SUMMARY - Teaching in the Age of AI and Algorithms
Artificial intelligence is transforming nearly every sector—and education is no exception. AI tutoring systems promise personalized instruction at scale. Generative AI produces essays, solves problems, and creates content that students once had to create themselves. Algorithms influence what information students encounter and how they're assessed. The emergence of powerful AI tools fundamentally challenges assumptions about what students need to learn, how they should demonstrate learning, and what teaching should look like. Navigating this transformation requires thoughtful engagement with both AI's possibilities and its problems.
AI's Educational Applications
Adaptive learning systems use AI to personalize instruction. Based on student performance, these systems adjust difficulty, suggest resources, and provide feedback. They promise individualized attention that human teachers can't provide to every student simultaneously. Commercial products implementing these approaches are already in many Canadian classrooms.
Intelligent tutoring systems go beyond adaptive content to engage in teaching interactions. They can explain concepts, respond to questions, provide worked examples, and guide problem-solving. Early systems were domain-specific; generative AI enables more flexible tutoring across subjects. Whether these systems genuinely tutor or merely simulate tutoring is debated.
AI assessment tools automate grading and feedback. Essay scoring algorithms evaluate writing; automated feedback systems respond to student work. These tools promise to reduce teacher workload while providing faster feedback. Questions arise about whether AI assessment captures what matters and whether it introduces biases.
Generative AI—ChatGPT and similar systems—creates text, code, images, and other content based on prompts. Students can use these tools to produce work that previously required their own thinking and effort. This fundamentally challenges assessment approaches that assumed produced work demonstrated student capability.
Challenges Generative AI Creates
Academic integrity is immediately challenged when AI can produce work indistinguishable from student work. Essays, reports, problem solutions, and code can be generated rather than created by students. Detection tools exist but are imperfect and create arms races with generation tools. Fundamental questions arise about what counts as student work when AI assistance is ubiquitous.
Assessment validity is questioned when AI changes what work demonstrates. If a student submits an AI-assisted essay, what does the essay demonstrate? Ability to prompt AI effectively? Skill in editing AI output? Nothing about writing ability itself? Assessment approaches developed before AI assumed certain relationships between work products and underlying abilities that AI disrupts.
Learning processes may be short-circuited if AI replaces student effort. Learning often happens through struggle—working through problems, revising drafts, constructing understanding through effort. If AI eliminates this struggle, what happens to learning? Students who let AI do their work may pass courses without developing capabilities courses were meant to develop.
Equity concerns arise from differential AI access. Students with internet access, technological literacy, and awareness of AI tools have advantages over those without. If AI-assisted work is judged against non-AI-assisted work, those with AI access appear to perform better. This creates new forms of inequality in educational systems.
What AI Changes About What Students Need
Knowledge acquisition may matter less when AI provides information instantly. Memorizing facts becomes less important when any fact can be retrieved immediately. This shifts emphasis from knowing content to other capabilities—though debates continue about whether some foundational knowledge remains necessary.
Higher-order thinking may matter more. Analysis, synthesis, evaluation, and creative thinking—capabilities at higher levels of cognitive taxonomies—become more important when lower-level recall and application are AI-assisted. Education might shift toward these capabilities, though they're harder to teach and assess.
AI literacy becomes essential. Understanding what AI can and can't do, how to use it effectively, how to evaluate its outputs, and what its limitations and biases are—these capabilities become necessary for navigating an AI-infused world. AI literacy might become foundational like traditional literacies.
Uniquely human capabilities may warrant greater emphasis. Relationship-building, emotional intelligence, ethical reasoning, physical skills, and creative vision remain distinctly human. Education might emphasize what humans do better than AI rather than competing with AI on tasks AI does well.
Possibilities for Teaching
AI assistants could support teachers. Automating routine tasks—grading objective assessments, providing basic feedback, generating practice problems—could free teachers for higher-value activities. Teachers might spend less time on administration and more on relationships, mentoring, and complex instruction that AI can't provide.
Personalization at scale might become feasible. AI tutoring could provide individualized instruction that human teachers can't offer to every student. Combined with human teachers who provide what AI can't, hybrid approaches might serve students better than either alone.
Shifting pedagogy to integrate AI productively is possible. Rather than banning AI, education might teach students to use AI well—as a tool that augments rather than replaces their capabilities. Learning to work with AI effectively could become an educational goal rather than threat.
Focus on application and creation over reproduction aligns with AI realities. If AI can reproduce content, assessments might emphasize applying knowledge in authentic contexts, creating genuinely original work, or demonstrating capabilities in ways AI can't fake. This pedagogical shift might actually improve education.
Concerning Possibilities
Deskilling of teaching could occur if AI takes over instructional functions. Teachers might become monitors of AI systems rather than professionals making pedagogical judgments. This deskilling would diminish the profession and possibly diminish educational quality that depends on human expertise.
Corporate control of education could expand through AI systems. Educational AI is developed primarily by corporations with commercial interests. Dependence on these systems gives corporations influence over education that should concern communities. Data extraction, algorithmic biases, and profit motives may not align with educational interests.
Surveillance and data collection through educational AI raise privacy concerns. AI systems collect data about student learning that can be used for purposes beyond education. Who owns this data, how it's used, and what consent students and families have are important questions often inadequately addressed.
Widening inequality could result if AI benefits compound existing advantages. Students with AI access, AI literacy, and support for AI use will navigate AI-infused environments better than those without. Without deliberate effort to ensure equitable AI access and capability, AI could widen rather than narrow educational gaps.
Policy and Practice Questions
How should AI tools be permitted in educational contexts? Blanket bans seem unrealistic and possibly counterproductive. Unrestricted use seems equally problematic. Thoughtful policies must distinguish contexts, specify appropriate uses, and adapt as AI evolves.
How should assessment adapt to AI realities? Assessments that AI can complete don't assess student capability. New assessment approaches—proctored in-person work, oral examinations, process documentation, authentic application—may be necessary. Assessment systems designed for pre-AI contexts require fundamental rethinking.
How should curriculum change in response to AI? What knowledge and skills remain important? What becomes less important? What new capabilities need emphasis? Curriculum review through AI lenses can identify needed changes, though rapid AI development makes stable curriculum difficult.
How should teachers be prepared for AI-infused education? Teacher education and professional development must address AI—both how to teach about AI and how to teach with AI. This preparation barely exists in current teacher education programs, leaving teachers to figure out AI implications on their own.
Questions for Reflection
How has AI already affected education in contexts you're familiar with? What changes have you observed or experienced?
What would appropriate AI use in education look like to you? Where should AI contribute, and where should humans remain central?
How should educators respond to generative AI's challenge to traditional assessment? What approaches would maintain meaningful assessment of student learning?
What should students learn about AI and learn to do with AI? What AI literacy should education develop?
What concerns do you have about AI in education, and how might they be addressed?