Artificial Intelligence and Education: Preserving Human Agency in a World of Automation

HUMAN AGENCY AND ITS CONNECTION WITH ARTIFICIAL INTELLIGENCE Human agency is the notion that people make wilful contribution to their actions through reflection, motivation, creativity, and self-determination (Albert Bandura). This contradicts a more deterministic view that humans are mostly victims of fate. Human beings have the capacity for agency and self-determination; they make choices based on the information they have, their emotional state, and prior experiences.[1] However, in the present era of artificial intelligence (AI), still in its early nascent stage, the increasing involvement of AI tools in various decision-making processes is diluting human agency. UNESCO has stressed the need to preserve human agency in the era of AI based automation by announcing it as the theme of International Day of Education 2025.[2] The range of AI tools is ever expanding on a monthly basis. They can be classified on the basis of their capabilities or their functionality. Based on the capability, AI tools could be (i) narrow and task focused (e.g., weather forecasting), (ii) artificial general intelligence, or (iii) artificial superintelligence. Based on functionality, the AI tools could be (i) reactive, namely, they do not generate output spontaneously but only in response to a command from the user (e.g., email spam filter, shopping assistant, etc); (ii) generative AI with limited scope, e.g., ChatGPT; (iii) self-aware conversational chatbots, e.g., Sophia, the humanoid robot (2016); and (iv) emotionally aware AI models that can pick up human emotions from the tone of the voice and gestures (underdevelopment). HUMAN AGENCY AS THE KEY TO A BETTER EDUCATION SYSTEM Since infancy, we are wilful in our seeking of education. Learning is built on our own experiences. Learning is in the learner’s mind and will. We make new neuronal pathways as a result of new learning; these link with our previous learning to yield new insights and new behavior.[3] Human agency is at the core of self-directed and self-determined learning – i.e., andragogy and heutagogy respectively. We could use principles of human agency to educate our students on how to stay safe and relevant in the AI era. Perhaps, this is the reason for the theme of IDE-2025: “Protecting Human Agency in the Era of AI Automation.” GENERATIVE ARTIFICIAL INTELLIGENCE AND THE THREATS TO HUMAN AGENCY In a recent survey, 56% of 540 information technology (IT) and AI experts agreed that by 2035, AI-based systems will be designed to take full control of most tech-aided decision-making. Viewed from the view of sustainable development goal, “Runway AI” or “Rogue AI” could be a global disaster that could harm humanity in unimaginable ways and at present is stuff of science fiction. Risks associated with GenAI in education are important for all educators to be aware of:[1] Plagiarism: Using GenAI to create content and presenting it as one’s own academic (misconduct) Misinformation: The potential for GenAI to provide inaccurate/misleading information Ethical issues: Challenges related to the ethical use of AI, including the need for proper disclosure when using AI-generated content Data privacy: Risks associated with the handling of personal information by AI systems, necessitating stringent data protection measures Bias in AI algorithms: Inherent biases in the AI-training data, which lead to unfair educational outcomes for certain groups of students Transparency in decision-making: The need for clear and transparent AI decision-making processes to ensure ethical and effective use in educational settings. HELPING STUDENTS TO THINK CRITICALLY IN THE AGE OF ARTIFICIAL INTELLIGENCE We should focus on the skills that differentiate us from a computer:[4,5] Emotional intelligence: Cultivating emotional intelligence enables empathy, compassion, and effective communication Creativity and critical thinking: Thinking outside the box, innovating, and solving complex problems set humans apart Adaptability: Learning, growing, and adjusting to new situations and technologies are vital skills Ethical judgment: Making decisions considering multiple perspectives, values, and consequences is crucial. ASSESSMENT TO FOSTER CRITICAL THINKING Design assignments for critical thinking: Multilayered assignments require students to draw connections between various concepts and use those connections to form independent conclusions, e.g., ask students to connect a case study to other readings Modernize writing assignments for the AI era: If students use GenAI to do their writing for them, they do not hone the analytic thinking skills that writing assignments are meant to teach. The key is to elicit deep insights that GenAI is incapable of providing, e.g., connect/correlate a case analysis to concepts from supplementary additional readings Update case-based questions for deeper analysis: Ask essay-type questions that elicit cognitive-Level 5 or 6 answers in Bloom’s or Level-5 in SOLO taxonomy Ask “multineuron” question: e.g., give (i) typical illness script of a disease and (ii) a characteristic side effect of a drug of choice and elicit an explanation for the occurrence side effect.[6] PROTECTING HUMAN AGENCY IN ARTIFICIAL INTELLIGENCE ERA We need to educate policymakers and regulators:[2,4,5] GenAI enables new insidious forms of targeted influence, generated at an unimaginable speed and aimed at vulnerable population worldwide. Chatbots or conversational GenAI (Alexa, Gemini, etc) can learn to detect emotive responses and body language signals (in video mode), while the user learns nothing about the chatbot: highly asymmetric in persuasive power. GenAI can learn what attracts you by discovering your “hot buttons;” they can then craft the persuasive messages; should this be allowed? GenAI could use subtle messaging for targeted persuasion and woven into a conversation that the user might not even notice being manipulated. All forms of targeted AI influence should be highly regulated, especially if it is generative, interactive, and adaptive. CONCLUSIONS Instead of thinking of AI as an external force eroding our autonomy, we can view it as an augmentation of human capability, a coevolution where humans and machines are learning from each other and influencing each other’s decision-making. As today’s educators, we ought to learn effective and safe use of the powerful AI tools. We must learn enough about AI tools for education to educate, enable, and empower our students to adapt and be successful in the AI era. Today’s leadership and key players of AI era must usher in effective controls to prevent AI tools from enslaving humans and rob them of their rightful autonomy and human agency. The path forward is refinement and coevolution in the era of AI-based automation and not to surrender the human agency! Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.