@kozlovski: why agents need VMs, not containers with David Crawshaw, ex-CTO & co-founder of Tailscale now co-founder and CEO of exe…
Summary
David Crawshaw, ex-CTO of Tailscale, discusses why virtual machines are a better abstraction than containers for AI agents, and shares insights on building a new cloud provider called exe.
View Cached Full Text
Cached at: 05/23/26, 08:16 PM
why agents need VMs, not containers
with David Crawshaw, ex-CTO & co-founder of Tailscale
now co-founder and CEO of exe - a new up-and-coming cloud provider
Timestamps
(0:00) why build a new cloud? (2:07) why Docker isn’t enough for agents (12:28) why AI-friendly is developer-friendly (20:32) why VMs are the right abstraction (and the serendipity of just dropping an idea prompt from your phone) (28:30) the exorbitant price of IOPS in the cloud (32:21) Cloud Discounts (33:40) the rise of self hosting (41:25) Shelly and AI ops agents (48:10) the hard problem with AI SREs (53:00) parting thoughts and early EC2’s noisy neighbor shenanigans
Similar Articles
I am building a cloud
Tailscale co-founder David Crawshaw announces Series A for exe.dev, a new cloud provider aiming to fix fundamental abstraction mismatches in current clouds like rigid VM sizing and limiting PaaS layers.
@walden_yan: If you're building your own cloud agent like Devin or Ramp Inspect, there's lots of great details here on setting up VM…
A deep dive into building cloud agents with Walden Yan (Cognition) and Cole Murray (OpenInspect), covering VM setup, computer use, memory, and the rise of async agents in the AI engineering landscape.
@kylejeong: Containers are fast. VMs are safe. Everyone building agent infra needs both. That's not just an opinion, it's the probl…
The author argues that building agent infrastructure requires both the speed of containers and the safety of VMs, highlighting AWS Firecracker as a solution that combines both.
@thsottiaux: We are continuing to invest in making agents work better on Windows. Highly recommend reading David's engineering post …
OpenAI is improving agent support on Windows by implementing a custom sandbox for Codex, addressing OS-level isolation challenges to ensure safe and efficient operation.
How do you make agents run for hours, and what architectures are actually agent-friendly?#deep-dive #vibe-coder-issues
The author explores two key challenges for AI coding agents: ensuring long-duration autonomous execution (hours) and designing agent-friendly architectures for local applications. They propose an explicit knowledge organization stage to manage messy context before planning and execution.