We built OpenClaw Ultron to replace 20 people at our company | E2246

304 snips

Feb 7, 2026

Alex Cheema, founder and CEO of Exo Labs, builds tools to run frontier AI on consumer Apple Silicon. They discuss running powerful models locally for data control and cost savings. Short demos cover automating recurring workflows, a dashboard that stores memories and files, clustering Macs for inference, and risks like prompt injection.

Ask episode

AI Snips

Chapters

Transcript

Episode notes

INSIGHT

One Replicant To Replace Many Chores

Jason explains the goal: build one OpenClaw replicant to handle ~20 employees' tasks and 100–200 skills.
The aim is to remove chores so humans focus on high-value investor and creative work.

INSIGHT

Local Models As Your Exocortex

Alex frames local models as an exocortex you should own for sovereignty and control.
Running models locally reduces lock-in and preserves control over evolving behavior and data.

ADVICE

Ship A Dashboard First

Start your OpenClaw project by building a dashboard to visualize memory, files, skills, and cron jobs.
A visual interface beats interacting only through chat and helps manage the system effectively.

Get the Snipd Podcast app to discover more snips from this episode

Get the app