Kochi University of Technology Academic Resource Repository

武内, 一晴

Japanese

皮膚兎錯覚現象を応用した 触覚方位知覚の検討

Paul Daniels

142
The Impact of AI on Modern Education
Paul Daniels 
(Received: May 31st, 2024) 
Department of Core Studies, Kochi University of Technology
185 Miyanokuchi, Tosayamada, Kami City, Kochi 782-8502, JAPAN
E-mail: daniels.paul@kochi-tech.ac.jp
Abstract: Artificial intelligence (AI) continues to revolutionize the way we work and learn, and its 
impact on education is unfolding. Recent research sheds light on how AI can best support both 
teachers and learners. Studies on generative AI and large language models (LLMs) indicate that 
educators are optimistic about AI’s productive role in education6,2). Additionally, research in the 
language learning field has demonstrated that AI is as good as, or even better than, humans at 
assessing writing skills1). To further investigate how AI can facilitate learning in general, a 6-month 
overseas research project was conducted in the winter of 2023. This project involved interviews with 
educators at leading educational institutions in America and a quantitative study to better understand 
how generative AI can be leveraged to optimize teaching and learning.
Overview of overseas research project
This 6-month overseas research project entailed an 
in-depth analysis of how AI and speech recognition 
technology are reshaping education in both America 
and Japan. Focused interviews with educators and 
students at high schools and universities were 
conducted to identify the challenges and opportunities 
of implementing AI in educational settings. The key 
objectives of this research project are addressed in the 
three main sections of this report: (1) general 
perspectives on the implications of using AI in 
education, (2) focused research on the implementation 
of generative AI in language instruction, and (3) an 
overview of the educational affiliations fostered 
between KUT and institutions in America. The 
conclusion of this report discusses future implications 
of AI in language instruction, specifically aimed at 
improving the English education initiatives at Kochi 
University of Technology.
1. Teacher perspectives and experiences
The first objective of this project was to investigate 
attitudes and implications of using AI in education. The 
primary focus of this component of the research was to 
assess the perspectives and experiences of both 
educators and students regarding the use of generative 
AI within educational contexts. Drawing upon data 
gathered from interviews conducted with educators and 
students at several institutions in the United States, I 
attempted to outline the beliefs, attitudes, and 
experiences of teachers and learners impacted by the 
introduction of generative AI tools.
Attitudes towards AI in education: The majority of 
the educators I interviewed held positive attitudes 
towards AI and advocated that AI holds promise as a 
learning tool. At the same time, most advocated for 
clear rules and policies regarding the use of generative 
AI at their educational institutions. Most stated that 
their institutions have already initiated policy reforms 
to include the acceptable use of generative AI. They 
143
also reported drafting guidance rules for students 
specifically for AI use in their particular courses.
Another key comment regarding AI was ̒ balance.  ̓
Several teachers stressed that they intended to balance 
the use of AI during the learning process. For example, 
one teacher allowed learners to use generative AI to 
suggest topics or rewrite paragraphs, but during exams, 
students were requested to use only paper and pencil. 
AI was seen as helpful for generating ideas and offering 
edits during the writing process, but for the final 
product, students were required to generate their own 
ideas.
Educators also believed that AI has the potential to 
expand access to education, particularly with AI’s 
ability to offer personalized learning. Therefore, 
implementing generative AI activities into the 
curriculum may help reduce educational costs and 
increase learner engagement.
Educators were also aware of the implications of 
ignoring the AI revolution. While most believed that AI 
was not going to replace their jobs, they were aware 
that professionals better educated about the use of AI 
would have expanded career opportunities. Likewise, 
they felt it was paramount that students learn how to 
use AI tools to boost their future job prospects.
Although the majority of the interviewed educators 
perceived AI positively, there were also serious 
concerns. The most significant concern was the 
unreliability of AI responses. Generative AI has a 
reputation for giving false and inaccurate information, 
as well as information that is out of date. Another 
concern among educators was the ethical use of AI 
tools, although this issue can be mitigated by drafting 
acceptable use policies and being aware of the 
limitations of these AI tools. Finally, concerns were 
voiced about the readiness of faculty and students. 
Some teachers and learners have had more exposure to 
AI tools and are more familiar with their strengths and 
weaknesses, while others have only used AI at a surface 
level and do not have a comprehensive grasp of its 
diverse uses and shortcomings.
Operationalizing generative AI: The interviews 
conducted as part of this project also shed light on how 
generative AI is being utilized by educators. The most 
common use of generative AI by teachers is to assist in 
developing teaching resources. Readings, visual aids, 
transcripts, summaries, and quiz items were several of 
the more common tasks that teachers accomplished 
with the assistance of AI chatbots. Teaching materials 
that often took hours to create could be developed in 
much shorter periods. More engaging and relevant 
teaching content could also be developed by 
simplifying the administrative processes involved in 
creating instructional materials.
Teachers also felt that AI could effectively be applied 
to personalized learning experiences. The efficiency of 
AI tools allows teachers to create a wider range of 
materials that could better match learners  ̓ levels, 
preferences, and pace. More importantly, AI could be 
leveraged to enhance personal assessment and 
feedback, particularly with large class sizes, to engage 
students more actively in the learning process. A more 
efficient learning process allows teachers and students 
to focus on the learning process rather than solely on 
the outcomes.
Outside of the classroom, generative AI was also 
found to be a valuable tool for summarizing, outlining, 
and semantic searching of research documents, 
particularly when synthesizing vast amounts of past 
research data related to a current project. AI can 
improve research by providing more efficient access to 
and sorting of large data sets.
Being a language teacher, I was interested in how AI 
was specifically being leveraged for language learning 
purposes. The most exciting area where AI is making a 
difference is with speaking practice. In an EFL context 
like Japan, language learners typically cannot converse 
with English speakers in the classroom. With the 
introduction of AI chatbots and virtual tutors, learners 
can hone their speaking and listening skills using open-
ended speaking tasks. Speaking practice with an AI 
chatbot can open up new opportunities for learners who 
would normally have little or no interaction in the 
target language.
For writing improvement, AI can provide 
personalized assistance with revisions, translation, and 
grammar. An AI-based grammar assistant can compile 
lists of common writing errors by class or by student 
and then generate customized lessons or quizzes to help 
students identify and correct these errors in the future. 
Automated scoring of both speaking and writing has 
also improved significantly with the widespread 
introduction of generative AI tools. Automated scoring 
of language production, guided by AI chatbots, is a 
144
disruptive innovation gaining attention as researchers 
try to interpret the complexities of AI and automated 
scoring 4, 5). Automated writing and speech scoring 
systems are now being used by schools to automatically 
assess writing and speaking quality by replicating the 
scoring procedures that human raters employ. With the 
improved AI tools available to teachers, automated 
scoring of both speaking and writing has become 
feasible. AI-enhanced automated scoring systems have 
been found to generate scores that positively correlate 
with human graders in writing 3).
2. Generative AI and language instruction
The second objective of this research was to evaluate 
the effectiveness of generative AI in education. During 
this overseas research experience, I completed a 
qualitative project comparing grammar scores 
generated by human raters against those generated by 
AI chatbots using 152 EFL student writing samples. 
This section gives a brief overview of the study, which 
was published in TEFL-Praxis, an international 
peer-reviewed journal. The study showcases the 
potential advantages of generative AI through the 
demonstration of automated essay scoring (AES) in 
English as a Foreign Language (EFL) settings. The 
three research questions of the study are stated below.
Research Q1: How similar are AI-generated 
grammar scores to human-generated grammar scores 
using the same set of writing samples?
Research Q2: Is the inter-rater reliability between 
human raters and AI raters similar when scoring the 
grammar of L2 learners?
Research Q3: How consistent are AI chatbots among 
themselves at scoring the grammar of student writing?
The 152 writing samples were gathered from 
Japanese engineering students. Three TEFL-certified 
English teachers scored each writing sample using a 
grammar scoring rubric. The writing samples were then 
fed into three AI chatbots—ChatGPT 3.5, Google Bard 
(currently rebranded as Gemini), and Microsoft Bing 
Chat (currently rebranded as Copilot)—using the same 
grammar scoring rubric to generate grammar scores for 
the writing samples.
The mean scores of the four human raters and the 
ChatGPT-generated scores were found to be positively 
correlated, r(152) = .55, p < .001. Likewise, a positive 
correlation was observed between the human and Bing 
chatbot scores, r(152) = .55, p < .001. These results 
indicate that both the ChatGPT and Bing chatbots 
assigned grammar scores comparable to those of the 
human raters. On the other hand, a weak relationship 
was observed between the human scores and the Bard 
chatbot scores, r(152) = .26.
The full research paper with detailed results can be 
found at: https://doi.org/10.5281/zenodo.10402530.
3.  Promotion of collaboration between 
KUT and academic institutions abroad
Collaboration efforts were explored at four 
universities and two technology companies in America. 
This collaboration provided valuable insights into how 
public institutions are capitalizing on current 
technological trends. This section provides a brief 
summary of the sites that were visited.
UCONN, https://uconn.edu/
The University of Connecticut, or UCONN, is a 
leading state institution that encompasses six campuses 
and 32,000 graduate and undergraduate students. 
UCONN is actively engaged in cutting-edge research 
across various fields, including biomedical sciences, 
engineering, agriculture, and social sciences. I met with 
faculty members from various departments to discuss 
the impact of AI on teaching and learning and to 
explore the possibilities of conducting student 
exchanges. Similar to other American universities, 
UCONN has extensive undergraduate exchange 
programs compared with graduate exchange programs. 
This is partly because undergraduate students have 
more time to study abroad. Additionally, the graduate 
program in science and engineering already has a large 
international population, reducing the need for 
international exchanges. Interestingly, the College of 
Engineering at UCONN has established an 
International Engineering Program where students can 
earn both a Bachelor of Science degree in engineering 
and a Bachelor of Arts degree in a language.
William & Mary, https://www.wm.edu/
William & Mary is a smaller but competitive public 
university located in Williamsburg, Virginia. Its 
graduate schools focus on STEM research. William & 
Mary boasts the second lowest student-to-faculty ratio 
in the country, at 12 students per instructor. I 
145
interviewed a thriving professor in the biomedical field 
who is currently coordinating a research project with a 
Japanese medical university in Tokyo. As with many 
public universities in America, research funding 
predominantly comes from government grants, and 
researchers must continually apply for grants to sustain 
their work. Additionally, graduate research in science 
and engineering is mainly at the doctoral level, and 
international students represent about one-third of these 
doctoral-level researchers. Therefore, research 
collaboration is typically conducted at the doctoral 
level.
Old Dominion University, https://www.odu.edu/
Old Dominion University (ODU) currently has an 
exchange agreement with Kochi University of 
Technology (KUT). In the past, KUT graduate students 
visited ODU for research collaboration. Undergraduate 
students at ODU and KUT have also participated in an 
online language exchange. During this visit, the KUT 
international relations office organized several meetings 
with engineering teachers at ODU. As with William & 
Mary, ODU’s science and engineering graduate 
programs consist mainly of international students at the 
doctoral level. There is less focus on master-level 
research in science and engineering. ODU is also 
focusing on the integration of big data and medical 
science. Recently, ODU joined forces with Eastern 
Virginia Medical School to establish a future health 
sciences center, enhancing research collaboration 
between the science and engineering and medical 
science departments.
University of California San Diego, https://ucsd.edu/
At UCSD, I met with a professor in the School of 
Medicine to learn more about the joint efforts between 
the School of Medicine, Computational Science and 
Engineering, and Bioengineering. Like other research 
institutions in America, UCSD has identified the 
advantages of integrating science and engineering with 
other departments. There was also discussion of student 
exchanges and research opportunities between KUT 
and UCSD.
WJE Associates, Inc., https://www.wje.com/
In New York City, I met the senior vice president of 
construction at Wiss, Janney, Elstner Associates, Inc. 
(WJE). This global company consists of engineers, 
architects, and materials scientists specializing in the 
investigation, testing, and design of repairs for 
buildings and bridges. The engineers gave an overview 
of their concrete research focusing on corrosion and 
cracking of concrete and masonry structures. 
Interestingly, there were several areas of research that 
overlapped with research being conducted at KUT, 
particularly regarding the durability of concrete made 
with municipal solid waste fly ash. Along with research 
themes, we also discussed their internship program, 
which is seeking architectural/architectural engineering, 
civil, or structural engineering students for a summer 
through fall Co-op at their New York, NY office. 
Additional information can be found at: 
https://www.wje.com/careers/job-opportunities.
Ambi Robotics, https://www.ambirobotics.com/
In California, I met with a research scientist at Ambi 
Robotics, who is working on AI-powered robotic 
solutions for parcel sortation. The company is a small 
start-up but has already deployed robots to sort 
packages at companies such as Pitney Bowes and OSM 
Worldwide. Ambi Robotics has a robust internship 
program, and strong communication skills in English 
are important. Additional internship information can be 
found at: https://www.ambirobotics.com/blog/interns-
of-ambi-robotics-summer-2023.
Implications
The report concludes with a synopsis of how 
generative AI can serve as an ally in tackling the 
challenges posed by growing classroom sizes, with 
insights into the present landscape and future horizons 
of AI integration in language courses at Kochi 
University of Technology. Generative AI presents both 
opportunities and challenges for educators. Educators 
and administrators need to draft guidelines that include 
acceptable and unacceptable uses of generative AI in 
education. Guidelines should be outlined at both the 
university-wide and course levels. The formation of an 
AI working group would be beneficial, along with 
dialogue between students, teachers, and administrators 
about how to best adapt to emerging generative AI 
tools. Learner access to AI must also be evaluated to 
ensure inclusive learning opportunities for all students. 
Over time, we can assume students and teachers will 
146
adapt to this disruptive technology as they have with 
other disruptive technologies in the past, such as the 
introduction of personal computers, the internet, and 
mobile devices.
References
1) B. Bridgeman, C. Trapani, & Y. Attali, “Comparison 
of human and machine scoring of essays: Differences 
by gender, ethnicity, and country”. Applied 
Measurement in Education, 25(1), 27-40, 2012. 
https://doi.org/10.1080/08957347.2012.635502.
2) A. Caines, L. Benedetto, S. Taslimipoor, C. Davis, Y. 
Gao, O. Andersen, & P. Buttery, “On the application 
of Large Language Models for language teaching 
and assessment technology”. arXiv preprint, 2023. 
https://doi.org/10.48550/arXiv.2307.08393.
3) P. Daniels, “Automated Grammar Scoring with AI 
Chatbots”. TEFL Praxis Journal, 2(1), 59-68, 2023. 
https://doi.org/10.5281/zenodo.10402530.
4) M. A. Hussein, H. Hassan, & M. Nassef, 
“Automated language essay scoring systems: A 
literature review”. PeerJ Computer Science, 5, 
e208, 2019. https://doi.org/10.7717/peerj-cs.208.
5) V. Kumar, & D. Boulanger, “Explainable automated 
essay scoring: Deep learning really has pedagogical 
value”. Frontiers in Education, Assessment, 
Testing and Applied Measurement, 5, 2020. 
https://doi.org/10.3389/feduc.2020.572367.
6) S. Polak, G. Schiavo, & M. Zancanaro, “Teachers  ̓
perspective on artificial intelligence education: 
An initial investigation”, CHI Conference on 
Human Factors in Computing Systems Extended 
Abstracts, pp. 1-7, 2022. New Orleans, USA. 
https://doi.org/10.1145/3491101.3519866.
147
AI が現代の教育に与える影響
Paul Daniels
（受領日：2024 年 5 月 31 日）
高知工科大学共通教育教室
〒 782-8502　高知県香美市土佐山田町宮ノ口 185
E-mail: danils.paul@kochi-tech.ac.jp
要約：人工知能（AI）は、私たちの働き方や学習方法に革命をもたらし続けており、教育への
影響も拡大しつつある。最近の研究では、AI が教師と学習者の双方をどのようにサポートする
のがベストなのかに光を当てている。生成 AI と大規模言語モデル（LLM）に関する研究は、
教育者が教育における AI の生産的役割について楽観的であることを示している6,2)。さらに、言
語学習分野の研究では、AI がライティングスキルの評価において人間と同等か、それ以上であ
ることが実証されている1)。AI が一般的な学習をどのように促進できるかをさらに調査するた
め、2023 年の冬に 6 ヶ月間の海外調査プロジェクトを実施した。このプロジェクトでは、アメ
リカの主要な教育機関の教育者へのインタビューと定量的な調査を行い、生成 AI をどのよう
に活用すれば教育と学習を最適化できるかをより深く理解することができた。


English

The Impact of AI on Modern Education

亀岡, 伸義

CNT forest film の電気特性に及ぼす CNT 構造の効果

Samita Lappayanant

Eﬀects of Interactive Experiences on Memory in a Virtual Museum

若松, 宗真

マルチスケール FEM による熱硬化 FRP の成形誘起変形 シミュレーション

⻄岡, 幸亮

放置林における CO2 固定量算出に向けた 計測と解析⼿法の構築

髙橋, 龍一

データ駆動型プロセッサの環状パイプライン構成法

市野 駿

気象の変化に対応したインフラサウンド波形の計測波形について

篠森　敬三

視覚・感性統合研究室　―設置の背景と目標・研究テーマと成果・将来展望（第I 期～第II 期）―

岡本, 怜士

Rust を用いてOTP アプリケーションを 構築するためのフレームワークの開発

Kochi University of Technology

Not a member yet

皮膚兎錯覚現象を応用した触覚方位知覚の検討

The Impact of AI on Modern Education

CNT forest film の電気特性に及ぼす CNT 構造の効果

Eﬀects of Interactive Experiences on Memory in a Virtual Museum

マルチスケール FEM による熱硬化 FRP の成形誘起変形シミュレーション

放置林における CO2 固定量算出に向けた計測と解析⼿法の構築

データ駆動型プロセッサの環状パイプライン構成法

気象の変化に対応したインフラサウンド波形の計測波形について

視覚・感性統合研究室　―設置の背景と目標・研究テーマと成果・将来展望（第I 期～第II 期）―

Rust を用いてOTP アプリケーションを構築するためのフレームワークの開発

2,359

2,367