Project: AI Impacts on Human Learning an Development

[Steven-Birkinbine]

Setting up specific discussion for the project AI Impacts on Human Development and Learning. Please use this thread to discuss this project and provide specific sources/research as needed

[ofrancon]

You might be interested by this work:

• AGI: Artificial General Intelligence for Education
• Fine-tuning ChatGPT for automatic scoring

[Steven-Birkinbine]

US Department of Education: https://www.ed.gov/sites/ed/files/documents/ai-report/ai-report.pdf

Microsoft reports reduced critical thinking in the abuse or overreliance on AI: https://www.microsoft.com/en-us/research/uploads/prod/2025/01/lee_2025_ai_critical_thinking_survey.pdf

The current state of education in the US: https://usafacts.org/topics/education/

Public education stats: https://www.census.gov/topics/public-sector/education-finances.html

[Steven-Birkinbine]

Alpha School and Unbound Academy are both using similar models of AI led classrooms with human instructors there for motivational and emotional support. I have reached out to both of them to see if they are willing to share their findings.

https://unbound.school/
https://alpha.school/

[makkzone]

It may help
https://www.sciencedirect.com/science/article/pii/S2666920X2100014X
https://elearningindustry.com/ai-in-stem-education-transforming-learning

[ngodard2]

Here is some resource on how important it is for human beings to talk to children for their brain development (vs. machines/ AI).
https://www.lena.org/resources/
https://www.lena.org/

[Steven-Birkinbine]

Progress Update - May 2025

Help Wanted: If anyone has any corporate or institutional contacts that would like to be involved in studying how AI is used in Education, please let me know and I will include them when I send out official requests next month.

Help Wanted: If there are any corporate sponsors that would like to be involved and may have or know how to build effective AI Agents for the purposes of education, I would love to discuss a partnership.

The Needs

Need Area	Description
Clarification on AI Usage	Instructors worldwide are clamoring for clear guidance from governing bodies on AI best practices, the role of technology, and the teacher’s role—especially since the AIs already exist.
Personalized Student Progress	There is an opportunity to allow students to advance at their own pace by tailoring learning to individual needs.
Teacher Education	There is a lack of education and training for teachers on how to effectively utilize novel teaching methodologies in the classroom.

---

The Context

What AI Can Do

Capability	Description
Collaborative AI Agents	Utilize specialized AI agents (e.g., History Bot, Math Bot, Feedback Bot) that work together so a student can ask a batch of bots and receive a deeper, more coherent response.
Personalized Learning	Create individualized learning pathways to enable each student to learn as fast as possible.
Data Network Effects	Leverage network effects where long-term student improvements increase as more students utilize the AI.

The Future of Classroom Design

Design Element	Description
Human-Centric & Learner-Driven	Transition towards education that prioritizes the learner’s experience by placing them at the center of the process.
Four Innovative Designs	Incorporate Experiential Learning, Microlearning, Adaptive Learning, and Game-Based Learning into classroom structures.
Streamlined Modular Certification (SMC)	Develop a certification program for teachers that includes a competency-based framework with portfolio assessments, regular calibrations, learning predictive analytics, and student-driven debate.
Rethinking Assessments	Move away from high-stakes exams towards ongoing, data-driven evaluations that capture understanding through in-class interactions.

Modern Problems

Problem Area	Description
Cheating	It can be easier to cheat than to do work, which may lead to increased evaluation time and efforts to detect dishonesty.
Critical Thinking Decline	The current utilization of AI may be damaging students’ critical thinking abilities.
Social Connectedness	A distinct lack of social connectedness is dehumanizing people and has led to what some describe as the "lonely generation."

---

Learner Personas

Learner Persona	Definition / Description
Active	Learns by engaging directly with activities, discussions, group work, simulations, or hands-on projects.
Reflective	Prefers to think things through by carefully processing and analyzing information before engaging.
Sequential	Best learns through a clear, step-by-step progression where each concept builds logically on the previous one.
Global	Prefers starting with the big picture before zooming into the details; benefits from overviews and thematic summaries.
Digital Native	Grown up with technology, these learners excel in fast-paced digital environments with multimedia and interactive experiences.
Digital Immigrant	Adapted to technology later in life; they thrive with blended experiences that integrate traditional methods with digital tools.
Intrinsic	Motivated by internal rewards—pursuing learning for its own interest, curiosity, and personal growth.
Extrinsic	Driven by external rewards such as grades, praise, and recognition.
Metacognitive	Learns best by being self-aware of their learning processes, planning, monitoring, and evaluating their strategies.
Experiential	Thrives on practical application through experiments, fieldwork, internships, or scenario-based learning.
Gamified	Engages and learns best through game-based elements such as points, badges, levels, and leaderboards that make learning feel like an interactive challenge.
Challenger	Energized by difficult tasks, debate, and complex problem solving; benefits from challenges that push them out of their comfort zone.

---

Expected Population

Population Group	Description
Undergrad Students	Undergraduates across the US, Canada, and EU.
Doctoral Students	Doctoral candidates across the US, Canada, and EU.

---

Other Topics of Research

Research Focus	Description
Predictive Model for Educational Help	Can we create a predictive model to identify where students need help and where AI can be most beneficial? (Inspired by Duolingo’s approach)
Appropriate Age for AI in Classrooms	Should AI be introduced only for children aged 13 or older? Some research in Early Childhood Education suggests that introducing technology earlier may be harmful.
Evaluating Student Readiness	How can an educational AI determine if a student is ready to progress beyond relying solely on self-reported data?
Evolving Ethical & Regulatory Frameworks	In what ways do ethical and regulatory frameworks need to evolve as AI becomes more autonomous in education?

---

Roadmap

Date	Milestones
June, 2025	- Define Survey Questions - Define Semi-structured Interview Questions - Identify schools/companies that would serve as good case studies - Finalize Letter of Introduction - Create initial AI usage guidance for classrooms (both Student and Teacher editions) using TPACK as best practice
July, 2025	- Reach out to companies and institutions to create a DMZ for learning about AI in education - Potential companies include: Duolingo, Khan Academy, LENA, LEAD, X-Culture, Complexly, Nebula
August, 2025	- Run a pilot focus group to ensure survey and interview data are useful; adjust protocols as needed - Apply various analysis techniques to maximize data viability
November, 2025	- Check in on student progress - Confirm there is no obvious harm or damage from the study - Adjust course as necessary to ensure involved students do not fall behind