A Retired Teacher’s Journey Into AI: From Skeptic to Advocate

Patricia Holman spent 34 years teaching high school history in suburban Ohio. She retired at 62 with a pension, a profound love of learning, and a deep suspicion of artificial intelligence. “I thought it was going to destroy education,” she says frankly. “I thought it was a cheating machine. I thought it was going to make students lazy and educators obsolete.” When her daughter installed ChatGPT on her laptop during a holiday visit in late 2022 and dared her to try it, Patricia sat down at the keyboard with her arms metaphorically crossed.

Two years later, Patricia runs a small online tutoring business that earns her between $2,200 and $3,100 per month — supplementing her pension comfortably and giving her a sense of purpose in retirement. Her tools include ChatGPT for lesson planning and content creation, Google NotebookLM for deep research assistance, and a simple online booking system. She has become an advocate for thoughtful AI adoption among educators, speaking at three regional teacher retirement conferences in 2024.

This is the story of her journey: the fears, the false starts, the breakthroughs, and the specific ways she learned to use AI tools to build something new in retirement.

The First Encounter: Skepticism Meets Curiosity

Initial Fears That Proved Partially Correct

Patricia’s first session with ChatGPT lasted about 45 minutes. She asked it history questions — her subject area, her home turf. She asked about the causes of World War I, the significance of the Magna Carta, and the economic factors behind the French Revolution. The answers were, she grudgingly admitted, “pretty good — like a smart student who has read the textbook but hasn’t done the primary sources.”

Then she tested it more aggressively. She asked about a specific local historical controversy she knew well — a dispute about the role of Ohio industrialists in shaping Progressive Era legislation. The AI’s answer was generic and missed the specific regional nuances entirely. “That,” she said with some satisfaction, “is exactly the kind of answer I’d give a C+ to. Technically not wrong but missing everything that makes the history interesting.”

This was her first genuine insight about AI: it is broad but not deep. It knows a great deal about well-documented topics and comparatively little about specialized, local, or nuanced areas. For a former teacher with 34 years of deep domain expertise, this meant AI was a complement, not a replacement.

The Learning Curve: Three Months of Dedicated Exploration

After that first session, Patricia gave herself a structured learning challenge: spend 30 minutes per day for 90 days genuinely exploring AI tools. She kept a handwritten journal — old habits die hard — of what she tried, what worked, what failed, and what surprised her.

Her exploration followed the natural arc of a curious, methodical educator. She started with ChatGPT, learning to write increasingly specific prompts. She progressed from “tell me about the Civil War” to “act as a Socratic tutor for a 10th-grade student who knows the basic timeline of the Civil War but does not understand the economic tensions between North and South. Ask questions rather than providing answers.” The shift in output quality was dramatic.

Month two brought her to Google NotebookLM, which quickly became her favorite tool. Unlike ChatGPT, which draws on its training data, NotebookLM allows users to upload their own documents — books, articles, PDFs — and then asks questions about those specific sources. For Patricia, who had decades of teaching materials, academic articles, and primary source documents, this was transformative. She could upload a set of primary sources on the New Deal and ask NotebookLM to help her identify connections between documents, generate discussion questions grounded in specific passages, or summarize the major scholarly debates in her source collection. Learn more about how NotebookLM works in our dedicated guide.

Month three was about failure and recalibration. She attempted to use AI to generate a full curriculum unit on the Cold War and was disappointed — the output was technically accurate but pedagogically flat, missing the narrative tension and conceptual scaffolding that makes complex history accessible to teenagers. “AI can write about history,” she concluded. “It cannot teach it. Teaching requires knowing the student.”

The Business Idea: Where AI Skills Met Market Need

The tutoring business idea crystallized from a conversation with a neighbor whose high school junior was struggling with AP United States History. Patricia tutored the student for a month as a favor and noticed something: modern students were using ChatGPT to help them study, but they were using it badly — asking it for answers rather than explanations, accepting its outputs uncritically, and missing the deeper conceptual understanding that standardized tests actually assess.

She realized she had a distinctive offering: she could teach students how to use AI tools effectively for learning, while simultaneously providing the deep subject-matter expertise and pedagogical skill that AI could not replicate. Her tutoring model was not “let me help you understand history.” It was “let me help you use AI to become a better independent learner — while I fill in the gaps that AI cannot fill.”

This positioning differentiated her from both traditional tutors (who competed primarily on price) and AI tools (which students were already using, badly). She was the bridge between the two.

Building the Tutoring Business: The First Six Months

Patricia launched her tutoring business in September 2023 with no marketing budget, a simple website built on a free template, and word-of-mouth from former colleagues and parents she had known through her school career. Her initial rate was $65 per hour — below market for a tutor of her qualifications, deliberately set low to build a client base quickly.

She offered two formats: one-on-one tutoring sessions via Zoom and small-group “AI-assisted study skills” workshops for two to four students. The workshops were priced at $35 per student per session and covered how to use ChatGPT and NotebookLM effectively for academic research, essay planning, and concept review — without cheating.

The workshop format proved unexpectedly popular. Parents were particularly interested: in a landscape of widespread concern about AI and academic integrity, a structured program that taught responsible AI use was reassuring. By month three she was running four workshops per week with three to four students each, alongside eight to ten one-on-one sessions.

Her monthly revenue by month six: approximately $2,800. She had raised her one-on-one rate to $80 per hour and workshop rates to $45 per student. Her working hours were about 20 per week — manageable and energizing rather than exhausting.

How AI Tools Transformed Her Lesson Preparation

The efficiency gains Patricia achieved through AI were most pronounced in lesson preparation. As a tutor working with multiple students on different topics and at different levels, she previously spent about two hours per new student preparing customized materials — review sheets, discussion questions, practice essay prompts. AI cut that time by about 70%.

Her current prep workflow for a new student: upload relevant course materials to NotebookLM, generate a summary of key concepts and common student misconceptions, use ChatGPT to create a diagnostic quiz and three practice essay prompts at the appropriate difficulty level, and spend 20 minutes reviewing and refining the AI-generated materials with her own pedagogical judgment.

The review-and-refinement step is non-negotiable. “The AI’s quiz questions are usually accurate but sometimes pedagogically naive — they test memorization instead of understanding,” Patricia says. “I look at every question and ask: does this question reveal whether a student actually understands the concept, or just whether they’ve memorized a fact?” This editorial layer is where her 34 years of classroom experience creates irreplaceable value.

Overcoming the Fear: Advice for Other Skeptics

Patricia speaks frequently about her transition from AI skeptic to AI advocate, and she is careful to honor the legitimacy of the skepticism. “The fears about AI in education are real,” she says. “Students do use it to cheat. It does produce plausible-sounding misinformation. It does threaten certain kinds of jobs. Those concerns are not irrational.”

Her reframing for skeptics: “The question is not whether AI is good or bad. The question is whether you want to engage with it on your own terms or have it happen to you. Ignoring it doesn’t make it go away — it just means someone else is shaping how it gets used.”

For educators specifically, she recommends starting with AI as a research assistant rather than a content generator. Use it to quickly review a topic you need to refresh, to generate alternative explanations of a concept you are struggling to communicate, or to identify common misconceptions about a topic that your students may hold. These use cases leverage AI’s strength (broad knowledge synthesis) while keeping your expertise firmly in control of the pedagogical decisions. Resources like our AI for teachers guide can provide additional starting points.

The Unexpected Benefits: Community and Purpose

Patricia is candid that the financial benefits of her tutoring business, while meaningful, are not the primary reason she loves it. The deeper benefit is purpose. After 34 years in a classroom, she missed teaching. She missed the moment when a concept clicked for a student, the satisfaction of watching a struggling learner find their footing, the intellectual engagement of discussing history with young people.

“Retirement is supposed to be relaxing,” she says. “And it is. But it was also a little empty. I missed being useful in the specific way that teaching makes you useful.” The tutoring business, enabled by AI tools that made it operationally feasible for a solo practitioner to serve a varied client base, restored that sense of purposeful contribution.

She has also found an unexpected community of educators navigating similar questions. The retirement conferences where she speaks have connected her with dozens of retired teachers who are curious about AI but intimidated. She has become an informal resource for this community — answering questions, sharing resources, and modeling what thoughtful late-career AI adoption looks like. Understanding AI ethics is a key part of this conversation she facilitates.

The Technology Stack: Simple and Effective

Patricia’s technology setup is deliberately simple. She uses ChatGPT’s paid plan ($20/month) for lesson planning and content creation, Google NotebookLM (free) for research-intensive prep work, Zoom for tutoring sessions ($15.99/month), a simple scheduling tool for booking management ($12/month), and Stripe for payment processing (transaction fee only). Her total monthly tool cost is approximately $50 — an almost trivially small overhead relative to her $2,200–$3,100 monthly revenue.

She deliberately avoids more complex tools. “I am not a tech person,” she says. “The simpler the better. If a tool requires more than an hour to learn, I don’t use it.” This philosophy of minimal viable technology is itself an important model for other retirees considering AI-enabled businesses: the most sophisticated tools are not necessarily the most useful ones.

For reference, the AI automation playbook covers how professionals across industries are using similar simple tool stacks to build sustainable businesses.

What Other Retired Educators Can Learn from Patricia’s Story

Patricia’s story contains several lessons that transfer to other retired professionals considering AI-enabled side businesses. First, domain expertise is an enduring competitive advantage — AI amplifies expertise but cannot manufacture it. Second, the most valuable applications of AI for service professionals are often in the behind-the-scenes work: preparation, research, content creation, and administrative tasks. Third, the skepticism that many experienced professionals feel toward AI is a feature, not a bug: it produces the critical evaluation mindset that prevents over-reliance on AI outputs.

For retired educators specifically, tutoring supplemented by AI is one of the highest-ROI applications of the combination. The educator brings subject expertise, pedagogical skill, and the ability to diagnose individual student needs — none of which AI can replicate. AI brings research efficiency, content generation, and availability. Together they create a service that neither could deliver alone.

Frequently Asked Questions

Is it too late to learn AI tools if you are over 60?

Absolutely not. The most popular AI tools — ChatGPT, NotebookLM, Claude — are designed for general audiences without technical backgrounds. Patricia learned them at 63 with no prior technical training. The learning curve is real but manageable, particularly for someone who brings the patience, intellectual curiosity, and structured learning habits that experienced educators typically possess. Starting with a specific, practical use case — rather than trying to “learn AI” in the abstract — makes the learning process more concrete and rewarding.

What subjects are most in demand for online tutoring?

High school and college-level STEM subjects (mathematics, chemistry, physics, biology) consistently command the highest rates and have the most robust demand. Advanced Placement courses in any subject have strong demand during the school year. Writing, history, and social studies tutoring is also in demand, particularly for college application essay coaching. Language tutoring — especially English as a Second Language — has strong demand year-round. The most sustainable tutoring businesses typically specialize in one or two subjects rather than positioning as generalists.

How do you address academic integrity concerns when teaching students to use AI?

This is the most important question in AI-assisted education. Patricia’s approach is to teach students that using AI to generate work they submit as their own is cheating — full stop. But using AI as a study tool (to get explanations, generate practice questions, test their understanding through dialogue) is legitimate and valuable. She also teaches students to verify AI outputs against authoritative sources, since AI tools can produce plausible-sounding but incorrect information. The goal is to develop AI literacy — the ability to use AI tools critically and effectively — not to use AI as a shortcut to avoid learning.

What are the best AI tools for educators and tutors?

ChatGPT is the most versatile starting point for educators — it can help with lesson planning, explanation generation, quiz creation, and feedback drafting. Google NotebookLM is particularly valuable for research-intensive subjects because it works with your own uploaded documents rather than its training data. Claude is Patricia’s preference for longer-form writing assistance because of what she describes as its more nuanced, careful writing style. For educators teaching STEM subjects, tools like Khan Academy’s Khanmigo (built on AI) are specifically designed for educational use and include strong safety guardrails.

How much can a retired teacher realistically earn from AI-assisted online tutoring?

Income varies significantly based on subject area, experience, rates, and hours committed. Tutors charging $60–$120 per hour for one-on-one sessions working 15–20 hours per week can earn $3,600–$9,600 per month. Patricia’s $2,200–$3,100 monthly income reflects about 20 hours per week at a mix of individual and group rates. The most experienced tutors in high-demand subjects (AP calculus, college admissions essay coaching, test preparation) charge $150–$300 per hour. AI tools enable retired educators to serve more students with less preparation time, improving both income and sustainability.

Sources

• Grokipedia
• Stanford HAI AI Index Report

Get free AI tips delivered daily → Subscribe to Beginners in AI

You May Also Like

• What Is Artificial Intelligence
• Best AI Tools for Beginners
• How to Use AI
• AI Tools Directory
• Best Free AI Courses