In a recent assignment for LRNT 524, I was asked to research and evaluate an instructional or learning design innovation. I chose to focus on a learning design innovation known as adapted learning in which technology is used to provide customized learning experiences for learners based on the needs of the individual by creating unique pathways and progression of learning material and activities (Leaders, 2022). As part of this discussion, I shared how artificial intelligence (AI) has progressed adapted learning, introducing tools such as ones that can detect the learners cognitive state based on conversational elements which it uses to guide the learner through productive conversations to enhance learning (Capuano & Caballé, 2020).
As a follow up blog post, I have been asked to explore a learning innovation and discuss it’s impact. Upon further research in adapted learning, I discovered deep learning which is a AI-based technology that attempts to mimic the human brain through image or object detection using a multi-layererd approach to make intelligent decisions (Han & Xu, 2020; IBM, n.d.). Deep learning is a tool that can enhance adapted learning through a complex analysis of objects that goes beyond a linear process of adaptation, similar to neural connections and processing in the human brain. After an initial exploration of the literature, I found that there were many interpretations of what deep learning is and that the term is often used interchangeably with machine learning and deep neural networks. Kavlakoglu (2020) at IBM contends that deep learning is a subset of machine learning with deep neural networks making up its algorithms. The lack of common language and understanding in the literature of what deep learning is made my research on this topic more difficult.
Furthermore, there does not seem to be a lot of literature about the use of deep learning technology in the context of learning in formal education. Deep learning is most known for its use in other sectors and innovations such as driverless cars where a car must learn how to detect a stop sign (Venkateswaran et al., 2021), a true focus on machine learning rather than human learning. However, I do see that deep learning can have a large impact on human learning, but perhaps in more of an ‘unlearning’ way. As deep learning technology advances in an effort to simulate processes of the human brain, less effort is required from humans to learn and complete tasks. This exclusive reliance on technology to make decisions and problem solve could be problematic for two reasons. Firstly, AI technology, including deep learning is still far from matching human intelligence, making outputs not always accurate (McClelland & Botvinick, 2020). And secondly, (and this is my own pondering) I am left wondering if the long-term use of AI, including deep learning technology, will change the architecture and performance of the human brain over time. With the use of AI driven deep neural networks, are we losing the opportunity to be develop our own neural networks in our own brains through problem solving?
I see some potential for deep learning technology in creating more meaningful learning experiences for culturally diverse learners. For example, this technology could adapt learning content and images to represent the culture and worldview of the learner while still meeting the learning outcomes. In that case, this could lead to greater inclusion and sense-making according to diverse worldviews. However, there could still be bias present as there is a wide range of diversity within a culture, which will have to be considered in the design and data collection for deep learning technology.
References
Baker, C. (2022, June 3). What is adaptive learning and can It work for business? Leaders. https://leaders.com/articles/innovation/adaptive-learning/
Capuano, N., & Caballé, S. (2020). Adaptive learning technologies. AI Magazine, 41(2), 96-98. https://doi.org/10.1609/aimag.v41i2.5317
Han, Z., & Xu, A. (2020). Ecological evolution path of smart education platform based on deep learning and image detection. Microprocessors and Microsystems, 80, 103343. https://doi.org/10.1016/j.micpro.2020.103343
IBM (n.d.). What is deep learning? https://www.ibm.com/topics/deep-learning
Kavlakoglu, E. (2022, May 27). AI vs. machine learning vs. deep learning vs. neural networks: What’s the difference? IBM. https://www.ibm.com/cloud/blog/ai-vs-machine-learning-vs-deep-learning-vs-neural-networks
McClelland, J. L., & Botvinick, M. M. (2020). Deep learning: Implications for human learning and memory. PsyArXiv. https://doi.org/10.31234/osf.io/3m5sb
Venkateswaran, C., Amudha, M., Ramachandran, M., Saravanan, V., Vennila, T. (2021). A study on artificial intelligence with machine learning and deep learning techniques. Data Analytics and Artificial Intelligence 1(1). https://secureservercdn.net/50.62.90.29/d8a.8cf.myftpupload.com/wp-content/uploads/2022/01/A-Study-on-Artificial-intelligence-with-Machine-learning-and-Deep-Learning-Techniques.pdf
Some great thoughts here Leah. I, too, am intrigued about the potential for deep learning in K-12 education. You brought up the concern of the long-term use of AI affecting our brain’s ability to problem solve. As an educator, I am hesitant to implement AI into my classroom until more research is done. Other challenges include data storage and privacy, which are real concerns for educators. Would you see these same challenges in the healthcare field as well?
Thanks for reading, Terry. I have spent a lot of time working in the area of developing physical literacy in the early years and studying how the brain develops, strengthens and prunes neural connections so it was a natural thought to wonder how this technology could impact our own brain development and plasticity. As for your question about data storage and privacy, although I am new to working in the healthcare sector, I know that this is a high priority and something that will always be at the forefront of any technology adoption and use. Yet the benefits of AI in healthcare are high so I am sure there are rigorous procedures to vet new tech.
Leah
You have done a great job further exploring an innovation that you found interesting. Your post is enjoyable and reflective. I like that you noted the challenge around terminology and definitions a barrier to finding literature. This is a common challenge when searching innovation as there may be work to do for the clarity you are seeking (a knowledge gap;)) I think an important differentiation to make here is “deep learning” vs “deep learning technology”. Use of these with consistency would make your post clearer. You have posed some great thought provoking questions and provided a strong example of use for international students – well done!
Hi Leeann,
Thanks for reading and the feedback. The term deep learning has been interesting in this context as I use this term often to describe meaningful learning from a human perspective, yet when I google deep learning, it is defined as a type of machine learning. Wild! I have never thought of deep learning beyond my own understanding and use of the term. It is good to know!
Leah