AI in Education: Revolutionizing Teaching and Learning

The Changing Landscape of Education with AI

The Evolution of Assessment

  • Oral exams are becoming less common in universities.
  • There's a question of whether the decline in speaking and the rise of writing will impact liberal arts education.
  • Change is constant in both industry and higher education.
  • Writing remains crucial for learning, despite the shifts.

Ethical Use of New Tools

  • New tools like AI can be used nefariously or ethically.
  • Example: PhotoMath initially seemed like a cheating tool.
  • Students began using PhotoMath to check their work.
  • PhotoMath became an instant feedback tool.
  • Students often find ethical uses for new technologies.

Transparency and Accountability

  • Traditional methods like Turnitin may become less reliable.
  • A transition toward transparency-based accountability is suggested.
  • This involves submitting everything from start to finish, including rough drafts and handwritten outlines, to show the entire writing process.

Ethical AI Use in Writing

  • High school students are using AI ethically in writing processes.
Research Phase
  • Instead of starting research with general online searches, students are using embedded chatbots like ChatGPT (paid version) or Consensus.
  • Consensus provides peer-reviewed journal articles and summaries.
  • Students interact with these tools by asking questions and follow-ups.
  • This allows them to enter the research phase with prior knowledge.
  • AI acts as a thought partner during research.
Outlining and Organization
  • Students gather research and create outlines.
  • They compare their outlines with AI-generated outlines.
  • This helps them identify potential organizational improvements.
Writing and Feedback
  • AI provides feedback on their writing.
  • Students rewrite confusing sections in their own style.

Comparison to Professional Writing

  • The AI-assisted writing process mirrors how college professors write journal articles.
  • This suggests that AI will become an integrated part of academic writing.

Cheating and Academic Integrity

  • Some students will use AI to cheat.
  • Cheating has always existed, even before AI.
  • Example: At ASU (Arizona State University), some football players had others write their essays.

Differentiated Instruction with AI

  • The focus is shifting to using AI for differentiated instruction, primarily on the teacher's side.
Building Conceptual Knowledge
  • AI can help explain complex concepts in simple language.
  • Example: The speaker's son, an engineering major, struggled with AP stats and understanding p-values.
  • They used a chatbot to explain p-values in simple language as if he were 12 years old.
  • AI can provide examples from relevant fields, like sports.
  • Students can use AI to explore topics and ask questions.
  • Example: The speaker's son used ChatGPT to question the cloning of a 14,000-year-old Dire Wolf and explore ethical questions related to cloning and animal intelligence.
  • He used AI to play the role of different philosophers (Spinoza, Descartes) to discuss epistemology, cosmology, and the nature of the universe.
Designing Language Scaffolds and Supports
  • AI can be used to design language scaffolds and supports for various subjects.
Reciprocal Teaching Example
  • Example: While listening to a course on the pyramids of Egypt, the speaker prompted his son to use ChatGPT to come up with theories on how the pyramids were built, asking questions and receiving feedback until he discovered the process on his own.
  • This reciprocal teaching approach encouraged critical thinking and deeper engagement.
Sentence Stems and Academic Language
  • AI can generate sentence stems to aid in discussions.
  • Example: Creating sentence stems for sixth-grade students learning about mummification.
  • These sentence stems can be tailored to different ELP (English Language Proficiency) levels.
  • Sentence stems can improve discourse, even for honors students, by promoting nuanced thinking.
Vocabulary Support
  • AI can create content vocabulary written in student-friendly language.
  • Models like Marzano vocabulary can be used.
Total Physical Response (TPR)
  • AI can suggest TPR activities to help students remember concepts.
  • Example: Creating TPRs for militarism and nationalism involves visual cues, definitions, and hand movements.
  • TPRs work because multimodal learning builds synaptic connections and improves retention.
Language Support and Scaffolding
  • AI can generate sentence stems and vocabulary lists for language support.

Modifying Text for Different Reading Levels

  • AI can adjust text to different grade levels, acting as a scaffold.
  • Example: An AP US History teacher used AI to modify primary source documents to an eighth-grade reading level.
  • The experimental group using modified texts performed better than the control group.
  • Reducing anxiety and building foundational knowledge are key benefits.

Gradual Increase in Complexity

  • Like the video game Super Mario, learning should start easy and gradually become more challenging.
  • Example: Modifying texts to lower grade levels initially, then gradually increasing the reading level.
  • Even complex texts like the Federalist Papers can be easier to understand with aids like the musical Hamilton.

Creating Leveled Readers

  • AI can create leveled readers from outlines.
  • Example: An eighth-grade teacher creates a station rotation activity about the Industrial Revolution and then uses AI to generate a fifth-grade reading level version.
  • This allows students to read independently at their appropriate level while discussing the same information.
  • Magazines and newspapers are generally written at a fourth to fifth-grade level, with The New York Times at a sixth-grade level.

Effective Study Techniques

  • Highlighting and underlining are ineffective study techniques.
  • Writing in the margins is marginally effective.
  • The most effective study techniques involve:
    • Identifying what you know and don't know.
    • Seeking trends in your understanding.
    • Asking clarifying questions.
    • Testing yourself progressively harder.
    • Teaching the material to someone else.

AI-Assisted Study Process

  1. Create a T-chart of what you know and don't know.
  2. Submit your work to a chatbot and ask for trends.
  3. Compare your perception with the AI's perception.
  4. Ask clarifying questions.
  5. Have the chatbot test you, starting with multiple-choice questions and progressing to open-ended questions.

Limitations of AI

  • AI excels at word problems and conceptual thinking but may struggle with basic numerical calculations.
  • It is a large language model, not a mathematical model.

AI for Tutorials

  • AI can create detailed tutorials with step-by-step instructions and screenshots.
  • Example: Using tutorials to support details with sample sentences.

Breaking Down Tasks

  • AI can help break down tasks for students with ADHD.
  • Convert large projects into daily to-do lists.
  • Use the endowed progress effect by creating early wins.
  • Include tasks like taking out the to-do list to build momentum and reduce cognitive load.