This is episode five of the Agents Unleashed podcast — your guide to filtering the signal from the noise in the fast-evolving world of agentic AI.

In this episode, host Thomas Maybrier goes hands-on with some of the newest demos in the agentic AI space, including the Open Computer project from Smol Agents, which combines Qwen2, a Linux sandbox, and Python-based agent reasoning.

He shares his firsthand experience with the demo, the UX tradeoffs, and the broader questions raised by giving AI agents access to your computer or desktop.

The episode also explores how companies like Microsoft and Google are bringing agentic behavior to the mainstream. Microsoft is embedding AI agents directly into Windows via Copilot Vision, while Google’s Gemini Pro is showing off powerful multimodal capabilities, including the ability to generate learning apps from YouTube videos.

This episode covers questions of privacy, control, and ownership. If agents are watching our screens and reading our files, who benefits? And how can users share in the upside of the AI economy?

Thomas offers one possible path forward: agentic ecosystems like Olas, where open-source agents and incentivized staking models shift the power dynamic back to the user.

Resources mentioned in this episode:

SmolAgents

Open Computer demo

Qwen2 from Alibaba

E2B Linux sandbox

Microsoft Copilot Vision

Google Gemini Pro + Flash

🔔 Don’t forget: Agents Unleashed is also a global event series on agentic AI — the next event is June 30th in Cannes, France. RSVP now!

Sponsored by Olas — the best way to own and monetize your AI agents. Learn more

🎵 Theme music: "Forward" by Grand Project on Pixabay