length, and number of tokens. Multiple linear regression models were used to compare LLM performance and assess metric interrelations.Results: First, we developed a web-based mental health platform using a user-centered design, structured into frontend, backend, and database layers. The system integrates therapeutic chat (GPT-4o and Llama 3.1-8B), psychological assessments (PHQ-9, GAD-7), CBT-based tasks, and an emergency alert system. The platform supports secure user authentication, data encryption, multilingual access, and session tracking. Second, GPT-4o outperformed Llama 3.1-8B in both quantitative and qualitative metrics, generating longer and more lexically diverse responses, using more tokens, and scoring higher in clarity, robustness, completeness, boundaries, and target language. However, it incurred higher costs, with no significant differences in tone, accuracy, or safety.Conclusion: Our study presents a conversational agent with multiple functionalities and shows that GPT-4o outperforms Llama 3.1-8B in performance, although at a higher cost. This platform could be used in future clinical trials or real-world implementation studies.

This dataset is separated into three parts. The text and images in all documents have been manually copied from the OpenAI webpage, and the formating has been recreated to the greatest extent.

1. DATASET1: Conversation history from Scenario Ontology Generation with OpenAI GPT-4 This dataset contains the conversation history where OpenAI GPT-4 is asked to analyze a text, is asked to create an ontology and then visualize it in a graph. The dataset contains all user input and the OpenAI GPT-4 output, including the graphical representations of the ontology. The dataset also contains a JSON file that was generated that describes the ontology.

The document contains a conversation log with OpenAI GPT-4 where a text is analyzed by the AI and an ontology graph and a JSON file describing the ontology is created.

1. DATASET2: Conversation history from Scenario Generation with OpenAI GPT-4 This dataset contains the conversation history where OpenAI GPT-4 is asked to generate a hypothetical scenario for the development of a new hypothetical multi-role tactical airlift. The dataset contains all user input and the OpenAI GPT-4 output, including the graphical representations of the ontology.

The document contains conversation log with OpenAI GPT-4 where a hypothetical scenario is generated.

1. DATASET3: CHAMP Operational Scenario generated from OpenAI GPT-4 OpenAI GPT-4 was used to generate a hypothetical scenario for the development of a hypothetical new multi-role airlift.

The document contains a hypothetical scenario that has been generated by OpenAI GPT-4

The dataset was originally published in DiVA and moved to SND in 2024.

Market Overview The global AI Skincare Advisor market is projected to reach a valuation of XXX million by 2033, exhibiting a CAGR of XX% during the forecast period. The increasing demand for personalized skincare recommendations and the growing adoption of AI technologies are driving this growth. These advisors use advanced algorithms to analyze an individual's skin type, concerns, and usage patterns, providing tailored skincare routines that enhance efficacy and minimize irritation. Market Dynamics Key drivers of the market include the rising prevalence of skin concerns, the increasing disposable income for skincare products, and the growing awareness of AI's benefits in skincare. Trends such as the integration of telehealth platforms and the rise of smart beauty devices are further contributing to market expansion. However, restraints like the high cost of AI technology and regulatory concerns regarding data privacy pose challenges. The market is segmented by type (online use, mobile app), application (beauty salon, home use), and region (North America, Europe, Asia Pacific). Major players in the market include Revieve, SKINMART, KIKO, Haut.AI, and Chat Gpt.

This dataset contains the conversation history where OpenAI GPT-4 is asked to generate a hypothetical scenario for the development of a new hypothetical multi-role tactical airlift. The dataset contains all user input and the OpenAI GPT-4 output, including the graphical representations of the ontology. The generated word-file, containing the generated scenario, is stored in a separate dataset: "CHAMP Operational Scenario generated from OpenAI GPT-4"

Document containing conversation log with OpenAI GPT-4 where a hypothetical scenario is generated. The text and images have been manually copied from the OpenAI webpage, and the formating has been recreated to the greatest extent

License: Creative Commons Attribution 4.0 International

The dataset was originally published in DiVA and moved to SND in 2024.

This study aims to explore students' associations with Artificial Intelligence (AI) and how these perceptions have evolved following the release of Chat GPT. A free word association test was conducted with 836 German high school students aged 10–20. Associations were collected before and after the release of Chat GPT, processed, cleaned, and inductively categorized into nine groups: technical association, assistance system, future, human, negative, positive, artificial, others, and no association. In total, 355 distinct terms were mentioned, with “robot” emerging as the most frequently cited, followed by “computer” and “Chat GPT,” indicating a strong connection between AI and technological applications. The release of Chat GPT had a significant impact on students' associations, with a marked increase in mentions of Chat GPT and related assistance systems, such as Siri and Snapchat AI. The results reveal a shift in students' perception of AI-from abstract, futuristic concepts to more immediate, application-based associations. Network analysis further demonstrated how terms were semantically clustered, emphasizing the prominence of assistance systems in students' conceptions. The findings underscore the importance of integrating AI education that fosters both critical reflection and practical understanding of AI, encouraging responsible engagement with the technology. These insights are crucial for shaping the future of AI literacy in schools and universities.

WildVision-Chat

WildVisvion-Chat is the publicly released chat data collected from WildVision-Arena. We hope our released data can promote the development of a multimodal language model.

  Models

the WildVision datasets contain user conversations with PaliGemma, GPT-4T, GPT-4o, Phi 3 vision, Gemini 1.5, Neva 22b, Claude 3 Haiku, Idefics2-8b, Qwen-VL plus, Claude 3.5 Sonnet, Qwen-VL max, Yi-VL plus, MiniCPM LLama3, Claude 3 Sonnet, Claude 3 Opus, and GPT-4 Vision preview.… See the full description on the dataset page: https://huggingface.co/datasets/WildVision/wildvision-chat.

Summary of Artifacts

This is the replication package for the paper titled 'Can Developers Prompt? A Controlled Experiment for Code Documentation Generation' that is part of the 40th IEEE International Conference on Software Maintenance and Evolution (ICSME), from October 6 to 11, 2024, located in Flagstaff, AZ, USA.

Full Abstract

Large language models (LLMs) bear great potential for automating tedious development tasks such as creating and maintaining code documentation. However, it is unclear to what extent developers can effectively prompt LLMs to create concise and useful documentation. We report on a controlled experiment with 20 professionals and 30 computer science students tasked with code documentation generation for two Python functions. The experimental group freely entered ad-hoc prompts in a ChatGPT-like extension of Visual Studio Code, while the control group executed a predefined few-shot prompt. Our results reveal that professionals and students were unaware of or unable to apply prompt engineering techniques. Especially students perceived the documentation produced from ad-hoc prompts as significantly less readable, less concise, and less helpful than documentation from prepared prompts. Some professionals produced higher quality documentation by just including the keyword Docstring in their ad-hoc prompts. While students desired more support in formulating prompts, professionals appreciated the flexibility of ad-hoc prompting. Participants in both groups rarely assessed the output as perfect. Instead, they understood the tools as support to iteratively refine the documentation. Further research is needed to understand which prompting skills and preferences developers have and which support they need for certain tasks.

Author Information

Citation Information

@inproceedings{kruse-icsme-2024,
author={Kruse, Hans-Alexander and Puhlf{\"u}r{\ss}, Tim and Maalej, Walid},
booktitle={2022 IEEE International Conference on Software Maintenance and Evolution},
title={Can Developers Prompt? A Controlled Experiment for Code Documentation Generation},
year={2024},
doi={tba},
}

Artifacts Overview

1. Preprint

The file kruse-icsme-2024-preprint.pdf is the preprint version of the official paper. You should read the paper in detail to understand the study, especially its methodology and results.

2. Results

The folder results includes two subfolders, explained in the following.

Demographics RQ1 RQ2

The subfolder Demographics RQ1 RQ2 provides Jupyter Notebook file evaluation.ipynb for analyzing (1) the experiment participants' submissions of the digital survey and (2) the ad-hoc prompts that the experimental group entered into their tool. Hence, this file provides demographic information about the participants and results for the research questions 1 and 2. Please refer to the README file inside this subfolder for installation steps of the Jupyter Notebook file.

RQ2

The subfolder RQ2 contains further subfolders with Microsoft Excel files specific to the results of research question 2:

• The subfolder UEQ contains three times the official User Experience Questionnaire (UEQ) analysis Excel tool, with data entered from all participants/students/professionals.
• The subfolder Open Coding contains three Excel files with the open-coding results for the free-text answers that participants could enter at the end of the survey to state additional positive and negative comments about their experience during the experiment. The Consensus file provides the finalized version of the open coding process.

3. Extension

The folder extension contains the code of the Visual Studio Code (VS Code) extension developed in this study to generate code documentation with predefined prompts. Please refer to the README file inside the folder for installation steps. Alternatively, you can install the deployed version of this tool, called Code Docs AI, via the https://marketplace.visualstudio.com/items?itemName=re-devtools.code-docs-ai" href="https://marketplace.visualstudio.com/items?itemName=re-devtools.code-docs-ai">VS Code Marketplace.

You can install the tool to generate code documentation with ad-hoc prompts directly via the https://marketplace.visualstudio.com/items?itemName=zhang-renyang.chat-gpt" href="https://marketplace.visualstudio.com/items?itemName=zhang-renyang.chat-gpt">VS Code Marketplace. We did not include the code of this extension in this replication package due to license conflicts (GNUv3 vs. MIT).

4. Survey

The folder survey contains PDFs of the digital survey in two versions:

• The file Survey.pdf contains the rendered version of the survey (how it was presented to participants).
• The file SurveyOptions.pdf is an export of the LimeSurvey web platform. Its main purpose is to provide the technical answer codes, e.g., AO01 and AO02, that refer to the rendered answer texts, e.g., Yes and No. This can help you if you want to analyze the CSV files inside the results folder (instead of using the Jupyter Notebook file), as the CSVs contain the answer codes, not the answer texts. Please note that an export issue caused page 9 to be almost blank. However, this problem is negligible as the question on this page only contained one free-text answer field.

5. Appendix

The folder appendix provides additional material about the study:

• The subfolder tool_screenshots contains screenshots of both tools.
• The file few_shots.txt lists the few shots used for the predefined prompt tool.
• The file test_functions.py lists the functions used in the experiment.

Revisions

License

See LICENSE file.

Artificial intelligence is a computer system which can replicate human behavior and largely supports human actions and interpretation, but not replace human responses. Over the past few decades, the field of artificial intelligence (AI) has experienced phenomenal development and expansion. We are surrounded by several instances of AI. The most typical examples include Chat GPT, Alexa, Google Maps, Autocorrect and Text Editors, e-payments, Virtual Travel Booking Agent, Social Media Monitoring, gaming, including chess matches involving computers versus human chess masters, Self driving Cars, Adaptive Cruise Control, Parking Assistance, and Facial Recognition for Biometrics such as Retinal Scans and Fingerprint scans. AI has application in different branches of Dentistry . The future of AI in dentistry seems to be very promising with developments in Comprehensive AI technologies, AI dental assistance and advancements in dental educational tools with the incorporation of Artificial Intelligence. Artificial Intelligence is still in its introductory stages in Dentistry and lot of research must be done in this direction. This review article attempts to highlight these points and lays an emphasis on how AI is driving dentistry in the present and will improve dental care in the future.

Objective: To examine the potential of artificial intelligence (AI) in gynecologic oncology decision making. Design: Feasibility study. Setting: Fictive. Participants: Fictitious case vignettes of gynecologic carcinomas. Methods: Fictitious case vignettes of gynecologic carcinomas were created and evaluated by physicians with varying levels of professional experience, as well as by language models including Chat-GPT 4.0, Google Gemini, and Bing-Copilot. Treatment approval decisions were based on standardized clinical and laboratory criteria. Results: Two cases of breast cancer, one case of ovarian cancer, one case of cervical cancer and one case of endometrial cancer were evaluated. All three language models were able to evaluate all clinical cases and make therapy-relevant suggestions, with Chat-GPT providing the most clear and concise recommendations that were in three cases totally consistent with physician assessments. Conclusions: The study demonstrates that AI models, such as Chat-GPT, can to some extent evaluate clinical cases, recognize clinical and/or laboratory abnormalities and make therapy-related suggestions. Despite high overall agreement, differences were predominantly noted in the more complex cases, rendering human interpretation necessary. The findings underscore the benefits of AI in terms of clarity, time efficiency, and cost-effectiveness. Future research should further explore the application of AI to real patient data and development of hybrid decision models to optimize integration into clinical practice. Limitations: Feasibility study with five fictitious case vignettes.

Comparison of Seconds to Output 500 Tokens, including reasoning model 'thinking' time; Lower is better by Model

Comparison of Represents the average of coding benchmarks in the Artificial Analysis Intelligence Index (LiveCodeBench & SciCode) by Model

Comprehensive comparison of Latency (Time to First Token) vs. Output Speed (Output Tokens per Second) by Model

Comprehensive comparison of Artificial Analysis Intelligence Index vs. Seconds to Output 500 Tokens, including reasoning model 'thinking' time by Model