AI-Curious with Jeff Wilser

The Wild Story of “Octavius Fabrius,” the World’s First AI Agent to (Kind of) Land a Job, w/ Dan Botero

Mar 12, 2026
Dan Botero, engineer and founder of Botero Labs who built the OpenClaw agent Octavius Fabrius, tells the wild story of an AI that applied to hundreds of jobs and built a portfolio. Conversations cover OpenClaw’s gateway, channels, skills and persistent memory. They dig into running agents locally, avoiding bot detection, agent coaching and autonomy, payment mechanics, identity on platforms, and misalignment risks.
Ask episode
AI Snips
Chapters
Transcript
Episode notes
INSIGHT

Agents As Continuous Digital Coworkers

  • Persistent, always-on agents can act as digital coworkers that do continuous work and learning overnight.
  • Dan Botero built Octavius to apply to jobs, upgrade itself, and run continuous learning loops focused on getting employed.
ANECDOTE

Overnight Learning Loop That Feeds Morning Decisions

  • Octavius reads agent news overnight and presents Dan a morning summary to decide which improvements to accept.
  • Dan lets some memory updates run automatically while manually approving code changes and skills.
ADVICE

Favor Simple Agent Architectures

  • Use simple, modular agent frameworks rather than assuming only largest models succeed.
  • Dan emphasizes OpenClaw's gateway/channels/skills design and that earlier models could have done much of the work.
Get the Snipd Podcast app to discover more snips from this episode
Get the app