Artificial intelligence in education

Artificial intelligence in education (often abbreviated as AIEd) is a subfield or related field of educational technology that addresses the involvement of artificial intelligence technology, such as generative AI chatbots, in or to create learning environments.^[1]

The field considers the ramifications and impacts of AI on existing educational infrastructure, as well as future possibilities and innovations. Considerations in the field include data-driven decision-making, AI ethics, data-privacy and AI literacy.^[2]

Use of artificial intelligence in education has raised concerns such as the environmental impact and existential risk of AI, as well as the potential for classroom misuse, challenges to learner agency and autonomy, and the perpetuation of misinformation and bias.^[3]

History

AIEd can be traced back to the 1960s, when educators and researchers began developing computer-based instruction systems, such as PLATO developed by University of Illinois.^[4]

In the 1970s and 1980s, intelligent tutoring systems (ITS) were being adapted for classroom instruction.

The International Artificial Intelligence in Education Society was founded in 1993.^[5]

In the 2020s, chatbots based on large language models, such as OpenAI's ChatGPT, were first released to the general public and rapidly became popular. LLMs' general-purpose capabilities triggered concerns about the potential for misuse and academic dishonesty. AI content detectors have been developed, although their accuracy is limited. Some schools banned LLMs, but many bans were later lifted.^[6]

The AI in education community has grown rapidly in the global north, driven by venture capital, big tech, and open educationalists.^[7]

Theory

AIEd applies theory from education studies, machine learning, and related fields.

Three paradigms of AIEd

One posited model suggests the following three paradigms for AI in education, which follow roughly from least to most learner-centered and from requiring least to most technical complexity from the AI systems:

AI-Directed, Learner-as-recipient: AIEd systems present a pre-set curriculum based on statistical patterns that do not adjust to learner’s feedback.

AI-Supported, Learner-as-collaborator: Systems that incorporate responsiveness to learner’s feedback through, for example, natural language processing, wherein AI can support knowledge construction.

AI-Empowered, Learner-as-leader: This model seeks to position AI as a supplement to human intelligence wherein learners take agency and AI provides consistent and actionable feedback.^[8]

Socio-technical imaginaries

Some scholars frame AI in education within the concept of the socio-technical imaginary, defined as collective visions and aspirations that shape societal transformations and governance through the interplay of technology and social norms.^[9]

This framing positions AI in the history of “emerging technologies” that have and will transform education, such as computing, the internet, or social media.^[10]

Applications

AI has been employed in educational settings through a wide range of tools.

AI-based tutoring systems

Intelligent tutors or Intelligent tutoring systems (ITS), such as SCHOLAR system in the 1970s, are designed to create an artificial interaction between a student and a teacher.^[11]

ITS have also been considered for accessibility purposes like supporting students in larger classes who may not be able to get direct attention from human instructors.^[11]

Customized learning platforms

Personalized AI platforms can tailor instructional environments to students' needs using algorithms to predict students' patterns and habits and making recommendations to improve performance.^[12] Many such platforms are app-based, for example Photomath, which purports to help students solve and understand equations.^[13]

Automated grading systems

Automation in student assessment and feedback could save time for educators. Systems make use of different rubric combinations to grade performances. These systems need oversight to prevent bias in scoring and raise concerns about labor equity and job replacement.^[12]

Generative AI

Large language models when designed and/or employed for educational contexts represent a significant area of interest to the AIEd field. Many of the above systems operate using natural language processing and the transformer architecture of generative AI platforms like OpenAI's ChatGPT, and Grok.^[14]

Educational uses of generative AI include assessment and feedback, instant machine translations, on-demand proof-reading and copy editing, intelligent tutoring or virtual assistants.^[7]

Perspectives

Some scholars have suggested that AI might automate procedural knowledge and expertise^[15] or even match or surpass human capacities on cognitive tasks. They advocate for the integration of AI across the curriculum and the development of AI Literacy.^[16] With higher education facilities finding themselves with an opportunity to create a path for themselves and their students by creating guidelines so that AI can incorporated into their curriculum.^[17] Others are more skeptical as AI faces an ethical challenge,^[18] where "fabricated responses" or "inaccurate information", politely referred to as "hallucinations"^[15] are generated and presented as fact. Some remain curious about societies tendency to put their faith in engineering achievements, and the systems of power and privilege^[19] that leads towards deterministic thinking.^[20] While others see copyright infringement^[21][22][23] or the introduction of harm, division and other social impacts, and advocate resistance to AI.^[24]

Commercial perspectives

AI companies that focus on education, are currently preoccupied with generative artificial intelligence (GAI), although data science and data analytics is another popular educational theme. These initiatives may be considered related to the AI boom and potential AI bubble.

Educator perspectives

Some educators and school administrations have found AI to improve the efficiency of their work. Some educators are concerned about job replacement.^[25][26] Some teachers lack trust and have a negative attitude towards the use of AI in education.^[27]

Student perspectives

Some studies have found students to be flexible with technology such as personalized feedback and self-paced learning, but still perceive reliability, privacy, and fairness are concerns.^[28][29]

Challenges and ethical concerns

The advancement and adoption of AI in education comes with criticisms and ethical challenges.

Over-reliance, Inaccuracy, and Academic Integrity

Some critics believe that reliance on the technology could lead to lesser creativity, critical thinking and problem solving abilities especially if students skip traditional methods. Reliance on generative artificial intelligence has been linked with reduced academic self-esteem and performance, and heightened learned helplessness.^[30] Algorithm errors and hallucinations are some of the common flaws in AI agents, making them less trustworthy or reliable.^[3] These findings further underscore concerns raised in prior studies regarding academic integrity in the context of AI use in academic settings.^[31]

Accessibility

Equal access to AI could be one of the areas that comes into consideration. As there may many low incomes and rural areas deprived of the platform use. This might widen the gap in terms of education access. Global efforts should be made to accessibility and train educators in those underprivileged areas.^[3][32]

Bias and fairness

AI agents might be trained on biased data according to different company driven agendas. Bias can come in different forms, some of which include: algorithmic, architectural, and machine-learning bias. There are many different kinds of bias that can be introduced to the AI during the machine-learning process.^[33] Common types of bias that occur during the machine learning process are: association bias, language bias, exclusion bias, marginalized bias, and sample bias.^[34] Since LLMs were created to produce human-like text, bias can easily, and unintentionally be introduced and reproduced.^[33] This might lead to knowledge which is fed to them in form of misinformation. There should be policies and check to maintain such bias practices.^[32] Critics also argue that AI's data processing and monitoring reinforce neoliberal approaches to education rather than addressing inequalities.^[35][36]

Data privacy

Data privacy is an ethical concern as most of the results are on trained data and it can be misused for various purposes. Additionally, there is a lack of transparency from developers, and compliance laws should make sure of the transparency and data privacy is intact.^[11][37]

See also

• ChatGPT in education
• Computational education
• Computing education
• Computers in the classroom
• List of chatbots