Did you know that we have LinkedIn, Instagram and X accounts
that you can follow?

Hi {{first_name|everyone}},

Perplexity wants AI to do more than finish a task, and its new Teammates are built to remember what still needs doing and follow through on it. After that, we’ve got a look at how an expert is using AI on serious work, a few handy tools, and some creative learning projects made with AI. Here’s the full rundown.

🤿 DEEP DIVE

Not just do the work, but remember what comes next.

There's a difference between asking AI to do something and giving it something to be responsible for.

And apparently, Perplexity wants to tackle the second one.

The company has launched something called Teammates, which are basically AI agents designed to take ownership of ongoing projects. Not just complete a task, hand you the result, and wait for another prompt.

The idea is that you give an agent a job, connect it to your tools, and it keeps track of what needs doing, even as priorities change.

And the distinction is actually pretty interesting.

Perplexity already has Computer, which can handle complicated projects using different AI models and connected tools.

But Teammates adds something on top of that. Instead of you having to organize every new task, the agent coordinates Computer sessions, remembers previous work, and can even hand tasks to other agents.

Take the sales example Perplexity gives.

Say a regular customer suddenly stops buying. A sales teammate could notice that, check their order history, look through support issues, and figure out what might have changed.

Then it could prepare a relevant offer or flag a problem that needs fixing before anyone reaches out.

And it doesn't just disappear after doing the research. It keeps sharing updates, requesting approvals, and following through until the work is finished.

That's a pretty different relationship from typing a prompt and getting an answer.

One detail I liked is how Teammates handles instructions learned along the way.

Perplexity gives an example where someone tells an operations teammate that a particular agency needs two weeks' notice.

The agent remembers that and factors it into future project schedules.

Simple example, but it explains the idea nicely. The agent isn't just remembering a conversation. It's supposed to use that information when planning future work.

Teammates can also work together. One might investigate bugs on a website, then pass its findings to another agent responsible for preparing fixes.

And because they can join Slack channels, actual colleagues can assign them tasks or ask for updates too.

There's still a human review process. Users can inspect completed work, approve actions, redirect an agent, or take over its desktop.

Each teammate also runs in its own separate environment, with access controlled through Computer's existing permissions.

For now, Teammates is launching on the web for Perplexity Max subscribers, with iOS support coming soon.

I think the clearest way to understand the whole thing is that Perplexity isn't just trying to make AI better at completing tasks.

It's trying to make AI responsible for remembering which tasks still need completing.

And that little difference is basically the whole story.

🤝 FROM OUR FRIENDS

A quick share we thought you might find useful

People assume that real businesses get built by other people with more resources. It's how great ideas get stuck in notes apps and conversations.

But you don't need six months and a developer's salary to validate your concept. You just need a weekend. Describe what you want to exist and Lovable builds it: a real site, a real product, ready to accept real payments from real customers. Not a mockup to show people. An actual thing they can use.

You don't need anyone's sign-off to get started. Take $5 off your first month.

Find out this weekend

_{^{Discount of $5 applies to first month of any paid plan. You need to sign up via this webpage in order to claim the discount. The discount applies to new and returning users without an active paid plan. Terms apply.}}

🔥 WHAT’S THE BUZZ

What people are building, sharing, and talking about

Linus Torvalds used Claude Code while developing an experimental improvement to the Linux kernel.

The work targets a performance bottleneck that occurs when multiple programs read the same file simultaneously. In tests on his own computer, the proposed changes increased throughput from approximately 10 million to between 30 million and 33 million reads per second.

Claude Code helped generate code, tests, reviews and a draft commit message, while Torvalds directed the work and performed the measurements.

This is a useful example of AI assisting someone who already understands their field deeply.

The takeaway isn't that AI replaces expertise. It's that experienced professionals can use AI to handle implementation work, investigate alternatives and accelerate tasks while retaining responsibility for the results.

That approach applies to software development, research, analysis and other knowledge-intensive work.

Nikita Bier sparked a discussion by arguing that traditional technical advantages are becoming less effective at protecting software businesses.

His reasoning is that AI makes it increasingly easy for competitors to recreate software features. He identified 3 advantages that remain difficult to copy: proprietary data, network effects and physical components.

This raises an interesting question for anyone building a business, creating software or developing professional skills.

If AI makes producing software easier, the value may increasingly come from understanding customers, building relationships, collecting useful knowledge and solving problems that aren't easily copied.

The same idea applies to individual careers. Knowing how to use AI matters, but understanding which problems are worth solving may become even more important.

⚒ TOOL SNAPSHOTS

Futuristic tools within AI, no-code, and productivity

👀 ON OUR RADAR

One more thing we thought you'd like

Get all the facts and latest updates on global affairs, politics and more from 1440’s Daily Digest: unbiased news from 100+ sources, read in five minutes. Join 4+ million readers. It's free.

Get the facts free

🤖 AI MADE THIS

Interesting and inspiring creations made with AI

CopperForgeAI used DeepSeek and Claude to help build a free collection of interactive 3D science lessons. Instead of just watching a demonstration, students can experiment with everything from chemical reactions to magnetic fields.

AI used:

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!