Anthropic Measures AI Work Impact

Hi everyone,

Anthropic just released an amazing report that everyone should read where they measure how AI is used at work. Also, Manus keeps shipping even after getting acquired by Meta! Now they have this phone app launch capability as well.

Let’s get right into it.

In this issue:

• 🤝 In Partnership: Turn ideas into videos

• 🤿Deep Dive: Anthropic tracks how AI changes jobs

• 🤝 Powered by: Your input helps improve the ads you see

• 🤿Deep Dive: Manus streamlines app testing

• ⚒️Tool Snapshots: Tools for AI, no-code, and productivity

• 🖼️AI Art: Examples of great and trending AI art

• 🎯Quick Favor: Fill out TIW survey form

🤝IN PARTNERSHIP WITH MEDEO

Bring Your Ideas to Life with a
Single Conversation

What if creating a professional video was as simple as chatting?

With Medeo, it is. Every video begins with a spark and instead of struggling with timelines or editing software, you describe your vision in natural language. Medeo handles the rest: script, visuals, sound, and editing.

Whether you imagine a cinematic moment, an anime short, a recreated scene, or an educational explainer, Medeo helps you shape it in a way that feels natural, fast, and effortless.

What becomes possible with Medeo:

Start with a thought. Watch it become a finished video.
With Medeo, you don’t just make content, you create stories that scale. Whether you’re a marketer, educator, or storyteller, Medeo gives you professional results without the complexity.

🤿 DEEP DIVE

Claude Speeds Up Complex Work More Than Simple Tasks

Intelligence: Anthropic's fourth Economic Index tracking real-world Claude usage found that more complex tasks get sped up most. College-degree-level work (16 years of education) accelerates 12x on Claude.ai versus 9x for high school-level tasks, though success rates drop slightly from 70% to 66% as complexity increases.

• The report introduced five "economic primitives" measuring task complexity, skill level, purpose, AI autonomy, and success across November 2025 conversations using Claude Sonnet 4.5.

• Claude maintains 50% success rates on tasks taking humans 3.5 hours via API and 19 hours on Claude.ai, substantially longer than METR's 2-hour benchmark, likely because users break complex tasks into smaller steps and select tasks they expect will work.

• Removing Claude-covered tasks from jobs would cause net deskilling as AI handles higher-education tasks (14.4 years average versus economy-wide 13.2 years), affecting technical writers, travel agents, and teachers most, though this assumes no labor market adjustments as the economy adapts.

• Accounting for task reliability cuts projected US labor productivity growth from 1.8 percentage points annually to 1.0-1.2 points, still returning productivity to late 1990s rates, with potential for much higher gains as models improve and workplace AI sophistication increases.

• Geographic patterns show lower-income countries use Claude mainly for education while higher GDP nations deploy it for work and personal tasks, with US state-level concentration decreasing so rapidly that the model predicts equalized usage within 2-5 years.

🤝POWERED BY BEEHIIV

Quick question about newsletter ads

We’re running a super short survey to see if our newsletter ads are being noticed. It takes about 20 seconds and there's just a few easy questions.

Your feedback helps us make smarter, better ads.

Take the survey now.

🤿 DEEP DIVE

Manus Simplifies Mobile App Testing Process

Intelligence: Manus launched App Sharing and Testing, a feature that automates the complex process of getting mobile apps into testers' hands by packaging Android apps into AAB format for Google Play and directly uploading iOS builds to TestFlight via App Store Connect.

• The feature handles app creation, build packaging, and TestFlight review submission automatically, turning what traditionally takes days into a guided workflow that completes in minutes.

• .For iOS apps, Manus creates the app in your developer account, packages the build, uploads to App Store Connect, and automatically sends a TestFlight invitation to your developer email.

• The feature doesn't publish apps directly to stores but focuses on the testing phase, providing guided workflows for inviting internal and external testers and demystifying the path to eventual launch.

• Manus addresses the bottleneck where creative work is done, but sharing becomes a friction point, letting product builders stay in creator mode rather than switching to a release engineer mindset for hours or days navigating platform-specific configurations.

• Available to all Manus users with the Develop Apps capability, the feature requires an Apple Developer account for iOS and a Google Play Developer account for Android, with users maintaining control over final publishing decisions through their respective developer portals.

⚒️ TOOL SNAPSHOTS

Futuristic tools within AI, no-code, and productivity

• 📲 Noodle Seed - Effortlessly create a branded AI app and boost sales.

• 🤖 OS Ninja - Transforming complex code into interactive learning adventures

• 🖥️ Updatest - Unified dashboard for seamless and secure Mac updates.

• 🧩 FlowGenie - Visualize and build workflows with ease, no code required.

• 📍 Tofu Maps - Create, save and share personal maps effortlessly.

🖼️ AI ART

Examples of great and trending AI art

🎯QUICK FAVOR

I’m updating how I write TIW and want to make sure it fits what you actually do all day.
If you’ve got 30 seconds, answer a few questions about your role, tools, and goals:
👉Fill out the TIW reader survey
It’ll help me send you fewer “meh, not for me” issues and more “oh, this is exactly what I needed” ones.

ℹ️ ABOUT US

The Intelligent Worker helps you to be more productive at work with AI, automation, no-code, and other technologies.

We like real, practical, and tangible use-cases and hate hand-wavy, theoretical, and abstract concepts that don’t drive real-world outcomes.

Our mission is to empower individuals, boost their productivity, and future-proof their careers.

We read all your comments - please provide your feedback!