Don't Worry About the Vase Podcast Claude Code, Cowork and Codex #6: Claude Code Auto Mode and Full Cowork Computer Use
Mar 25, 2026
Quoted contributors like Adam Wolf, Dean W. Ball and Gary Tan offer short takes on Anthropic updates. They unpack Claude Code Auto Mode, Claude Cowork controlling keyboard/mouse/screen, and risks of full computer access. Conversation covers agentic coding tradeoffs, code review automation, Dispatch workflows, and fast product iteration with practical tips and cautionary notes.
AI Snips
Chapters
Transcript
Episode notes
Prefer A Desktop For Persistent Agent Tasks
- Keep the machine you let Claude fully control as a desktop that's always on and plugged in.
- Persistent, powered machines like Mac Mini style desktops work better for long-running agent tasks than laptops.
Not Everything Should Be Done By Voice Or Agents
- Human interaction patterns with computers often compress into macros, shortcuts, and clear commands better than freeform natural language.
- Zvi argues many tasks still favor direct use or better interfaces rather than full agent substitution.
SuperApp Combines Code Models And UI
- Combining coding models and UI (super app) makes sense to create a smooth bridge into agentic coding workflows.
- Anthropic already integrates Claude Code and Cowork; OpenAI plans a Codex+ChatGPT+Atlas desktop SuperApp.