“Learn AI” Is Bad Advice. Learn This Instead

An analytical guide to six practical skills that will gain value as AI advances. The video outlines designing and managing agentic workflows and local models; building distribution and demand; robotics engineering that combines hardware, AI and manufacturing; curating short-form content and niche signals; becoming a builder-distributor who ships products and markets them; and leading IRL communities. Each skill includes concrete starter projects, success metrics and reps to learn quickly, arguing that combining these capabilities creates durable career leverage.

– Agent design & local models: Build, manage and evaluate agent “employees” with tools, memory, permissions and approval rules; starter rep: create a daily briefing agent that pulls calendar, notes and links.

– Distribution marketing: Map where attention lives, craft 20 hooks and translate one insight into tweets, videos, headlines and landing pages to create demand before you build.

– Robotics & hardware: Prototype low-cost arms, integrate vision+models, collect demos and learn supplier sourcing; first rep: teach a robot one boring task and document failures.

– Curation & short-form content: Become a reliable signal-filter—yap with a clear take, run a 7-day curation sprint, and build a taste file of high-signal inputs.

– Builder-distributor & IRL community: Close the loop by shipping tiny AI products and distributing them (48-hour loops), and host small recurring IRL rooms that turn conversations into networks and assets.

Quotes:

Build a daily briefing agent for yourself.

The next decade will reward people who can move atoms around.

Pick one and get dangerous.