Hey everyone,

Hope you had a great week!

I've been diving deep into the AI and productivity space lately, and honestly, there's so much happening right now that it's hard to keep up.

I wanted to share some resources and articles I came across this week that got me thinking—mostly around Notion AI, agentic workflows, and how we're all learning to work differently with these tools.

Some of these are super tactical, others are more big-picture thinking. Either way, I found something interesting in each one, and I think you will too.

Let's jump in.

Connect your email, and you’ll instantly get a CRM with enriched customer insights and a platform that grows with your business.

With AI at the core, Attio lets you:

Join industry leaders like Granola, Taskrabbit, Flatfile and more.

👉 Try Attio Pro for free

Notion dropped their first major update of 2026, and it's pretty stacked.

What I'm most excited about? AI Notes on mobile that actually work in the background. You can start a transcription while commuting or walking, and it keeps running even when you lock your phone. That's huge for anyone who's tried to take meeting notes on the go.

They're also giving us full Agent support on phones now, which means you can kick off tasks and let them run while you do other stuff. Plus, you can now switch between GPT-5.2, Claude Opus 4.5, and Gemini 3 while keeping your context intact.

I love that Notion is positioning itself as model-agnostic—your memory lives in the workspace, not the model.

Oh, and there's a People Directory now for Enterprise teams, improved Jira integration, and some serious performance boosts (27% faster on Windows). If you're using Notion for team collaboration or automation, this update is worth checking out.

🔗 Read more

Okay, the title is a bit clickbait-y, but this video tutorial is actually fascinating.

Francesco D’Alessio built an entire content planning system using Claude Co-work and markdown files instead of traditional project management tools.

The setup uses markdown files as persistent "brains" that remember rules and briefs better than context windows alone.

What I found interesting is the voice-first approach—he starts by talking through problems using a voice app before building anything.

That helps create better-structured systems from the start. The whole thing syncs with Google Drive for backup, and you can update rules just by talking to it naturally.

It's a cool peek at how people are experimenting with AI for personal workflows, especially if you're curious about alternatives to database-heavy systems.

🔗 Watch the Video

If you're building workflows with Notion agents or thinking about AI security, this one's important.

Notion published a detailed breakdown of how they're protecting against prompt injection—basically when malicious commands hidden in content try to manipulate AI systems.

There’s a few layers to this: detection systems for hidden commands in files, admin controls to restrict web access, and user approval workflows for suspicious links.

I appreciate their honesty about limitations—they're clear that no system is perfect, kind of like email protection against phishing.

The Shared Responsibility Model is worth understanding too. Notion handles system security and agent safety, but users need to manage access, policies, and external content settings. Essential reading if you're deploying agents in your workspace.

🔗 How Notion 3.0 protects against prompt injection risks

This podcast with Reid Hoffman and Dan Shipper is packed with predictions about where AI is heading. Hoffman's big call?

By the end of 2026, thriving companies will be recording every meeting and using agents to amplify their work processes.

Bold statement, but he makes a compelling case.

He also talks about orchestration—managing groups of agents—as the critical new skill emerging this year. We're moving from "coding agents" to "agents for everything," and he expects 10-100x more people will experience what it's like to have AI doing productive work while they grab coffee.

There's also a fascinating discussion about the 9-to-5 work model becoming extinct (personally not sure about this prediction). Not because we'll work less, but because work becomes more entrepreneurial—variable intensity with agents running in parallel. Worth a listen if you're thinking about how AI changes organizational operations.

🔗 Watch the conversation

This article has been read 170 million times (!!!) on X.

This one's less about AI and more about identity-level transformation. Dan Koe argues that New Year's resolutions fail because they're surface-level behavior changes rather than identity shifts.

The bodybuilder doesn't "grind" to eat healthy—they can't imagine living any other way.

He provides a one-day protocol for psychological excavation: morning journaling to uncover your "anti-vision" (the life you don't want), daytime pattern interrupts to break autopilot, and evening synthesis to crystallize goals. The framework treats life like a video game with missions, boss fights, and quests.

Productivity tips and methodologies can get stale. Many writers nerd out on systems but neglect underlying identity patterns that Dan Koe dives deep on in this article.

The gamification framework at the end is immediately applicable to how we organize goals and projects.

🔗 Read the full article

Claude just launched Cowork—basically Claude Code but with a visual interface for non-developers.

A wild stat: the Claude team build Cowork in just a week and a half. Insane.

What's cool is it runs locally on your computer and can access files and browse your logged-in Chrome sessions (Gmail, analytics, social feeds).

The async workflow is genuinely different from typical chatbots—you queue up tasks like leaving notes for a colleague instead of waiting for each response.

Honestly, I’ve briefly tested it to see how it manipulates files and exploring different use cases, but I’m super excited about AI tools that go beyond chat.

No other company is offering a GUI-wrapped agentic assistant with local computer access for non-developers. If you're on Claude Pro and curious about async AI work, it's worth testing out!

🔗 Read more

Another one from Every.to (sorry, it’s my favorite AI publication by far), this comprehensive guide is for anyone building AI-powered applications.

It introduces "agent-native" architecture—where AI agents with atomic tools operate in loops to achieve outcomes rather than executing pre-written code.

The five core principles are gold: Parity (agents can do anything users can), Granularity (tools as primitives, features as outcomes), Composability (new features via prompts), Emergent Capability (agents accomplish unanticipated tasks), and Improvement Over Time (context accumulation and prompt refinement).

What I love is the practical implementation patterns—file-based architectures, mobile checkpointing, avoiding common anti-patterns. If you're setting up custom agents in Notion or building automation workflows, this framework helps you think about making your workspace "agent-ready."

🔗 Read the guide

That's it for this week! A lot of these articles intersect around a central theme: we're learning to work differently with AI. Whether it's async agents that run while you're away, markdown-based workflows, or agent-native architecture—the common thread is moving from "AI as chatbot" to "AI as colleague."

Let me know what resonated with you. I'm always curious what you're experimenting with in your own workflows.

Until next time,
Dave

ps. if you’ve noticed a different look, it’s because we’ve moved to Beehiiv!