ChatGPT no longer sits idle in a browser tab waiting for questions. For Samantha McLean, it’s now a calendar-aware assistant that manages her schedule, flags priorities, and automates repetitive tasks—all without being asked. Her morning routine starts with an email from the AI summarizing her day, listing pending work, and suggesting where she can delegate tasks.
“ChatGPT has been in my calendar,” she says. “It acts as my assistant. It knows what I’ve got going on.”
This shift marks the end of the chatbot era, she argued in the opening session of Transform XIV, a 30-day AI training program she leads. The new phase isn’t about typing prompts into a box—it’s about handing off work to AI agents that operate across desktop tools, not just web interfaces. “It’s less typing things into a box and more handing things off,” she says.
Three Rules for Useful AI
An AI assistant needs three things to work effectively, according to McLean: training on how to respond, personal context, and automation for repetitive tasks. The first rule, how it responds, starts with filtering out awkward phrasing. She cites examples like an AI-generated LinkedIn post claiming a rent roll’s value “can shift overnight,” or using phrases like “what keeps me up at night” from a system that doesn’t sleep. “You wouldn’t say that out loud to someone’s face, right?”
Read Also: AGNT names Leo Pareja as new CEO
Her solution is a simple scale: 1 for a casual text to a friend, 10 for a formal contract. She aims for a 6, polished but still human-sounding. The goal isn’t perfection; it’s avoiding the robotic tone that plagues generic chatbot outputs.
Context Over Commands
The second requirement is context, what McLean calls “just a fancy word for background.” She demonstrates this with a fill-in-the-blanks exercise from her program, where users define their professional identity, key skills, and weaknesses. Her own example reads: “I’m a magazine editor turned AI product builder who likes to chase a big idea until it’s real.” Her weaknesses are “taking on too much at once and letting the fine print slip.”
Building this “business brain” takes time. McLean has spent six months refining hers, starting with basic integrations like Canva before expanding to tools like Descript. She advises beginners to begin small, connecting one tool at a time, rather than overhauling their workflows overnight. “Context is the fingerprints of your thinking,” she says.
This middle step explains why AI assistants outperform chatbots. A system that knows your calendar, tools, and communication style can act independently, while a generic chatbot remains a passive question-answering tool.
Read Also: Fixed Fees Shield Australian Agents’ Revenue as Demand Falls
Automation and the ‘Tribble’ Effect
The third rule is repetition. Once an AI handles a task well, McLean says, it should do it automatically, daily or weekly, without manual input. Her own assistant now scans AI news every morning, then generates social media posts, video ideas, lesson plans for her Elite Agent Insiders community, and drafts for her Substack newsletter. “Put the good stuff on repeat,” she advises.
She calls this “tribble mode,” a nod to Star Trek’s self-replicating creatures. A single AI agent can spawn specialized helpers, each handling a subset of tasks, before consolidating results. The analogy works: tribbles multiply when ignored, but they’re less intimidating than gremlins. The point is scalability. What starts as one automated task can expand into a network of semi-independent workflows.
McLean’s transition from chatbot user to AI delegator hinges on these three principles. The shift isn’t just technical; it’s about rethinking how work gets done. Instead of asking an AI to solve problems on demand, she treats it as an extension of her team, one that learns her habits, anticipates needs, and handles the grind so she doesn’t have to.
