AI is personalizing education and creating new communication channels across learning environments. Students are better able to learn and remember new information. Students are also achieving higher grades, in part, due to the integration of academic AIs. However, these systems are most effective with the guidance of quality educators.
For awhile, education birthed the simplified model of a single instructor overseeing a single cohort of thirty students. Each student advanced through the content at their own pace, and it was assumed students would fill knowledge gaps on their own.
Thanks to advancements in technology, this model is shifting. AI learning systems can evaluate a student in real time and determine the right learning materials to present to that student. Now, education practices can evolve away from the traditional model and towards a student-centered approach.
This article explains the current state of education and the technology's potential, the effect of the technology based on current literature, and what education practitioners can expect from the technology.
How is Learning Technology Systems Optimized?
The premise of an adaptive learning system is to assess the user through objective learning content. The system then personalizes the next learning challenge. This content is unconstrained and can be limitless.
Khanmigo, Duolingo and MATHia by Carnegie Learning are just some AI based tools that are geared towards education. To elaborate on the example of Carnegie Learning, their AI analyzes students’ knowledge of math and associated concepts, finds knowledge gaps, and helps students advance from their current level of understanding. Carnegie Learning claims that students who use MATHia retain learned concepts better than students who participate in traditional classroom instruction.
Adaptive learning systems offer more than just content. They diagnose learner understanding and offer an appropriate, context-specific solution.
What are the advantages of AI in education?
Tailored Learning Experiences for Every Student
In a traditional classroom of 25-35 students, a teacher cannot provide individualized instruction. But AI adapts to the learning of each student in real time and uses that data to design a unique learning experience for that student. A teacher may spend hours designing a lesson for one student, but the AI does it the moment the student interacts with it.
This is extremely helpful for both the bored, advanced student, and the struggling, unmotivated student. Both of these students benefit from instruction that is designed for their actual level, instead of the assumed level by the curriculum.
Real Time Feedback
Timely feedback is one of the many accelerators of learning established by education research. However, feedback that is not given in real time, or close to the time of learning, becomes useless.
Examples include Grammarly and Turnitin. The former offers writing guidance and the latter offers real-time editing. The real-time guidance provided by these applications shortens the feedback loop that is almost impossible to overcome with traditional learning.
With Respect to Accessibility and Inclusion in AI
Compared to traditional tech, AI offers learning tools that are broadly inclusive with respect to challenges that learning disabilities and barriers to learning present. For example, real-time captioning, translation, and learning support tools are available in Microsoft's Immersive Reader. These tools ensure that students who cannot access the content remain able to participate in the learning.
The tools support English language learners by allowing them to interact with learning materials at grade level, and building their language proficiency.
What Are the Limitations of AI in Education?
An honest discussion of the role of AI in education must address its shortfalls.
- AI cannot replicate the human connection. It cannot provide support for motivation, emotional challenges, mentorship, and the type of constructive feedback offered by teachers in the context of a student's situation. Students that lack self-regulation, or that require additional support to remain interested in learning, find the absence of other tools that AI is unable to provide, a major challenge.
- Data privacy is a real issue. Significant behavioral and performance data is collected by adaptive learning platforms. These tools should not be implemented, especially in learning environments with children, without sufficient data governance policies.
- There are still challenges to equity. AI-based learning systems necessitate the internet, and the infrastructure is deficient to such an extent in many rural and low-income areas that such systems may in fact increase rather than decrease achievement gaps.
- There is a danger of losing the ability to learn. AI may automate learning to the extent that the user becomes reliant on it. The advantage of AI may quickly become the reverse, and it may even become a learning detriment, if AI is employed as a learning substitute rather than a learning scaffold.
How are Schools and Universities Using AI?
There is a broad spectrum in the extent of use. Some use AI without hesitation and others have adopted AI with clear frameworks to guide use. Arizona State University has built partnerships with OpenAI to support AI-based learning in Augmented Tutors, so students can learn in an AI-supported environment.
In K-12 education, AI is being tested, as in Universities, in a variety of applications. Some examples include AI-supported grading, flexible reading instruction, and prediction models for student failure. These assist in managing student-related data to predict failure based on a student’s engagement and academic record.
AI is also finding use in lesson preparation by teachers. AI can generate practice questions, create lesson variants, and even draft grading rubrics. This allows for more instructional time to be available for students by reducing the lesson preparation burden on teachers.
What Are the Current Findings on AI and Educational Outcomes?
An emerging trend in the literature has very recently begun to demonstrate a positive relationship between students and the use of AI in educational settings. Traditionally, the use of a tutor was said to have a positive relationship on the educational growth of a student. An example of this is described in a 2023 edition of Science that suggests that students paired with an AI tutor utilizing the Socratic method gained a positive educational growth of two standard deviations than students placed in a traditional educational classroom.
The use of AI in instructional systems and adaptive learning has demonstrated positive growth on achievement in educational systems and has shown the most growth in the areas of math and science. In their 2016 meta-analysis, Kulik and Fletcher suggested that the use of instructional systems and AI in the classroom demonstrated a positive growth on student achievement and an effect size of approximately 0.66 when compared to traditional teaching methods.
It is important to remember, when utilizing instructional systems that employ AI, the effect size is dependent on the subject matter, the population of the student, and the quality of the instruction employed.
The Future of AI in the Classroom
AI should assist, rather than replace, teachers. Having AI perform routine tasks that do not require critical thinking, such as providing students with practice problems or progress assessments, allows teachers to focus on the more critical and thoughtful elements of their job. Utilizing AI in this way also allows teachers to build relationships with students and develop their intellectual curiosity. Schools that utilize AI as an instructional supplement rather than a replacement are seeing the most positive outcomes from their instructional systems.
Preparing Students for an AI-Driven World
Understanding AI and its implications is a foundational skill for the future workforce and contributes to positive learning outcome arguments for AI in education.
Integrating AI into the classroom prepares students to engage with an AI-informed world, while teaching them to critique and/or harness AI systems to perform tasks, assess the quality and accuracy of AI output, and understand the boundaries and implications of AI technology.
What Will Happen Next
AI and education are just beginning to converge. The capabilities of AI and the integration of sophisticated systems into educational technology frameworks are expanding rapidly. We can expect advancements in AI and educational technology that serves users and provides conversational interfaces for students, supports real-time surveillance and assessment systems, and offers predictive modeling feedback.
The most benefit will go to educational systems with thoughtful technology integration that includes realistic assessments of learner outcomes and strives to continually improve the quality of teaching and learning.
FAQs
What does adaptive learning mean in the context of education?
Adaptive learning describes a method in which educational systems based on AI adjust the learning content and instructional style based on the real-time learning needs of the student. AI is used in adaptive learning systems to create uniquely defined learning pathways and flexible learning goals.
Can AI replace teachers?
AI lacks the ability to intellectually and emotionally engage and understand students, which are the core components of teaching. AI may efficiently perform basic teaching functions such as presenting information or giving assessments. However, the role of AI should be to enhance the teaching process, not take it over.
Are AI learning tools effective for all students?
AI is more or less effective depending on the subject, the learning style, and the way in which the AI is integrated into the learning process, but can be particularly beneficial for subjects like math, with a less flexible learning process. Not all students will experience the same degree of improvement, especially if they have an inadequate support system or lack stable access to the internet.
What are the privacy risks of AI in education?
AI in education can track and record a student’s activity in the learning program and their responses to the program. Learning companies that use AI have the potential to violate student’s privacy by using their data in ways that are unregulated and unethical, especially if the student is a minor. Before using a new learning program, schools should check that the program complies with data privacy regulations and incorporates protective measures like FERPA.
How can students avoid AI dependence when using AI tools?
Most effectively, students can use AI as a learning aid to help them think critically by engaging with the AI and the information it provides instead of using the AI to do the thinking for them. Students should engage with the information and the AI to think critically and analytically instead of accepting the AI answer as the end point.