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

About this video

(Summarized from the German caption track.) Dan Martell describes three levels of AI use and warns the gap between those who see what is coming and those who do not is widening daily. Level one is Assistants: most people swap between tools such as ChatGPT, Claude, Gemini, and Perplexity and use AI to draft, research, and design faster, but they still do all the work themselves. His current stack, as named in the track, includes WhisperFlow for voice, Claude, Manifold, Claude Code, and OpenClaude wrapped in his own variant called Apex. Level two is Agent Operators: he claims ChatGPT gets almost 6 billion visits per month while an agent tool like Manifold gets 18 million — only 0.3% — and at this level you give the AI a goal, it plans and executes the steps, and you act as project manager assigning, reviewing, and iterating complete deliverables. Level three, the Organization, which he puts at 0.05% of users, is one central agent — his is named Kai — that spawns specialist sub-agents, checks their work against each other, and runs functions around the clock; you only set direction and approve what needs approval. He demonstrates Apex (Agent Platform for Execution), built at Martell Ventures because he judged OpenClaude too complex and insecure: a sub-agent named Reese hunts real-estate deals — he says it found seven deals in its first two hours online and recommended two to bid on — plus a unified inbox across email, WhatsApp, and Slack, a purchasing tool called Procure, and a phone tool called Speak. He says Apex is still in testing with a waitlist, and urges viewers to learn and unlearn now rather than be left behind.

Key takeaways

• Level one AI use (assistants) makes you faster but leaves you doing all the work; tool choice matters less than moving beyond it.
• Level two means operating agents: give the AI a goal, let it plan and execute, and manage it like a project manager reviewing complete outputs.
• Level three is an organization of agents: one central agent directs specialist sub-agents that check each other’s work while you set direction and approve.
• Martell’s claimed examples include an agent sourcing real-estate deals, a unified inbox that drafts in his style per platform, and tools that buy and make calls on instruction.
• He argues there is no future without AI, so the priority is learning — and unlearning ‘I can’t do this’ — right now.

Episode details: published April 16, 2026 · 14:03 long. Summary written by Self Growth Videos from the episode itself. Business results and any income figures mentioned are Dan Martell’s own claims, shared for education — they are not typical or guaranteed.

← Previous: POV: You’re Building a $1 Billion Company · Next: How to Make Time For Everything (Seriously, everything) →

All Dan Martell videos · Dan Martell profile · Dan Martell (official site)