The AI Daily Brief: "How to Use Claude's Massive New Upgrades"
Host: Nathaniel Whittemore (NLW)
Date: March 25, 2026
Episode Overview
NLW explores the wave of groundbreaking upgrades recently released for Anthropic’s CLAUDE, including both Claude Code and Cowork. Focusing on practical uses, he details new features—like remote control, Dispatch, channels, scheduled tasks, and full computer control—positioning Claude as an always-on, delegation-ready AI work companion. The episode compiles real user experiences and expert commentary, illustrating how these upgrades shift the core paradigm of digital labor.
Key Discussion Points and Insights
1. Claude’s Update Lineage: From Models to Full-Orchestration Agents
- Model Evolution: The advancement from Opus 4.5/GPT 5.2 to Opus 4.6 and GPT 5.4 marked a jump in capability (05:10).
- OpenClaw’s Influence: OpenClaw's launch initiated trends toward agentic, persistent AI tools that can be delegated complex workflows. Anthropic’s "Claudification" emulates and advances these (06:12).
2. Remote Control: Seamless Desktop-to-Mobile Tasking
- Feature Description: Allows users to start a CLAUDE Code session on the desktop and seamlessly continue from their mobile device (08:45).
- Technical Flow: Initiate via terminal, receive a session URL or QR code, pick up work on another device.
- Key Benefit: Maintains a local session using the machine’s full environment (files, servers, tools).
- Notable Quote:
“Claude remote control is extremely nice, can edit on macOS or iOS and Claude app on my production server from anywhere.”
— Peter Levels (12:10) - Deeper Shift:
“That’s not a productivity feature, that’s a relationship shift. You stop thinking of it as a tool you operate and start thinking of it as something you delegate to and check in with—different mental model entirely.”
— Gagan Soluda (13:05)
3. Dispatch: Persistent, Orchestrated Conversations Across Devices
- Feature Description: Persistent Claude Cowork thread that spans desktop and phone; continuously aware of workflows and context (16:25).
- Capabilities: Assign complex tasks, receive updates, and interact with multiple parallel sessions, all from a single conversational thread.
- Illustrative User Story: Pavel Huron directs multiple knowledge and dev tasks while out with his kids—Claude executes in parallel, only requiring brief direction.
- Key Insight:
“The model isn’t grind during gaps. It’s design your day differently because the work runs without you sitting in front of it.”—Pavel Huron (20:28) - Notable Quote:
“Dispatch is not Claude chat on your phone. Dispatch is an orchestrator…Your phone is the command chair. Your desktop does the heavy lifting.”
— Pavel Huron (21:55) - Security Note:
“Covers 90% of what I was trying to use OpenClaw for, but feels far less likely to upload my entire drive to a malware site...better integration with Gmail, browsers, etc.”
— Ethan Moloch (24:03)
4. Channels: Event-Driven Claude Code Sessions via Telegram, Discord, and More
- Feature Description: Claude Code can now be controlled via “Channels” (e.g., Telegram/Discord MCP servers) triggering and responding to events in real time (27:13).
- Event Push Model: Allows events like CI failures, payloads, or chat messages to automatically trigger responses from Claude, even when the user isn’t present.
- Flexibility: Particularly aimed at developers for hackable, custom event workflows.
- Notable Quote:
“Channels are essentially MCP servers that push events into a Claude code session, letting Claude react to the outside world beyond the terminal.”
— Dario (31:50) - Use Case Example: Connecting Sentry monitoring to Claude so it can react to code or production errors automatically.
5. Scheduled Tasks: Local and Cloud Scheduling
- Types of Scheduling:
- Cowork Scheduled Tasks: For routine briefings, report generations, or updates (34:25).
- Local Scheduled Tasks (Claude Code Desktop): E.g., checking logs every few hours, creating PRs.
- Recurring Cloud-Based Tasks: Run even when user's device is offline (35:45).
- Quote & Use Case:
“Set…a schedule and a prompt. Claude runs it via cloud infra on your schedule so you don’t need to keep Claude code running on your local machine.”
— Noah Zwieben (36:10)
6. Computer Use: Claude Can Now Operate Your Entire Desktop
- Major Upgrade: Claude can now control mouse, keyboard, and screen, using any application as if it were a human operator (38:11).
- How It Works: Uses connectors first when available; otherwise, directly manipulates desktop environment to complete tasks.
- Implications:
- Delegated end-to-end task execution, even in legacy/non-API apps.
- Begin work from your phone, complete it automatically on your desktop while away.
- Notable Quotes:
- “Claude can now control your entire computer with one prompt forever…Not a sandbox, not a simulation. You’re real, Jarvis. This is absolutely insane. The LLM era is over.”
— Sweebirtabo (41:05) - “It figures out what you actually meant when you said handle this for me. That’s the shift.”
— Gagan Soluda (43:00) - “Claude is no longer just a tool that uses the computer, it operates inside it as a true execution layer. It can replicate everything a human does…”
— Bilal (43:45)
- “Claude can now control your entire computer with one prompt forever…Not a sandbox, not a simulation. You’re real, Jarvis. This is absolutely insane. The LLM era is over.”
- Enterprise Implication:
“Most work requires hopping between multiple applications…Now we will have agents that are the equivalent of having an expert programmer…This opens up the broadest set of agentic use cases.”—Aaron Levy, Box (44:35)
7. Quality of Life Improvements and Additional Upgrades
- Projects in Cowork: New project management concept for improved workflow organization (47:18).
- Code Review: Automated multi-agent code review for bug hunting (47:55).
- Massive Context Windows: 1-million-token context window now available for deep analysis in Opus/Sonnet models (48:25).
- App Integrations: Improved chart/diagram creation, enhanced Excel and PowerPoint add-ins, and skill sharing for automated workflows (49:10).
- Expanded Access: Memory and connectors are now available on the free plan; plugin marketplace for enterprises.
- Productivity Impact:
“A very different software development process is possible with large strategic implications.”
— Ethan Malik (50:51)
Memorable Moments & Closing Thoughts
- “This is kind of why you’re going to want to go hack at all of these to understand how they fit your own use cases.” (33:09)
- “Effectively the idea is an always-on, context maintaining, persistent, interactive orchestration experience where work is happening all the time even when you’re not doing it.” (33:35)
- “Lots of change coming in the world of work agents and it’s going to get pretty wild.” (46:00)
Quick Reference Checklist (Timestamps):
- Remote Control: 08:45–15:30
- Dispatch: 16:25–24:04
- Channels: 27:13–33:09
- Scheduled Tasks: 34:25–36:10
- Computer Use: 38:11–46:00
- Quality of Life Updates: 47:18–50:51
A detailed checklist with links can be found at Play AIDaily Brief AI.
Summary
Anthropic’s recent upgrades to Claude are redefining what it means to work with AI—transforming it from a reactive tool to a proactive, context-rich agent capable of orchestrating complex flows, operating across your devices, and even navigating your desktop like a human assistant. These developments herald a new AI-powered paradigm for productivity, delegation, and the structure of digital work itself.