Sam Finnegan-Dehn has two jobs. By day, he works in fundraising for a charity based in London. By night—and on weekends—he tutors university students in math and philosophy from his home office. The side hustle lets him leverage degrees in philosophy while sharing his love of the subject with clients. But here's the thing nobody tells you about tutoring: meeting with students is maybe 20% of the actual work. There's lesson planning, reading material curation, assignment creation, invoicing, client progress tracking, and keeping up with new research—all stacked on top of a full-time gig.

The Notebook AI Workflow

Finnegan-Dehn turned to AI to handle what he calls the "secretarial role" across his digital life. He describes using Notion AI as essentially having a second memory that connects ideas scattered across various notebooks and tabs. After experimenting with Claude and ChatGPT, he settled on Notion AI specifically because it integrates directly with his existing tutoring workflow—no copying and pasting between apps, no friction. The system handles meeting recordings (with client consent), automated summaries for refining teaching strategies, goal-setting with step-by-step breakdowns from "North Star" objectives, lesson note drafting, invoicing, and social media post generation. If the AI's summary flags that a particular technique isn't landing with a student, he adjusts his approach accordingly. It's not glamorous work—it's the kind of rote task that eats hours if you let it.

Real Results From Real Businesses

Notion AI isn't alone in this space. Grandma's Quilt Shop in Yuma, Arizona uses Rain, a software suite tailored for craft companies, to generate inventory descriptions and pricing for their fabric designs. The owners claim the tool cuts item-listing time by 60 to 80 percent. That's not incremental improvement—that's a fundamental shift in how small retailers can compete with larger operations that have dedicated staff for catalog management. These examples illustrate something important: AI tools don't need to be revolutionary to provide value. They just need to handle the administrative overhead that prevents solo operators and micro-businesses from scaling.

The Fine Print Nobody Talks About

Here's where I get skeptical—because every tool has its price, both literal and otherwise. Notion AI's add-on runs $20 per month. Finnegan-Dehn himself describes some of the platform's idiosyncrasies as "clunky" at times. And here's what the marketing won't tell you: these systems feed on your data to function. The convenience of integrated notebooks means you're building dependency on a specific ecosystem before you've even validated whether it works for your use case. As with all new tools, small business owners should carefully assess how potential gains measure against the cost of just doing the job themselves—and factor in the real cost of data lock-in.

Security Considerations for Sensitive Operations

Our reporting has covered the documented risks that online AI models have in leaking sensitive data. There have been numerous reports about how AI companies collect your inputs when you query their chatbots. Even if your business doesn't handle obviously sensitive information, there can be things you'd prefer not to share publicly—client names, financial details, proprietary processes. For these cases, running open-source LLMs locally on laptops or small desktops is increasingly viable and worth considering instead of handing data to proprietary services like ChatGPT or Claude. The performance gap between local models and hosted options has narrowed significantly over the past year.

Key Takeaways

  • Evaluate AI tools based on workflow integration, not feature lists—Finnegan-Dehn chose Notion AI because it fit his existing notes system
  • Start with low-risk administrative tasks: invoicing, scheduling, document syncing before attempting creative or strategic work
  • Calculate true cost including monthly fees, learning curve time, and potential data lock-in
  • Consider local models for any information you'd prefer not to share with third-party AI providers

The Bottom Line

The real opportunity here isn't AI replacing human creativity—it's AI absorbing the administrative cruft that prevents one-person operations from operating at scale. But there's a catch: most of these tools require you to buy into their ecosystem first, which means you're making a bet on a platform before you've proven it works for your use case. That's not necessarily wrong, but it's worth being honest about what you're trading away. Privacy-conscious businesses should treat local models as the default, with cloud AI as the exception rather than the rule.