AI in Education
-
Understand, predict, and guide the impact of generative AI in education
The sudden, widespread, and largely unplanned adoption of genAI tools in education is unprecedented. Historically, innovations find their way slowly to classrooms, after long periods of testing and refinement in factories, offices, and homes. In contrast, teaching and learning have quickly emerged as the primary use cases for genAI tools to the extent that leading companies like OpenAI are quickly retooling their business plans for the education market.
In real time, teachers and students in every field of study are hastily improvising and negotiating new norms and expectations surrounding the use of AI tools. Administrators, faculty, and students alike are bombarded by ads making untested claims about AI's promotion of learning.
Using quantitative, qualitative and innovative approaches, our research seeks answers to these core questions:
*How can we minimize the risks and maximize the rewards of AI-powered learning?
*How can we distinguish between learning-enhancing and learning-inhibiting AI-powered learning practices?
*How can we formulate and communicate new norms and expectations around AI use that reflect and reinforce discipline-specific values and goals?
The AI In Education team responds to this critical moment by considering the current, likely, and ideal impacts of GenAI tools on education from four complementary vantages:
Understanding users: This subteam seeks to understand the attitudes and behaviors and competencies of students, faculty, and administrators as they relate to the application of GenAI tools in educational settings.
Evaluating tools: This subteam explores the capabilities and limitations of generative AI tools for educational purposes, including spin-off technologies such as AI detection tools.
Assessing impacts: This subteam performs technological assessments to identify the current and likely risks and benefits of AI on student learning.
Innovating EdTech: This subteam innovates new educational technology aimed at maximizing the benefits and/or mitigating the risks of generative AI in education.
Methods and Technologies
- Experiment design
- Qualitative research (interviews, focus groups, participant observations, case studies)
- Quantitative analysis (survey design and implementation, data analysis, comparative performance assessment)
- AI-powered coding platforms (Cursor, Github Copilot)
- Python NLP libraries (NLTK, spaCy)
- GenAI detection tools (ZeroGPT, GPTZero, Turnitin)
- Large language models (Claude, Gemini, ChatGPT)
- Development (prototyping, fail-testing, integration testing)
Areas of Interest
- Computer Science
- Education
- Educational Psychology
- Ethics and Policy
- Human-Computer Interaction
- Mathematics
- Website and App Development
- Writing Studies
Subteams
- Understanding users
- Evaluating tools
- Assessing impacts
- Innovating EdTech
Partners
- Expository Writing Program
- NYU Writing Center