From pen and paper to artificial intelligence: Understanding the limits of educational technologies

There are efforts to improve educational data management and sharing practices in Africa through the adoption of technologies. [iStockphoto]

Educational technologies provide learners, teachers, and schools with many benefits. These benefits include features that provide learners with practice opportunities and feedback; virtual teaching and learning that allows learning beyond school walls; learning management systems that help teachers provide instruction and organize information; technologies that give learners opportunities to participate in different ways (e.g., sound, touch); and accessibility features such as text to speech, and voice-overs.

Globally, learning technologies exist in the context of pervasive inequalities. The digital divide is characterized by differences in access to technology as well as in how people use technology. Countries in Africa have embraced education technologies in an effort to innovate classrooms and practices from basic education to institutions of higher learning.

They have drawn up digital learning policies and reformed curriculums to incorporate courses in digital literacy and coding. When it comes to educational technologies, having the newest technology or having a high number of technological devices or platforms is seen by many as a sign of success and advancement.

What people are talking about these days?

Artificial intelligence (AI) uses stored information and special formulas (algorithms) to complete mundane tasks based on individual habits and preferences. An individual's information can include age, gender, and preferences. For example, when shopping online, AI embedded in the website can use a customer's past shopping habits and preferences to suggest highly personalized options.

Similarly, streaming services such as Netflix use an individual's profile to recommend shows one is likely to find interesting. In education, AI can improve access and instruction for learners, through tailoring material and instruction and assessment practices based on their performance and learning styles.

Additionally, AI can be used for automatic invigilation (proctoring) and checking for plagiarism to ensure that learners do not consult unauthorized material or present un-original work. The COVID-19 pandemic increased the adoption of educational technologies and AI can improve the effectiveness of these technologies and the learning experience of learners.

ChatGPT, an example of an AI chatbot (a computer program that mimics a human conversation) developed by OpenAI, generates human-like text and interacts in a conversational way. Public conversations about this tool's application in life and education have included a combination of excitement and concerns.

Can technology solve all the problems in education?

Technology enthusiasts and developers oversell educational technologies and their potential benefits. However, a keen perspective finds that educational technologies fall short of the marketed promises. For instance, research is still sparse on whether they improve learning outcomes or whether they have the capacity to increase inclusion and equity.

Arguably, the educational community could benefit from asking different questions and being realistic about the limits of educational technologies in order to further harness their potential.

As Africa expands its digital technology blueprint in various sectors, many observers and onlookers continue to remark at the impressive potential and speed of digitization. However, one question that could benefit stakeholders including policymakers, teachers, school leaders and school administrators calls for a deeper analysis and understanding of technology adoption.

For starters, technologists have largely debunked a fear that educational technologies will get rid of teachers as AI takes over routine tasks that humans can accomplish. The reassurance that teachers will keep their jobs, falls short because, notwithstanding, increased AI may lead to loss of jobs and income in different fields.

Further, the threats raise unforeseen challenges where humans need to work alongside robots, forgetting that human beings are not machines and cannot perform at the same level. Take higher education; with prospects of AI becoming popular, it is likely that certain tasks will be automated leading to career displacement.

While AI is becoming more common and can be used to help make things easier, AI systems are only as good as the data it's fed. Therefore, it is important that the information used to develop AI systems is from a similar context as the target implementation area, accurate, and relevant to avoid harm and biases in its decisions.

This is a huge challenge in Africa more so in the education sector where data is often unstructured and dispersed, that is, the data is rarely properly organized. The reason for this disorganization stems from the fact that institutions struggle with data management and sharing practices, leading to insufficient machine-readable data for training educational tools powered by AI.

This leads to where most educational technologies adopted in Africa run on models trained using incomplete/unrepresentative data or data from abroad. Such technologies are likely to reinforce existing disparities in learning outcomes as they are often decontextualized and seldom informed by the local African context.

AI-powered educational technologies are also susceptible to the biases of the developers which can affect the usefulness of the technologies and may entrance existing inequalities. Also, if left unchecked, learners can pass off AI-generated work as their own, which can undermine the integrity of the learning process. As a result, stakeholders should initiate in-depth conversations on technology adoption involving a healthy level of skepticism and analysis about short and long-term implications.

How do we engage in critical conversations about educational technologies?

Critical thinking and willingness to challenge deeply held assumptions and beliefs about the magic and neutrality of technology are necessary to uncover aspects of concern that surround educational technologies.

Practically, teacher training should include topics that explore the benefits of technology and also increase teachers' awareness of how to use educational technologies in classrooms to promote inclusion and harness all students' creativity.

Another way to engage more concretely is for users including teachers to participate in the design process, especially content development that emphasizes a focus on the uniqueness of student learning and needs.

Moreover, content is needed that promotes diverse student voices and characters. In line with engaging in the design and development process, stakeholders need to clarify the aims of initiatives that encourage innovation. For instance, beyond the contribution to the world of work, what are the additional benefits and aims of the coding program?

Lastly, teachers and school leaders at all levels should engage in conversations on the use of educational technologies that prioritize sound instructional principles. Is the focus more on automated rote learning or is the focus on developing creativity and ensuring all students, particularly girls are encouraged to participate in STEM fields.

Additionally, how can we challenge learners to verify information all around them? Engaging the youth at an early age holds promise in questioning the ways that technologies are embraced in society, as technology can be used for positive social change.

There are efforts to improve educational data management and sharing practices in Africa through the adoption of technologies and initiatives like the World Bank Open Data and RELI-microdata portal among others. However, we need to systematically target more institutions to share their data and make it machine-readable for efficient deployment of AI-powered technologies.

There is immense potential for the transformation of the education sector in Africa using technology given the abundant potential and talent. Educational stakeholders can leverage education technologies including AI-powered varieties to improve access to education, equity, inclusion and learning outcomes of learners through intentional and sustained efforts.

Elisheba Kiru and Kennedy Kamau.

Dr Kiru is a Post-Doctoral Research Scientist & Mr Kamau is a Data Scientist.