AI in education: Leveraging technology for human potential
9 Jul 2025 15:50h - 16:05h
AI in education: Leveraging technology for human potential
Session at a glance
Summary
Kevin Mills, leading education partnerships at OpenAI, delivered a presentation at a UN gathering about the transformative role of artificial intelligence in education. Mills shared his personal journey from a college intern at UNESCO twenty years ago to his current position, explaining how his early exposure to global education challenges shaped his career mission to help people reach their full potential. He described OpenAI’s evolution from a small research lab focused on ensuring artificial general intelligence benefits humanity to a company serving over half a billion weekly ChatGPT users, including over a million organizations using their enterprise products.
Mills revealed that 40% of ChatGPT users are under 24 years old, with learning being the platform’s number one use case across all demographics, making OpenAI effectively the world’s largest education platform. He outlined OpenAI’s two-pronged approach to educational responsibility: building AI literacy through initiatives like ChatGPT Lab and OpenAI Academy, and partnering with educational institutions worldwide for responsible AI adoption. The company has moved from experimental pilots to widespread institutional adoption, with universities like Harvard, University of Pennsylvania, and Arizona State University implementing custom GPTs for personalized tutoring and campus operations.
Mills highlighted major partnerships including the California State University system serving 500,000 users and a countrywide agreement with Estonia. While acknowledging concerns about students potentially shortcutting their learning, he emphasized that when used correctly as a learning assistant, AI shows powerful educational benefits supported by research studies. Mills concluded by positioning AI as a tool to unlock human potential, envisioning a future where every educator receives needed support and every student, regardless of background, can access quality education and reach their full potential.
Keypoints
**Major Discussion Points:**
– **Personal journey and mission in education**: Kevin Mills shares his 20-year career path from a college intern at UNESCO to leading education partnerships at OpenAI, driven by his belief that education is the core solution to global challenges
– **ChatGPT’s explosive growth and educational impact**: The platform has reached over 500 million weekly users in just 2.5 years, with 40% of users under 24 and learning being the number one use case, making OpenAI inadvertently the world’s largest education platform
– **OpenAI’s dual educational strategy**: Building AI literacy through community initiatives like ChatGPT Lab and OpenAI Academy, while partnering with educational institutions and governments for responsible AI adoption
– **Institutional transformation and partnerships**: Universities are moving from experimentation to widespread adoption of AI tools, with examples from Harvard, University of Pennsylvania, and ASU showing how custom GPTs are revolutionizing teaching, student services, and research
– **Challenges and responsible implementation**: Addressing the risk of students using AI to shortcut learning rather than enhance it, emphasizing the need for research-backed best practices and proper implementation to ensure AI serves as a learning tool rather than a replacement for critical thinking
**Overall Purpose:**
The discussion serves as a keynote presentation at a UN gathering where Mills aims to demonstrate OpenAI’s commitment to education, showcase the transformative potential of AI in learning environments, and advocate for responsible AI adoption in educational institutions worldwide.
**Overall Tone:**
The tone is consistently optimistic and inspirational throughout, with Mills maintaining an enthusiastic and visionary approach. He balances this optimism with measured acknowledgment of challenges, but the overall message remains hopeful about AI’s potential to democratize education and unlock human potential globally. The tone doesn’t significantly shift, remaining professional yet passionate from beginning to end.
Speakers
– Kevin Mills: Leading education partnerships at OpenAI; previously worked in government, university administration, and education technology for 12 years; helped build MOOC platforms like Udacity and Coursera; former college intern at the U.S. State Department
Additional speakers:
– Doreen: Role/title not specified, but mentioned in context of discussing AI skills expectations from employers
Full session report
# Comprehensive Report: AI’s Transformative Role in Education – A UN Gathering Presentation
## Executive Summary
Kevin Mills, leading education partnerships at OpenAI, delivered a comprehensive presentation at a United Nations gathering that explored the transformative potential of artificial intelligence in global education. Drawing from his extensive 20-year career journey from a college intern at UNESCO to his current leadership role, Mills articulated a vision for how AI can democratise access to quality education worldwide. The presentation revealed striking statistics about ChatGPT’s educational impact, outlined OpenAI’s strategic approach to educational partnerships, and addressed both the opportunities and challenges inherent in AI-assisted learning.
## Speaker Background and Personal Mission
Mills began his presentation by establishing his credentials and personal motivation, sharing his journey from “a lowly college intern at the U.S. State Department” and UNESCO twenty years ago to his current position leading education partnerships at OpenAI. His background spans the last 12 years working in education technology, including roles helping build MOOC platforms such as Udacity and Coursera.
The speaker’s personal mission emerged from a realisation at a UNESCO conference about the future of education two decades earlier, where he concluded that education serves as the core solution to major global challenges including peace and prosperity. This foundational belief has driven his career-long commitment to helping as many people as possible reach their full potential through global education initiatives. Mills positioned this personal mission as directly aligned with OpenAI’s broader objectives.
## OpenAI’s Evolution and Current Scale
Mills provided context for OpenAI’s transformation from a small research laboratory focused on ensuring artificial general intelligence benefits humanity to a major technology platform. He noted that ChatGPT broke records “going from zero to 100 million monthly active users in just two months” and now serves over half a billion weekly users. The company now supports over one million organisations using their enterprise products.
A particularly striking revelation was that 40% of ChatGPT users are under 24 years of age, with learning identified as the platform’s number one use case across all demographics and regions. Mills disclosed that between 10% and 40% of every message sent on the ChatGPT platform relates to education and learning, effectively making OpenAI the world’s largest education platform.
## Strategic Approach to Educational Responsibility
OpenAI has developed a two-pronged strategy to address their role as a major educational platform. The first component focuses on building AI literacy from the ground up through intentional community initiatives. These include ChatGPT Lab, which creates student communities around AI learning, and the OpenAI Academy, a free online platform that “launched a couple months ago” designed to build AI literacy and skills across all demographics. The company has also joined the UN AI Skills Coalition to expand access to AI education resources through established UN platforms.
The second strategic component involves partnering directly with educational institutions and governments to ensure that “AI adoption is done responsibly and in a way that has long-term benefits for society.” This approach recognises that successful AI integration requires institutional support and proper implementation frameworks.
## Institutional Transformation and Partnerships
Mills described a significant shift in how educational institutions approach AI technology, moving from experimental pilots to widespread adoption as core infrastructure. Universities are now integrating AI across three key areas: “teaching and learning,” “campus operations,” and “research and scientific discovery.”
Several major partnerships illustrate this institutional transformation. Harvard University has implemented custom GPTs in the business school, which Mills described as “like ChatGPT but customized to a specific use case” for personalised tutoring. The University of Pennsylvania has developed AI systems for campus operations. Arizona State University moved forward with “over 250 different projects.” The California State University system represents one of the largest implementations, serving “all 500,000 students, faculty, and staff” across the network.
Mills announced “our first country partnership with the government of Estonia to bring ChatGPT EDU to schools across the country,” noting that they expect “many more of these countrywide announcements in the coming months.”
## Challenges and Responsible Implementation
Mills acknowledged significant challenges in AI educational implementation. The primary concern centres on students using AI to shortcut their learning rather than engage with materials for critical thinking development. He emphasised that true learning requires friction and struggle, and if students offload all work to AI tools, they will not develop essential skills or critical thinking capabilities.
However, Mills argued that when used correctly as a learning assistant and tutor, AI shows powerful positive results in education. He referenced what he called “Magic Chats or Magic Prompts” as specific terminology used by OpenAI in their educational approach.
To address these challenges, OpenAI has embarked on research collaborations with academic partners to conduct longitudinal studies and determine best practices for AI use in education. Mills announced a partnership with the American Federation of Teachers, announced “just yesterday,” to launch a new AI institute focused on proper implementation guidance.
## Workforce Implications and Skills Development
Mills highlighted an important consideration regarding workforce preparation, referencing “what Doreen was talking about” regarding employer expectations. He noted that using AI tools has become an essential skill for the global labour market, with employers now expecting students graduating from high school or college to be proficient in AI skills.
## Practical Educational Benefits
The presentation included specific examples of AI’s positive impact on educational experiences. Faculty members reported that AI assistance with routine questions freed up office hours for more meaningful interactions with students, allowing them to focus on career aspirations and personal development rather than basic content clarification.
Mills described AI as representing “the holy grail of edtech, essentially offering every student a personalized education, a world-class tutor for everyone in the world.” This vision positions AI as a democratising force that can provide equitable access to quality educational resources globally.
## Research and Evidence-Based Approach
Throughout the presentation, Mills emphasised the importance of research-driven implementation. OpenAI is conducting academic partnerships to establish evidence-based best practices and disseminate findings to educators and students as quickly as possible. The company’s research efforts focus on answering fundamental questions about the right way to use AI in education.
## Vision for the Future
Mills concluded with a vision that reframes AI development in terms of human flourishing. He emphasised that OpenAI’s ultimate goal is to use AI to unlock human potential, creating a world where every educator receives the support they need and every student, regardless of background or circumstances, can access quality education and reach their full potential.
## Conclusion
Mills’ presentation at the UN gathering provided a comprehensive overview of AI’s current and potential role in education, combining personal narrative, statistical evidence, practical examples, and future vision. The discussion moved beyond product promotion to engage with fundamental questions about learning, technology, and human potential.
The presentation addressed both transformative potential and significant challenges whilst advocating for research-based, responsible implementation. By positioning AI as a tool for unlocking human potential, Mills articulated a vision that addresses both technological possibilities and educational values.
The scale of ChatGPT’s educational usage establishes the importance of addressing AI integration responsibly. As educational institutions worldwide grapple with AI adoption, the approaches outlined in this presentation provide guidance for navigating this transformative period in educational technology.
Session transcript
Kevin Mills: Hello. It’s an incredible honor to be here with you today. The last UN gathering I attended was almost exactly 20 years ago. I wasn’t on stage. I wasn’t speaking. In fact, I was probably manning the door at the men’s restroom or something like that. I was a lowly college intern at the U.S. State Department, tasked with helping put on a conference on the future of education at UNESCO headquarters in Paris. And while I don’t remember so much about what I was actually doing at that conference, what stood out to me was that was the moment where I had the epiphany where I understood what I wanted to be when I grew up, where I wanted to dedicate myself. Because I had come to college thinking I wanted to be a diplomat, to wrestle with these tough global challenges like peace and prosperity. But what I came away with at that conference was that the core solution to so many of those big challenges was education. And so I remember actually writing in my journal at the time that I decided I wanted my career to be in the field of global education, and I wanted my personal mission statement to be to help as many people as possible reach their full potential. In the 20 years since then, that has been my North Star in every career decision I’ve made. It led me to work in government, in university administration, and for the last 12 years working in education technology, helping build MOOC platforms like Udacity and Coursera that have been so important in closing a global skills gap and democratizing access to the world’s best education. And when ChatGPT launched in November of 2022, I knew that this would have a profound impact on the future of education because it finally felt like we were on the cusp of the holy grail of edtech, essentially offering every student a personalized education, a world-class tutor for everyone in the world. And so today I’m so honored to be here with you, leading education partnerships at OpenAI, and sharing with you today a little bit about why education is so important to OpenAI, and also what we’re doing to help shape the future of AI in education. To start, I wanted to share a little bit about our history. OpenAI actually started as a small research lab dedicated to this mission, to ensure that artificial general intelligence benefits all of humanity. Now you may not have heard about us or AGI 10 years ago, but I bet you did in November of 2022 when we launched what we called then a low-key research preview that ended up changing the course of history. In the two and a half years since ChatGPT first launched, we’ve now grown to have over half a billion users use ChatGPT every single week, and that number is growing rapidly. On top of that, we have over a million organizations, companies, small businesses, schools, and universities using our enterprise products like ChatGPT EDU. You may think that that is fast, and you’d be right. This transformation is happening much quicker than past technological transformations, where usually there were decades between the initial discovery and widespread adoption. AI is much faster. Even by modern internet standards, ChatGPT broke all the records, going from zero to 100 million monthly active users in just two months. And while this growth story is largely known now, what I think is less understood is the important role education and students have played in that growth story. So I want to share a few stats with you. 40%. That’s the percentage of ChatGPT users under the age of 24. And can you guess what the number one use case for ChatGPT is? It’s learning. Across every demographic, every region, between 10% and 40% of every message sent on the ChatGPT platform is related to education and learning. So with half a billion users, we’re already the world’s largest education platform. And not only is it a tool for learning all sorts of skills, using the tool itself is now an essential skill for the global labor market. Employers now expect students graduating from high school or college to be proficient in AI skills, much like what Doreen was talking about. And so knowing this, the responsibilities and opportunities in education for open AI, how should we be thinking about this as a company? Well, we’re focused on two primary areas. The first is building AI literacy from the ground up. We want as many people around the world to have access to this technology and to know how to use it well. And the second aspect of it is partnering with education institutions and governments around the world to ensure that AI adoption is done responsibly and in a way that has long-term benefits for society. I’ll start by sharing a bit about our AI literacy efforts. So we all know students like to learn from one another. And one of the coolest things about AI technology is that so much of the innovation actually happens after the model has left the lab, and it’s how users learn to use it. And that’s the same in education. And so the approach we have taken at OpenAI is to build an intentional community of students from around the world. We call it ChatGPT Lab, and these students come together and just share how they’re using ChatGPT to further their education. They share specific prompts and use cases, and then we help curate and collect those and then amplify those voices to share to students around the world who can similarly benefit from those learnings. We call these Magic Chats or Magic Prompts, and we have hundreds of them gathered from these students that we share on social media so that students can apply these same lessons in their own schooling and their learning journey. A second element of our focus on the AI literacy has been the launch of what we call OpenAI Academy. This launched a couple months ago. I encourage you to check it out. It’s similar to a MOOC platform in that it has many, many online free courses on AI literacy and skill building for all sorts of personas, from youth to elderly to educators and students. And I’m excited that we just joined the UN AI Skills Coalition earlier today, and so this will be available on the UN platforms as well. Now while these efforts have been powerful and strong results, they aren’t enough and we can’t do this alone. And that’s why myself and my team are so focused on having deep relationships with educational institutions and governments around the world. We launched ChatGPTEDU about one year ago to do these partnerships. And when we launched, it was a very different world. We found that universities, colleges, schools were just in an experiment phase, doing small pilots, testing the water, trying to understand what to make of this hype around AI. Fast forward to today and it’s a very different story. Universities, colleges, and schools are going from experimentation to action and from pilots to widespread adoption because these universities and colleges realize that AI is no longer a nice to have. It’s now a core part of the modern university infrastructure and students and faculty expect an institution to have an AI tool. And the universities that are doing this are imagining all areas of the campus experience transforming. From teaching and learning, which is probably the most prominent example where AI can be used as a personalized tutor for students and a teaching assistant for faculty. It’s used to enhance efficiencies and have campus operations run more smoothly and is powering and accelerating research and scientific discovery at these institutions. Just to make it more practical to share a few examples, Harvard was one of our earliest partners with ChatGPTEDU. They used our product to really rethink their curriculum in the business school using a type of our technology called a custom GPT. Which as it sounds is like ChatGPT but customized to a specific use case. In this case, professors uploaded their syllabi, their lecture notes, text to basically build a custom GPT that could serve as a 24-7 teaching assistant for their students. So they could ask questions at any time. It’s a good thing because it turned out that the most popular time to ask those questions was midnight and 3 a.m. Not exactly your typical office hours. Speaking of office hours, faculty reported that they loved what it did to their office hours because now they got to use that time to get to know their students and their career aspirations rather than just answer questions about the content. Similarly, at University of Pennsylvania, they took that custom GPT technology, applied it not only into the classroom but in student services across the campus including their career management GPT which students found very helpful in navigating their career service resources. At ASU, they actually did a bottoms-up approach, inviting proposals from around the campus on ideas on how best to use ChatGPT EDU. They ended up moving forward with over 250 different projects ranging from a student mental health aid to a language learning aid to research assistance for faculty. And more recently, we’ve gone beyond partnering with individual campuses to even entire systems of higher education. Like our recently announced partnership with the California State University system to bring ChatGPT EDU to all 500,000 students, faculty, and staff. And earlier this year, we announced our first country partnership with the government of Estonia to bring ChatGPT EDU to schools across the country. And I believe that we’ll have many more of these countrywide announcements in the coming months. Now while there is so much potential and promise with this technology in education, there are real challenges that we have to be clear-eyed about. Probably the most prominent and top of mind for many of us is the risk that students use this to shortcut their own learning. We know that true learning takes friction. It takes struggle. You have to engage with the materials. And if students offload all of that work to a tool like ChatGPT, they will not learn those skills and they will not gain that critical thinking. That said, when ChatGPT is used correctly, as a learning assistant and as a tutor, the results are powerful. And study after study confirms this. And so the question is not so much, is AI good or bad for education? It’s a tool. And if used correctly as a tool, it can be powerfully good in education. And that’s a responsibility we at OpenAI take seriously. To answer that question, what is the right way to use AI in education? And so we’ve embarked on a number of research collaborations with academic partners recently to answer just that question and to have it research-backed and with longitudinal studies. And the second focus we have is getting that information and disseminating it out to professors and students around the world as quickly as possible so that they can be sure to be implementing it in the correct way. Just yesterday, in the United States, we announced a partnership with the American Federation of Teachers to launch a new AI institute to do just that. And I anticipate we’ll do more of that all around the world. Because at the end of the day, we’re not here just to build better machines or smarter models. We’re here to use AI to unlock human potential. That is why I chose to enter this field of education 20 years ago at that UNESCO conference. I’m sure it’s what drives many of you in the education community. We envision a world where every educator has the support needed to focus on what matters most, and where every student, no matter where they’re born or what circumstances they’re born into, get the education and the support they need to thrive and to reach their full potential. This is the opportunity in front of us, and I’m excited to unlock this new vision with all of you. Thank you so much.
Kevin Mills
Speech speed
147 words per minute
Speech length
2050 words
Speech time
835 seconds
Education is the core solution to major global challenges like peace and prosperity (Kevin Mills)
Explanation
Mills argues that while he initially wanted to be a diplomat to tackle global challenges, he realized at a UNESCO conference 20 years ago that education is the fundamental solution to major world problems. This epiphany shaped his career direction and personal mission.
Evidence
Personal experience at UNESCO conference in Paris 20 years ago where he had this realization while working as a college intern at the U.S. State Department
Major discussion point
Education as foundation for solving global challenges
Topics
Development | Sociocultural
Agreed with
Agreed on
Education as fundamental solution to global challenges
Personal mission to help as many people as possible reach their full potential through global education (Kevin Mills)
Explanation
Mills describes how his personal mission statement, developed 20 years ago, has been his North Star in every career decision. This mission has guided him through various roles in government, university administration, and education technology.
Evidence
Written in his journal at the time of the UNESCO conference; led him to work in government, university administration, and 12 years in education technology building MOOC platforms like Udacity and Coursera
Major discussion point
Personal commitment to global education access
Topics
Development | Sociocultural
ChatGPT represents the holy grail of edtech – offering personalized education and world-class tutoring for everyone (Kevin Mills)
Explanation
Mills argues that ChatGPT’s launch in November 2022 marked a breakthrough moment in educational technology. He believes it finally makes possible the long-sought goal of providing personalized education and world-class tutoring to every student globally.
Evidence
ChatGPT’s launch in November 2022 and its rapid adoption in educational contexts
Major discussion point
AI as transformative educational technology
Topics
Sociocultural
ChatGPT has grown to over half a billion weekly users and over a million organizations using enterprise products (Kevin Mills)
Explanation
Mills presents the rapid growth statistics of ChatGPT to demonstrate the unprecedented speed of AI adoption. He emphasizes that this transformation is happening much faster than previous technological shifts, which typically took decades between discovery and widespread adoption.
Evidence
Over half a billion users use ChatGPT weekly; over a million organizations use enterprise products like ChatGPT EDU; ChatGPT went from zero to 100 million monthly active users in just two months, breaking all records
Major discussion point
Unprecedented speed of AI adoption
Topics
Development | Economic
40% of ChatGPT users are under age 24, with learning being the number one use case across all demographics (Kevin Mills)
Explanation
Mills highlights the significant role that young people and education play in ChatGPT’s success. He emphasizes that learning is the primary use case across all user demographics and regions, making education central to the platform’s value proposition.
Evidence
40% of ChatGPT users are under age 24; learning is the number one use case; 10-40% of every message sent on ChatGPT platform is related to education and learning across every demographic and region
Major discussion point
Education as primary driver of AI platform usage
Topics
Sociocultural
OpenAI has become the world’s largest education platform with 10-40% of messages related to education and learning (Kevin Mills)
Explanation
Mills argues that OpenAI has inadvertently become the world’s largest education platform due to the massive scale of educational usage. With half a billion users and such a high percentage of education-related messages, the platform’s educational impact is unprecedented.
Evidence
With half a billion users and 10-40% of messages being education-related across all demographics and regions
Major discussion point
Scale of AI’s educational impact
Topics
Sociocultural | Development
Building AI literacy from the ground up through intentional student communities like ChatGPT Lab (Kevin Mills)
Explanation
Mills describes OpenAI’s approach to building AI literacy by creating communities where students share innovative uses of ChatGPT for education. This peer-to-peer learning model leverages the fact that much AI innovation happens after the model leaves the lab, through user experimentation.
Evidence
ChatGPT Lab community where students share prompts and use cases; hundreds of ‘Magic Chats’ or ‘Magic Prompts’ collected and shared on social media
Major discussion point
Community-driven AI literacy development
Topics
Sociocultural | Development
Launching OpenAI Academy as a free online platform for AI literacy and skill building across all demographics (Kevin Mills)
Explanation
Mills presents OpenAI Academy as a comprehensive solution for AI education, similar to MOOC platforms. The platform offers free courses designed for various user groups, from youth to elderly, educators to students, making AI literacy accessible to all.
Evidence
OpenAI Academy launched a couple months ago with many free online courses on AI literacy and skill building for all personas; joining the UN AI Skills Coalition to make it available on UN platforms
Major discussion point
Democratizing AI education access
Topics
Sociocultural | Development
Joining the UN AI Skills Coalition to expand access to AI education resources (Kevin Mills)
Explanation
Mills announces OpenAI’s participation in the UN AI Skills Coalition as part of their commitment to global AI literacy. This partnership will help distribute their educational resources through UN platforms, expanding international reach.
Evidence
Joined the UN AI Skills Coalition earlier that day; OpenAI Academy courses will be available on UN platforms
Major discussion point
International cooperation for AI education
Topics
Development | Sociocultural
Universities are transitioning from experimentation to widespread adoption of AI as core infrastructure (Kevin Mills)
Explanation
Mills describes a fundamental shift in how educational institutions view AI technology. Universities have moved beyond small pilots and testing phases to treating AI as essential infrastructure that students and faculty expect, similar to other core campus technologies.
Evidence
When ChatGPT EDU launched a year ago, universities were in experiment phase doing small pilots; now they’re going from experimentation to action and pilots to widespread adoption; AI is now core part of modern university infrastructure
Major discussion point
Institutional transformation in AI adoption
Topics
Sociocultural | Development
AI is transforming teaching, learning, campus operations, and research across educational institutions (Kevin Mills)
Explanation
Mills outlines the comprehensive impact of AI across all areas of university operations. He describes AI serving multiple roles: as personalized tutors for students, teaching assistants for faculty, efficiency tools for campus operations, and accelerators for research and scientific discovery.
Evidence
AI used as personalized tutor for students and teaching assistant for faculty; enhances campus operations efficiency; powers and accelerates research and scientific discovery
Major discussion point
Comprehensive AI integration in higher education
Topics
Sociocultural
Successful partnerships range from individual campuses like Harvard and ASU to entire systems like California State University and country-wide implementations like Estonia (Kevin Mills)
Explanation
Mills provides concrete examples of AI implementation at various scales to demonstrate the broad adoption of ChatGPT EDU. These examples show progression from individual institutional partnerships to system-wide and national implementations.
Evidence
Harvard used custom GPTs for 24/7 teaching assistants with syllabi and lecture notes; University of Pennsylvania applied it to student services including career management; ASU implemented over 250 projects from student mental health aid to research assistance; California State University partnership covers 500,000 students, faculty, and staff; Estonia country partnership for schools nationwide
Major discussion point
Scalable AI implementation models
Topics
Sociocultural | Development
The primary risk is students using AI to shortcut their learning rather than engage with materials for critical thinking development (Kevin Mills)
Explanation
Mills acknowledges the most significant concern about AI in education: that students might use it to avoid the necessary struggle and friction that true learning requires. He emphasizes that if students offload all their work to AI, they won’t develop essential skills and critical thinking abilities.
Evidence
True learning takes friction and struggle; students must engage with materials; if they offload all work to ChatGPT, they will not learn skills or gain critical thinking
Major discussion point
Balancing AI assistance with authentic learning
Topics
Sociocultural | Human rights
Agreed with
Agreed on
AI technology requires proper implementation to be beneficial in education
When used correctly as a learning assistant and tutor, AI shows powerful positive results in education (Kevin Mills)
Explanation
Mills argues that the key issue isn’t whether AI is inherently good or bad for education, but rather how it’s implemented. He emphasizes that when used appropriately as a supportive tool rather than a replacement for learning, AI can be highly beneficial for educational outcomes.
Evidence
Study after study confirms powerful results when ChatGPT is used correctly as learning assistant and tutor; the question is not if AI is good or bad, but how to use it correctly as a tool
Major discussion point
Proper implementation of AI in education
Topics
Sociocultural
Agreed with
Agreed on
AI technology requires proper implementation to be beneficial in education
OpenAI is conducting research collaborations to determine best practices and disseminate proper implementation methods globally (Kevin Mills)
Explanation
Mills describes OpenAI’s commitment to evidence-based AI implementation in education through academic research partnerships. The company is conducting longitudinal studies to establish best practices and then rapidly sharing these findings with educators worldwide to ensure proper implementation.
Evidence
Research collaborations with academic partners using longitudinal studies; partnership with American Federation of Teachers to launch AI institute; plans for similar initiatives worldwide
Major discussion point
Research-driven approach to AI in education
Topics
Sociocultural | Development
Agreed with
Agreed on
Need for research-driven approach to AI in education
Using AI to unlock human potential rather than just building better machines (Kevin Mills)
Explanation
Mills emphasizes that OpenAI’s ultimate goal transcends technological advancement to focus on human empowerment. He connects this mission back to his original motivation for entering education, emphasizing that technology should serve to enhance human capabilities rather than replace them.
Evidence
Personal mission from 20 years ago at UNESCO conference; focus on unlocking human potential drives the education community
Major discussion point
Human-centered approach to AI development
Topics
Sociocultural | Human rights
Creating a world where every educator has needed support and every student gets education to reach their full potential regardless of circumstances (Kevin Mills)
Explanation
Mills presents his vision for AI’s role in democratizing education globally. He envisions AI providing educators with the support they need to focus on what matters most while ensuring that every student, regardless of their background or circumstances, has access to quality education and support.
Evidence
Vision connects to personal mission of helping people reach full potential; goal of supporting educators and students regardless of birth circumstances
Major discussion point
Democratizing quality education through AI
Topics
Development | Sociocultural | Human rights
Agreements
Agreement points
Education as fundamental solution to global challenges
Speakers
– Kevin Mills
Arguments
Education is the core solution to major global challenges like peace and prosperity (Kevin Mills)
Summary
Mills presents education as the foundational approach to addressing major world problems like peace and prosperity, based on his realization at a UNESCO conference 20 years ago
Topics
Development | Sociocultural
AI technology requires proper implementation to be beneficial in education
Speakers
– Kevin Mills
Arguments
The primary risk is students using AI to shortcut their learning rather than engage with materials for critical thinking development (Kevin Mills)
When used correctly as a learning assistant and tutor, AI shows powerful positive results in education (Kevin Mills)
Summary
Mills acknowledges that AI’s value in education depends entirely on proper implementation – it can be harmful if used to avoid learning friction, but powerful when used as a supportive tool
Topics
Sociocultural | Human rights
Need for research-driven approach to AI in education
Speakers
– Kevin Mills
Arguments
OpenAI is conducting research collaborations to determine best practices and disseminate proper implementation methods globally (Kevin Mills)
Summary
Mills emphasizes the importance of evidence-based implementation through academic research partnerships and longitudinal studies to establish and share best practices
Topics
Sociocultural | Development
Similar viewpoints
Consistent focus on human empowerment and democratizing access to quality education as the ultimate goal, with technology serving as a means to unlock human potential rather than an end in itself
Speakers
– Kevin Mills
Arguments
Personal mission to help as many people as possible reach their full potential through global education (Kevin Mills)
Using AI to unlock human potential rather than just building better machines (Kevin Mills)
Creating a world where every educator has needed support and every student gets education to reach their full potential regardless of circumstances (Kevin Mills)
Topics
Development | Sociocultural | Human rights
Unexpected consensus
Acknowledgment of AI risks in education by AI company representative
Speakers
– Kevin Mills
Arguments
The primary risk is students using AI to shortcut their learning rather than engage with materials for critical thinking development (Kevin Mills)
Explanation
It is notable that Mills, representing OpenAI, openly acknowledges the significant risks of AI in education and emphasizes that improper use can be detrimental to learning. This transparent discussion of limitations by a company representative is unexpected and demonstrates responsible positioning
Topics
Sociocultural | Human rights
Overall assessment
Summary
Since this transcript contains only one speaker (Kevin Mills), there are no inter-speaker agreements or disagreements. However, Mills demonstrates internal consistency in his arguments, showing alignment between his personal mission, OpenAI’s approach, and the broader goals of democratizing education through AI. His arguments form a coherent narrative from personal motivation to practical implementation.
Consensus level
The presentation shows strong internal coherence in Mills’ vision for AI in education, with consistent emphasis on human empowerment, responsible implementation, and evidence-based approaches. The lack of other speakers limits assessment of broader consensus, but Mills’ balanced approach acknowledging both opportunities and risks suggests a framework that could build consensus among educators and policymakers.
Differences
Different viewpoints
Unexpected differences
Overall assessment
Summary
No disagreements identified – single speaker presentation
Disagreement level
This transcript contains only one speaker (Kevin Mills from OpenAI) presenting his views on AI in education without any opposing viewpoints or debate. The presentation is structured as a monologue sharing OpenAI’s perspective, achievements, and vision for AI in education. While Mills acknowledges challenges and risks (such as students using AI to shortcut learning), he presents these as problems to be solved rather than points of contention with other speakers. The lack of multiple perspectives or debate means there are no disagreements to analyze, though this also means important counterarguments or alternative approaches to AI in education are not represented in this discussion.
Partial agreements
Partial agreements
Similar viewpoints
Consistent focus on human empowerment and democratizing access to quality education as the ultimate goal, with technology serving as a means to unlock human potential rather than an end in itself
Speakers
– Kevin Mills
Arguments
Personal mission to help as many people as possible reach their full potential through global education (Kevin Mills)
Using AI to unlock human potential rather than just building better machines (Kevin Mills)
Creating a world where every educator has needed support and every student gets education to reach their full potential regardless of circumstances (Kevin Mills)
Topics
Development | Sociocultural | Human rights
Takeaways
Key takeaways
Education is identified as the core solution to major global challenges, with AI technology representing a transformative opportunity to democratize access to personalized, world-class education
OpenAI has rapidly become the world’s largest education platform with over 500 million weekly users, 40% of whom are under 24, and learning being the primary use case across all demographics
The education sector is rapidly transitioning from experimental AI pilots to widespread institutional adoption, with AI now considered essential infrastructure rather than optional technology
AI literacy has become a critical skill for the modern workforce, with employers expecting graduates to be proficient in AI tools
Successful AI implementation in education requires proper guidance and research-backed best practices to ensure students use AI as a learning assistant rather than a shortcut that bypasses critical thinking development
OpenAI’s vision focuses on using AI to unlock human potential and create equitable access to quality education regardless of geographic or socioeconomic circumstances
Resolutions and action items
OpenAI joined the UN AI Skills Coalition to expand access to AI education resources through UN platforms
Launch of research collaborations with academic partners to determine best practices for AI use in education through longitudinal studies
Partnership announced with the American Federation of Teachers to launch a new AI institute for proper implementation guidance
Commitment to disseminate research findings and best practices to professors and students globally as quickly as possible
Expansion of institutional partnerships from individual campuses to entire university systems and country-wide implementations
Unresolved issues
The fundamental challenge of preventing students from using AI to shortcut their learning and bypass the necessary friction and struggle required for true skill development and critical thinking
The need for comprehensive research to definitively answer ‘what is the right way to use AI in education’ across different contexts and educational levels
Scaling responsible AI implementation globally while maintaining quality and educational integrity
Balancing rapid technological advancement with the time needed for proper educator training and institutional adaptation
Suggested compromises
Positioning AI as a tool that can be ‘powerfully good’ when used correctly as a learning assistant and tutor, rather than viewing it as inherently good or bad for education
Focusing on research-backed implementation methods to ensure AI enhances rather than replaces the learning process
Combining rapid technological deployment with careful attention to proper usage guidelines and educator support
Thought provoking comments
40%. That’s the percentage of ChatGPT users under the age of 24. And can you guess what the number one use case for ChatGPT is? It’s learning. Across every demographic, every region, between 10% and 40% of every message sent on the ChatGPT platform is related to education and learning. So with half a billion users, we’re already the world’s largest education platform.
Speaker
Kevin Mills
Reason
This comment is profoundly insightful because it reframes the entire conversation about AI in education. Rather than discussing AI as a potential tool for education, Mills reveals that ChatGPT has already become the world’s largest education platform organically, driven by user behavior rather than institutional adoption. This challenges the traditional view of educational platforms and institutions.
Impact
This revelation shifts the discussion from ‘how might AI be used in education’ to ‘how do we responsibly manage AI’s already dominant role in education.’ It establishes the urgency and scale of the challenge, making subsequent discussions about partnerships and responsible implementation more compelling and necessary.
We know that true learning takes friction. It takes struggle. You have to engage with the materials. And if students offload all of that work to a tool like ChatGPT, they will not learn those skills and they will not gain that critical thinking. That said, when ChatGPT is used correctly, as a learning assistant and as a tutor, the results are powerful.
Speaker
Kevin Mills
Reason
This comment is thought-provoking because it directly addresses the fundamental tension in AI-assisted education. Mills acknowledges the core pedagogical principle that learning requires effort and struggle, while simultaneously arguing for AI’s educational value. This honest acknowledgment of the potential for AI to undermine learning demonstrates intellectual honesty and depth of understanding about educational theory.
Impact
This comment represents a crucial turning point in the presentation, moving from promotional content about AI’s benefits to a more nuanced, balanced discussion. It validates educators’ concerns while positioning the conversation toward finding the right balance, making the subsequent discussion about research partnerships and proper implementation more credible and necessary.
And not only is it a tool for learning all sorts of skills, using the tool itself is now an essential skill for the global labor market. Employers now expect students graduating from high school or college to be proficient in AI skills.
Speaker
Kevin Mills
Reason
This observation is insightful because it identifies a meta-skill requirement that creates a compelling argument for AI integration in education. It’s not just that AI can help students learn; students must learn AI to be employable. This creates a circular necessity that transforms AI from optional educational technology to essential curriculum.
Impact
This comment strengthens the argument for institutional AI adoption by connecting it directly to student outcomes and employability. It moves the discussion beyond theoretical benefits to practical necessity, making resistance to AI integration seem potentially harmful to students’ future prospects.
Speaking of office hours, faculty reported that they loved what it did to their office hours because now they got to use that time to get to know their students and their career aspirations rather than just answer questions about the content.
Speaker
Kevin Mills
Reason
This comment is particularly insightful because it illustrates how AI can enhance rather than replace human connection in education. Instead of dehumanizing education, the technology freed up faculty time for more meaningful human interactions. This addresses a common concern about AI replacing human elements in education by showing how it can actually strengthen them.
Impact
This example provides concrete evidence of AI’s positive impact on the human aspects of education, countering fears about dehumanization. It helps shift the narrative from AI as a threat to human connection to AI as an enabler of deeper human relationships in educational settings.
Because at the end of the day, we’re not here just to build better machines or smarter models. We’re here to use AI to unlock human potential.
Speaker
Kevin Mills
Reason
This closing statement is thought-provoking because it reframes the entire AI development enterprise in humanistic terms. Rather than focusing on technological advancement for its own sake, Mills positions AI development as fundamentally about human flourishing and potential realization, connecting back to his personal mission statement from 20 years earlier.
Impact
This comment serves as a philosophical capstone that ties together all previous discussions under a unified human-centered vision. It elevates the conversation from technical implementation details to a broader moral and social purpose, leaving the audience with a sense of shared mission rather than just product promotion.
Overall assessment
These key comments shaped the discussion by transforming it from a typical product presentation into a nuanced exploration of AI’s role in education. Mills skillfully moved the conversation through several phases: establishing AI’s already dominant presence in education, acknowledging legitimate concerns about learning effectiveness, providing concrete examples of positive human impact, and concluding with a unifying humanistic vision. The most impactful aspect was his willingness to address potential negative consequences honestly, which lent credibility to his positive claims and created a more sophisticated dialogue about responsible AI implementation in education. Rather than simply promoting OpenAI’s products, these comments collectively built a case for thoughtful, research-backed integration of AI in educational systems worldwide.
Follow-up questions
What is the right way to use AI in education?
Speaker
Kevin Mills
Explanation
This is a critical question that Mills explicitly states OpenAI is working to answer through research collaborations with academic partners, as it determines whether AI will be beneficial or harmful in educational settings
How can we ensure AI adoption is done responsibly in educational institutions?
Speaker
Kevin Mills
Explanation
Mills mentions this as one of OpenAI’s two primary focus areas when partnering with education institutions and governments, indicating it requires ongoing research and development
How can we prevent students from using AI to shortcut their own learning while still gaining the benefits?
Speaker
Kevin Mills
Explanation
Mills identifies this as the most prominent challenge, noting that true learning requires friction and struggle, and if students offload all work to ChatGPT they won’t develop critical thinking skills
What are the long-term effects of AI tutoring on student learning outcomes?
Speaker
Kevin Mills
Explanation
Mills mentions that OpenAI has embarked on research collaborations with academic partners to conduct longitudinal studies to answer questions about proper AI use in education
How can research-backed information about proper AI use in education be effectively disseminated to professors and students worldwide?
Speaker
Kevin Mills
Explanation
Mills states this as OpenAI’s second focus area in their research efforts, emphasizing the need to get research findings to educators and students as quickly as possible for proper implementation
Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.
Related event
