99% of People Have No Idea What’s About to Happen With AI

Key Concepts

• AI Assistance (Level 1): Using AI as a smart tool to speed up manual tasks.
• Agent Operators (Level 2): Managing AI agents that execute end-to-end workflows.
• AI Organization (Level 3): A hierarchical system where a "master" AI manages specialized sub-agents to run entire business or personal operations.
• Agentic AI: AI systems capable of planning, executing, and iterating on tasks autonomously.
• Apex (Agent Platform for Execution): A proprietary, secure platform for managing AI agents and workflows.

1. The Three Levels of AI Evolution

The speaker categorizes the current AI landscape into three distinct tiers, noting that the vast majority of users are stuck at the entry level.

• Level 1: The Assistant (The "Kiddie Pool"):
  • Function: AI acts as a smart assistant for research, design, and writing.
  • User Role: The human still performs the work; the AI simply accelerates the process.
  • Tools: Claude, ChatGPT, Gemini, Perplexity, Whisperflow, Claude Code.
  • Limitation: High "touch points"—the user must interact with multiple tools for every individual task.
• Level 2: The Agent Operator:
  • Function: The AI is given a goal, plans the steps, executes them, and iterates based on feedback.
  • User Role: The human acts as a project manager, configuring the agent and checking outputs.
  • Tools: Manis, Claude Co-work.
  • Adoption: Only ~0.3% of ChatGPT users are currently utilizing agentic tools like Manis.
• Level 3: The Organization (The Frontier):
  • Function: A "master" AI (e.g., the speaker's agent, "Kai") manages a network of specialized sub-agents.
  • User Role: The human provides high-level direction and final approval. The AI handles the "how," the execution, and the cross-departmental communication.
  • Adoption: Extremely rare (~0.05% of users).

2. Real-World Applications and Case Studies

The speaker demonstrates how Level 3 AI is currently being applied at his company, Martell Ventures:

• Real Estate Portfolio Management: An agent named "Reese" scans the internet 24/7 for off-market deals, analyzes them, and presents the best options to the human for approval.
• Unified Inbox Management: AI agents monitor Slack, WhatsApp, email, and social media. They learn the user's specific writing style to draft responses and surface only high-priority items, effectively eliminating the "mental load" of notifications.
• Procurement (Procure): An agent handles purchasing tasks (e.g., buying electronics) by researching, finding the best price, and executing the transaction using secure, limited-access virtual credit cards.
• Voice-Activated Operations (Speak): A voice-interface tool allows the user to call their master AI ("Kai") to get briefings, dictate messages, and manage schedules in real-time.

3. Methodologies and Frameworks

The transition from Level 1 to Level 3 is defined by a shift in Touch Points and Autonomy:

| Feature | Level 1 (Assistant) | Level 2 (Operator) | Level 3 (Organization) | | :--- | :--- | :--- | :--- | | Touch Points | Many (per task) | Few (per job type) | One (master agent) | | AI Capability | Helps do the work | Completes the work | Manages the entire operation | | Human Role | Doing the work | Managing the tool | Setting the direction |

4. Key Arguments and Perspectives

• The "Gap" Argument: The speaker argues that the gap between those who understand agentic AI and those who don't is widening daily. He compares the current moment to the dawn of the internet or Bitcoin, suggesting that those who fail to adapt will be left behind.
• The "Unlearning" Requirement: A major barrier to entry is the psychological belief that one is "not technical enough." The speaker emphasizes that the future is about "execution at the speed of talk," making technical coding skills less relevant than the ability to direct AI.
• Efficiency Gains: By moving to Level 3, the speaker claims he can achieve "10 times more with 10 times less people," significantly reducing the emotional and mental burden of business management.

5. Notable Quotes

• "What's coming with AI is bigger than the internet, bigger than smartphones, bigger than Bitcoin."
• "In level three, your job becomes deciding what to do. The AI agent decides how to do it."
• "Don't be that person selling newspapers in the era of the internet."

6. Synthesis and Conclusion

The core takeaway is that AI is rapidly evolving from a passive tool (Level 1) to an autonomous workforce (Level 3). The speaker advocates for a shift in mindset: stop trying to "use" AI tools and start "managing" AI agents. By building or utilizing platforms like his "Apex" (Agent Platform for Execution), individuals can offload the entirety of their operational and administrative burdens, allowing them to focus solely on high-level strategy and decision-making. The future is not five years away; it is happening now for those willing to learn and unlearn traditional work habits.