Working with AI Agents: ADHD Manifesto, Burnout Risks, and a Ladder of Eight Orchestration Levels

A live-streamed discussion between a productivity coach and an AI practitioner who wrote the "ADHD Manifesto" about what happens when AI agents become your daily co-workers. The conversation moves from the psychological side (attention deficits, cognitive fuel, burnout) to the practical side (local hardware, agent hierarchies, security), based on their own hands-on experience over the past year.

■ AI Agents and Your Mind
- Why agent-based work turns "attention deficit" into a job requirement, and how the function of stopping an impulse becomes the most valuable skill
- The "task cone" risk: small five-minute tasks that quietly eat your whole day
- Underestimating effort in a new paradigm, the half-life of skills, and the practice of cutting losses
- Decision fatigue, the cost of running many projects, and the value of "burying" projects on purpose

■ Workflow and Team Practices
- Building a personal agent system with an orchestrator, memory, deferred tasks, and agents that check your meetings and promises
- Local models and hardware: running ComfyUI on a Mac Studio overnight, routing requests between local and cloud, and why hardware matters more than the OS
- Using agents for personal learning and education, including tutoring and lesson generation
- The cost of ideas in the agent era, and a possible niche for family-oriented chatbots

■ The Ladder of AI Mastery
- Eight levels of working with AI, from replacing Google with a chatbot up to a full "corporation" of agents with onboarding, audits, and security routines
- Why the fifth level (running your own infrastructure) is the key barrier, and the vendor lock-in risk of hosted platforms
- Compound context: why the accumulated record of your work grows more valuable over time

■ Risks and Future Shifts
- Cybersecurity concerns as the cost of attacks drops, and the need to protect keys and agent infrastructure
- Changing etiquette around instant replies, and how AI spreads through a team

This is aimed at people who already use AI tools at work and want to move beyond casual chat, as well as anyone worried about burnout or about skills becoming obsolete. It offers concrete practices — deferral buttons, loss-cutting, personal context building — and a framework for judging which level of agent use is right for you. Watching it gives you language and mental models for deciding what to automate, what to keep, and how much cognitive fuel each new project will actually cost.