In partnership with

Ever wished you could just tell your laptop to do something while you're away?

Login or Subscribe to participate

I'll be honest, I've been feeling a little embarrassed opening this newsletter lately.

Because every other week, it's Claude, or Anthropic or something Anthropic shipped that I can't stop thinking about.

Across developer tools, desktop automation, the API, models, all shipping in parallel, 74 releases in 52 days is a firehose.

And I know some of you are tired of hearing about Claude. I get it.

But I'd be doing you a disservice if I ignored what's happening just to be contrarian.

So yes. This week is Claude again.

Anthropic dropped something called Dispatch and here’s what I tested. But before that, some catchup:

NEWS NEWS NEWS

TOOLS of the Week

1. Tinker: Shopify's free AI creative toolkit that can generate product photos, videos, 3D models, social ads, virtual try-ons, and more, all from your phone. No design experience needed, no subscription required.

2. Expect: One command scans your code changes, generates a test plan, and runs it against a live browser automatically. Built for AI coding workflows where you're shipping fast and don't want to write tests by hand.

3. Paper Snapshot: A Chrome extension that lets you capture any element on a webpage and paste it directly into Paper as an editable layer without screenshots or copy-pasting. Useful if you're mocking up redesigns of live sites without starting from scratch.

What is Dispatch?

The core idea is almost simple: you type something on your phone, Claude does it on your computer.

It can open apps, browse Chrome, creates files and send messages while you're away from your desk.

I know. Every AI demo sounds like this and then disappoints you in real life. So before I explain anything, here's what it did for me.

I asked Claude to go to X, scroll through my feed, find 5–10 high-performing posts from the last 24 hours, write a full script for each one, and put everything into a PDF in my downloads folder.

It came back to a PDF with 7 posts, 7 scripts, and 7 virality scorecards sitting in my downloads and I hadn't even touched my laptop once.

Midway through, Claude decided the search results were too broad and narrowed them on its own. Nobody told it to do that.

That's an hour of work done with one prompt.

Setting it up (actually 2 minutes)

1. Download the Claude desktop app

2. Open it > click Cowork at the top > click Dispatch in the sidebar

3. It shows you a QR code, scan it with Claude on your phone

4. Go into Settings and turn on Computer Use

Without Computer Use on, Claude only works through its built-in connectors but with it on, it can take over your mouse, keyboard, and screen.

That's what makes the magic happen.

What I tested

Task 1: Slack message without leaving the couch

Typed on my phone: "open my Slack and text [name] what's the schedule for tomorrow?"

Claude asked permission to use Slack and I tapped allow.

Slack opened on its own, found the contact, clicked into the DM, started typing letter by letter like someone was sitting at my desk.

Asked me to confirm before sending and done.

Task 2: content research in one shot

The X research task I described above.

Output:

Each script came with the source post, engagement numbers, why it was performing, and a full ready-to-record script with hooks, B-roll directions, and on-screen text.

Script 5 alone had 4.9K likes and 743 reposts. Claude found that, understood why it was working, and wrote a script around it.

Check it out 👇

Viral_Scripts_March31.pdf

Viral_Scripts_March31.pdf

74.56 KBPDF File

Task 3: Scheduling

I scheduled that same task to run every alternate morning at 9am.

Now before I sit down, the PDF is there and the team and I save 3-4 hours every week.

Someone on Twitter described it as: "combine this with /schedule and you've basically got a background worker that can interact with any app on a cron job. That's not an AI assistant anymore, that's infrastructure."

And I couldn’t agree more.

My Take

Dispatch is fun.

But I've been sitting with it for a few days and I'm genuinely not sure how to feel about it beyond that, and I think that's the honest answer for something this early.

It's a research preview. Forming a strong opinion and committing to it feels premature, especially when Anthropic is shipping at the pace they are.

They're clearly experimenting, seeing what sticks, figuring it out in public. But it doesn't feel haphazard.

There are decisions here that feel deliberate, like the permission flow, the way it escalates from connectors to computer use, and the scheduling piece.

Where I landed: Dispatch works really well for a specific kind of task like exploratory or repetitive stuff and grunt work you'd normally have to sit through.

If the task doesn't need you to be actively present, you can hand it off and check in from your phone.

For anything more complex, it still feels too raw. It asked for permissions more often than I'd like.

Also worth knowing: if you're on a Max plan, watch your quota.

Some users reported a single Dispatch prompt eating through 10% of their monthly allowance as agentic tasks cost significantly more than regular chat.

And for now it's Mac only.

But I had a lot of fun with this. That alone makes me excited to see where it goes. What about you?

Until next time, (It'll probably be Claude again. I'm not sorry)
Vaibhav 🤝🏻

If you read till here, you might find this interesting

#AD 1

The Free Newsletter Fintech Execs Actually Read

Most coverage tells you what happened. Fintech Takes is the free newsletter that tells you why it matters. Each week, I break down the trends, deals, and regulatory shifts shaping the industry — minus the spin. Clear analysis, smart context, and a little humor so you actually enjoy reading it. Subscribe free.

#AD 2

Protect online privacy from the very first click

Your digital footprint starts before you can even walk.

In today’s data economy, “free” inboxes from Google and Microsoft, like Gmail and Outlook, are funded by data collection. Emails can be analyzed to personalize ads, train algorithms, and build long-term behavioral profiles to sell to third-party data brokers.

From family updates, school registrations, medical reports, to financial service emails, social media accounts, job applications, a digital identity can take shape long before someone understands what privacy means.

Privacy shouldn’t begin when you’re old enough to manage your settings. It should be the default from the start.

Proton Mail takes a different approach: no ads, no tracking, no data profiling — just private communication by default. Because the next generation deserves technology that protects them, not profiles them.

Reply

Avatar

or to participate

Keep Reading