scripod.com
Vibe Check: Claude Cowork Is Claude Code for the Rest of Us

Overview

Shownote

Highlights

Transcript

Chapters

Pins

Vibe Check: Claude Cowork Is Claude Code for the Rest of Us

AI & I

1 DAYS AGO
AI & I

AI & I

1 DAYS AGO
OverviewShownoteHighlightsTranscriptChaptersPins

Shownote

Anthropic just dropped Claude Cowork—essentially Claude Code for everyone, not just engineers—and we got to chat about it with a product engineer at Anthropic who helped build it. In this live Vibe Check, Dan Shipper and Kieran Klaassen explore the new in...

Highlights

In this live session, Dan Shipper and Kieran Klaassen dive into Claude Cowork, a new experimental interface from Anthropic that reimagines how users interact with AI. Joined by Felix Rieseberg, a product engineer at Anthropic, they explore its core features and design philosophy in real time, focusing on how it enables non-technical users to delegate complex, long-running tasks to an AI agent.
03:36
Claude Cowork drafted usable remarks from Gmail with minimal edits needed.
07:39
Claude Cowork generates a detailed taxonomy from 'The Outsider' book
11:39
Claude Cowork fuses chat and code features to enable non-technical users to run complex, long-term tasks via browser automation.
28:23
Claude Cowork enables intelligent folder selection during file organization tasks.
34:32
Character limits hinder detailed responses in the brainstorming process.
37:57
The separate tab enables rapid iteration by running locally and involving users early.
46:32
Power users value deep product familiarity over marginal productivity gains in UI/UX improvements.
49:41
Emergent capabilities arise when agents combine generalizable tools in unexpected ways.
1:02:19
Felix Rieseberg built it in a week and a half, which is now considered normal.
1:08:28
Skills in Claude Cowork are more than prompts — they're hackable workflows for tasks like 3D printing and proofreading.

Chapters

What is Claude Cowork
00:00
First demo: competitor analysis
02:36
Email drafting that sounds like me
03:33
Calendar audit running for an hour
06:18
Book taxonomy demo
07:39
PostHog analytics via Chrome browsing
08:42
Chat vs Code vs Cowork: when to use what
14:36
Felix from Anthropic joins
31:06
Why they built it in a week and a half
36:39
Design decision: why a separate tab
37:57
Skills as the primary hackable surface
43:57
Agent-native architecture principles
49:36
The origin story of skills at Anthropic
56:57
Our final rating
1:03:00

Transcript

Dan Shipper: If you're a non-technical person, you are used to a world where you send a prompt and then you get a response within a couple of minutes. And once you send a prompt or a chat, you can't do anything else with that AI. This is built for working ...