Introducing this Spark
This Spark activity demonstrates how sole reliance on generative AI tools without domain knowledge often leads to friction and flawed business decision-making. Students compare their performance using traditional hands-on methods against varying degrees of AI automation to discover how to use AI as an effective augment without sacrificing human knowledge.
Learning Outcome
Students will be able to solve forecasting problems using AI as an augmenter rather than a complete replacement. They will also be able to "explain how ... forecasting ... translate[s] raw consumer data into business insights" (from course objective 1).
Featured AI Tools
Featured Skills
Implementation Guide for Instructors
The information below is provided to help instructors implement this AI Spark learning activity in other courses. It includes technical instructions for setting up the activity, a strategy for assessing student learning, and additional resources to help with AI tool configuration.
This pass/fail activity requires students to complete a business demand forecasting exercise using only Google Gemini to direct spreadsheet modifications. Building on a previous manual homework assignment, students test complete reliance on AI to observe where automation excels (speed, formula drafting) and where it causes friction (formatting, context awareness, advanced statistical logic). By forcing Gemini to choose an unfamiliar forecasting technique, students experience "driving without a license"—evaluating whether AI recommendations can be trusted when the user lacks domain mastery. The assignment is designed to foster critical thinking and "useful" engagement with AI tools, exploring both benefits and drawbacks (should they be used to "outsource" the student's responsibility).
Step 1. Starter Spreadsheet: Create a starter spreadsheet modeled after an already-submitted homework assignment. The template can be accessed here: Bike Rental Demand Starter Sheet. Set the spreadsheet permissions to "View Only" so students must make a copy. Keep the dataset size modest to prevent context window constraints with the model.
Step 2. Canvas Assignment Configuration: Set the Canvas submission type to File Upload, restricting accepted formats to .xlsx and .pdf.
Step 3. Upload Reference Materials: Post reference materials, such as lecture slides or PDF reference guides, in Canvas. Students can upload these documents alongside data files to give Gemini domain context.
Step 4. Grading Expectations: Configure the assignment as Complete/Incomplete (Pass/Fail) to incentivize experimentation rather than cosmetic perfection.
Graded as Pass/Fail. Lowering the stakes encourages experimentation and reduces stress over output formatting. Assessment focuses on a completed spreadsheet deliverable, a transcript of interaction with AI, and a 1-2 page analytical reflection that strengthens the students' understanding of forecasting as a tool and the associated constraints of AI use.
Operational Notes:
- Large datasets can cause output issues or hallucinations in Gemini; keep sample sizes focused.
- Transcript Export Options: Students can either ask Gemini to generate a summary transcript of the session (preferred) or use the Share & Export option in Gemini to export the complete chat.
Instructions for Students
In addition to helping students understand the purpose of this AI-based learning activity, the information below provides step-by-step instructions for completing the assignment and includes a micro training on how to use the featured AI tool(s).
In this activity, you will use Google Gemini using your FSU Google credentials. Gemini is a generative AI assistant capable of analyzing structured data, generating spreadsheet formulas, suggesting analytical models, and explaining complex concepts in real time. Gemini integrates natively with Google Workspace, allowing minimal friction when reading data files and constructing spreadsheet workflows. This exercise tests Gemini's strengths and limits in real-world business forecasting tasks.
Activity Goal: Build a clean, well-formatted demand forecast spreadsheet using only Gemini’s assistance and guidance, then reflect on the strengths and friction points of relying entirely on AI.
- Access the Starter Sheet: Make a personal copy of the template linked here: Bike Rental Demand.
- Initiate Gemini Session: Open gemini.google.com and log in with your FSU account. Upload the spreadsheet file (and relevant course lecture notes if desired for context) into your chat session.
- Strict Automation Rules:
- Rule 1: Do not manipulate cells, formulas, or formatting manually unless Gemini explicitly instructs you to do so.
- Rule 2: You must prompt your way to a complete forecast spreadsheet (including working formulas, color-coding, time series plots, and formatted tables). Be very specific in your prompting.
- Rule 3: Ask Gemini to select and implement at least one forecast methodology you have NOT yet learned in class to simulate driving the tool without prior training (Tip: use lecture notes for new model ideas).
-
Draft the Reflection: After completing the spreadsheet, write a 1- to 2-page reflection addressing the four required sections:
Section 1: Unfamiliar Forecasting Model: Identify the model Gemini selected that was not covered in class. Detail Gemini's reasoning for selecting this model. Do you trust the model's results? Can you explain the mechanics behind it?
Section 2: Where AI Excels: Compare this process to Homework 1. Where did Gemini streamline or speed up your work? How is GenAI well suited for the task of forecasting on time-series data?
Section 3: Where AI Creates Friction: What parts of this assignment would have been easier, faster, or less frustrating to do on your own (without AI) based on your knowledge from class? What are Gemini's current weaknesses when handling data and designing spreadsheets?
Section 4: Workplace Integration: How will tools like Gemini fit into professional workflows? Which specialized knowledge tasks do you believe are best suited for AI enhancement? Which tasks still demand human expertise?
- Export Deliverables: Save your final Google Sheet as an Excel file (.xlsx) and ask Gemini to create a summary transcript of your chat session as a PDF document (.pdf) Ask Gemini to create a formatted transcript of your interaction as a PDF and submit.
Helpful Prompting Tips:
- Position your browser windows side-by-side (Gemini on one side, Google Sheets/Excel on the other). If Gemini is struggling to create new spreadsheets, you may ask for help via email.
- Prompt Gemini to explain its decisions and logic step-by-step before implementing formulas.
- Be specific with your formatting requirements (e.g., "Provide working Excel/Google Sheets formulas and clear color coding between model comparisons").
- Do not enter personal identifying information.
What to Submit:
- LASTNAME_FIRSTNAME_AILAB.xlsx: Completed Google Sheet forecast.
- LASTNAME_FIRSTNAME_AIREFLECTION.pdf: 1- to 2-page reflection covering the four subheadings above.
- LAST_FIRST_TRANSCRIPT.pdf: PDF export of Gemini chat transcript.
What is Google Gemini? Google Gemini is an AI-powered productivity tool designed to enhance your work and learning experience across the Google Workspace. Read more at Google Gemini.
- How to login to Google Gemini:
- Navigate to Google Gemini
- Enter your FSU email address (fsuid@fsu.edu)
- Enter your FSU password
- Confirm on Duo, if needed
- Select the I understand button to acknowledge the Terms of Service (first time only)
- Always access the tool via your FSU institutional account; personal accounts will default to an unconfigured model that produces generic, overly harsh, or ungrounded responses.