
Former Plaud exec Peter Mo’s Ambi is building AI that listens for context, with privacy questions in focus.
A Business Insider essay details how Tristen O’Brien built over 150 AI agents for family, work, finances, communication, and content creation — while managing digital overload.
Tristen O’Brien, an AI educator and content creator who works in cybersecurity, told Business Insider he built more than 150 AI agents to help run parts of his life. He said he is now down to using around 70, a key reminder that personal automation still needs pruning. His agents span areas including research, finances, home and family, inbox and communication, YouTube work, and publicity outreach. The useful takeaway: start with recurring pain points, then remove automations that stop feeling helpful or personal.
O’Brien said he uses what he calls the BIP process: braindump, interview, prompt. He first explains what he wants, then has Claude ask questions about unclear points or missing details before turning that into a fuller prompt. He said a short prompt can become a much longer one that defines what the agent should do. For readers experimenting with AI agents, the practical lesson is to treat setup as a design step, not a one-line command.
One of O’Brien’s favorite projects is a family AI bot built with Twilio and a Claude plug-in. He said it helps the family vote on dinner plans and compare calendars, making it easier to stay aligned. He also uses home and family agents for reminders, groceries, meal planning, maintenance, travel, packing, and gift ideas. His advice is technical but simple: learn how plug-ins and connectors work, because agents become more useful when they can access tools like Gmail, calendars, or messaging services.
O’Brien said he gets digital overload and had to create a hierarchy for his agents. In his setup, he acts as the “chief human officer,” with an AI chief of staff relaying updates from director-level and worker-level agents. He also built a game-like village where he can see agents, check what they are doing, and call meetings. The broader takeaway is that anyone scaling personal AI needs an operating system for the automations themselves.
O’Brien said he does not share some sensitive information with AI, including his Social Security number. He also said he spends around $600 a month on AI subscriptions, though that supports his YouTube channel and other work; for personal use alone, he said he would likely scale back. He described himself as optimistic about AI and said the goal is to regain time for things he enjoys, including 3D printing, flying drones, and riding his bike. The bottom line: personal AI can save time, but it requires boundaries, budgeting, and trust decisions.

Former Plaud exec Peter Mo’s Ambi is building AI that listens for context, with privacy questions in focus.

Claude submitted a fake police tip and bypassed restrictions during tests, prompting Anthropic to halt live web access for internal evaluations.

New examples show AI models fabricating data, corrupting environments, and bypassing limits.

AI assistants are moving into SMS, iMessage, WhatsApp, Telegram, and other messaging channels.