his report was compiled from a collective intelligence gathering of World in 2050’s (W2050) Senior Fellows (Education & Work Committee). The meeting took place under the Chatham House Rule, so ideas will not be directly attributed to any specific Fellow. W2050 Senior Fellows attending the committee meeting were: Katherine Blanchard, Henry Anumudu, Manjula Dissanayake, and Dr. Noah W. Sobe.
EdTech has been a hot button topic for some time now, and the focus on generative AI in 2023 has only amplified interest in EdTech. Proponents see a plethora of ways that AI can improve education outcomes. AI’s potential for transforming education is real, but so are the dangers—so we must move forward with care and intentionality. It is with an eye toward care and intentionality that members of W2050’s Senior Fellows committee on education and work met to discuss how AI could—and how it should—impact education in 2024 and beyond.
AI Means the Education Transformation is Here, Ready or Not
AI is bringing about a new reckoning with performativity in education. From rethinking how to measure student learning at a time Chat-GPT can write papers to ideas about how AI can take on some teaching tasks, there’s a lot to digest. But the fundamental underlying issues that we think about with AI in the classroom are the same issues that education stakeholders have been grappling with for years. Learning by memorization has been outdated for years thanks to search engines. What labor markets need from education has shifted. Our understanding of how to support teachers has evolved. What students want and need from education themselves has also been at the heart of thinking about the education transformation for several years now.
In short, the education transformation has been underway for years, but AI is poised to accelerate it at speeds we are not prepared for. It is incumbent on us to ensure that transformation is undertaken with intentionality and understanding of how AI can disrupt our education systems, so we get the disruptions that will be productive and healthy for future generations.
AI Solutions Can Exacerbate, Alleviate Inequalities
Bias in AI training data sets is a well-documented phenomenon, so how students and teachers interact with AI—and how those AI solutions are trained—will be crucial. Access to funding for appropriate and effective AI solutions is also a well-known challenge. Less discussed are challenges around making sure that AI tools are suited to the environment where they are deployed—a solution designed for relatively affluent Western systems will inevitably miss the specific needs of children in more marginal communities. This is particularly problematic given AI is trained on data sets which generally reflect more mainstream, affluent situations, posing a danger that we could create new inequalities.
There is also a need to ensure students, teachers, and administrators themselves receive the digital literacy education they need to use their AI tools appropriately and safely. For instance, in many marginalized communities across the Global South, in particular many social media users are largely unaware of the plethora of issues plaguing these platforms. Thus, users in these communities are more likely to fall prey to mis- and disinformation as well as have their data harvested and used in ways that may be harmful. With AI in the classroom, this danger is exponentially higher.
Ensuring AI alleviates inequalities rather than exacerbating old ones and creating new ones will require care and intention in how we innovate and how we use those innovations.
Priorities for an Effective, Inclusive, AI-Powered Education Transformation
Rethinking Assessments: Generative AI can produce things that look like student performance. Some such uses may be appropriate, others less so. AI can also help teachers assess students in new ways (more on this analytical power below). This means education stakeholders must fundamentally rethink what are the measurables we assess, and how we assess them, when it comes to understanding student progress and success.
AI as Teacher Support: One anxiety over AI in the classroom is that it could be used to replace teachers in problematic ways. Finding ways to instead innovate AI tools that help teachers perform back-end tasks such as grading, lesson planning, and administrative work could give teachers more time to engage in the act of teaching and relationships—and, crucially, for professional development. This fundamentally flips the script on how we envision using AI in education settings and is a more appropriate use of AI, because AI cannot give students the sort of human interaction that is such a part of the education process.
AI’s Analytical Power: Pilot projects using AI as an analytical tool found that using technology to keep track of student attendance helped teachers identify potentially harmful trends and protect against things like child marriage, domestic abuse, and other out-of-school issues which can impact student attendance. Similarly, AI could help analyze other classroom behaviors to help identify struggles students may be experiencing that a teacher on their own, especially in a full classroom, may struggle to catch—such as neurodiversity, learning disabilities, or even simply being more or less advanced in certain parts of the curriculum.
AI and Personalization: Adopting AI as a tool for teacher support and for its analytical powers in the classroom—rather than for teaching—creates powerful opportunities for personalized education. In this way, teachers can adjust for student needs more readily, given they have more time and more analytical tools. This lets education systems not only better support learner needs, but also gives opportunities for students to better choose their own priorities.
Educating for AI: AI tools cannot simply be introduced into an education ecosystem as a prescriptive medicine which will fix what ails that system. Teachers, administrators, and students all require basic levels of digital literacy to understand both the basic potential and weaknesses of AI tools. They will also need specific training on the tools which have been chosen for their particular needs (which, in turn, requires acting with great intentionality).
Adopting AI with Intent: All of the above considerations, alongside the dangers of creating new inequalities, mean that we must exercise a very human trait which AI lacks: discernment. How training data sets are chosen for what AI tool and for which contextual situation, how funding is allocated, what tools are made available for who, and importantly what solutions are developed in the first place are key considerations that require a robust and inclusive consultation process with education stakeholders—student, parent, teacher, administrator, or policymaker—all over the world.
a global affairs media network
AI for Education, Educating for AI
Photo by Giu Vicente on Unsplash
January 9, 2024
W2050 Senior Fellows discussed what AI means for the education transformation and equitable access to/benefit from education. They found that AI means the education transformation is happening now, ready or not, and we need to reframe how we think about tech and education to be truly prepared.
T
his report was compiled from a collective intelligence gathering of World in 2050’s (W2050) Senior Fellows (Education & Work Committee). The meeting took place under the Chatham House Rule, so ideas will not be directly attributed to any specific Fellow. W2050 Senior Fellows attending the committee meeting were: Katherine Blanchard, Henry Anumudu, Manjula Dissanayake, and Dr. Noah W. Sobe.
EdTech has been a hot button topic for some time now, and the focus on generative AI in 2023 has only amplified interest in EdTech. Proponents see a plethora of ways that AI can improve education outcomes. AI’s potential for transforming education is real, but so are the dangers—so we must move forward with care and intentionality. It is with an eye toward care and intentionality that members of W2050’s Senior Fellows committee on education and work met to discuss how AI could—and how it should—impact education in 2024 and beyond.
AI Means the Education Transformation is Here, Ready or Not
AI is bringing about a new reckoning with performativity in education. From rethinking how to measure student learning at a time Chat-GPT can write papers to ideas about how AI can take on some teaching tasks, there’s a lot to digest. But the fundamental underlying issues that we think about with AI in the classroom are the same issues that education stakeholders have been grappling with for years. Learning by memorization has been outdated for years thanks to search engines. What labor markets need from education has shifted. Our understanding of how to support teachers has evolved. What students want and need from education themselves has also been at the heart of thinking about the education transformation for several years now.
In short, the education transformation has been underway for years, but AI is poised to accelerate it at speeds we are not prepared for. It is incumbent on us to ensure that transformation is undertaken with intentionality and understanding of how AI can disrupt our education systems, so we get the disruptions that will be productive and healthy for future generations.
AI Solutions Can Exacerbate, Alleviate Inequalities
Bias in AI training data sets is a well-documented phenomenon, so how students and teachers interact with AI—and how those AI solutions are trained—will be crucial. Access to funding for appropriate and effective AI solutions is also a well-known challenge. Less discussed are challenges around making sure that AI tools are suited to the environment where they are deployed—a solution designed for relatively affluent Western systems will inevitably miss the specific needs of children in more marginal communities. This is particularly problematic given AI is trained on data sets which generally reflect more mainstream, affluent situations, posing a danger that we could create new inequalities.
There is also a need to ensure students, teachers, and administrators themselves receive the digital literacy education they need to use their AI tools appropriately and safely. For instance, in many marginalized communities across the Global South, in particular many social media users are largely unaware of the plethora of issues plaguing these platforms. Thus, users in these communities are more likely to fall prey to mis- and disinformation as well as have their data harvested and used in ways that may be harmful. With AI in the classroom, this danger is exponentially higher.
Ensuring AI alleviates inequalities rather than exacerbating old ones and creating new ones will require care and intention in how we innovate and how we use those innovations.
Priorities for an Effective, Inclusive, AI-Powered Education Transformation
Rethinking Assessments: Generative AI can produce things that look like student performance. Some such uses may be appropriate, others less so. AI can also help teachers assess students in new ways (more on this analytical power below). This means education stakeholders must fundamentally rethink what are the measurables we assess, and how we assess them, when it comes to understanding student progress and success.
AI as Teacher Support: One anxiety over AI in the classroom is that it could be used to replace teachers in problematic ways. Finding ways to instead innovate AI tools that help teachers perform back-end tasks such as grading, lesson planning, and administrative work could give teachers more time to engage in the act of teaching and relationships—and, crucially, for professional development. This fundamentally flips the script on how we envision using AI in education settings and is a more appropriate use of AI, because AI cannot give students the sort of human interaction that is such a part of the education process.
AI’s Analytical Power: Pilot projects using AI as an analytical tool found that using technology to keep track of student attendance helped teachers identify potentially harmful trends and protect against things like child marriage, domestic abuse, and other out-of-school issues which can impact student attendance. Similarly, AI could help analyze other classroom behaviors to help identify struggles students may be experiencing that a teacher on their own, especially in a full classroom, may struggle to catch—such as neurodiversity, learning disabilities, or even simply being more or less advanced in certain parts of the curriculum.
AI and Personalization: Adopting AI as a tool for teacher support and for its analytical powers in the classroom—rather than for teaching—creates powerful opportunities for personalized education. In this way, teachers can adjust for student needs more readily, given they have more time and more analytical tools. This lets education systems not only better support learner needs, but also gives opportunities for students to better choose their own priorities.
Educating for AI: AI tools cannot simply be introduced into an education ecosystem as a prescriptive medicine which will fix what ails that system. Teachers, administrators, and students all require basic levels of digital literacy to understand both the basic potential and weaknesses of AI tools. They will also need specific training on the tools which have been chosen for their particular needs (which, in turn, requires acting with great intentionality).
Adopting AI with Intent: All of the above considerations, alongside the dangers of creating new inequalities, mean that we must exercise a very human trait which AI lacks: discernment. How training data sets are chosen for what AI tool and for which contextual situation, how funding is allocated, what tools are made available for who, and importantly what solutions are developed in the first place are key considerations that require a robust and inclusive consultation process with education stakeholders—student, parent, teacher, administrator, or policymaker—all over the world.
