🎙️ How I AI: How the engineer behind Claude Cowork actually uses Claude Cowork & What launched at Google I/O 2026

2026-05-25 · Source: Lenny's Newsletter · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Robotics & Autonomous Systems · Depth: Intermediate, extended

Summary

Felix Rieseberg, Anthropic's engineering lead for Claude Cowork, demonstrates practical AI applications, including building a 3D floor planner from a 2D house plan, leveraging email as a personal inventory database, and creating live, auto-updating personal dashboards. He also showcased a \$20 hardware "Claude buddy" for task approvals. Rieseberg emphasizes overcoming the psychological barrier to AI adoption and debugging workflows rather than the model when errors occur. Concurrently, Google I/O 2026 introduced several AI advancements, such as Gemini 3.5 Flash, an agentic coding model reportedly four times faster than competitors, and Antigravity 2.0, an IDE aiming for feature parity with Claude Code. Other launches included Omni for production-quality video, Flow for video editing, and Stitch and Pomelli for design and marketing. A significant critique noted was Google's persistent "launch-to-availability gap," with many announced features not yet functional.

Key takeaway

For AI Engineers and Product Managers evaluating new tools, recognize that Anthropic's Claude offers robust personal and professional automation, particularly with live artifacts and workflow debugging. Your focus should shift from model capabilities to creative problem abstraction. Be wary of Google's "launch-to-availability gap" at events like I/O 2026; prioritize solutions that are proven functional over announced features, even if they promise advanced capabilities like Gemini 3.5 Flash or Omni video.

Key insights

Overcome psychological barriers to AI adoption by abstracting tasks and treating AI as a collaborative problem-solver.

Principles

• Go one abstraction layer up for tedious tasks.
• Debug AI workflows, not the model itself.
• Opus interprets intent; Sonnet handles well-scoped problems.

Method

To integrate AI, identify annoying, non-creative tasks, then ask the AI to perform them, and subsequently to automate the process further.

In practice

• Use email as a personal inventory database.
• Build live, auto-updating personal dashboards.
• Connect low-cost hardware for AI interaction.

Topics

• AI Workflows
• Claude Cowork
• Google I/O 2026
• Gemini 3.5 Flash
• Live Artifacts
• AI Adoption Barriers
• Hardware Integration

Best for: Machine Learning Engineer, NLP Engineer, Entrepreneur, AI Engineer, AI Product Manager, Director of AI/ML

Related on AIssential

• Anthropic’s Claude Cowork finally lands on Windows — and it wants to automate your workday — Anthropic's Claude Cowork for Windows expands AI agent capabilities, intensifying competition and reshaping enterprise software.
• Anthropic Releases Claude Cowork: Is Claude Winning the Enterprise AI Race? — Claude Cowork offers autonomous AI task execution and file management, enhancing enterprise productivity across diverse workflows.
• Cowork and plugins for teams across the enterprise — Anthropic's Cowork updates enable deep enterprise customization and control over Claude's AI agent capabilities via private plugin marketplaces.
• Gemini is Now the Best All-in-One AI & More AI Use Cases — Leading AI developers are rapidly advancing multimodal models, agentic capabilities, and user-friendly applications.
• Anthropic Shipped Claude Opus 4.8 Yesterday. — Claude Opus 4.8 enhances AI honesty and introduces Dynamic Workflows for complex, large-scale coding tasks.
• How the engineer behind Claude Cowork actually uses Claude | Felix Rieseberg (Anthropic) — AI's true power lies in automating tedious tasks and enabling creative problem-solving, especially when users provide rich context and trust its autonomous execution.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Lenny's Newsletter.