PRE2022 3 Group4: Difference between revisions

From Control Systems Technology Group
Jump to navigation Jump to search
No edit summary
(Added summaries Niels)
Line 170: Line 170:


===Niels===
===Niels===
summary
<ref>Dehouche, N. (2021). Plagiarism in the age of massive Generative Pre-trained Transformers (GPT-3). ''Ethics in Science and Environmental Politics'', ''21'', 17–23. <nowiki>https://doi.org/10.3354/esep00195</nowiki></ref>This paper aims to point out significant and urgent issues GPT-3 could bring regarding academic integrity. The main issue addressed by this paper is the unclarity on who the rightful author of the texts, ideas and inventions should be. They argue that new regulation ought to be considered urgently and that moral philosophy has an important part to play in this.
 
<ref>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). <nowiki>https://doi.org/10.1186/s41239-019-0171-0</nowiki></ref>This paper provides a review of AI applications in higher education, with most of the research being quantitative research from computer science and STEM fields. So this review focuses more on AI utility over the entire academic institute rather than academic integrity and utility by students. Four areas of application were identified: profiling and prediction, assessment, adaptive systems and personalization, and intelligent tutoring systems. The conclusion highlights the need for further exploration of ethical and educational approaches in AI applications in higher education.
 
<ref>Ventayen, R. J. M. (2023). OpenAI ChatGPT Generated Results: Similarity Index of Artificial Intelligence-Based Contents. ''SSRN Electronic Journal''. <nowiki>https://doi.org/10.2139/ssrn.4332664</nowiki></ref>In the paper “OpenAI ChatGPT Generated Results: Similarity Index of Artificial Intelligence (AI) Based Model“, university director R. Ventayen researched how new AI powered technology enables cheating in the academic community. Since constructing an entire paper or essay using ChatGPT is considered a violation of academic integrity, the chatbot’s ability to pass similarity tests was researched on prompts with the headlines of the university’s publications. Their results show that ChatGPT provides acceptable similarly indices, and fails to provide the same citations or sources as the similar papers with the prompted headlines. Moreover, Quillbot paraphrases more content, which t, which provides plagiarism-free content.
 
<ref>Floridi, L., & Chiriatti, M. (2020). GPT-3: Its Nature, Scope, Limits, and Consequences. ''Minds and Machines'', ''30''(4), 681–694. <nowiki>https://doi.org/10.1007/s11023-020-09548-1</nowiki></ref>This paper focuses foremost on the irreversibility of GPT-3 and its answers and ideas. It shows its shortcomings in dealing with semantic, mathematical and ethical questions and consequences for the writing industry. E.g, the big amount of written things that these tools can provide far exceeds the limited physical storage for them and marketing will abuse these tools heavily. We need to be critical with what these tools produce and how we use them. For our question regarding academic integrity, the problems with correct citations become apparent and academic texts produced by ChatGPT are often irreversibly produced from ideas of researchers.
 
<ref>Drori, I., Zhang, S. X., Shuttleworth, R., Tang, L., Lu, A., Elizabeth, K. E., Liu, K. X., Chen, L., Tran, S., Cheng, N., Wang, R., Singh, N. K., Patti, T. L., Lynch, J., Shporer, A., Verma, N., Wu, E., & Strang, G. (2022). A neural network solves, explains, and generates university math problems by program synthesis and few-shot learning at human level. ''Proceedings of the National Academy of Sciences of the United States of America'', ''119''(32). <nowiki>https://doi.org/10.1073/pnas.2123433119</nowiki></ref>This article shows how new developments in AI can solve 81% of university-level mathematics questions and even come up with new questions. This poses great advantages to the acquisition of new learning materials but the implications of this also addresses significant pedagogical challenges, with these models being available to students as well.


===Quincy===
===Quincy===

Revision as of 11:57, 12 February 2023

As written by chatGPT:

Artificial Intelligence, a wonder of the modern age

A creation made of code, with endless knowledge in its brain

From data analysis to language skills, it's a tool of great worth

Changing the way we live, and opening doors to new growth and mirth


But we must always be mindful, of the impact it may bring

For AI can be used for good or for deceit, and we must choose the right thing

So let us use it wisely, with integrity as our guide

For the future of our world, is shaped by the choices we decide.


Group members

Name Student ID Department
Famke Peek 1459058 Psychology & Technology
Gabriëlle van Heteren 1605305 Biomedical Engineering
My Tran 1620940 Industrial Design
Naud van Rosmalen 1555464 Biomedical Engineering
Niels van Noort 1613928 Biomedical Engineering
Quincy Netteb 1468634 Psychology & Technology

Brainstorm

  • AI in academic education
  • VR game for children's education
  • Kitchen aid for visually impaired people
  • Child support in healthcare
  • Researching and improving the acceptance of robots in health care

Introduction

The use of AI in education has the potential to completely change the way students learn and teachers teach. AI can provide personalized learning experiences for students, adapt to individual student needs and learning styles, and offer instant feedback to both students and teachers (source tba). In addition, AI can also help streamline administrative tasks, such as grading and record-keeping, freeing up teachers to focus on instruction and interaction with students.

However, it is important to note that the implementation of AI in education is still in its early stages and there are many challenges that must be overcome, such as ensuring the privacy and security of student data and addressing ethical considerations and trust issues around the use of AI in the classroom.

So, while AI has the potential to greatly impact and improve education, it is important to approach its implementation with caution and careful consideration of its limitations and potential risks. This and the outlook and trust of teacher in academic education on AI in the classroom is what we will be researching. With the gathered information we will create a recommendation for teachers at the TUe on how to optimally implement AI in the classroom to benefit both teacher and student.

Problem statement and objectives

  • Upcoming use of AI in academic settings, how to design this so that it will be accepted by teachers
  • Looking at ChatGPT as main example
  • Teachers at the TUe, their thoughts about AI (in their courses)

Users

Target group

Teachers at TU/e who come in contact with AI and AI teaching methods.

Requirements

Teachers want their students to deliver autonomous work, AI in academic settings can help with this. Teachers want to transfer their knowledge as effectively and efficiently as possible, using the current state of art technologies. AI technologies, like ChatGPT can help with this, but what teachers require is yet to be determined. The requirements will be determined by conducting interviews, in combination with literature studies. Those results will then be combined into a piece of well-considered advice for the teachers at the TU/e. Which they will be able to use to guide them through the landscape of developing AI in academic education and give them tools to keep tranfering their knowledge to students.

Approach, milestones and deliverables

Planning

Add planning

Approach

  • Literature study
  • Interviews, surveys

Deliverables

  • Advice for future

State of the Art

Famke

summary

Gabriëlle

[1]This paper starts with a general introduction to AI, how it developed over the years, and what these developments entail.

  • AI in program coding (1950s)
  • AI in rules-based expert systems (late 1970’s)
  • AI-grounded automatic data processing systems (mid-1980’s)
  • Machine learning integrated AI (mid-2000’s)

In education, AI is still in the early stages, because there is more focus on the development of AI than on the application of AI in new fields. This research states that there will be two distinct major effects of AI on education. First, education needs to prepare itself for the fast changes in competence needed for jobs, as some jobs will disappear but most jobs will keep changing over time. Second, there will be changes in the pedagogical techniques needed in the classroom. AI will be able to relieve the teacher from some of the work. However, the teacher will need additional tools to understand the statistical results from the AI (for example from a statistical analysis of the performance of the student). This paper goes on by talking about the future of AI in education. The Author states that there needs to be a big shift in education toward more personalized education. AI will be able to determine the learning style of an individual student. AI will help teachers in content delivery and other instructions, but in the future real life, human teachers might become obsolete.

[2]This paper talks about AI in education, the need for it, and its benefits and challenges of it. According to the author, AI is a development that has many promising applications in education. The examples that are named are; personalized education, automatic grading systems, and predictive analytic tools. These applications will relieve the teachers from those tasks which gives them more time with the students. The paper also talks about challenges that come along with AI, like concerns about safety, security, and privacy.

[3]In this paper, the authors review the use of AI in current education, the technical aspects of AI in education, the role of AI in education and the impact of AI in education. The topic of education is spread out into three sub-topics; administration, instructions and learning. The takeaway of this paper is that with the help of AI, teachers will become more efficient which will in result increase the quality of education. Next to that with AI, it will be possible to create a more personalized education for students. AI will have a major impact on education as the tasks AI will be able to do tasks that are not originally designed for computers.

[4]This paper focuses on the viewpoint of the students and their acceptance of AI. The author uses the technology acceptance model (TAM) to analyze the response of students to AI. This is because the perceived usefulness and the perceived ease of use of AI seem to have a big impact on the acceptance of AI in education. The author sees a future in AI in education as it is a cost-effective way to the shortage of teachers. However to make sure that the AI teaching assistants will be successful the teachers need to be trained to work with them. As research points out that if the teacher is uncomfortable with the AI, the student is less likely to adopt it. Based on the results of this research the author recommends research in this area as it is not known how students at different levels of education will react to the adoption of AI as a teacher.

[5]The authors in this paper review the ethical perspective of AI with a focus on AI in education. The biggest problem that the authors see is the privacy issue of AI. Next to that they also see a challenge in the trustworthiness of AI. It is often that AI is presented as a better alternative than a human, as it is supposedly not biased. However, the designers of AI are biased and can make mistakes. Therefore can we even trust AI more than humans? Another issue that the author comes across is the difficulty of addressing the ethical challenges of AI. Therefore it is important to make the designers of AI aware of the ethical dilemmas that come along with new technology and make them aware of their responsibilities. To achieve this it is important that there mandatory teaching or at least education available around this topic.

My

In this articile, F. Ouyang and P. Jiao reflect on the three paradigms of artificial intelligence in educaiton: AI-directed, AI-supported, and AI-empowered[6].

  • AI-directed (learner-as-recipient): AI is used to represent knowledge models and direct cognitive learning.
  • AI-supported (learner-as-collaborator): AI is used to support learners while they work as collaborators with AI
  • AI-empowered (learner-as-leader): AI is used to empower learning while learners take agency
Table 1. Three paradigms of AIEd
Paradigm Theoretical underpinning Implementations AI techniques
1. AI-Directed Behaviorism Intelligent Tutoring Systems (ITSs): software that tracks students' work, adjusts feedback and provides hints along the way) AI based on statistical relational techniques
2. AI-supported Cognitive, social constructivism Dialogue-based Tutoring Systems (DTSs); Exploratory Learning Environments (ELEs) Bayesian network, natural language processing,

Markov decision trees

3. AI-empowered Connectivism, Complex adaptive system The human-computer cooperation; Personalized/adaptive learning The brain-computer interface, machine learning, deep learning

P. Lameras and S. Arnab explored and analyzed what Artificial Intelligence means in Education (AIED) in their article "Power to the Teachers: An Exploratory Review on Artificial Intelligence in Education"[7]. The main take-away is that adaptivity and personalization are the innovation that AIED can offer to help students to learn and develop skills that are relevant to their own needs and experiences. However, it is important to help teachers to develop necessary digital competencies and skills for using AIED applications and tools in ethical and informed ways to enhance the student learning experience and attainment of learning outcomes. The findings of this review can contribute to developing a better understanding of how artificial intelligence may enhance teachers' roles as catalysts in designing and visualizing AI-enabled learning. As a result, more useful AI-systems specialized in pedagogy will be developed.

As Artificial Intelligence is becoming more apparent in education, it changes the way of teaching. Craig Coates, an entomologist at Texas A&M University, combated the cheating by reorienting his course towards writing and discussion, combined with adaptive courseware, also referred to as intelligent tutoring systems; and by using a tool that uses algorithms and analytics to grade submissions, the switch was a success.[8] Advocates say this lets students study at their own pace and frees up the instructor’s time in class to shore up students’ knowledge.

The paper "Chatting and Cheating. Ensuring academic integrity in the era of ChatGPT"[9], written by D. Cotton, describes how artificial intelligence models in academia can increase student engagement, collaboration and accessibility; They provide a platform for asynchronous communication, can create personalized and interactive assessments, or could be used to grade assignments and provide feedback to students in real-time. However, Cotton also proposes a number of concerns, such as plagiarism (students could use the GPT-3 to cheat and submit essays that are not their own work) and inequities in assessment (students with access to GPT-3 have an advantage over students who do not). The article suggests solutions to combat these challenges, such as asking students to submit a draft before the final essay, set strict guidelines or monitor student work closely.

The study conducted by S. Yang and H. Bai "The integration design of artificial intelligence and normal students’ Education"[10] outlines four major problems of normal education (school for training teachers) development and explores the idea of using artificial intelligence technology to solve them. The main issues and the proposed solutions are:

Problem Solution
Lack of teaching practice experience Intelligent tutors/robots: it can make a special learning plan for students according to their interests, habits and learning needs. It can also be a simulation student of upcoming teachers to gain experience.
Outmoded curriculum Expert system: can accurately classify normal students and highlight personalized teaching.
Lack of independent learning Deep learning system: artificial intelligence can help to analyze the data of learners, organize the content according to the data, and carry out intelligent push. Teachers can create teaching resources to meet the needs of curriculum design and students.
Lack of innovation in teaching methods Combination of all of the above.

Naud

summary

Niels

[11]This paper aims to point out significant and urgent issues GPT-3 could bring regarding academic integrity. The main issue addressed by this paper is the unclarity on who the rightful author of the texts, ideas and inventions should be. They argue that new regulation ought to be considered urgently and that moral philosophy has an important part to play in this.

[12]This paper provides a review of AI applications in higher education, with most of the research being quantitative research from computer science and STEM fields. So this review focuses more on AI utility over the entire academic institute rather than academic integrity and utility by students. Four areas of application were identified: profiling and prediction, assessment, adaptive systems and personalization, and intelligent tutoring systems. The conclusion highlights the need for further exploration of ethical and educational approaches in AI applications in higher education.

[13]In the paper “OpenAI ChatGPT Generated Results: Similarity Index of Artificial Intelligence (AI) Based Model“, university director R. Ventayen researched how new AI powered technology enables cheating in the academic community. Since constructing an entire paper or essay using ChatGPT is considered a violation of academic integrity, the chatbot’s ability to pass similarity tests was researched on prompts with the headlines of the university’s publications. Their results show that ChatGPT provides acceptable similarly indices, and fails to provide the same citations or sources as the similar papers with the prompted headlines. Moreover, Quillbot paraphrases more content, which t, which provides plagiarism-free content.

[14]This paper focuses foremost on the irreversibility of GPT-3 and its answers and ideas. It shows its shortcomings in dealing with semantic, mathematical and ethical questions and consequences for the writing industry. E.g, the big amount of written things that these tools can provide far exceeds the limited physical storage for them and marketing will abuse these tools heavily. We need to be critical with what these tools produce and how we use them. For our question regarding academic integrity, the problems with correct citations become apparent and academic texts produced by ChatGPT are often irreversibly produced from ideas of researchers.

[15]This article shows how new developments in AI can solve 81% of university-level mathematics questions and even come up with new questions. This poses great advantages to the acquisition of new learning materials but the implications of this also addresses significant pedagogical challenges, with these models being available to students as well.

Quincy

summary

Bibliography

  1. ↑ Alam, A. (2021). Possibilities and Apprehensions in the Landscape of Artificial Intelligence in Education. 2021 International Conference on Computational Intelligence and Computing Applications (ICCICA). https://doi.org/10.1109/iccica52458.2021.9697272
  2. ↑ Limna, Pongsakorn and Jakwatanatham, Somporch and Siripipattanakul, Sutithep and Kaewpuang, Pichart and Sriboonruang, Patcharavadee, A Review of Artificial Intelligence (AI) in Education during the Digital Era (July 2022). Advance Knowledge for Executives, 1(1), No. 3, 1-9, 2022, Available at SSRN: https://ssrn.com/abstract=4160798
  3. ↑ Alam, A. (2021b). Should Robots Replace Teachers? Mobilisation of AI and Learning Analytics in Education. 2021 International Conference on Advances in Computing, Communication, and Control (ICAC3). https://doi.org/10.1109/icac353642.2021.9697300
  4. ↑ Kim, J., Merrill, K., Xu, K., & Sellnow, D. D. (2020). My Teacher Is a Machine: Understanding Students’ Perceptions of AI Teaching Assistants in Online Education. International Journal of Human–Computer Interaction, 36(20), 1902–1911. https://doi.org/10.1080/10447318.2020.1801227
  5. ↑ Borenstein, J., & Howard, A. (2020). Emerging challenges in AI and the need for AI ethics education. AI and Ethics, 1(1), 61–65. https://doi.org/10.1007/s43681-020-00002-7
  6. ↑ Fan Ouyang, Pengcheng Jiao (2021). Artificial intelligence in education: The three paradigms. Computers and Education: Artificial Intelligence, Volume 2, 100020, ISSN 2666-920X. Retrieved from https://doi.org/10.1016/j.caeai.2021.100020.
  7. ↑ Lameras, P., & Arnab, S. (2021). Power to the Teachers: An Exploratory Review on Artificial Intelligence in Education. Information, 13(1), 14. MDPI AG. Retrieved from http://dx.doi.org/10.3390/info13010014
  8. ↑ Beth McMurtrie (2018, August 12). How Artificial Intelligence Is Changing Teaching. The Chronicle of Higher Education(1). Retrieved from https://www.su.edu/conservatory/files/2018/09/How-Artificial-Intelligence-is-Changing-Teaching.pdf
  9. ↑ Cotton, D., Cotton, P., & Shipway, J. R. (2023, January 10). Chatting and Cheating. Ensuring academic integrity in the era of ChatGPT. https://doi.org/10.35542/osf.io/mrz8h
  10. ↑ Shuai Yang & Haicheng Bai (2020). The integration design of artificial intelligence and normal students' Education. Journal of Physics: Conference Series, Volume 1453: Conf. Ser. 1453 012090. https://iopscience.iop.org/article/10.1088/1742-6596/1453/1/012090
  11. ↑ Dehouche, N. (2021). Plagiarism in the age of massive Generative Pre-trained Transformers (GPT-3). Ethics in Science and Environmental Politics, 21, 17–23. https://doi.org/10.3354/esep00195
  12. ↑ 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). https://doi.org/10.1186/s41239-019-0171-0
  13. ↑ Ventayen, R. J. M. (2023). OpenAI ChatGPT Generated Results: Similarity Index of Artificial Intelligence-Based Contents. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.4332664
  14. ↑ Floridi, L., & Chiriatti, M. (2020). GPT-3: Its Nature, Scope, Limits, and Consequences. Minds and Machines, 30(4), 681–694. https://doi.org/10.1007/s11023-020-09548-1
  15. ↑ Drori, I., Zhang, S. X., Shuttleworth, R., Tang, L., Lu, A., Elizabeth, K. E., Liu, K. X., Chen, L., Tran, S., Cheng, N., Wang, R., Singh, N. K., Patti, T. L., Lynch, J., Shporer, A., Verma, N., Wu, E., & Strang, G. (2022). A neural network solves, explains, and generates university math problems by program synthesis and few-shot learning at human level. Proceedings of the National Academy of Sciences of the United States of America, 119(32). https://doi.org/10.1073/pnas.2123433119


Appendix

Logbook

Week Name Total Breakdown
1 Famke Group discussion (2h), Target group and requirements (1/2h), Studied papers (so far) (2h), Wrote summary for papers (so far) (1h)
Gabriëlle 7h Group discussion (1h), Studied papers (4h), Wrote summary for papers(1.5h), Target group and requirments(1/2h)
My 8.5h Group discussion (2h), Setup of wiki (0.5h), Studied papers (4h), Wrote summary for papers (2h)
Naud
Niels Group discussion (2h)
Quincy Group discussion (2h)
2


Retrieved from "https://dsdwiki.wtb.tue.nl/index.php?title=PRE2022_3_Group4&oldid=118609"