Predictions for 2025:
AI, Asset Orientations, and Collaborative Learning
Introduction:
Since it’s late-January, it’s too late to say “Happy New Year” and too late to make end-of-year predictions. Unless, however, the predictions are for end-of-year 2025! So let's leap ahead and explore what 2025 might hold for us!
My Predictions for 2025:
AI-Driven Asset Oriented Individualization: By the end of 2025, AI will finally address the need to leverage the diverse backgrounds of students, moving beyond current implementations of “personalized learning” to a truly asset-oriented approach.
AI and Student Collaboration: AI will become as integral to fostering student collaboration as it currently is in personalization efforts.
Why 2025 and Not 2024?
Presently, the educational landscape is focused on hyper-individualized learning models, powered by general AI models. I expect that 2024 will be the year of disillusionment, and by the end of 2025 we'll witness meaningful and transformative changes that are more aligned with the true needs of students and teachers.
Asset-Oriented AI:
What do we mean by "asset-oriented AI"? It's a shift from focusing solely on remedying student “deficits” to recognizing and building upon their strengths. This includes their unique cultural and linguistic skills. Unfortunately, many current "personalized" learning environments fall short in this regard. They often ignore the rich diversity in student backgrounds, and often even inadvertently perpetuate biases. However, there's hope that by 2025, new Generative AI models, trained on more diverse datasets, will pave the way for truly asset-oriented learning experiences.
The steps getting us from where we are today to this vision for 2025:
Disillusionment and Hope: Disappointment in current GenAI-based tutoring solutions will give rise to the search for alternatives.
Recognition of Alternatives: Insights from research will guide us towards new approaches in AI-based educational technology (for instance, this paper on Asset-based instruction and assessment).
Implementation of New Models: The cost and effort of creating customized AI models is decreasing at an exponential rate. GenAI models trained on customized datasets will soon enable truly inclusive and strength-based personalized tutors.
The result will be AI-driven personalized tutors that finally live up to their promise of supporting all students. However, there is only so much that personalized tutors can do, bringing me to the second prediction.
Supporting Student Collaboration with AI:
The second key prediction for 2025 is the elevation of AI in supporting student collaboration. Over the next year, we can expect the limitations of hyper-individualized learning environments to become apparent to all.
This will increase recognition of the importance of collaborative problem-solving and group sensemaking. These skills are highly valued in today's workforce, have significant academic and social benefits to students, and fulfill a true teacher need. Innovations like OKO’s support of small group collaboration, and SRI Education's research on automating the detection of different types of collaborative behavior show promising avenues for AI supporting student collaboration. As more innovators identify opportunities in the collaborative learning space we can expect exciting advancements, from the obvious (such as providing real-time language translation) to the truly innovative (starting with, and moving beyond, the prior examples).
Conclusion:
These predictions are not just about technological advancements. They are about reshaping how EdTech companies view learning, as they recognize their role in leveraging the diversity of student strengths and fostering collaborative skills essential for the modern world. Stay tuned, and let’s see what 2025 holds!
