In partnership with
🧠 THAT ONE AI - This Week’s Signal
Here’s what’s shaping AI right now - without the noise:
🧠 Meta launches Muse Spark from its rebuilt AI stack
⚡ Wade Foster shows what an AI Chief of Staff really looks like
⚒️ Anthropic removes the hardest part of building agents
🧰 Tools worth testing this week
🧠 Meta Rebuilds Its Stack for Muse Spark
Meta has unveiled Muse Spark, the first major model release from Alexandr Wang and Meta Superintelligence Labs.
What makes this interesting is not just the benchmark performance — it’s the claim that Meta rebuilt its AI stack from scratch to ship it.
That includes:
New infrastructure
New architecture
Fresh data pipelines
According to Wang, the result is a model that competes closely with today’s leading systems across several benchmarks.
The bigger signal: Meta isn’t just iterating.
It’s rebuilding its foundations for what comes next.
Smart starts here.
You don't have to read everything — just the right thing. 1440's daily newsletter distills the day's biggest stories from 100+ sources into one quick, 5-minute read. It's the fastest way to stay sharp, sound informed, and actually understand what's happening in the world. Join 4.5 million readers who start their day the smart way.
⚡ How Wade Foster Built an AI Chief of Staff Using the Zapier SDK
Zapier CEO Wade Foster shared one of the most grounded real-world AI workflows we’ve seen.
Using the Zapier SDK inside Cursor, he built what’s essentially an AI Chief of Staff — without servers, custom auth, or complex infrastructure.
The setup powers daily workflows like:
Morning briefs from Gmail, Calendar, and task tools
Slack thread analysis
Meeting follow-up drafts
Hiring packet reviews
Sunday customer outreach
Writing style checks
Google Docs feedback
The key insight isn’t the number of use cases.
It’s the simplicity.
The SDK already handles auth across 9,000+ apps, while Cursor writes the scripts based on plain-language instructions.
The result: a non-engineer CEO now runs 100+ automations and scripts that improve week after week.
This is one of the clearest examples yet of how AI agents are shifting from “interesting demos” into personal operating systems for work.
⚒️ Anthropic Turns Agent Building Into a Few Clicks
Anthropic just opened public beta access for Claude Managed Agents, a new platform that removes the backend complexity of deploying agents.
Instead of building infra from scratch, users simply define:
The task
The tools
The guardrails
Claude handles the rest:
Runtime
Security
Access controls
Long-running memory
Agents can now stay active for hours without losing state, and a preview coordination mode even lets one agent hand off subtasks to others.
The bigger unlock here is speed. Early teams like Rakuten, Notion, Asana, and Sentry are already using it, with Rakuten reportedly rolling agents across departments in roughly a week each.
This is Anthropic’s clearest push yet toward making agent deployment feel like launching software, not engineering infrastructure.
That One AI 🧰 TOOLBOX
A few tools quietly worth exploring:
🔗 Zapier SDK → auth + orchestration layer for thousands of apps
💻 Cursor Agents → launch and control coding agents remotely
📩 Lessie → automate outreach and follow-ups
📝 Fonic → turn messy work into interactive reports