AI and Democratization of Higher Education

• In Event: Digital Learning in Higher Education: Examining Environments, Engagement, and Empowerment

Proposal

Whether AI can democratize our higher education or not depends on whether the technology can move in the right direction for collective human benefits. The main contributions are an urgent call to generate greater awareness of the inbalanced power in the AI system and offer some educational insights such that the benefits of AI can be shared by many people. It is essential to consider wider social, ethical, and political contexts for the deployment and development of AI and to reactivate the lost core humanistic values that can easily be flattened and denied in the current AI system. It is not a matter whether or not human beings are in the loop regarding AI; it is a matter of inbalanced power between AI technology companies and consumers.

The current structure of AI social-technology system is based on an economics of scale, efficiency, multidisciplinary expertise, giant capital investment, and an intimate integration among a few technology companies such as the collaboration between Open AI and Microsoft. The integration of AI into the higher education workflow is not just a technological shift; it requires a cultural and organization transformation. Many societal ethics and core human values such as mercy, dignity, and genuine human connections are reduced to AI’s values of optimization of human preferences (Russell, 2022; Tasioulas, 2022).

Research Questions:
1. What is our collective vision of a desirable and achievable global higher education system that leverages AI while centering on core human values?
2. Will the idea of having humans in the loop increase human control upon the AI system?
3. How should our global higher education system equip our students to know how to work closely with AI and also how to develop good practices that better incorporate broader values for human flourishing?

Methodology

I build on extensive literature review about AI system and also reflections about decades of teaching experience of using AI in higher education in Asia, U.S. and Europe.

Findings
Answer to research Q.1: Countries need to be well-prepared for digital transformations with relevant infrastructures, institutional environments and AI-complementary skills for harnessing the benefits of AI. AI can automate and augment human capabilities to scale up effective and low-cost solutions quickly. From a social technological system perspective (Bommansani et al., 2023), the public is educated enough to know what AI is and is incentivized enough to balance the dominant perspectives of a select few rich technology companies. Policy-makers and regulators are recommended to introduce all technical, social, economic, and scientific dimensions of AI systems in society through the following five key tasks: clarify what AI is and focus on the right risks and opportunities (i.e., demystification); create a functional ecosystem to make AI work (i.e., contextualization); involve diverse stakeholders from civil society to address relevant values and interests affected by using AI technology (i.e., engagement); develop a directive framework (i.e., regulation); and engage wisely with other global actors (i.e., societies) (Sheikh et al., 2023b). These five tasks steer the process of co-development between technology and society. The progress of AI and the shared benefits of AI are most likely to be maximized in democratic societies that allow many stakeholders to contribute and control the direction of AI’s development (Acemoglu and Johnson, 2023).
Answer to research Q. 2 Many stakeholders of education in the U.S. adopt “humans in the loop” as a key criteria for educational use of AI (U.S. Department of Education, 2023:53). However, it is not a matter of whether or not human beings are in the loop regarding AI; it is a matter of an imbalanced power relationship between AI technology companies and consumers (Tschider, 2023). Most AI companies do not invest substantial time in user interface/user experience/human factors design with respect to AI and can easily attribute the AI design problems to human beings. Indeed, human beings are invited to interrupt the AI process that we do not fully understand and may interrupt when they are not needed or fail to perform when they do. In addition, keeping humans in the loop may be used to function as a “stop gap” for actual effective standards or legal developments. Humans are not often trained to interact with AI effectively, because creators may not be able to anticipate unexpected behaviors or hallucinations. Humans may defer to technology when they don’t have an equivalent expertise and sometimes they will have an automation bias.
Answer to research Q.3: Educators are encouraged to help students to know AI and AI ethics, and cultivate an environment for responsible usage of AI. Educators and students should question the paradigm of maximizing output when our personhood is under attack in the new world of AI systems. We must not let AI be the “opium of personhood.” Eshelman, Lam, and Cook (2012) invite and encourage educators and students to fully experience personhood, mindfulness, and meditation in their own lives such that they can be better equipped to engage with evolving technologies. Educators need to incorporate a process view and increase the consciousness of our students and peers toward our being and becoming (Vaill, 1989).
AI system gradually challenges our democratic values and our expectations for what high quality higher education look like in a thriving society. With better public education about what AI is and a broad engagement of the public about the direction of AI, we can rebalance the power between AI technology companies and consumers. We have to know what AI is and understand its limitations from a social-technological system perspective. The effectiveness of AI depends on the context that supports AI. We need to be conscious of our core human values such as genuine human contact and the dignity of work that gradually are being replaced by AI as it mainly focuses on economics of scale, efficiency and lower costs. We need safe AI rather than an AI safe culture!

Author

• Maria Lai-Ling Lam, LCC International University