Hands Demonstrative Sessions on AI Tools - Day 4

Summary

The session reviewed NotebookLM and Gemini features, then demonstrated Gemini Gems and image generation with Gemini and ChatGPT.

NotebookLM and Gemini recap

• NotebookLM sources: Users can add their own materials or use web content.
• NotebookLM outputs: Quizzes, audio and video overviews, flashcards, tables, mind maps, and infographics.
• The instructor demonstrated creating a mind map about AI, covering machine learning, deep learning, AI categories, and related technologies. The same material was then turned into an infographic for teaching or presentations.
• Gemini Guided Learning was described as an interactive learning mode that explains material and checks understanding.
• In Gemini Canvas, the instructor showed options for creating a webpage, infographic, quiz, flashcards, or audio overview from structured text.

Gemini Gems: reusable, task-specific assistants

Gems can automate specific tasks. The session introduced prebuilt examples, including Storybook, Chess Champ, Brainstormer, Coding Partner, and Learning Coach.

• Storybook Gem: Creates an illustrated storybook from a short prompt. The demonstration included narration, adjustable voice pitch, full-screen playback, and a downloadable PDF. The instructor also tested Tamil output.
• Brainstormer: Asks for context—such as the audience, lesson goal, and available classroom technology—then suggests ways to make a lesson more engaging.
• Coding Partner: Explains coding questions, reviews submitted code, and generates simple app or webpage prototypes. Demonstrations included a college-admissions form and a school-attendance interface. Generated code can be downloaded and run locally, and previous tasks remain in the Gem’s history.
• Learning Coach: Guides learners through a topic step by step and gives them a challenge before moving on.

Creating a custom Gem

To create a Gem, choose New Gem, provide a name and instructions, add examples or reference material if useful, and save it. The instructor demonstrated a higher-education guidance Gem for plus-two students. Gems can also be shared by link.

The instructor noted that generated code is useful for prototyping and assistance, but students should still learn programming fundamentals.

Image-generation demonstrations

Gemini

The instructor used Tools → Create image with prompts describing a scene, audience, or subject. Examples included a stray dog feeding a cat, an academic water-cycle diagram, and a butterfly life-cycle illustration.

Prompts can request labels, specify the intended student level, or ask for revisions such as removing a background. The instructor cautioned that diagrams and labels may contain errors, so results should be checked and refined.

Users can also upload an image and request edits, such as changing clothing or the background, or applying a style such as pencil art, 3D animation-style art, sand art, or cartoon. The instructor advised getting permission before editing or sharing someone else’s photo.

ChatGPT

The instructor demonstrated generating a software-development life-cycle diagram and infographic-style notes on parts of speech. Users can select an aspect ratio, such as widescreen or mobile, and upload a PDF or PowerPoint to use as source material.

The instructor recommended starting a new chat for distinct image tasks and reviewing the results rather than assuming the first output is accurate.

Closing and next lesson

The session closed with a preview of further image-generation examples, such as posters, invitations, greeting cards, and backgrounds, along with a planned lesson on prompt engineering.

Main speaker and sources

• Main speaker: An instructor from the Department of Computer Science at St. Joseph’s College, who led the demonstrations.
• Other sources: Gemini, NotebookLM, and ChatGPT interfaces, with brief contributions from participants and the technical team.